id	sid	tid	token	lemma	pos
kim-ai-2020	1	1	Bohyun	Bohyun	NNP
kim-ai-2020	1	2	Kim	Kim	NNP
kim-ai-2020	1	3	,	,	,
kim-ai-2020	1	4	AI	AI	NNP
kim-ai-2020	1	5	and	and	CC
kim-ai-2020	1	6	Its	-PRON-	PRP$
kim-ai-2020	1	7	Moral	Moral	NNP
kim-ai-2020	1	8	Concerns	Concerns	NNPS
kim-ai-2020	1	9	AI	AI	NNP
kim-ai-2020	1	10	and	and	CC
kim-ai-2020	1	11	Its	-PRON-	PRP$
kim-ai-2020	1	12	Moral	Moral	NNP
kim-ai-2020	1	13	Concerns	Concerns	NNPS
kim-ai-2020	1	14	Bohyun	Bohyun	NNP
kim-ai-2020	1	15	Kim	Kim	NNP
kim-ai-2020	1	16	1	1	CD
kim-ai-2020	1	17	.	.	.
kim-ai-2020	2	1	Automating	automate	VBG
kim-ai-2020	2	2	Decisions	Decisions	NNPS
kim-ai-2020	2	3	and	and	CC
kim-ai-2020	2	4	Actions	Actions	NNPS
kim-ai-2020	2	5	The	the	DT
kim-ai-2020	2	6	goal	goal	NN
kim-ai-2020	2	7	of	of	IN
kim-ai-2020	2	8	artificial	artificial	JJ
kim-ai-2020	2	9	intelligence	intelligence	NN
kim-ai-2020	2	10	(	(	-LRB-
kim-ai-2020	2	11	AI	AI	NNP
kim-ai-2020	2	12	)	)	-RRB-
kim-ai-2020	2	13	as	as	IN
kim-ai-2020	2	14	a	a	DT
kim-ai-2020	2	15	discipline	discipline	NN
kim-ai-2020	2	16	is	be	VBZ
kim-ai-2020	2	17	to	to	TO
kim-ai-2020	2	18	create	create	VB
kim-ai-2020	2	19	an	an	DT
kim-ai-2020	2	20	artificial	artificial	JJ
kim-ai-2020	2	21	system	system	NN
kim-ai-2020	2	22	–	–	:
kim-ai-2020	2	23	whether	whether	IN
kim-ai-2020	2	24	it	-PRON-	PRP
kim-ai-2020	2	25	be	be	VB
kim-ai-2020	2	26	a	a	DT
kim-ai-2020	2	27	piece	piece	NN
kim-ai-2020	2	28	of	of	IN
kim-ai-2020	2	29	software	software	NN
kim-ai-2020	2	30	or	or	CC
kim-ai-2020	2	31	a	a	DT
kim-ai-2020	2	32	machine	machine	NN
kim-ai-2020	2	33	with	with	IN
kim-ai-2020	2	34	a	a	DT
kim-ai-2020	2	35	physical	physical	JJ
kim-ai-2020	2	36	body	body	NN
kim-ai-2020	2	37	–	–	:
kim-ai-2020	2	38	that	that	DT
kim-ai-2020	2	39	is	be	VBZ
kim-ai-2020	2	40	as	as	RB
kim-ai-2020	2	41	intelligent	intelligent	JJ
kim-ai-2020	2	42	as	as	IN
kim-ai-2020	2	43	a	a	DT
kim-ai-2020	2	44	human	human	NN
kim-ai-2020	2	45	in	in	IN
kim-ai-2020	2	46	its	-PRON-	PRP$
kim-ai-2020	2	47	performance	performance	NN
kim-ai-2020	2	48	,	,	,
kim-ai-2020	2	49	either	either	CC
kim-ai-2020	2	50	broadly	broadly	RB
kim-ai-2020	2	51	in	in	IN
kim-ai-2020	2	52	all	all	DT
kim-ai-2020	2	53	areas	area	NNS
kim-ai-2020	2	54	of	of	IN
kim-ai-2020	2	55	human	human	JJ
kim-ai-2020	2	56	activities	activity	NNS
kim-ai-2020	2	57	or	or	CC
kim-ai-2020	2	58	narrowly	narrowly	RB
kim-ai-2020	2	59	in	in	IN
kim-ai-2020	2	60	a	a	DT
kim-ai-2020	2	61	specific	specific	JJ
kim-ai-2020	2	62	activity	activity	NN
kim-ai-2020	2	63	,	,	,
kim-ai-2020	2	64	such	such	JJ
kim-ai-2020	2	65	as	as	IN
kim-ai-2020	2	66	playing	play	VBG
kim-ai-2020	2	67	chess	chess	NN
kim-ai-2020	2	68	or	or	CC
kim-ai-2020	2	69	driving	driving	NN
kim-ai-2020	2	70	.	.	.
kim-ai-2020	3	1	[	[	-LRB-
kim-ai-2020	3	2	footnoteRef:0	footnoteref:0	NN
kim-ai-2020	3	3	]	]	-RRB-
kim-ai-2020	3	4	The	the	DT
kim-ai-2020	3	5	actual	actual	JJ
kim-ai-2020	3	6	capability	capability	NN
kim-ai-2020	3	7	of	of	IN
kim-ai-2020	3	8	most	most	JJS
kim-ai-2020	3	9	AI	AI	NNP
kim-ai-2020	3	10	systems	system	NNS
kim-ai-2020	3	11	remained	remain	VBD
kim-ai-2020	3	12	far	far	RB
kim-ai-2020	3	13	below	below	IN
kim-ai-2020	3	14	this	this	DT
kim-ai-2020	3	15	ambitious	ambitious	JJ
kim-ai-2020	3	16	goal	goal	NN
kim-ai-2020	3	17	for	for	IN
kim-ai-2020	3	18	a	a	DT
kim-ai-2020	3	19	long	long	JJ
kim-ai-2020	3	20	time	time	NN
kim-ai-2020	3	21	.	.	.
kim-ai-2020	4	1	But	but	CC
kim-ai-2020	4	2	with	with	IN
kim-ai-2020	4	3	recent	recent	JJ
kim-ai-2020	4	4	successes	success	NNS
kim-ai-2020	4	5	with	with	IN
kim-ai-2020	4	6	machine	machine	NN
kim-ai-2020	4	7	learning	learning	NN
kim-ai-2020	4	8	and	and	CC
kim-ai-2020	4	9	deep	deep	JJ
kim-ai-2020	4	10	learning	learning	NN
kim-ai-2020	4	11	,	,	,
kim-ai-2020	4	12	the	the	DT
kim-ai-2020	4	13	performance	performance	NN
kim-ai-2020	4	14	of	of	IN
kim-ai-2020	4	15	some	some	DT
kim-ai-2020	4	16	AI	AI	NNP
kim-ai-2020	4	17	programs	program	NNS
kim-ai-2020	4	18	has	have	VBZ
kim-ai-2020	4	19	started	start	VBN
kim-ai-2020	4	20	surpassing	surpass	VBG
kim-ai-2020	4	21	that	that	DT
kim-ai-2020	4	22	of	of	IN
kim-ai-2020	4	23	humans	human	NNS
kim-ai-2020	4	24	.	.	.
kim-ai-2020	5	1	In	in	IN
kim-ai-2020	5	2	2016	2016	CD
kim-ai-2020	5	3	,	,	,
kim-ai-2020	5	4	an	an	DT
kim-ai-2020	5	5	AI	AI	NNP
kim-ai-2020	5	6	program	program	NN
kim-ai-2020	5	7	developed	develop	VBN
kim-ai-2020	5	8	with	with	IN
kim-ai-2020	5	9	the	the	DT
kim-ai-2020	5	10	deep	deep	JJ
kim-ai-2020	5	11	learning	learning	NN
kim-ai-2020	5	12	method	method	NN
kim-ai-2020	5	13	,	,	,
kim-ai-2020	5	14	AlphaGo	AlphaGo	NNP
kim-ai-2020	5	15	,	,	,
kim-ai-2020	5	16	astonished	astonish	VBD
kim-ai-2020	5	17	even	even	RB
kim-ai-2020	5	18	its	-PRON-	PRP$
kim-ai-2020	5	19	creators	creator	NNS
kim-ai-2020	5	20	by	by	IN
kim-ai-2020	5	21	winning	win	VBG
kim-ai-2020	5	22	four	four	CD
kim-ai-2020	5	23	out	out	IN
kim-ai-2020	5	24	of	of	IN
kim-ai-2020	5	25	five	five	CD
kim-ai-2020	5	26	Go	Go	NNP
kim-ai-2020	5	27	matches	match	NNS
kim-ai-2020	5	28	with	with	IN
kim-ai-2020	5	29	the	the	DT
kim-ai-2020	5	30	eighteen	eighteen	CD
kim-ai-2020	5	31	-	-	HYPH
kim-ai-2020	5	32	time	time	NN
kim-ai-2020	5	33	world	world	NN
kim-ai-2020	5	34	champion	champion	NN
kim-ai-2020	5	35	,	,	,
kim-ai-2020	5	36	 	  	_SP
kim-ai-2020	5	37	Sedol	Sedol	NNP
kim-ai-2020	5	38	Lee	Lee	NNP
kim-ai-2020	5	39	.	.	.
kim-ai-2020	6	1	[	[	-LRB-
kim-ai-2020	6	2	footnoteRef:1	footnoteRef:1	NNP
kim-ai-2020	6	3	]	]	-RRB-
kim-ai-2020	6	4	In	in	IN
kim-ai-2020	6	5	2020	2020	CD
kim-ai-2020	6	6	,	,	,
kim-ai-2020	6	7	Google	Google	NNP
kim-ai-2020	6	8	’s	’s	NNP
kim-ai-2020	6	9	DeepMind	DeepMind	NNP
kim-ai-2020	6	10	unveiled	unveil	VBD
kim-ai-2020	6	11	Atari57	Atari57	NNP
kim-ai-2020	6	12	,	,	,
kim-ai-2020	6	13	a	a	DT
kim-ai-2020	6	14	deep	deep	JJ
kim-ai-2020	6	15	reinforcement	reinforcement	NN
kim-ai-2020	6	16	learning	learn	VBG
kim-ai-2020	6	17	algorithm	algorithm	NN
kim-ai-2020	6	18	that	that	WDT
kim-ai-2020	6	19	reached	reach	VBD
kim-ai-2020	6	20	superhuman	superhuman	JJ
kim-ai-2020	6	21	levels	level	NNS
kim-ai-2020	6	22	of	of	IN
kim-ai-2020	6	23	play	play	NN
kim-ai-2020	6	24	in	in	IN
kim-ai-2020	6	25	57	57	CD
kim-ai-2020	6	26	classic	classic	JJ
kim-ai-2020	6	27	Atari	Atari	NNP
kim-ai-2020	6	28	games	game	NNS
kim-ai-2020	6	29	.	.	.
kim-ai-2020	7	1	[	[	-LRB-
kim-ai-2020	7	2	footnoteRef:2	footnoteRef:2	VBN
kim-ai-2020	7	3	]	]	-RRB-
kim-ai-2020	7	4	[	[	-LRB-
kim-ai-2020	7	5	0	0	CD
kim-ai-2020	7	6	:	:	:
kim-ai-2020	7	7	Note	note	VB
kim-ai-2020	7	8	that	that	IN
kim-ai-2020	7	9	by	by	IN
kim-ai-2020	7	10	‘	'	``
kim-ai-2020	7	11	as	as	RB
kim-ai-2020	7	12	intelligent	intelligent	JJ
kim-ai-2020	7	13	as	as	IN
kim-ai-2020	7	14	a	a	DT
kim-ai-2020	7	15	human	human	NN
kim-ai-2020	7	16	,	,	,
kim-ai-2020	7	17	’	'	''
kim-ai-2020	7	18	I	-PRON-	PRP
kim-ai-2020	7	19	simply	simply	RB
kim-ai-2020	7	20	mean	mean	VBP
kim-ai-2020	7	21	human	human	JJ
kim-ai-2020	7	22	-	-	HYPH
kim-ai-2020	7	23	level	level	NN
kim-ai-2020	7	24	AI	ai	NN
kim-ai-2020	7	25	,	,	,
kim-ai-2020	7	26	or	or	CC
kim-ai-2020	7	27	more	more	RBR
kim-ai-2020	7	28	accurately	accurately	RB
kim-ai-2020	7	29	,	,	,
kim-ai-2020	7	30	AI	AI	NNP
kim-ai-2020	7	31	at	at	IN
kim-ai-2020	7	32	human	human	JJ
kim-ai-2020	7	33	-	-	HYPH
kim-ai-2020	7	34	level	level	NN
kim-ai-2020	7	35	performance	performance	NN
kim-ai-2020	7	36	.	.	.
kim-ai-2020	8	1	By	by	IN
kim-ai-2020	8	2	‘	'	``
kim-ai-2020	8	3	as	as	RB
kim-ai-2020	8	4	intelligent	intelligent	JJ
kim-ai-2020	8	5	as	as	IN
kim-ai-2020	8	6	a	a	DT
kim-ai-2020	8	7	human	human	NN
kim-ai-2020	8	8	,	,	,
kim-ai-2020	8	9	’	'	''
kim-ai-2020	8	10	I	-PRON-	PRP
kim-ai-2020	8	11	do	do	VBP
kim-ai-2020	8	12	not	not	RB
kim-ai-2020	8	13	mean	mean	VB
kim-ai-2020	8	14	general(/strong	general(/strong	.
kim-ai-2020	8	15	)	)	-RRB-
kim-ai-2020	8	16	AI	AI	NNP
kim-ai-2020	8	17	.	.	.
kim-ai-2020	9	1	General	General	NNP
kim-ai-2020	9	2	AI	AI	NNP
kim-ai-2020	9	3	–	–	:
kim-ai-2020	9	4	also	also	RB
kim-ai-2020	9	5	known	know	VBN
kim-ai-2020	9	6	as	as	IN
kim-ai-2020	9	7	‘	'	``
kim-ai-2020	9	8	artificial	artificial	JJ
kim-ai-2020	9	9	general	general	JJ
kim-ai-2020	9	10	intelligence	intelligence	NN
kim-ai-2020	9	11	(	(	-LRB-
kim-ai-2020	9	12	AGI	AGI	NNP
kim-ai-2020	9	13	)	)	-RRB-
kim-ai-2020	9	14	’	'	''
kim-ai-2020	9	15	and	and	CC
kim-ai-2020	9	16	‘	'	``
kim-ai-2020	9	17	strong	strong	JJ
kim-ai-2020	9	18	AI	AI	NNP
kim-ai-2020	9	19	’	'	''
kim-ai-2020	9	20	–	–	:
kim-ai-2020	9	21	refers	refer	VBZ
kim-ai-2020	9	22	to	to	IN
kim-ai-2020	9	23	the	the	DT
kim-ai-2020	9	24	ability	ability	NN
kim-ai-2020	9	25	to	to	TO
kim-ai-2020	9	26	adapt	adapt	VB
kim-ai-2020	9	27	to	to	TO
kim-ai-2020	9	28	achieve	achieve	VB
kim-ai-2020	9	29	any	any	DT
kim-ai-2020	9	30	goals	goal	NNS
kim-ai-2020	9	31	;	;	:
kim-ai-2020	9	32	it	-PRON-	PRP
kim-ai-2020	9	33	does	do	VBZ
kim-ai-2020	9	34	not	not	RB
kim-ai-2020	9	35	refer	refer	VB
kim-ai-2020	9	36	to	to	IN
kim-ai-2020	9	37	human	human	JJ
kim-ai-2020	9	38	-	-	HYPH
kim-ai-2020	9	39	level	level	NN
kim-ai-2020	9	40	performance	performance	NN
kim-ai-2020	9	41	in	in	IN
kim-ai-2020	9	42	achieving	achieve	VBG
kim-ai-2020	9	43	one	one	CD
kim-ai-2020	9	44	particular	particular	JJ
kim-ai-2020	9	45	goal	goal	NN
kim-ai-2020	9	46	.	.	.
kim-ai-2020	10	1	An	an	DT
kim-ai-2020	10	2	AI	AI	NNP
kim-ai-2020	10	3	system	system	NN
kim-ai-2020	10	4	developed	develop	VBD
kim-ai-2020	10	5	to	to	TO
kim-ai-2020	10	6	perform	perform	VB
kim-ai-2020	10	7	one	one	CD
kim-ai-2020	10	8	specific	specific	JJ
kim-ai-2020	10	9	type	type	NN
kim-ai-2020	10	10	of	of	IN
kim-ai-2020	10	11	activity	activity	NN
kim-ai-2020	10	12	or	or	CC
kim-ai-2020	10	13	activities	activity	NNS
kim-ai-2020	10	14	in	in	IN
kim-ai-2020	10	15	one	one	CD
kim-ai-2020	10	16	specific	specific	JJ
kim-ai-2020	10	17	domain	domain	NN
kim-ai-2020	10	18	is	be	VBZ
kim-ai-2020	10	19	called	call	VBN
kim-ai-2020	10	20	a	a	DT
kim-ai-2020	10	21	‘	'	``
kim-ai-2020	10	22	narrow	narrow	JJ
kim-ai-2020	10	23	(	(	-LRB-
kim-ai-2020	10	24	/weak	/weak	NFP
kim-ai-2020	10	25	)	)	-RRB-
kim-ai-2020	10	26	AI	AI	NNP
kim-ai-2020	10	27	’	'	''
kim-ai-2020	10	28	system	system	NN
kim-ai-2020	10	29	.	.	.
kim-ai-2020	10	30	]	]	-RRB-
kim-ai-2020	11	1	[	[	-LRB-
kim-ai-2020	11	2	1	1	CD
kim-ai-2020	11	3	:	:	:
kim-ai-2020	11	4	AlphaGo	alphago	PRP
kim-ai-2020	11	5	can	can	MD
kim-ai-2020	11	6	be	be	VB
kim-ai-2020	11	7	said	say	VBN
kim-ai-2020	11	8	to	to	TO
kim-ai-2020	11	9	be	be	VB
kim-ai-2020	11	10	“	"	``
kim-ai-2020	11	11	as	as	RB
kim-ai-2020	11	12	intelligent	intelligent	JJ
kim-ai-2020	11	13	as	as	IN
kim-ai-2020	11	14	humans	human	NNS
kim-ai-2020	11	15	,	,	,
kim-ai-2020	11	16	”	"	''
kim-ai-2020	11	17	but	but	CC
kim-ai-2020	11	18	only	only	RB
kim-ai-2020	11	19	in	in	IN
kim-ai-2020	11	20	playing	play	VBG
kim-ai-2020	11	21	Go	Go	NNP
kim-ai-2020	11	22	,	,	,
kim-ai-2020	11	23	where	where	WRB
kim-ai-2020	11	24	it	-PRON-	PRP
kim-ai-2020	11	25	exceeds	exceed	VBZ
kim-ai-2020	11	26	human	human	JJ
kim-ai-2020	11	27	capability	capability	NN
kim-ai-2020	11	28	.	.	.
kim-ai-2020	12	1	So	so	CC
kim-ai-2020	12	2	,	,	,
kim-ai-2020	12	3	it	-PRON-	PRP
kim-ai-2020	12	4	does	do	VBZ
kim-ai-2020	12	5	not	not	RB
kim-ai-2020	12	6	qualify	qualify	VB
kim-ai-2020	12	7	as	as	IN
kim-ai-2020	12	8	general	general	JJ
kim-ai-2020	12	9	/	/	SYM
kim-ai-2020	12	10	strong	strong	JJ
kim-ai-2020	12	11	AI	AI	NNP
kim-ai-2020	12	12	in	in	IN
kim-ai-2020	12	13	spite	spite	NN
kim-ai-2020	12	14	of	of	IN
kim-ai-2020	12	15	its	-PRON-	PRP$
kim-ai-2020	12	16	human	human	JJ
kim-ai-2020	12	17	-	-	HYPH
kim-ai-2020	12	18	level	level	NN
kim-ai-2020	12	19	intelligence	intelligence	NN
kim-ai-2020	12	20	in	in	IN
kim-ai-2020	12	21	Go	go	NN
kim-ai-2020	12	22	-	-	HYPH
kim-ai-2020	12	23	playing	playing	NN
kim-ai-2020	12	24	.	.	.
kim-ai-2020	13	1	It	-PRON-	PRP
kim-ai-2020	13	2	is	be	VBZ
kim-ai-2020	13	3	to	to	TO
kim-ai-2020	13	4	be	be	VB
kim-ai-2020	13	5	noted	note	VBN
kim-ai-2020	13	6	that	that	IN
kim-ai-2020	13	7	general(/strong	general(/strong	NN
kim-ai-2020	13	8	)	)	-RRB-
kim-ai-2020	13	9	AI	AI	NNP
kim-ai-2020	13	10	and	and	CC
kim-ai-2020	13	11	narrow(/weak	narrow(/weak	JJ
kim-ai-2020	13	12	)	)	-RRB-
kim-ai-2020	13	13	AI	AI	NNP
kim-ai-2020	13	14	signify	signify	VB
kim-ai-2020	13	15	the	the	DT
kim-ai-2020	13	16	difference	difference	NN
kim-ai-2020	13	17	in	in	IN
kim-ai-2020	13	18	the	the	DT
kim-ai-2020	13	19	scope	scope	NN
kim-ai-2020	13	20	of	of	IN
kim-ai-2020	13	21	AI	AI	NNP
kim-ai-2020	13	22	capability	capability	NN
kim-ai-2020	13	23	.	.	.
kim-ai-2020	14	1	General(/strong	General(/strong	NFP
kim-ai-2020	14	2	)	)	-RRB-
kim-ai-2020	14	3	AI	AI	NNP
kim-ai-2020	14	4	is	be	VBZ
kim-ai-2020	14	5	a	a	DT
kim-ai-2020	14	6	broader	broad	JJR
kim-ai-2020	14	7	concept	concept	NN
kim-ai-2020	14	8	than	than	IN
kim-ai-2020	14	9	human	human	NN
kim-ai-2020	14	10	-	-	HYPH
kim-ai-2020	14	11	like	like	JJ
kim-ai-2020	14	12	intelligence	intelligence	NN
kim-ai-2020	14	13	,	,	,
kim-ai-2020	14	14	either	either	CC
kim-ai-2020	14	15	with	with	IN
kim-ai-2020	14	16	its	-PRON-	PRP$
kim-ai-2020	14	17	carbon	carbon	NN
kim-ai-2020	14	18	-	-	HYPH
kim-ai-2020	14	19	based	base	VBN
kim-ai-2020	14	20	substrate	substrate	NN
kim-ai-2020	14	21	or	or	CC
kim-ai-2020	14	22	with	with	IN
kim-ai-2020	14	23	human	human	NN
kim-ai-2020	14	24	-	-	HYPH
kim-ai-2020	14	25	like	like	JJ
kim-ai-2020	14	26	understanding	understanding	NN
kim-ai-2020	14	27	that	that	WDT
kim-ai-2020	14	28	relies	rely	VBZ
kim-ai-2020	14	29	on	on	IN
kim-ai-2020	14	30	what	what	WP
kim-ai-2020	14	31	we	-PRON-	PRP
kim-ai-2020	14	32	regard	regard	VBP
kim-ai-2020	14	33	as	as	IN
kim-ai-2020	14	34	uniquely	uniquely	RB
kim-ai-2020	14	35	human	human	JJ
kim-ai-2020	14	36	cognitive	cognitive	JJ
kim-ai-2020	14	37	states	state	NNS
kim-ai-2020	14	38	such	such	JJ
kim-ai-2020	14	39	as	as	IN
kim-ai-2020	14	40	consciousness	consciousness	NN
kim-ai-2020	14	41	,	,	,
kim-ai-2020	14	42	qualia	quale	NNS
kim-ai-2020	14	43	,	,	,
kim-ai-2020	14	44	emotions	emotion	NNS
kim-ai-2020	14	45	,	,	,
kim-ai-2020	14	46	and	and	CC
kim-ai-2020	14	47	so	so	RB
kim-ai-2020	14	48	on	on	RB
kim-ai-2020	14	49	.	.	.
kim-ai-2020	15	1	For	for	IN
kim-ai-2020	15	2	more	more	RBR
kim-ai-2020	15	3	helpful	helpful	JJ
kim-ai-2020	15	4	descriptions	description	NNS
kim-ai-2020	15	5	of	of	IN
kim-ai-2020	15	6	common	common	JJ
kim-ai-2020	15	7	terms	term	NNS
kim-ai-2020	15	8	in	in	IN
kim-ai-2020	15	9	AI	AI	NNP
kim-ai-2020	15	10	,	,	,
kim-ai-2020	15	11	see	see	VB
kim-ai-2020	15	12	Tegmark	Tegmark	NNP
kim-ai-2020	15	13	2017	2017	CD
kim-ai-2020	15	14	,	,	,
kim-ai-2020	15	15	39	39	CD
kim-ai-2020	15	16	.	.	.
kim-ai-2020	16	1	For	for	IN
kim-ai-2020	16	2	more	more	JJR
kim-ai-2020	16	3	on	on	IN
kim-ai-2020	16	4	the	the	DT
kim-ai-2020	16	5	match	match	NN
kim-ai-2020	16	6	between	between	IN
kim-ai-2020	16	7	AlphaGo	AlphaGo	NNP
kim-ai-2020	16	8	and	and	CC
kim-ai-2020	16	9	Sedol	Sedol	NNP
kim-ai-2020	16	10	Lee	Lee	NNP
kim-ai-2020	16	11	,	,	,
kim-ai-2020	16	12	see	see	VB
kim-ai-2020	16	13	Koch	Koch	NNP
kim-ai-2020	16	14	2016	2016	CD
kim-ai-2020	16	15	.	.	.
kim-ai-2020	16	16	]	]	-RRB-
kim-ai-2020	17	1	[	[	-LRB-
kim-ai-2020	17	2	2	2	CD
kim-ai-2020	17	3	:	:	:
kim-ai-2020	17	4	Deep	deep	JJ
kim-ai-2020	17	5	reinforcement	reinforcement	NN
kim-ai-2020	17	6	learning	learning	NN
kim-ai-2020	17	7	is	be	VBZ
kim-ai-2020	17	8	a	a	DT
kim-ai-2020	17	9	type	type	NN
kim-ai-2020	17	10	of	of	IN
kim-ai-2020	17	11	deep	deep	JJ
kim-ai-2020	17	12	learning	learning	NN
kim-ai-2020	17	13	that	that	WDT
kim-ai-2020	17	14	is	be	VBZ
kim-ai-2020	17	15	goal	goal	NN
kim-ai-2020	17	16	-	-	HYPH
kim-ai-2020	17	17	oriented	orient	VBN
kim-ai-2020	17	18	and	and	CC
kim-ai-2020	17	19	reward	reward	NN
kim-ai-2020	17	20	-	-	HYPH
kim-ai-2020	17	21	based	base	VBN
kim-ai-2020	17	22	.	.	.
kim-ai-2020	18	1	See	see	VB
kim-ai-2020	18	2	Heaven	Heaven	NNP
kim-ai-2020	18	3	2020	2020	CD
kim-ai-2020	18	4	.	.	.
kim-ai-2020	18	5	]	]	-RRB-
kim-ai-2020	19	1	Early	early	JJ
kim-ai-2020	19	2	symbolic	symbolic	JJ
kim-ai-2020	19	3	AI	AI	NNP
kim-ai-2020	19	4	systems	system	NNS
kim-ai-2020	19	5	determined	determine	VBD
kim-ai-2020	19	6	their	-PRON-	PRP$
kim-ai-2020	19	7	outputs	output	NNS
kim-ai-2020	19	8	based	base	VBN
kim-ai-2020	19	9	upon	upon	IN
kim-ai-2020	19	10	given	give	VBN
kim-ai-2020	19	11	rules	rule	NNS
kim-ai-2020	19	12	and	and	CC
kim-ai-2020	19	13	logical	logical	JJ
kim-ai-2020	19	14	inference	inference	NN
kim-ai-2020	19	15	.	.	.
kim-ai-2020	20	1	AI	AI	NNP
kim-ai-2020	20	2	algorithms	algorithm	NNS
kim-ai-2020	20	3	in	in	IN
kim-ai-2020	20	4	these	these	DT
kim-ai-2020	20	5	rule	rule	NN
kim-ai-2020	20	6	-	-	HYPH
kim-ai-2020	20	7	based	base	VBN
kim-ai-2020	20	8	systems	system	NNS
kim-ai-2020	20	9	,	,	,
kim-ai-2020	20	10	also	also	RB
kim-ai-2020	20	11	known	know	VBN
kim-ai-2020	20	12	as	as	IN
kim-ai-2020	20	13	good	good	JJ
kim-ai-2020	20	14	old	old	JJ
kim-ai-2020	20	15	-	-	HYPH
kim-ai-2020	20	16	fashioned	fashioned	JJ
kim-ai-2020	20	17	AI	AI	NNP
kim-ai-2020	20	18	(	(	-LRB-
kim-ai-2020	20	19	GOFAI	GOFAI	NNP
kim-ai-2020	20	20	)	)	-RRB-
kim-ai-2020	20	21	,	,	,
kim-ai-2020	20	22	were	be	VBD
kim-ai-2020	20	23	pre	pre	VBN
kim-ai-2020	20	24	-	-	VBN
kim-ai-2020	20	25	determined	determined	JJ
kim-ai-2020	20	26	,	,	,
kim-ai-2020	20	27	predictable	predictable	JJ
kim-ai-2020	20	28	,	,	,
kim-ai-2020	20	29	and	and	CC
kim-ai-2020	20	30	transparent	transparent	JJ
kim-ai-2020	20	31	.	.	.
kim-ai-2020	21	1	On	on	IN
kim-ai-2020	21	2	the	the	DT
kim-ai-2020	21	3	other	other	JJ
kim-ai-2020	21	4	hand	hand	NN
kim-ai-2020	21	5	,	,	,
kim-ai-2020	21	6	machine	machine	NN
kim-ai-2020	21	7	learning	learning	NN
kim-ai-2020	21	8	,	,	,
kim-ai-2020	21	9	another	another	DT
kim-ai-2020	21	10	approach	approach	NN
kim-ai-2020	21	11	in	in	IN
kim-ai-2020	21	12	AI	AI	NNP
kim-ai-2020	21	13	,	,	,
kim-ai-2020	21	14	enables	enable	VBZ
kim-ai-2020	21	15	an	an	DT
kim-ai-2020	21	16	AI	ai	NN
kim-ai-2020	21	17	algorithm	algorithm	NN
kim-ai-2020	21	18	to	to	TO
kim-ai-2020	21	19	evolve	evolve	VB
kim-ai-2020	21	20	.	.	.
kim-ai-2020	22	1	It	-PRON-	PRP
kim-ai-2020	22	2	learns	learn	VBZ
kim-ai-2020	22	3	to	to	TO
kim-ai-2020	22	4	adjust	adjust	VB
kim-ai-2020	22	5	itself	-PRON-	PRP
kim-ai-2020	22	6	to	to	TO
kim-ai-2020	22	7	identify	identify	VB
kim-ai-2020	22	8	certain	certain	JJ
kim-ai-2020	22	9	patterns	pattern	NNS
kim-ai-2020	22	10	through	through	IN
kim-ai-2020	22	11	the	the	DT
kim-ai-2020	22	12	so	so	RB
kim-ai-2020	22	13	-	-	HYPH
kim-ai-2020	22	14	called	call	VBN
kim-ai-2020	22	15	‘	'	``
kim-ai-2020	22	16	training	training	NN
kim-ai-2020	22	17	’	'	''
kim-ai-2020	22	18	process	process	NN
kim-ai-2020	22	19	,	,	,
kim-ai-2020	22	20	which	which	WDT
kim-ai-2020	22	21	relies	rely	VBZ
kim-ai-2020	22	22	on	on	IN
kim-ai-2020	22	23	a	a	DT
kim-ai-2020	22	24	large	large	JJ
kim-ai-2020	22	25	amount	amount	NN
kim-ai-2020	22	26	of	of	IN
kim-ai-2020	22	27	data	datum	NNS
kim-ai-2020	22	28	and	and	CC
kim-ai-2020	22	29	statistics	statistic	NNS
kim-ai-2020	22	30	.	.	.
kim-ai-2020	23	1	Deep	deep	JJ
kim-ai-2020	23	2	learning	learning	NN
kim-ai-2020	23	3	,	,	,
kim-ai-2020	23	4	one	one	CD
kim-ai-2020	23	5	of	of	IN
kim-ai-2020	23	6	the	the	DT
kim-ai-2020	23	7	widely	widely	RB
kim-ai-2020	23	8	-	-	HYPH
kim-ai-2020	23	9	used	use	VBN
kim-ai-2020	23	10	techniques	technique	NNS
kim-ai-2020	23	11	in	in	IN
kim-ai-2020	23	12	machine	machine	NN
kim-ai-2020	23	13	learning	learning	NN
kim-ai-2020	23	14	,	,	,
kim-ai-2020	23	15	further	further	JJ
kim-ai-2020	23	16	refines	refine	NNS
kim-ai-2020	23	17	this	this	DT
kim-ai-2020	23	18	training	training	NN
kim-ai-2020	23	19	process	process	NN
kim-ai-2020	23	20	using	use	VBG
kim-ai-2020	23	21	a	a	DT
kim-ai-2020	23	22	‘	'	``
kim-ai-2020	23	23	neural	neural	JJ
kim-ai-2020	23	24	network	network	NN
kim-ai-2020	23	25	.	.	.
kim-ai-2020	23	26	’[footnoteRef:3	’[footnoteref:3	ADD
kim-ai-2020	23	27	]	]	-RRB-
kim-ai-2020	23	28	Machine	machine	NN
kim-ai-2020	23	29	learning	learning	NN
kim-ai-2020	23	30	and	and	CC
kim-ai-2020	23	31	deep	deep	JJ
kim-ai-2020	23	32	learning	learning	NN
kim-ai-2020	23	33	have	have	VBP
kim-ai-2020	23	34	brought	bring	VBN
kim-ai-2020	23	35	significant	significant	JJ
kim-ai-2020	23	36	improvements	improvement	NNS
kim-ai-2020	23	37	to	to	IN
kim-ai-2020	23	38	the	the	DT
kim-ai-2020	23	39	performance	performance	NN
kim-ai-2020	23	40	of	of	IN
kim-ai-2020	23	41	AI	AI	NNP
kim-ai-2020	23	42	systems	system	NNS
kim-ai-2020	23	43	in	in	IN
kim-ai-2020	23	44	areas	area	NNS
kim-ai-2020	23	45	such	such	JJ
kim-ai-2020	23	46	as	as	IN
kim-ai-2020	23	47	translation	translation	NN
kim-ai-2020	23	48	,	,	,
kim-ai-2020	23	49	speech	speech	NN
kim-ai-2020	23	50	recognition	recognition	NN
kim-ai-2020	23	51	,	,	,
kim-ai-2020	23	52	and	and	CC
kim-ai-2020	23	53	detecting	detect	VBG
kim-ai-2020	23	54	objects	object	NNS
kim-ai-2020	23	55	and	and	CC
kim-ai-2020	23	56	predicting	predict	VBG
kim-ai-2020	23	57	their	-PRON-	PRP$
kim-ai-2020	23	58	movements	movement	NNS
kim-ai-2020	23	59	.	.	.
kim-ai-2020	24	1	Some	some	DT
kim-ai-2020	24	2	people	people	NNS
kim-ai-2020	24	3	assume	assume	VBP
kim-ai-2020	24	4	that	that	IN
kim-ai-2020	24	5	machine	machine	NN
kim-ai-2020	24	6	learning	learn	VBG
kim-ai-2020	24	7	completely	completely	RB
kim-ai-2020	24	8	replaced	replace	VBD
kim-ai-2020	24	9	GOFAI	GOFAI	NNP
kim-ai-2020	24	10	,	,	,
kim-ai-2020	24	11	but	but	CC
kim-ai-2020	24	12	this	this	DT
kim-ai-2020	24	13	is	be	VBZ
kim-ai-2020	24	14	a	a	DT
kim-ai-2020	24	15	misunderstanding	misunderstanding	NN
kim-ai-2020	24	16	.	.	.
kim-ai-2020	25	1	Symbolic	symbolic	JJ
kim-ai-2020	25	2	reasoning	reasoning	NN
kim-ai-2020	25	3	and	and	CC
kim-ai-2020	25	4	machine	machine	NN
kim-ai-2020	25	5	learning	learning	NN
kim-ai-2020	25	6	are	be	VBP
kim-ai-2020	25	7	two	two	CD
kim-ai-2020	25	8	distinct	distinct	JJ
kim-ai-2020	25	9	but	but	CC
kim-ai-2020	25	10	not	not	RB
kim-ai-2020	25	11	mutually	mutually	RB
kim-ai-2020	25	12	exclusive	exclusive	JJ
kim-ai-2020	25	13	approaches	approach	NNS
kim-ai-2020	25	14	,	,	,
kim-ai-2020	25	15	and	and	CC
kim-ai-2020	25	16	they	-PRON-	PRP
kim-ai-2020	25	17	can	can	MD
kim-ai-2020	25	18	be	be	VB
kim-ai-2020	25	19	used	use	VBN
kim-ai-2020	25	20	together	together	RB
kim-ai-2020	25	21	(	(	-LRB-
kim-ai-2020	25	22	Knight	Knight	NNP
kim-ai-2020	25	23	2019a	2019a	CD
kim-ai-2020	25	24	)	)	-RRB-
kim-ai-2020	25	25	.	.	.
kim-ai-2020	26	1	[	[	-LRB-
kim-ai-2020	26	2	3	3	CD
kim-ai-2020	26	3	:	:	:
kim-ai-2020	26	4	Machine	machine	NN
kim-ai-2020	26	5	learning	learning	NN
kim-ai-2020	26	6	and	and	CC
kim-ai-2020	26	7	deep	deep	JJ
kim-ai-2020	26	8	learning	learning	NN
kim-ai-2020	26	9	have	have	VBP
kim-ai-2020	26	10	gained	gain	VBN
kim-ai-2020	26	11	momentum	momentum	NN
kim-ai-2020	26	12	as	as	IN
kim-ai-2020	26	13	the	the	DT
kim-ai-2020	26	14	cost	cost	NN
kim-ai-2020	26	15	of	of	IN
kim-ai-2020	26	16	high	high	JJ
kim-ai-2020	26	17	-	-	HYPH
kim-ai-2020	26	18	performance	performance	NN
kim-ai-2020	26	19	computing	computing	NN
kim-ai-2020	26	20	has	have	VBZ
kim-ai-2020	26	21	gone	go	VBN
kim-ai-2020	26	22	down	down	RB
kim-ai-2020	26	23	and	and	CC
kim-ai-2020	26	24	large	large	JJ
kim-ai-2020	26	25	data	data	NN
kim-ai-2020	26	26	sets	set	NNS
kim-ai-2020	26	27	have	have	VBP
kim-ai-2020	26	28	become	become	VBN
kim-ai-2020	26	29	available	available	JJ
kim-ai-2020	26	30	.	.	.
kim-ai-2020	27	1	For	for	IN
kim-ai-2020	27	2	example	example	NN
kim-ai-2020	27	3	,	,	,
kim-ai-2020	27	4	the	the	DT
kim-ai-2020	27	5	data	datum	NNS
kim-ai-2020	27	6	in	in	IN
kim-ai-2020	27	7	the	the	DT
kim-ai-2020	27	8	ImageNet	ImageNet	NNP
kim-ai-2020	27	9	contains	contain	VBZ
kim-ai-2020	27	10	more	more	JJR
kim-ai-2020	27	11	than	than	IN
kim-ai-2020	27	12	14	14	CD
kim-ai-2020	27	13	million	million	CD
kim-ai-2020	27	14	hand	hand	NN
kim-ai-2020	27	15	-	-	HYPH
kim-ai-2020	27	16	annotated	annotate	VBN
kim-ai-2020	27	17	images	image	NNS
kim-ai-2020	27	18	.	.	.
kim-ai-2020	28	1	The	the	DT
kim-ai-2020	28	2	ImageNet	ImageNet	NNP
kim-ai-2020	28	3	data	datum	NNS
kim-ai-2020	28	4	have	have	VBP
kim-ai-2020	28	5	been	be	VBN
kim-ai-2020	28	6	used	use	VBN
kim-ai-2020	28	7	for	for	IN
kim-ai-2020	28	8	the	the	DT
kim-ai-2020	28	9	well	well	RB
kim-ai-2020	28	10	-	-	HYPH
kim-ai-2020	28	11	known	know	VBN
kim-ai-2020	28	12	annual	annual	JJ
kim-ai-2020	28	13	AI	AI	NNP
kim-ai-2020	28	14	competition	competition	NN
kim-ai-2020	28	15	for	for	IN
kim-ai-2020	28	16	object	object	NN
kim-ai-2020	28	17	detection	detection	NN
kim-ai-2020	28	18	and	and	CC
kim-ai-2020	28	19	image	image	NN
kim-ai-2020	28	20	classification	classification	NN
kim-ai-2020	28	21	at	at	IN
kim-ai-2020	28	22	large	large	JJ
kim-ai-2020	28	23	scale	scale	NN
kim-ai-2020	28	24	from	from	IN
kim-ai-2020	28	25	2010	2010	CD
kim-ai-2020	28	26	to	to	IN
kim-ai-2020	28	27	2017	2017	CD
kim-ai-2020	28	28	.	.	.
kim-ai-2020	29	1	See	see	VB
kim-ai-2020	29	2	http://www.image-net.org/challenges/LSVRC/.	http://www.image-net.org/challenges/lsvrc/.	NN
kim-ai-2020	29	3	]	]	-RRB-
kim-ai-2020	30	1	With	with	IN
kim-ai-2020	30	2	their	-PRON-	PRP$
kim-ai-2020	30	3	limited	limit	VBN
kim-ai-2020	30	4	intelligence	intelligence	NN
kim-ai-2020	30	5	and	and	CC
kim-ai-2020	30	6	fully	fully	RB
kim-ai-2020	30	7	deterministic	deterministic	JJ
kim-ai-2020	30	8	nature	nature	NN
kim-ai-2020	30	9	,	,	,
kim-ai-2020	30	10	early	early	JJ
kim-ai-2020	30	11	rule	rule	NN
kim-ai-2020	30	12	-	-	HYPH
kim-ai-2020	30	13	based	base	VBN
kim-ai-2020	30	14	symbolic	symbolic	JJ
kim-ai-2020	30	15	AI	AI	NNP
kim-ai-2020	30	16	systems	system	NNS
kim-ai-2020	30	17	raised	raise	VBD
kim-ai-2020	30	18	few	few	JJ
kim-ai-2020	30	19	ethical	ethical	JJ
kim-ai-2020	30	20	concerns	concern	NNS
kim-ai-2020	30	21	.	.	.
kim-ai-2020	31	1	[	[	-LRB-
kim-ai-2020	31	2	footnoteRef:4	footnoteRef:4	NNS
kim-ai-2020	31	3	]	]	-RRB-
kim-ai-2020	31	4	AI	AI	NNP
kim-ai-2020	31	5	systems	system	NNS
kim-ai-2020	31	6	that	that	IN
kim-ai-2020	31	7	near	near	RB
kim-ai-2020	31	8	or	or	CC
kim-ai-2020	31	9	surpass	surpass	VB
kim-ai-2020	31	10	human	human	JJ
kim-ai-2020	31	11	capability	capability	NN
kim-ai-2020	31	12	,	,	,
kim-ai-2020	31	13	on	on	IN
kim-ai-2020	31	14	the	the	DT
kim-ai-2020	31	15	other	other	JJ
kim-ai-2020	31	16	hand	hand	NN
kim-ai-2020	31	17	,	,	,
kim-ai-2020	31	18	are	be	VBP
kim-ai-2020	31	19	likely	likely	JJ
kim-ai-2020	31	20	to	to	TO
kim-ai-2020	31	21	be	be	VB
kim-ai-2020	31	22	given	give	VBN
kim-ai-2020	31	23	the	the	DT
kim-ai-2020	31	24	autonomy	autonomy	NN
kim-ai-2020	31	25	to	to	TO
kim-ai-2020	31	26	make	make	VB
kim-ai-2020	31	27	their	-PRON-	PRP$
kim-ai-2020	31	28	own	own	JJ
kim-ai-2020	31	29	decisions	decision	NNS
kim-ai-2020	31	30	without	without	IN
kim-ai-2020	31	31	humans	human	NNS
kim-ai-2020	31	32	,	,	,
kim-ai-2020	31	33	even	even	RB
kim-ai-2020	31	34	when	when	WRB
kim-ai-2020	31	35	their	-PRON-	PRP$
kim-ai-2020	31	36	workings	working	NNS
kim-ai-2020	31	37	are	be	VBP
kim-ai-2020	31	38	not	not	RB
kim-ai-2020	31	39	entirely	entirely	RB
kim-ai-2020	31	40	transparent	transparent	JJ
kim-ai-2020	31	41	,	,	,
kim-ai-2020	31	42	and	and	CC
kim-ai-2020	31	43	some	some	DT
kim-ai-2020	31	44	of	of	IN
kim-ai-2020	31	45	those	those	DT
kim-ai-2020	31	46	decisions	decision	NNS
kim-ai-2020	31	47	are	be	VBP
kim-ai-2020	31	48	distinctively	distinctively	RB
kim-ai-2020	31	49	moral	moral	JJ
kim-ai-2020	31	50	in	in	IN
kim-ai-2020	31	51	character	character	NN
kim-ai-2020	31	52	.	.	.
kim-ai-2020	32	1	As	as	IN
kim-ai-2020	32	2	humans	human	NNS
kim-ai-2020	32	3	,	,	,
kim-ai-2020	32	4	we	-PRON-	PRP
kim-ai-2020	32	5	are	be	VBP
kim-ai-2020	32	6	trained	train	VBN
kim-ai-2020	32	7	to	to	TO
kim-ai-2020	32	8	recognize	recognize	VB
kim-ai-2020	32	9	situations	situation	NNS
kim-ai-2020	32	10	that	that	WDT
kim-ai-2020	32	11	demand	demand	VBP
kim-ai-2020	32	12	moral	moral	JJ
kim-ai-2020	32	13	decision	decision	NN
kim-ai-2020	32	14	-	-	HYPH
kim-ai-2020	32	15	making	making	NN
kim-ai-2020	32	16	.	.	.
kim-ai-2020	33	1	But	but	CC
kim-ai-2020	33	2	how	how	WRB
kim-ai-2020	33	3	would	would	MD
kim-ai-2020	33	4	an	an	DT
kim-ai-2020	33	5	AI	ai	NN
kim-ai-2020	33	6	system	system	NN
kim-ai-2020	33	7	be	be	VB
kim-ai-2020	33	8	able	able	JJ
kim-ai-2020	33	9	to	to	TO
kim-ai-2020	33	10	do	do	VB
kim-ai-2020	33	11	so	so	RB
kim-ai-2020	33	12	,	,	,
kim-ai-2020	33	13	and	and	CC
kim-ai-2020	33	14	should	should	MD
kim-ai-2020	33	15	they	-PRON-	PRP
kim-ai-2020	33	16	be	be	VB
kim-ai-2020	33	17	allowed	allow	VBN
kim-ai-2020	33	18	to	to	IN
kim-ai-2020	33	19	?	?	.
kim-ai-2020	34	1	With	with	IN
kim-ai-2020	34	2	self	self	NN
kim-ai-2020	34	3	-	-	HYPH
kim-ai-2020	34	4	driving	drive	VBG
kim-ai-2020	34	5	cars	car	NNS
kim-ai-2020	34	6	and	and	CC
kim-ai-2020	34	7	autonomous	autonomous	JJ
kim-ai-2020	34	8	weapons	weapon	NNS
kim-ai-2020	34	9	systems	system	NNS
kim-ai-2020	34	10	under	under	IN
kim-ai-2020	34	11	active	active	JJ
kim-ai-2020	34	12	development	development	NN
kim-ai-2020	34	13	and	and	CC
kim-ai-2020	34	14	testing	testing	NN
kim-ai-2020	34	15	,	,	,
kim-ai-2020	34	16	these	these	DT
kim-ai-2020	34	17	are	be	VBP
kim-ai-2020	34	18	no	no	RB
kim-ai-2020	34	19	longer	long	RBR
kim-ai-2020	34	20	idle	idle	JJ
kim-ai-2020	34	21	questions	question	NNS
kim-ai-2020	34	22	.	.	.
kim-ai-2020	35	1	[	[	-LRB-
kim-ai-2020	35	2	4	4	CD
kim-ai-2020	35	3	:	:	:
kim-ai-2020	35	4	For	for	IN
kim-ai-2020	35	5	an	an	DT
kim-ai-2020	35	6	excellent	excellent	JJ
kim-ai-2020	35	7	history	history	NN
kim-ai-2020	35	8	of	of	IN
kim-ai-2020	35	9	AI	AI	NNP
kim-ai-2020	35	10	research	research	NN
kim-ai-2020	35	11	,	,	,
kim-ai-2020	35	12	see	see	VB
kim-ai-2020	35	13	chapter	chapter	NN
kim-ai-2020	35	14	1	1	CD
kim-ai-2020	35	15	,	,	,
kim-ai-2020	35	16	“	"	``
kim-ai-2020	35	17	What	what	WP
kim-ai-2020	35	18	is	be	VBZ
kim-ai-2020	35	19	Artificial	Artificial	NNP
kim-ai-2020	35	20	Intelligence	Intelligence	NNP
kim-ai-2020	35	21	,	,	,
kim-ai-2020	35	22	”	"	''
kim-ai-2020	35	23	of	of	IN
kim-ai-2020	35	24	Boden	Boden	NNP
kim-ai-2020	35	25	2016	2016	CD
kim-ai-2020	35	26	,	,	,
kim-ai-2020	35	27	1	1	CD
kim-ai-2020	35	28	-	-	SYM
kim-ai-2020	35	29	20	20	CD
kim-ai-2020	35	30	.	.	.
kim-ai-2020	35	31	]	]	-RRB-
kim-ai-2020	36	1	2	2	LS
kim-ai-2020	36	2	.	.	.
kim-ai-2020	37	1	The	the	DT
kim-ai-2020	37	2	Trolley	Trolley	NNP
kim-ai-2020	37	3	Problem	Problem	NNP
kim-ai-2020	37	4	Recent	recent	JJ
kim-ai-2020	37	5	advances	advance	NNS
kim-ai-2020	37	6	of	of	IN
kim-ai-2020	37	7	AI	AI	NNP
kim-ai-2020	37	8	,	,	,
kim-ai-2020	37	9	such	such	JJ
kim-ai-2020	37	10	as	as	IN
kim-ai-2020	37	11	autonomous	autonomous	JJ
kim-ai-2020	37	12	cars	car	NNS
kim-ai-2020	37	13	,	,	,
kim-ai-2020	37	14	have	have	VBP
kim-ai-2020	37	15	brought	bring	VBN
kim-ai-2020	37	16	new	new	JJ
kim-ai-2020	37	17	interest	interest	NN
kim-ai-2020	37	18	to	to	IN
kim-ai-2020	37	19	the	the	DT
kim-ai-2020	37	20	trolley	trolley	NN
kim-ai-2020	37	21	problem	problem	NN
kim-ai-2020	37	22	,	,	,
kim-ai-2020	37	23	a	a	DT
kim-ai-2020	37	24	thought	thought	NN
kim-ai-2020	37	25	experiment	experiment	NN
kim-ai-2020	37	26	introduced	introduce	VBN
kim-ai-2020	37	27	by	by	IN
kim-ai-2020	37	28	Philippa	Philippa	NNP
kim-ai-2020	37	29	Foot	Foot	NNP
kim-ai-2020	37	30	in	in	IN
kim-ai-2020	37	31	1967	1967	CD
kim-ai-2020	37	32	.	.	.
kim-ai-2020	38	1	In	in	IN
kim-ai-2020	38	2	the	the	DT
kim-ai-2020	38	3	standard	standard	JJ
kim-ai-2020	38	4	version	version	NN
kim-ai-2020	38	5	of	of	IN
kim-ai-2020	38	6	this	this	DT
kim-ai-2020	38	7	problem	problem	NN
kim-ai-2020	38	8	,	,	,
kim-ai-2020	38	9	a	a	DT
kim-ai-2020	38	10	runaway	runaway	JJ
kim-ai-2020	38	11	trolley	trolley	NN
kim-ai-2020	38	12	barrels	barrel	NNS
kim-ai-2020	38	13	down	down	IN
kim-ai-2020	38	14	a	a	DT
kim-ai-2020	38	15	track	track	NN
kim-ai-2020	38	16	where	where	WRB
kim-ai-2020	38	17	five	five	CD
kim-ai-2020	38	18	unsuspecting	unsuspecting	JJ
kim-ai-2020	38	19	people	people	NNS
kim-ai-2020	38	20	are	be	VBP
kim-ai-2020	38	21	standing	stand	VBG
kim-ai-2020	38	22	.	.	.
kim-ai-2020	39	1	You	-PRON-	PRP
kim-ai-2020	39	2	happen	happen	VBP
kim-ai-2020	39	3	to	to	TO
kim-ai-2020	39	4	be	be	VB
kim-ai-2020	39	5	standing	stand	VBG
kim-ai-2020	39	6	next	next	RB
kim-ai-2020	39	7	to	to	IN
kim-ai-2020	39	8	a	a	DT
kim-ai-2020	39	9	lever	lever	NN
kim-ai-2020	39	10	that	that	WDT
kim-ai-2020	39	11	switches	switch	VBZ
kim-ai-2020	39	12	the	the	DT
kim-ai-2020	39	13	trolley	trolley	NN
kim-ai-2020	39	14	onto	onto	IN
kim-ai-2020	39	15	a	a	DT
kim-ai-2020	39	16	different	different	JJ
kim-ai-2020	39	17	track	track	NN
kim-ai-2020	39	18	,	,	,
kim-ai-2020	39	19	where	where	WRB
kim-ai-2020	39	20	there	there	EX
kim-ai-2020	39	21	is	be	VBZ
kim-ai-2020	39	22	only	only	RB
kim-ai-2020	39	23	one	one	CD
kim-ai-2020	39	24	person	person	NN
kim-ai-2020	39	25	.	.	.
kim-ai-2020	40	1	Those	those	DT
kim-ai-2020	40	2	who	who	WP
kim-ai-2020	40	3	are	be	VBP
kim-ai-2020	40	4	on	on	IN
kim-ai-2020	40	5	either	either	DT
kim-ai-2020	40	6	track	track	NN
kim-ai-2020	40	7	will	will	MD
kim-ai-2020	40	8	be	be	VB
kim-ai-2020	40	9	killed	kill	VBN
kim-ai-2020	40	10	if	if	IN
kim-ai-2020	40	11	the	the	DT
kim-ai-2020	40	12	trolley	trolley	NN
kim-ai-2020	40	13	heads	head	VBZ
kim-ai-2020	40	14	their	-PRON-	PRP$
kim-ai-2020	40	15	way	way	NN
kim-ai-2020	40	16	.	.	.
kim-ai-2020	41	1	Should	Should	MD
kim-ai-2020	41	2	you	-PRON-	PRP
kim-ai-2020	41	3	pull	pull	VB
kim-ai-2020	41	4	the	the	DT
kim-ai-2020	41	5	lever	lever	NN
kim-ai-2020	41	6	,	,	,
kim-ai-2020	41	7	so	so	IN
kim-ai-2020	41	8	that	that	IN
kim-ai-2020	41	9	the	the	DT
kim-ai-2020	41	10	runaway	runaway	JJ
kim-ai-2020	41	11	trolley	trolley	NN
kim-ai-2020	41	12	would	would	MD
kim-ai-2020	41	13	kill	kill	VB
kim-ai-2020	41	14	one	one	CD
kim-ai-2020	41	15	person	person	NN
kim-ai-2020	41	16	instead	instead	RB
kim-ai-2020	41	17	of	of	IN
kim-ai-2020	41	18	five	five	CD
kim-ai-2020	41	19	?	?	.
kim-ai-2020	42	1	Unlike	unlike	IN
kim-ai-2020	42	2	people	people	NNS
kim-ai-2020	42	3	,	,	,
kim-ai-2020	42	4	a	a	DT
kim-ai-2020	42	5	machine	machine	NN
kim-ai-2020	42	6	does	do	VBZ
kim-ai-2020	42	7	not	not	RB
kim-ai-2020	42	8	panic	panic	VB
kim-ai-2020	42	9	or	or	CC
kim-ai-2020	42	10	freeze	freeze	VB
kim-ai-2020	42	11	and	and	CC
kim-ai-2020	42	12	simply	simply	RB
kim-ai-2020	42	13	follows	follow	VBZ
kim-ai-2020	42	14	and	and	CC
kim-ai-2020	42	15	executes	execute	VBZ
kim-ai-2020	42	16	the	the	DT
kim-ai-2020	42	17	given	give	VBN
kim-ai-2020	42	18	instruction	instruction	NN
kim-ai-2020	42	19	.	.	.
kim-ai-2020	43	1	So	so	RB
kim-ai-2020	43	2	,	,	,
kim-ai-2020	43	3	this	this	DT
kim-ai-2020	43	4	opens	open	VBZ
kim-ai-2020	43	5	up	up	RP
kim-ai-2020	43	6	the	the	DT
kim-ai-2020	43	7	possibility	possibility	NN
kim-ai-2020	43	8	of	of	IN
kim-ai-2020	43	9	building	build	VBG
kim-ai-2020	43	10	an	an	DT
kim-ai-2020	43	11	AI	AI	NNP
kim-ai-2020	43	12	-	-	HYPH
kim-ai-2020	43	13	powered	power	VBN
kim-ai-2020	43	14	trolley	trolley	NN
kim-ai-2020	43	15	that	that	WDT
kim-ai-2020	43	16	can	can	MD
kim-ai-2020	43	17	act	act	VB
kim-ai-2020	43	18	morally	morally	RB
kim-ai-2020	43	19	.	.	.
kim-ai-2020	44	1	[	[	-LRB-
kim-ai-2020	44	2	footnoteRef:5	footnoteRef:5	NNP
kim-ai-2020	44	3	]	]	-RRB-
kim-ai-2020	44	4	The	the	DT
kim-ai-2020	44	5	question	question	NN
kim-ai-2020	44	6	itself	-PRON-	PRP
kim-ai-2020	44	7	remains	remain	VBZ
kim-ai-2020	44	8	,	,	,
kim-ai-2020	44	9	however	however	RB
kim-ai-2020	44	10	.	.	.
kim-ai-2020	45	1	Should	Should	MD
kim-ai-2020	45	2	one	one	CD
kim-ai-2020	45	3	program	program	NN
kim-ai-2020	45	4	the	the	DT
kim-ai-2020	45	5	AI	ai	NN
kim-ai-2020	45	6	-	-	HYPH
kim-ai-2020	45	7	powered	power	VBN
kim-ai-2020	45	8	trolley	trolley	NN
kim-ai-2020	45	9	to	to	TO
kim-ai-2020	45	10	swerve	swerve	VB
kim-ai-2020	45	11	or	or	CC
kim-ai-2020	45	12	stay	stay	VB
kim-ai-2020	45	13	on	on	RP
kim-ai-2020	45	14	course	course	NN
kim-ai-2020	45	15	?	?	.
kim-ai-2020	46	1	[	[	-LRB-
kim-ai-2020	46	2	5	5	CD
kim-ai-2020	46	3	:	:	:
kim-ai-2020	46	4	Programming	programming	NN
kim-ai-2020	46	5	here	here	RB
kim-ai-2020	46	6	does	do	VBZ
kim-ai-2020	46	7	not	not	RB
kim-ai-2020	46	8	exclusively	exclusively	RB
kim-ai-2020	46	9	mean	mean	VB
kim-ai-2020	46	10	a	a	DT
kim-ai-2020	46	11	deep	deep	JJ
kim-ai-2020	46	12	learning	learning	NN
kim-ai-2020	46	13	or	or	CC
kim-ai-2020	46	14	machine	machine	NN
kim-ai-2020	46	15	learning	learning	NN
kim-ai-2020	46	16	approach	approach	NN
kim-ai-2020	46	17	.	.	.
kim-ai-2020	46	18	]	]	-RRB-
kim-ai-2020	47	1	Different	different	JJ
kim-ai-2020	47	2	moral	moral	JJ
kim-ai-2020	47	3	theories	theory	NNS
kim-ai-2020	47	4	,	,	,
kim-ai-2020	47	5	such	such	JJ
kim-ai-2020	47	6	as	as	IN
kim-ai-2020	47	7	virtue	virtue	NN
kim-ai-2020	47	8	ethics	ethic	NNS
kim-ai-2020	47	9	,	,	,
kim-ai-2020	47	10	contractarianism	contractarianism	NN
kim-ai-2020	47	11	,	,	,
kim-ai-2020	47	12	and	and	CC
kim-ai-2020	47	13	moral	moral	JJ
kim-ai-2020	47	14	relativism	relativism	NN
kim-ai-2020	47	15	,	,	,
kim-ai-2020	47	16	take	take	VBP
kim-ai-2020	47	17	different	different	JJ
kim-ai-2020	47	18	positions	position	NNS
kim-ai-2020	47	19	.	.	.
kim-ai-2020	48	1	Here	here	RB
kim-ai-2020	48	2	,	,	,
kim-ai-2020	48	3	I	-PRON-	PRP
kim-ai-2020	48	4	will	will	MD
kim-ai-2020	48	5	consider	consider	VB
kim-ai-2020	48	6	utilitarianism	utilitarianism	NN
kim-ai-2020	48	7	and	and	CC
kim-ai-2020	48	8	deontology	deontology	NN
kim-ai-2020	48	9	since	since	IN
kim-ai-2020	48	10	their	-PRON-	PRP$
kim-ai-2020	48	11	tenets	tenet	NNS
kim-ai-2020	48	12	are	be	VBP
kim-ai-2020	48	13	relatively	relatively	RB
kim-ai-2020	48	14	straightforward	straightforward	JJ
kim-ai-2020	48	15	and	and	CC
kim-ai-2020	48	16	most	most	JJS
kim-ai-2020	48	17	AI	AI	NNP
kim-ai-2020	48	18	developers	developer	NNS
kim-ai-2020	48	19	are	be	VBP
kim-ai-2020	48	20	likely	likely	JJ
kim-ai-2020	48	21	to	to	TO
kim-ai-2020	48	22	look	look	VB
kim-ai-2020	48	23	towards	towards	IN
kim-ai-2020	48	24	those	those	DT
kim-ai-2020	48	25	two	two	CD
kim-ai-2020	48	26	moral	moral	JJ
kim-ai-2020	48	27	theories	theory	NNS
kim-ai-2020	48	28	for	for	IN
kim-ai-2020	48	29	guidance	guidance	NN
kim-ai-2020	48	30	and	and	CC
kim-ai-2020	48	31	insight	insight	NN
kim-ai-2020	48	32	.	.	.
kim-ai-2020	49	1	Utilitarianism	Utilitarianism	NNP
kim-ai-2020	49	2	argues	argue	VBZ
kim-ai-2020	49	3	that	that	IN
kim-ai-2020	49	4	the	the	DT
kim-ai-2020	49	5	utility	utility	NN
kim-ai-2020	49	6	of	of	IN
kim-ai-2020	49	7	an	an	DT
kim-ai-2020	49	8	action	action	NN
kim-ai-2020	49	9	is	be	VBZ
kim-ai-2020	49	10	what	what	WP
kim-ai-2020	49	11	makes	make	VBZ
kim-ai-2020	49	12	an	an	DT
kim-ai-2020	49	13	action	action	NN
kim-ai-2020	49	14	moral	moral	JJ
kim-ai-2020	49	15	.	.	.
kim-ai-2020	50	1	In	in	IN
kim-ai-2020	50	2	this	this	DT
kim-ai-2020	50	3	view	view	NN
kim-ai-2020	50	4	,	,	,
kim-ai-2020	50	5	what	what	WP
kim-ai-2020	50	6	generates	generate	VBZ
kim-ai-2020	50	7	the	the	DT
kim-ai-2020	50	8	greatest	great	JJS
kim-ai-2020	50	9	amount	amount	NN
kim-ai-2020	50	10	of	of	IN
kim-ai-2020	50	11	good	good	JJ
kim-ai-2020	50	12	is	be	VBZ
kim-ai-2020	50	13	the	the	DT
kim-ai-2020	50	14	most	most	RBS
kim-ai-2020	50	15	moral	moral	JJ
kim-ai-2020	50	16	thing	thing	NN
kim-ai-2020	50	17	to	to	TO
kim-ai-2020	50	18	do	do	VB
kim-ai-2020	50	19	.	.	.
kim-ai-2020	51	1	If	if	IN
kim-ai-2020	51	2	one	one	PRP
kim-ai-2020	51	3	regards	regard	VBZ
kim-ai-2020	51	4	five	five	CD
kim-ai-2020	51	5	human	human	JJ
kim-ai-2020	51	6	lives	life	NNS
kim-ai-2020	51	7	as	as	IN
kim-ai-2020	51	8	a	a	DT
kim-ai-2020	51	9	greater	great	JJR
kim-ai-2020	51	10	good	good	NN
kim-ai-2020	51	11	than	than	IN
kim-ai-2020	51	12	one	one	CD
kim-ai-2020	51	13	,	,	,
kim-ai-2020	51	14	then	then	RB
kim-ai-2020	51	15	one	one	CD
kim-ai-2020	51	16	acts	act	VBZ
kim-ai-2020	51	17	morally	morally	RB
kim-ai-2020	51	18	by	by	IN
kim-ai-2020	51	19	pulling	pull	VBG
kim-ai-2020	51	20	the	the	DT
kim-ai-2020	51	21	lever	lever	NN
kim-ai-2020	51	22	and	and	CC
kim-ai-2020	51	23	diverting	divert	VBG
kim-ai-2020	51	24	the	the	DT
kim-ai-2020	51	25	trolley	trolley	NN
kim-ai-2020	51	26	to	to	IN
kim-ai-2020	51	27	the	the	DT
kim-ai-2020	51	28	other	other	JJ
kim-ai-2020	51	29	track	track	NN
kim-ai-2020	51	30	.	.	.
kim-ai-2020	52	1	By	by	IN
kim-ai-2020	52	2	contrast	contrast	NN
kim-ai-2020	52	3	,	,	,
kim-ai-2020	52	4	deontology	deontology	NN
kim-ai-2020	52	5	claims	claim	VBZ
kim-ai-2020	52	6	that	that	IN
kim-ai-2020	52	7	what	what	WP
kim-ai-2020	52	8	determines	determine	VBZ
kim-ai-2020	52	9	whether	whether	IN
kim-ai-2020	52	10	an	an	DT
kim-ai-2020	52	11	action	action	NN
kim-ai-2020	52	12	is	be	VBZ
kim-ai-2020	52	13	morally	morally	RB
kim-ai-2020	52	14	right	right	JJ
kim-ai-2020	52	15	or	or	CC
kim-ai-2020	52	16	wrong	wrong	JJ
kim-ai-2020	52	17	is	be	VBZ
kim-ai-2020	52	18	not	not	RB
kim-ai-2020	52	19	its	-PRON-	PRP$
kim-ai-2020	52	20	utility	utility	NN
kim-ai-2020	52	21	but	but	CC
kim-ai-2020	52	22	moral	moral	JJ
kim-ai-2020	52	23	rules	rule	NNS
kim-ai-2020	52	24	.	.	.
kim-ai-2020	53	1	If	if	IN
kim-ai-2020	53	2	an	an	DT
kim-ai-2020	53	3	action	action	NN
kim-ai-2020	53	4	is	be	VBZ
kim-ai-2020	53	5	in	in	IN
kim-ai-2020	53	6	accordance	accordance	NN
kim-ai-2020	53	7	with	with	IN
kim-ai-2020	53	8	those	those	DT
kim-ai-2020	53	9	rules	rule	NNS
kim-ai-2020	53	10	,	,	,
kim-ai-2020	53	11	then	then	RB
kim-ai-2020	53	12	the	the	DT
kim-ai-2020	53	13	action	action	NN
kim-ai-2020	53	14	is	be	VBZ
kim-ai-2020	53	15	morally	morally	RB
kim-ai-2020	53	16	right	right	JJ
kim-ai-2020	53	17	.	.	.
kim-ai-2020	54	1	Otherwise	otherwise	RB
kim-ai-2020	54	2	,	,	,
kim-ai-2020	54	3	it	-PRON-	PRP
kim-ai-2020	54	4	is	be	VBZ
kim-ai-2020	54	5	morally	morally	RB
kim-ai-2020	54	6	wrong	wrong	JJ
kim-ai-2020	54	7	.	.	.
kim-ai-2020	55	1	If	if	IN
kim-ai-2020	55	2	not	not	RB
kim-ai-2020	55	3	to	to	TO
kim-ai-2020	55	4	kill	kill	VB
kim-ai-2020	55	5	another	another	DT
kim-ai-2020	55	6	human	human	NN
kim-ai-2020	55	7	being	being	NN
kim-ai-2020	55	8	is	be	VBZ
kim-ai-2020	55	9	one	one	CD
kim-ai-2020	55	10	of	of	IN
kim-ai-2020	55	11	those	those	DT
kim-ai-2020	55	12	moral	moral	JJ
kim-ai-2020	55	13	rules	rule	NNS
kim-ai-2020	55	14	,	,	,
kim-ai-2020	55	15	then	then	RB
kim-ai-2020	55	16	killing	kill	VBG
kim-ai-2020	55	17	someone	someone	NN
kim-ai-2020	55	18	is	be	VBZ
kim-ai-2020	55	19	morally	morally	RB
kim-ai-2020	55	20	wrong	wrong	JJ
kim-ai-2020	55	21	even	even	RB
kim-ai-2020	55	22	if	if	IN
kim-ai-2020	55	23	it	-PRON-	PRP
kim-ai-2020	55	24	is	be	VBZ
kim-ai-2020	55	25	to	to	TO
kim-ai-2020	55	26	save	save	VB
kim-ai-2020	55	27	more	more	JJR
kim-ai-2020	55	28	lives	life	NNS
kim-ai-2020	55	29	.	.	.
kim-ai-2020	56	1	Note	note	VB
kim-ai-2020	56	2	that	that	IN
kim-ai-2020	56	3	these	these	DT
kim-ai-2020	56	4	are	be	VBP
kim-ai-2020	56	5	highly	highly	RB
kim-ai-2020	56	6	simplified	simplified	JJ
kim-ai-2020	56	7	accounts	account	NNS
kim-ai-2020	56	8	of	of	IN
kim-ai-2020	56	9	utilitarianism	utilitarianism	NN
kim-ai-2020	56	10	and	and	CC
kim-ai-2020	56	11	deontology	deontology	NN
kim-ai-2020	56	12	.	.	.
kim-ai-2020	57	1	The	the	DT
kim-ai-2020	57	2	good	good	NN
kim-ai-2020	57	3	in	in	IN
kim-ai-2020	57	4	utilitarianism	utilitarianism	NN
kim-ai-2020	57	5	can	can	MD
kim-ai-2020	57	6	be	be	VB
kim-ai-2020	57	7	interpreted	interpret	VBN
kim-ai-2020	57	8	in	in	IN
kim-ai-2020	57	9	many	many	JJ
kim-ai-2020	57	10	different	different	JJ
kim-ai-2020	57	11	ways	way	NNS
kim-ai-2020	57	12	,	,	,
kim-ai-2020	57	13	and	and	CC
kim-ai-2020	57	14	the	the	DT
kim-ai-2020	57	15	issue	issue	NN
kim-ai-2020	57	16	of	of	IN
kim-ai-2020	57	17	conflicting	conflict	VBG
kim-ai-2020	57	18	moral	moral	JJ
kim-ai-2020	57	19	rules	rule	NNS
kim-ai-2020	57	20	is	be	VBZ
kim-ai-2020	57	21	a	a	DT
kim-ai-2020	57	22	perennial	perennial	JJ
kim-ai-2020	57	23	problem	problem	NN
kim-ai-2020	57	24	that	that	IN
kim-ai-2020	57	25	deontological	deontological	JJ
kim-ai-2020	57	26	ethics	ethic	NNS
kim-ai-2020	57	27	grapples	grapple	VBZ
kim-ai-2020	57	28	with	with	IN
kim-ai-2020	57	29	.	.	.
kim-ai-2020	58	1	[	[	-LRB-
kim-ai-2020	58	2	footnoteRef:6	footnoteref:6	NN
kim-ai-2020	58	3	]	]	-RRB-
kim-ai-2020	58	4	For	for	IN
kim-ai-2020	58	5	our	-PRON-	PRP$
kim-ai-2020	58	6	purpose	purpose	NN
kim-ai-2020	58	7	,	,	,
kim-ai-2020	58	8	however	however	RB
kim-ai-2020	58	9	,	,	,
kim-ai-2020	58	10	these	these	DT
kim-ai-2020	58	11	simplified	simplify	VBN
kim-ai-2020	58	12	accounts	account	NNS
kim-ai-2020	58	13	are	be	VBP
kim-ai-2020	58	14	sufficient	sufficient	JJ
kim-ai-2020	58	15	to	to	TO
kim-ai-2020	58	16	highlight	highlight	VB
kim-ai-2020	58	17	the	the	DT
kim-ai-2020	58	18	aspects	aspect	NNS
kim-ai-2020	58	19	in	in	IN
kim-ai-2020	58	20	which	which	WDT
kim-ai-2020	58	21	the	the	DT
kim-ai-2020	58	22	utilitarian	utilitarian	NN
kim-ai-2020	58	23	and	and	CC
kim-ai-2020	58	24	the	the	DT
kim-ai-2020	58	25	deontological	deontological	JJ
kim-ai-2020	58	26	position	position	NN
kim-ai-2020	58	27	appeal	appeal	VBP
kim-ai-2020	58	28	to	to	IN
kim-ai-2020	58	29	and	and	CC
kim-ai-2020	58	30	go	go	VB
kim-ai-2020	58	31	against	against	IN
kim-ai-2020	58	32	our	-PRON-	PRP$
kim-ai-2020	58	33	moral	moral	JJ
kim-ai-2020	58	34	intuition	intuition	NN
kim-ai-2020	58	35	.	.	.
kim-ai-2020	59	1	[	[	-LRB-
kim-ai-2020	59	2	6	6	CD
kim-ai-2020	59	3	:	:	:
kim-ai-2020	59	4	For	for	IN
kim-ai-2020	59	5	an	an	DT
kim-ai-2020	59	6	overview	overview	NN
kim-ai-2020	59	7	,	,	,
kim-ai-2020	59	8	see	see	VB
kim-ai-2020	59	9	Sinnott	Sinnott	NNP
kim-ai-2020	59	10	-	-	HYPH
kim-ai-2020	59	11	Armstrong	Armstrong	NNP
kim-ai-2020	59	12	2019	2019	CD
kim-ai-2020	59	13	and	and	CC
kim-ai-2020	59	14	Alexander	Alexander	NNP
kim-ai-2020	59	15	and	and	CC
kim-ai-2020	59	16	Moore	Moore	NNP
kim-ai-2020	59	17	2016	2016	CD
kim-ai-2020	59	18	.	.	.
kim-ai-2020	59	19	]	]	-RRB-
kim-ai-2020	60	1	If	if	IN
kim-ai-2020	60	2	a	a	DT
kim-ai-2020	60	3	trolley	trolley	NN
kim-ai-2020	60	4	can	can	MD
kim-ai-2020	60	5	not	not	RB
kim-ai-2020	60	6	be	be	VB
kim-ai-2020	60	7	stopped	stop	VBN
kim-ai-2020	60	8	,	,	,
kim-ai-2020	60	9	saving	save	VBG
kim-ai-2020	60	10	five	five	CD
kim-ai-2020	60	11	lives	life	NNS
kim-ai-2020	60	12	over	over	IN
kim-ai-2020	60	13	one	one	NN
kim-ai-2020	60	14	seems	seem	VBZ
kim-ai-2020	60	15	to	to	TO
kim-ai-2020	60	16	be	be	VB
kim-ai-2020	60	17	a	a	DT
kim-ai-2020	60	18	right	right	JJ
kim-ai-2020	60	19	thing	thing	NN
kim-ai-2020	60	20	to	to	TO
kim-ai-2020	60	21	do	do	VB
kim-ai-2020	60	22	.	.	.
kim-ai-2020	61	1	Utilitarianism	Utilitarianism	NNP
kim-ai-2020	61	2	appears	appear	VBZ
kim-ai-2020	61	3	to	to	TO
kim-ai-2020	61	4	get	get	VB
kim-ai-2020	61	5	things	thing	NNS
kim-ai-2020	61	6	right	right	JJ
kim-ai-2020	61	7	in	in	IN
kim-ai-2020	61	8	this	this	DT
kim-ai-2020	61	9	respect	respect	NN
kim-ai-2020	61	10	.	.	.
kim-ai-2020	62	1	However	however	RB
kim-ai-2020	62	2	,	,	,
kim-ai-2020	62	3	it	-PRON-	PRP
kim-ai-2020	62	4	is	be	VBZ
kim-ai-2020	62	5	hard	hard	JJ
kim-ai-2020	62	6	to	to	TO
kim-ai-2020	62	7	dispute	dispute	VB
kim-ai-2020	62	8	that	that	IN
kim-ai-2020	62	9	killing	kill	VBG
kim-ai-2020	62	10	people	people	NNS
kim-ai-2020	62	11	is	be	VBZ
kim-ai-2020	62	12	wrong	wrong	JJ
kim-ai-2020	62	13	.	.	.
kim-ai-2020	63	1	If	if	IN
kim-ai-2020	63	2	killing	killing	NN
kim-ai-2020	63	3	is	be	VBZ
kim-ai-2020	63	4	morally	morally	RB
kim-ai-2020	63	5	wrong	wrong	JJ
kim-ai-2020	63	6	no	no	RB
kim-ai-2020	63	7	matter	matter	RB
kim-ai-2020	63	8	what	what	WP
kim-ai-2020	63	9	,	,	,
kim-ai-2020	63	10	deontology	deontology	NN
kim-ai-2020	63	11	seems	seem	VBZ
kim-ai-2020	63	12	to	to	TO
kim-ai-2020	63	13	make	make	VB
kim-ai-2020	63	14	more	more	JJR
kim-ai-2020	63	15	sense	sense	NN
kim-ai-2020	63	16	.	.	.
kim-ai-2020	64	1	With	with	IN
kim-ai-2020	64	2	moral	moral	JJ
kim-ai-2020	64	3	theories	theory	NNS
kim-ai-2020	64	4	,	,	,
kim-ai-2020	64	5	things	thing	NNS
kim-ai-2020	64	6	seem	seem	VBP
kim-ai-2020	64	7	to	to	TO
kim-ai-2020	64	8	get	get	VB
kim-ai-2020	64	9	more	more	RBR
kim-ai-2020	64	10	confusing	confusing	JJ
kim-ai-2020	64	11	.	.	.
kim-ai-2020	65	1	Consider	consider	VB
kim-ai-2020	65	2	the	the	DT
kim-ai-2020	65	3	case	case	NN
kim-ai-2020	65	4	in	in	IN
kim-ai-2020	65	5	which	which	WDT
kim-ai-2020	65	6	one	one	NN
kim-ai-2020	65	7	freezes	freeze	VBZ
kim-ai-2020	65	8	and	and	CC
kim-ai-2020	65	9	fails	fail	VBZ
kim-ai-2020	65	10	to	to	TO
kim-ai-2020	65	11	pull	pull	VB
kim-ai-2020	65	12	the	the	DT
kim-ai-2020	65	13	level	level	NN
kim-ai-2020	65	14	.	.	.
kim-ai-2020	66	1	According	accord	VBG
kim-ai-2020	66	2	to	to	IN
kim-ai-2020	66	3	utilitarianism	utilitarianism	NN
kim-ai-2020	66	4	,	,	,
kim-ai-2020	66	5	this	this	DT
kim-ai-2020	66	6	would	would	MD
kim-ai-2020	66	7	be	be	VB
kim-ai-2020	66	8	morally	morally	RB
kim-ai-2020	66	9	wrong	wrong	JJ
kim-ai-2020	66	10	because	because	IN
kim-ai-2020	66	11	it	-PRON-	PRP
kim-ai-2020	66	12	fails	fail	VBZ
kim-ai-2020	66	13	to	to	TO
kim-ai-2020	66	14	maximize	maximize	VB
kim-ai-2020	66	15	the	the	DT
kim-ai-2020	66	16	greatest	great	JJS
kim-ai-2020	66	17	good	good	JJ
kim-ai-2020	66	18	,	,	,
kim-ai-2020	66	19	i.e.	i.e.	FW
kim-ai-2020	67	1	human	human	JJ
kim-ai-2020	67	2	lives	life	NNS
kim-ai-2020	67	3	.	.	.
kim-ai-2020	68	1	But	but	CC
kim-ai-2020	68	2	how	how	WRB
kim-ai-2020	68	3	far	far	RB
kim-ai-2020	68	4	should	should	MD
kim-ai-2020	68	5	one	one	PRP
kim-ai-2020	68	6	go	go	VB
kim-ai-2020	68	7	to	to	TO
kim-ai-2020	68	8	maximize	maximize	VB
kim-ai-2020	68	9	the	the	DT
kim-ai-2020	68	10	good	good	NN
kim-ai-2020	68	11	?	?	.
kim-ai-2020	69	1	Suppose	suppose	VB
kim-ai-2020	69	2	there	there	EX
kim-ai-2020	69	3	is	be	VBZ
kim-ai-2020	69	4	a	a	DT
kim-ai-2020	69	5	very	very	RB
kim-ai-2020	69	6	large	large	JJ
kim-ai-2020	69	7	person	person	NN
kim-ai-2020	69	8	on	on	IN
kim-ai-2020	69	9	a	a	DT
kim-ai-2020	69	10	footbridge	footbridge	NN
kim-ai-2020	69	11	over	over	IN
kim-ai-2020	69	12	the	the	DT
kim-ai-2020	69	13	trolley	trolley	NN
kim-ai-2020	69	14	track	track	NN
kim-ai-2020	69	15	and	and	CC
kim-ai-2020	69	16	one	one	NN
kim-ai-2020	69	17	pushes	push	VBZ
kim-ai-2020	69	18	that	that	DT
kim-ai-2020	69	19	person	person	NN
kim-ai-2020	69	20	off	off	IN
kim-ai-2020	69	21	the	the	DT
kim-ai-2020	69	22	footbridge	footbridge	NN
kim-ai-2020	69	23	onto	onto	IN
kim-ai-2020	69	24	the	the	DT
kim-ai-2020	69	25	track	track	NN
kim-ai-2020	69	26	,	,	,
kim-ai-2020	69	27	thus	thus	RB
kim-ai-2020	69	28	stopping	stop	VBG
kim-ai-2020	69	29	the	the	DT
kim-ai-2020	69	30	trolley	trolley	NN
kim-ai-2020	69	31	and	and	CC
kim-ai-2020	69	32	saving	save	VBG
kim-ai-2020	69	33	the	the	DT
kim-ai-2020	69	34	five	five	CD
kim-ai-2020	69	35	people	people	NNS
kim-ai-2020	69	36	.	.	.
kim-ai-2020	70	1	Would	Would	MD
kim-ai-2020	70	2	this	this	DT
kim-ai-2020	70	3	count	count	VB
kim-ai-2020	70	4	as	as	IN
kim-ai-2020	70	5	a	a	DT
kim-ai-2020	70	6	right	right	JJ
kim-ai-2020	70	7	thing	thing	NN
kim-ai-2020	70	8	to	to	TO
kim-ai-2020	70	9	do	do	VB
kim-ai-2020	70	10	?	?	.
kim-ai-2020	71	1	Utilitarianism	Utilitarianism	NNP
kim-ai-2020	71	2	may	may	MD
kim-ai-2020	71	3	argue	argue	VB
kim-ai-2020	71	4	that	that	DT
kim-ai-2020	71	5	.	.	.
kim-ai-2020	72	1	But	but	CC
kim-ai-2020	72	2	in	in	IN
kim-ai-2020	72	3	real	real	JJ
kim-ai-2020	72	4	life	life	NN
kim-ai-2020	72	5	,	,	,
kim-ai-2020	72	6	many	many	JJ
kim-ai-2020	72	7	might	may	MD
kim-ai-2020	72	8	consider	consider	VB
kim-ai-2020	72	9	throwing	throw	VBG
kim-ai-2020	72	10	a	a	DT
kim-ai-2020	72	11	person	person	NN
kim-ai-2020	72	12	morally	morally	RB
kim-ai-2020	72	13	wrong	wrong	JJ
kim-ai-2020	72	14	but	but	CC
kim-ai-2020	72	15	pulling	pull	VBG
kim-ai-2020	72	16	the	the	DT
kim-ai-2020	72	17	lever	lever	NN
kim-ai-2020	72	18	morally	morally	RB
kim-ai-2020	72	19	permissible	permissible	JJ
kim-ai-2020	72	20	.	.	.
kim-ai-2020	73	1	[	[	-LRB-
kim-ai-2020	73	2	footnoteRef:7	footnoteref:7	NN
kim-ai-2020	73	3	]	]	-RRB-
kim-ai-2020	73	4	[	[	-LRB-
kim-ai-2020	73	5	7	7	CD
kim-ai-2020	73	6	:	:	:
kim-ai-2020	73	7	For	for	IN
kim-ai-2020	73	8	an	an	DT
kim-ai-2020	73	9	empirical	empirical	JJ
kim-ai-2020	73	10	study	study	NN
kim-ai-2020	73	11	on	on	IN
kim-ai-2020	73	12	this	this	DT
kim-ai-2020	73	13	,	,	,
kim-ai-2020	73	14	see	see	VB
kim-ai-2020	73	15	Cushman	Cushman	NNP
kim-ai-2020	73	16	,	,	,
kim-ai-2020	73	17	Young	Young	NNP
kim-ai-2020	73	18	,	,	,
kim-ai-2020	73	19	and	and	CC
kim-ai-2020	73	20	Hauser	Hauser	NNP
kim-ai-2020	73	21	2006	2006	CD
kim-ai-2020	73	22	.	.	.
kim-ai-2020	74	1	For	for	IN
kim-ai-2020	74	2	the	the	DT
kim-ai-2020	74	3	results	result	NNS
kim-ai-2020	74	4	of	of	IN
kim-ai-2020	74	5	a	a	DT
kim-ai-2020	74	6	similar	similar	JJ
kim-ai-2020	74	7	survey	survey	NN
kim-ai-2020	74	8	that	that	WDT
kim-ai-2020	74	9	involves	involve	VBZ
kim-ai-2020	74	10	an	an	DT
kim-ai-2020	74	11	autonomous	autonomous	JJ
kim-ai-2020	74	12	car	car	NN
kim-ai-2020	74	13	instead	instead	RB
kim-ai-2020	74	14	of	of	IN
kim-ai-2020	74	15	a	a	DT
kim-ai-2020	74	16	trolley	trolley	NN
kim-ai-2020	74	17	,	,	,
kim-ai-2020	74	18	see	see	VB
kim-ai-2020	74	19	Bonnefon	Bonnefon	NNP
kim-ai-2020	74	20	,	,	,
kim-ai-2020	74	21	Shariff	Shariff	NNP
kim-ai-2020	74	22	,	,	,
kim-ai-2020	74	23	and	and	CC
kim-ai-2020	74	24	Rahwan	Rahwan	NNP
kim-ai-2020	74	25	2016	2016	CD
kim-ai-2020	74	26	.	.	.
kim-ai-2020	74	27	]	]	-RRB-
kim-ai-2020	75	1	The	the	DT
kim-ai-2020	75	2	problem	problem	NN
kim-ai-2020	75	3	with	with	IN
kim-ai-2020	75	4	utilitarianism	utilitarianism	NN
kim-ai-2020	75	5	is	be	VBZ
kim-ai-2020	75	6	that	that	IN
kim-ai-2020	75	7	it	-PRON-	PRP
kim-ai-2020	75	8	treats	treat	VBZ
kim-ai-2020	75	9	the	the	DT
kim-ai-2020	75	10	good	good	JJ
kim-ai-2020	75	11	as	as	IN
kim-ai-2020	75	12	something	something	NN
kim-ai-2020	75	13	inherently	inherently	RB
kim-ai-2020	75	14	quantifiable	quantifiable	JJ
kim-ai-2020	75	15	,	,	,
kim-ai-2020	75	16	comparable	comparable	JJ
kim-ai-2020	75	17	,	,	,
kim-ai-2020	75	18	calculable	calculable	JJ
kim-ai-2020	75	19	,	,	,
kim-ai-2020	75	20	and	and	CC
kim-ai-2020	75	21	additive	additive	NNP
kim-ai-2020	75	22	.	.	.
kim-ai-2020	76	1	But	but	CC
kim-ai-2020	76	2	not	not	RB
kim-ai-2020	76	3	all	all	DT
kim-ai-2020	76	4	considerations	consideration	NNS
kim-ai-2020	76	5	that	that	WDT
kim-ai-2020	76	6	we	-PRON-	PRP
kim-ai-2020	76	7	have	have	VBP
kim-ai-2020	76	8	to	to	TO
kim-ai-2020	76	9	factor	factor	VB
kim-ai-2020	76	10	into	into	IN
kim-ai-2020	76	11	moral	moral	JJ
kim-ai-2020	76	12	decision	decision	NN
kim-ai-2020	76	13	-	-	HYPH
kim-ai-2020	76	14	making	making	NN
kim-ai-2020	76	15	are	be	VBP
kim-ai-2020	76	16	measurable	measurable	JJ
kim-ai-2020	76	17	in	in	IN
kim-ai-2020	76	18	numbers	number	NNS
kim-ai-2020	76	19	.	.	.
kim-ai-2020	77	1	What	what	WP
kim-ai-2020	77	2	if	if	IN
kim-ai-2020	77	3	the	the	DT
kim-ai-2020	77	4	five	five	CD
kim-ai-2020	77	5	people	people	NNS
kim-ai-2020	77	6	on	on	IN
kim-ai-2020	77	7	the	the	DT
kim-ai-2020	77	8	track	track	NN
kim-ai-2020	77	9	are	be	VBP
kim-ai-2020	77	10	helpless	helpless	JJ
kim-ai-2020	77	11	babies	baby	NNS
kim-ai-2020	77	12	or	or	CC
kim-ai-2020	77	13	murderers	murderer	NNS
kim-ai-2020	77	14	who	who	WP
kim-ai-2020	77	15	just	just	RB
kim-ai-2020	77	16	escaped	escape	VBD
kim-ai-2020	77	17	from	from	IN
kim-ai-2020	77	18	the	the	DT
kim-ai-2020	77	19	prison	prison	NN
kim-ai-2020	77	20	?	?	.
kim-ai-2020	78	1	Would	Would	MD
kim-ai-2020	78	2	or	or	CC
kim-ai-2020	78	3	should	should	MD
kim-ai-2020	78	4	that	that	DT
kim-ai-2020	78	5	affect	affect	VB
kim-ai-2020	78	6	our	-PRON-	PRP$
kim-ai-2020	78	7	decision	decision	NN
kim-ai-2020	78	8	?	?	.
kim-ai-2020	79	1	Some	some	DT
kim-ai-2020	79	2	of	of	IN
kim-ai-2020	79	3	us	-PRON-	PRP
kim-ai-2020	79	4	would	would	MD
kim-ai-2020	79	5	surely	surely	RB
kim-ai-2020	79	6	hesitate	hesitate	VB
kim-ai-2020	79	7	saving	save	VBG
kim-ai-2020	79	8	the	the	DT
kim-ai-2020	79	9	lives	life	NNS
kim-ai-2020	79	10	of	of	IN
kim-ai-2020	79	11	five	five	CD
kim-ai-2020	79	12	murderers	murderer	NNS
kim-ai-2020	79	13	by	by	IN
kim-ai-2020	79	14	sacrificing	sacrifice	VBG
kim-ai-2020	79	15	one	one	CD
kim-ai-2020	79	16	innocent	innocent	JJ
kim-ai-2020	79	17	baby	baby	NN
kim-ai-2020	79	18	.	.	.
kim-ai-2020	80	1	But	but	CC
kim-ai-2020	80	2	what	what	WP
kim-ai-2020	80	3	if	if	IN
kim-ai-2020	80	4	things	thing	NNS
kim-ai-2020	80	5	were	be	VBD
kim-ai-2020	80	6	different	different	JJ
kim-ai-2020	80	7	and	and	CC
kim-ai-2020	80	8	we	-PRON-	PRP
kim-ai-2020	80	9	were	be	VBD
kim-ai-2020	80	10	comparing	compare	VBG
kim-ai-2020	80	11	five	five	CD
kim-ai-2020	80	12	school	school	NN
kim-ai-2020	80	13	children	child	NNS
kim-ai-2020	80	14	versus	versus	IN
kim-ai-2020	80	15	one	one	CD
kim-ai-2020	80	16	baby	baby	NN
kim-ai-2020	80	17	or	or	CC
kim-ai-2020	80	18	five	five	CD
kim-ai-2020	80	19	babies	baby	NNS
kim-ai-2020	80	20	versus	versus	IN
kim-ai-2020	80	21	one	one	CD
kim-ai-2020	80	22	schoolchild	schoolchild	NN
kim-ai-2020	80	23	?	?	.
kim-ai-2020	81	1	No	no	DT
kim-ai-2020	81	2	one	one	NN
kim-ai-2020	81	3	can	can	MD
kim-ai-2020	81	4	say	say	VB
kim-ai-2020	81	5	for	for	IN
kim-ai-2020	81	6	sure	sure	JJ
kim-ai-2020	81	7	what	what	WP
kim-ai-2020	81	8	is	be	VBZ
kim-ai-2020	81	9	the	the	DT
kim-ai-2020	81	10	morally	morally	RB
kim-ai-2020	81	11	right	right	JJ
kim-ai-2020	81	12	action	action	NN
kim-ai-2020	81	13	in	in	IN
kim-ai-2020	81	14	those	those	DT
kim-ai-2020	81	15	cases	case	NNS
kim-ai-2020	81	16	.	.	.
kim-ai-2020	82	1	[	[	-LRB-
kim-ai-2020	82	2	footnoteRef:8	footnoteref:8	NN
kim-ai-2020	82	3	]	]	-RRB-
kim-ai-2020	82	4	[	[	-LRB-
kim-ai-2020	82	5	8	8	CD
kim-ai-2020	82	6	:	:	:
kim-ai-2020	82	7	For	for	IN
kim-ai-2020	82	8	an	an	DT
kim-ai-2020	82	9	attempt	attempt	NN
kim-ai-2020	82	10	to	to	TO
kim-ai-2020	82	11	identify	identify	VB
kim-ai-2020	82	12	moral	moral	JJ
kim-ai-2020	82	13	principles	principle	NNS
kim-ai-2020	82	14	behind	behind	IN
kim-ai-2020	82	15	our	-PRON-	PRP$
kim-ai-2020	82	16	moral	moral	JJ
kim-ai-2020	82	17	intuition	intuition	NN
kim-ai-2020	82	18	in	in	IN
kim-ai-2020	82	19	different	different	JJ
kim-ai-2020	82	20	versions	version	NNS
kim-ai-2020	82	21	of	of	IN
kim-ai-2020	82	22	the	the	DT
kim-ai-2020	82	23	trolley	trolley	NN
kim-ai-2020	82	24	problem	problem	NN
kim-ai-2020	82	25	and	and	CC
kim-ai-2020	82	26	other	other	JJ
kim-ai-2020	82	27	similar	similar	JJ
kim-ai-2020	82	28	cases	case	NNS
kim-ai-2020	82	29	,	,	,
kim-ai-2020	82	30	see	see	VB
kim-ai-2020	82	31	Thomson	Thomson	NNP
kim-ai-2020	82	32	1976	1976	CD
kim-ai-2020	82	33	.	.	.
kim-ai-2020	82	34	]	]	-RRB-
kim-ai-2020	83	1	While	while	IN
kim-ai-2020	83	2	the	the	DT
kim-ai-2020	83	3	utilitarian	utilitarian	JJ
kim-ai-2020	83	4	position	position	NN
kim-ai-2020	83	5	appears	appear	VBZ
kim-ai-2020	83	6	less	less	RBR
kim-ai-2020	83	7	persuasive	persuasive	JJ
kim-ai-2020	83	8	in	in	IN
kim-ai-2020	83	9	light	light	NN
kim-ai-2020	83	10	of	of	IN
kim-ai-2020	83	11	these	these	DT
kim-ai-2020	83	12	considerations	consideration	NNS
kim-ai-2020	83	13	,	,	,
kim-ai-2020	83	14	deontology	deontology	NN
kim-ai-2020	83	15	does	do	VBZ
kim-ai-2020	83	16	n’t	not	RB
kim-ai-2020	83	17	fare	fare	VB
kim-ai-2020	83	18	too	too	RB
kim-ai-2020	83	19	well	well	RB
kim-ai-2020	83	20	,	,	,
kim-ai-2020	83	21	either	either	RB
kim-ai-2020	83	22	.	.	.
kim-ai-2020	84	1	Deontology	deontology	NN
kim-ai-2020	84	2	emphasizes	emphasize	VBZ
kim-ai-2020	84	3	one	one	PRP
kim-ai-2020	84	4	’s	’s	NN
kim-ai-2020	84	5	duty	duty	NN
kim-ai-2020	84	6	to	to	TO
kim-ai-2020	84	7	observe	observe	VB
kim-ai-2020	84	8	moral	moral	JJ
kim-ai-2020	84	9	rules	rule	NNS
kim-ai-2020	84	10	.	.	.
kim-ai-2020	85	1	But	but	CC
kim-ai-2020	85	2	what	what	WP
kim-ai-2020	85	3	if	if	IN
kim-ai-2020	85	4	those	those	DT
kim-ai-2020	85	5	moral	moral	JJ
kim-ai-2020	85	6	rules	rule	NNS
kim-ai-2020	85	7	conflict	conflict	NN
kim-ai-2020	85	8	with	with	IN
kim-ai-2020	85	9	one	one	CD
kim-ai-2020	85	10	another	another	DT
kim-ai-2020	85	11	?	?	.
kim-ai-2020	86	1	Between	between	IN
kim-ai-2020	86	2	the	the	DT
kim-ai-2020	86	3	two	two	CD
kim-ai-2020	86	4	moral	moral	JJ
kim-ai-2020	86	5	rules	rule	NNS
kim-ai-2020	86	6	,	,	,
kim-ai-2020	86	7	“	"	``
kim-ai-2020	86	8	do	do	VB
kim-ai-2020	86	9	not	not	RB
kim-ai-2020	86	10	kill	kill	VB
kim-ai-2020	86	11	a	a	DT
kim-ai-2020	86	12	person	person	NN
kim-ai-2020	86	13	”	"	''
kim-ai-2020	86	14	and	and	CC
kim-ai-2020	86	15	“	"	``
kim-ai-2020	86	16	save	save	VB
kim-ai-2020	86	17	lives	life	NNS
kim-ai-2020	86	18	,	,	,
kim-ai-2020	86	19	”	"	''
kim-ai-2020	86	20	which	which	WDT
kim-ai-2020	86	21	one	one	PRP
kim-ai-2020	86	22	should	should	MD
kim-ai-2020	86	23	trump	trump	VB
kim-ai-2020	86	24	the	the	DT
kim-ai-2020	86	25	other	other	JJ
kim-ai-2020	86	26	?	?	.
kim-ai-2020	87	1	The	the	DT
kim-ai-2020	87	2	conflict	conflict	NN
kim-ai-2020	87	3	among	among	IN
kim-ai-2020	87	4	values	value	NNS
kim-ai-2020	87	5	is	be	VBZ
kim-ai-2020	87	6	common	common	JJ
kim-ai-2020	87	7	in	in	IN
kim-ai-2020	87	8	life	life	NN
kim-ai-2020	87	9	,	,	,
kim-ai-2020	87	10	and	and	CC
kim-ai-2020	87	11	deontology	deontology	NN
kim-ai-2020	87	12	faces	face	VBZ
kim-ai-2020	87	13	difficulty	difficulty	NN
kim-ai-2020	87	14	in	in	IN
kim-ai-2020	87	15	guiding	guide	VBG
kim-ai-2020	87	16	how	how	WRB
kim-ai-2020	87	17	an	an	DT
kim-ai-2020	87	18	intelligent	intelligent	JJ
kim-ai-2020	87	19	agent	agent	NN
kim-ai-2020	87	20	is	be	VBZ
kim-ai-2020	87	21	to	to	TO
kim-ai-2020	87	22	act	act	VB
kim-ai-2020	87	23	in	in	IN
kim-ai-2020	87	24	a	a	DT
kim-ai-2020	87	25	tricky	tricky	JJ
kim-ai-2020	87	26	situation	situation	NN
kim-ai-2020	87	27	like	like	IN
kim-ai-2020	87	28	the	the	DT
kim-ai-2020	87	29	trolley	trolley	NN
kim-ai-2020	87	30	problem	problem	NN
kim-ai-2020	87	31	.	.	.
kim-ai-2020	88	1	[	[	-LRB-
kim-ai-2020	88	2	footnoteRef:9	footnoteRef:9	NNP
kim-ai-2020	88	3	]	]	-RRB-
kim-ai-2020	88	4	[	[	-LRB-
kim-ai-2020	88	5	9	9	CD
kim-ai-2020	88	6	:	:	:
kim-ai-2020	88	7	Some	some	DT
kim-ai-2020	88	8	moral	moral	JJ
kim-ai-2020	88	9	philosophers	philosopher	NNS
kim-ai-2020	88	10	doubt	doubt	VBP
kim-ai-2020	88	11	the	the	DT
kim-ai-2020	88	12	value	value	NN
kim-ai-2020	88	13	of	of	IN
kim-ai-2020	88	14	our	-PRON-	PRP$
kim-ai-2020	88	15	moral	moral	JJ
kim-ai-2020	88	16	intuition	intuition	NN
kim-ai-2020	88	17	in	in	IN
kim-ai-2020	88	18	constructing	construct	VBG
kim-ai-2020	88	19	a	a	DT
kim-ai-2020	88	20	moral	moral	JJ
kim-ai-2020	88	21	theory	theory	NN
kim-ai-2020	88	22	.	.	.
kim-ai-2020	89	1	See	see	VB
kim-ai-2020	89	2	Singer	singer	NN
kim-ai-2020	89	3	2005	2005	CD
kim-ai-2020	89	4	,	,	,
kim-ai-2020	89	5	for	for	IN
kim-ai-2020	89	6	example	example	NN
kim-ai-2020	89	7	.	.	.
kim-ai-2020	90	1	But	but	CC
kim-ai-2020	90	2	a	a	DT
kim-ai-2020	90	3	moral	moral	JJ
kim-ai-2020	90	4	theory	theory	NN
kim-ai-2020	90	5	that	that	IN
kim-ai-2020	90	6	clashes	clash	NNS
kim-ai-2020	90	7	with	with	IN
kim-ai-2020	90	8	common	common	JJ
kim-ai-2020	90	9	moral	moral	JJ
kim-ai-2020	90	10	intuition	intuition	NN
kim-ai-2020	90	11	is	be	VBZ
kim-ai-2020	90	12	unlikely	unlikely	JJ
kim-ai-2020	90	13	to	to	TO
kim-ai-2020	90	14	be	be	VB
kim-ai-2020	90	15	sought	seek	VBN
kim-ai-2020	90	16	out	out	RP
kim-ai-2020	90	17	as	as	IN
kim-ai-2020	90	18	a	a	DT
kim-ai-2020	90	19	guide	guide	NN
kim-ai-2020	90	20	to	to	IN
kim-ai-2020	90	21	making	make	VBG
kim-ai-2020	90	22	an	an	DT
kim-ai-2020	90	23	ethical	ethical	JJ
kim-ai-2020	90	24	decision	decision	NN
kim-ai-2020	90	25	.	.	.
kim-ai-2020	90	26	]	]	-RRB-
kim-ai-2020	91	1	3	3	LS
kim-ai-2020	91	2	.	.	.
kim-ai-2020	92	1	Understanding	understand	VBG
kim-ai-2020	92	2	What	what	WP
kim-ai-2020	92	3	Ethics	Ethics	NNP
kim-ai-2020	92	4	Has	have	VBZ
kim-ai-2020	92	5	to	to	TO
kim-ai-2020	92	6	Offer	offer	VB
kim-ai-2020	92	7	Due	due	IN
kim-ai-2020	92	8	to	to	IN
kim-ai-2020	92	9	the	the	DT
kim-ai-2020	92	10	high	high	JJ
kim-ai-2020	92	11	stakes	stake	NNS
kim-ai-2020	92	12	involved	involve	VBN
kim-ai-2020	92	13	,	,	,
kim-ai-2020	92	14	AI	AI	NNP
kim-ai-2020	92	15	-	-	HYPH
kim-ai-2020	92	16	powered	power	VBN
kim-ai-2020	92	17	military	military	JJ
kim-ai-2020	92	18	robots	robot	NNS
kim-ai-2020	92	19	and	and	CC
kim-ai-2020	92	20	autonomous	autonomous	JJ
kim-ai-2020	92	21	weapons	weapon	NNS
kim-ai-2020	92	22	systems	system	NNS
kim-ai-2020	92	23	can	can	MD
kim-ai-2020	92	24	present	present	VB
kim-ai-2020	92	25	the	the	DT
kim-ai-2020	92	26	moral	moral	JJ
kim-ai-2020	92	27	dilemma	dilemma	NN
kim-ai-2020	92	28	in	in	IN
kim-ai-2020	92	29	the	the	DT
kim-ai-2020	92	30	trolley	trolley	NN
kim-ai-2020	92	31	problem	problem	NN
kim-ai-2020	92	32	more	more	RBR
kim-ai-2020	92	33	convincingly	convincingly	RB
kim-ai-2020	92	34	.	.	.
kim-ai-2020	93	1	Suppose	suppose	VB
kim-ai-2020	93	2	that	that	IN
kim-ai-2020	93	3	some	some	DT
kim-ai-2020	93	4	engineers	engineer	NNS
kim-ai-2020	93	5	,	,	,
kim-ai-2020	93	6	following	follow	VBG
kim-ai-2020	93	7	utilitarianism	utilitarianism	NN
kim-ai-2020	93	8	and	and	CC
kim-ai-2020	93	9	interpreting	interpret	VBG
kim-ai-2020	93	10	victory	victory	NN
kim-ai-2020	93	11	as	as	IN
kim-ai-2020	93	12	the	the	DT
kim-ai-2020	93	13	ultimate	ultimate	JJ
kim-ai-2020	93	14	good	good	JJ
kim-ai-2020	93	15	/	/	SYM
kim-ai-2020	93	16	utility	utility	NN
kim-ai-2020	93	17	,	,	,
kim-ai-2020	93	18	wish	wish	VBP
kim-ai-2020	93	19	to	to	TO
kim-ai-2020	93	20	program	program	VB
kim-ai-2020	93	21	an	an	DT
kim-ai-2020	93	22	unmanned	unmanned	JJ
kim-ai-2020	93	23	aerial	aerial	JJ
kim-ai-2020	93	24	vehicle	vehicle	NN
kim-ai-2020	93	25	(	(	-LRB-
kim-ai-2020	93	26	UAV	UAV	NNP
kim-ai-2020	93	27	)	)	-RRB-
kim-ai-2020	93	28	to	to	TO
kim-ai-2020	93	29	autonomously	autonomously	RB
kim-ai-2020	93	30	drop	drop	VB
kim-ai-2020	93	31	bombs	bomb	NNS
kim-ai-2020	93	32	in	in	IN
kim-ai-2020	93	33	order	order	NN
kim-ai-2020	93	34	to	to	TO
kim-ai-2020	93	35	maximize	maximize	VB
kim-ai-2020	93	36	the	the	DT
kim-ai-2020	93	37	chances	chance	NNS
kim-ai-2020	93	38	of	of	IN
kim-ai-2020	93	39	victory	victory	NN
kim-ai-2020	93	40	.	.	.
kim-ai-2020	94	1	That	that	DT
kim-ai-2020	94	2	may	may	MD
kim-ai-2020	94	3	result	result	VB
kim-ai-2020	94	4	in	in	IN
kim-ai-2020	94	5	sacrificing	sacrifice	VBG
kim-ai-2020	94	6	a	a	DT
kim-ai-2020	94	7	greater	great	JJR
kim-ai-2020	94	8	number	number	NN
kim-ai-2020	94	9	of	of	IN
kim-ai-2020	94	10	civilians	civilian	NNS
kim-ai-2020	94	11	than	than	IN
kim-ai-2020	94	12	necessary	necessary	JJ
kim-ai-2020	94	13	,	,	,
kim-ai-2020	94	14	and	and	CC
kim-ai-2020	94	15	many	many	JJ
kim-ai-2020	94	16	will	will	MD
kim-ai-2020	94	17	consider	consider	VB
kim-ai-2020	94	18	this	this	DT
kim-ai-2020	94	19	to	to	TO
kim-ai-2020	94	20	be	be	VB
kim-ai-2020	94	21	morally	morally	RB
kim-ai-2020	94	22	wrong	wrong	JJ
kim-ai-2020	94	23	.	.	.
kim-ai-2020	95	1	Now	now	RB
kim-ai-2020	95	2	imagine	imagine	VB
kim-ai-2020	95	3	different	different	JJ
kim-ai-2020	95	4	engineers	engineer	NNS
kim-ai-2020	95	5	who	who	WP
kim-ai-2020	95	6	,	,	,
kim-ai-2020	95	7	adopting	adopt	VBG
kim-ai-2020	95	8	deontology	deontology	NN
kim-ai-2020	95	9	and	and	CC
kim-ai-2020	95	10	following	follow	VBG
kim-ai-2020	95	11	the	the	DT
kim-ai-2020	95	12	moral	moral	JJ
kim-ai-2020	95	13	principle	principle	NN
kim-ai-2020	95	14	of	of	IN
kim-ai-2020	95	15	not	not	RB
kim-ai-2020	95	16	killing	kill	VBG
kim-ai-2020	95	17	people	people	NNS
kim-ai-2020	95	18	,	,	,
kim-ai-2020	95	19	program	program	VB
kim-ai-2020	95	20	a	a	DT
kim-ai-2020	95	21	UAV	UAV	NNP
kim-ai-2020	95	22	to	to	TO
kim-ai-2020	95	23	autonomously	autonomously	RB
kim-ai-2020	95	24	act	act	VB
kim-ai-2020	95	25	in	in	IN
kim-ai-2020	95	26	a	a	DT
kim-ai-2020	95	27	manner	manner	NN
kim-ai-2020	95	28	that	that	WDT
kim-ai-2020	95	29	minimizes	minimize	VBZ
kim-ai-2020	95	30	casualties	casualty	NNS
kim-ai-2020	95	31	.	.	.
kim-ai-2020	96	1	This	this	DT
kim-ai-2020	96	2	may	may	MD
kim-ai-2020	96	3	lead	lead	VB
kim-ai-2020	96	4	to	to	IN
kim-ai-2020	96	5	defeat	defeat	NN
kim-ai-2020	96	6	on	on	IN
kim-ai-2020	96	7	the	the	DT
kim-ai-2020	96	8	battlefield	battlefield	NN
kim-ai-2020	96	9	,	,	,
kim-ai-2020	96	10	because	because	IN
kim-ai-2020	96	11	minimizing	minimize	VBG
kim-ai-2020	96	12	casualties	casualty	NNS
kim-ai-2020	96	13	may	may	MD
kim-ai-2020	96	14	not	not	RB
kim-ai-2020	96	15	be	be	VB
kim-ai-2020	96	16	always	always	RB
kim-ai-2020	96	17	advantageous	advantageous	JJ
kim-ai-2020	96	18	to	to	IN
kim-ai-2020	96	19	winning	win	VBG
kim-ai-2020	96	20	a	a	DT
kim-ai-2020	96	21	war	war	NN
kim-ai-2020	96	22	.	.	.
kim-ai-2020	97	1	These	these	DT
kim-ai-2020	97	2	considerations	consideration	NNS
kim-ai-2020	97	3	suggest	suggest	VBP
kim-ai-2020	97	4	that	that	IN
kim-ai-2020	97	5	philosophical	philosophical	JJ
kim-ai-2020	97	6	insights	insight	NNS
kim-ai-2020	97	7	from	from	IN
kim-ai-2020	97	8	utilitarianism	utilitarianism	NN
kim-ai-2020	97	9	and	and	CC
kim-ai-2020	97	10	deontology	deontology	NN
kim-ai-2020	97	11	may	may	MD
kim-ai-2020	97	12	provide	provide	VB
kim-ai-2020	97	13	little	little	JJ
kim-ai-2020	97	14	practical	practical	JJ
kim-ai-2020	97	15	guidance	guidance	NN
kim-ai-2020	97	16	on	on	IN
kim-ai-2020	97	17	how	how	WRB
kim-ai-2020	97	18	to	to	TO
kim-ai-2020	97	19	program	program	VB
kim-ai-2020	97	20	autonomous	autonomous	JJ
kim-ai-2020	97	21	AI	AI	NNP
kim-ai-2020	97	22	systems	system	NNS
kim-ai-2020	97	23	to	to	TO
kim-ai-2020	97	24	act	act	VB
kim-ai-2020	97	25	morally	morally	RB
kim-ai-2020	97	26	.	.	.
kim-ai-2020	98	1	Ethicists	ethicist	NNS
kim-ai-2020	98	2	seek	seek	VBP
kim-ai-2020	98	3	abstract	abstract	JJ
kim-ai-2020	98	4	principles	principle	NNS
kim-ai-2020	98	5	that	that	WDT
kim-ai-2020	98	6	can	can	MD
kim-ai-2020	98	7	be	be	VB
kim-ai-2020	98	8	generalized	generalize	VBN
kim-ai-2020	98	9	.	.	.
kim-ai-2020	99	1	For	for	IN
kim-ai-2020	99	2	this	this	DT
kim-ai-2020	99	3	reason	reason	NN
kim-ai-2020	99	4	,	,	,
kim-ai-2020	99	5	they	-PRON-	PRP
kim-ai-2020	99	6	are	be	VBP
kim-ai-2020	99	7	interested	interested	JJ
kim-ai-2020	99	8	in	in	IN
kim-ai-2020	99	9	borderline	borderline	NN
kim-ai-2020	99	10	cases	case	NNS
kim-ai-2020	99	11	that	that	WDT
kim-ai-2020	99	12	reveal	reveal	VBP
kim-ai-2020	99	13	subtle	subtle	JJ
kim-ai-2020	99	14	differences	difference	NNS
kim-ai-2020	99	15	in	in	IN
kim-ai-2020	99	16	our	-PRON-	PRP$
kim-ai-2020	99	17	moral	moral	JJ
kim-ai-2020	99	18	intuition	intuition	NN
kim-ai-2020	99	19	and	and	CC
kim-ai-2020	99	20	varying	vary	VBG
kim-ai-2020	99	21	moral	moral	JJ
kim-ai-2020	99	22	theories	theory	NNS
kim-ai-2020	99	23	.	.	.
kim-ai-2020	100	1	Their	-PRON-	PRP$
kim-ai-2020	100	2	goal	goal	NN
kim-ai-2020	100	3	is	be	VBZ
kim-ai-2020	100	4	to	to	TO
kim-ai-2020	100	5	define	define	VB
kim-ai-2020	100	6	what	what	WP
kim-ai-2020	100	7	is	be	VBZ
kim-ai-2020	100	8	moral	moral	JJ
kim-ai-2020	100	9	and	and	CC
kim-ai-2020	100	10	investigate	investigate	VB
kim-ai-2020	100	11	how	how	WRB
kim-ai-2020	100	12	moral	moral	JJ
kim-ai-2020	100	13	reasoning	reasoning	NN
kim-ai-2020	100	14	works	work	NNS
kim-ai-2020	100	15	or	or	CC
kim-ai-2020	100	16	should	should	MD
kim-ai-2020	100	17	work	work	VB
kim-ai-2020	100	18	.	.	.
kim-ai-2020	101	1	By	by	IN
kim-ai-2020	101	2	contrast	contrast	NN
kim-ai-2020	101	3	,	,	,
kim-ai-2020	101	4	engineers	engineer	NNS
kim-ai-2020	101	5	and	and	CC
kim-ai-2020	101	6	programmers	programmer	NNS
kim-ai-2020	101	7	pursue	pursue	VBP
kim-ai-2020	101	8	practical	practical	JJ
kim-ai-2020	101	9	solutions	solution	NNS
kim-ai-2020	101	10	to	to	IN
kim-ai-2020	101	11	real	real	JJ
kim-ai-2020	101	12	-	-	HYPH
kim-ai-2020	101	13	life	life	NN
kim-ai-2020	101	14	problems	problem	NNS
kim-ai-2020	101	15	and	and	CC
kim-ai-2020	101	16	look	look	VB
kim-ai-2020	101	17	for	for	IN
kim-ai-2020	101	18	guidelines	guideline	NNS
kim-ai-2020	101	19	that	that	WDT
kim-ai-2020	101	20	will	will	MD
kim-ai-2020	101	21	help	help	VB
kim-ai-2020	101	22	with	with	IN
kim-ai-2020	101	23	implementing	implement	VBG
kim-ai-2020	101	24	those	those	DT
kim-ai-2020	101	25	solutions	solution	NNS
kim-ai-2020	101	26	.	.	.
kim-ai-2020	102	1	Their	-PRON-	PRP$
kim-ai-2020	102	2	focus	focus	NN
kim-ai-2020	102	3	is	be	VBZ
kim-ai-2020	102	4	on	on	IN
kim-ai-2020	102	5	creating	create	VBG
kim-ai-2020	102	6	a	a	DT
kim-ai-2020	102	7	set	set	NN
kim-ai-2020	102	8	of	of	IN
kim-ai-2020	102	9	constraints	constraint	NNS
kim-ai-2020	102	10	and	and	CC
kim-ai-2020	102	11	if	if	IN
kim-ai-2020	102	12	-	-	HYPH
kim-ai-2020	102	13	then	then	RB
kim-ai-2020	102	14	statements	statement	NNS
kim-ai-2020	102	15	,	,	,
kim-ai-2020	102	16	which	which	WDT
kim-ai-2020	102	17	will	will	MD
kim-ai-2020	102	18	allow	allow	VB
kim-ai-2020	102	19	a	a	DT
kim-ai-2020	102	20	machine	machine	NN
kim-ai-2020	102	21	to	to	TO
kim-ai-2020	102	22	identify	identify	VB
kim-ai-2020	102	23	and	and	CC
kim-ai-2020	102	24	process	process	VB
kim-ai-2020	102	25	morally	morally	RB
kim-ai-2020	102	26	relevant	relevant	JJ
kim-ai-2020	102	27	considerations	consideration	NNS
kim-ai-2020	102	28	,	,	,
kim-ai-2020	102	29	so	so	IN
kim-ai-2020	102	30	that	that	IN
kim-ai-2020	102	31	it	-PRON-	PRP
kim-ai-2020	102	32	can	can	MD
kim-ai-2020	102	33	determine	determine	VB
kim-ai-2020	102	34	and	and	CC
kim-ai-2020	102	35	execute	execute	VB
kim-ai-2020	102	36	an	an	DT
kim-ai-2020	102	37	action	action	NN
kim-ai-2020	102	38	that	that	WDT
kim-ai-2020	102	39	is	be	VBZ
kim-ai-2020	102	40	not	not	RB
kim-ai-2020	102	41	only	only	RB
kim-ai-2020	102	42	rational	rational	JJ
kim-ai-2020	102	43	but	but	CC
kim-ai-2020	102	44	also	also	RB
kim-ai-2020	102	45	ethical	ethical	JJ
kim-ai-2020	102	46	in	in	IN
kim-ai-2020	102	47	the	the	DT
kim-ai-2020	102	48	given	give	VBN
kim-ai-2020	102	49	situation	situation	NN
kim-ai-2020	102	50	.	.	.
kim-ai-2020	103	1	[	[	-LRB-
kim-ai-2020	103	2	footnoteRef:10	footnoteref:10	CD
kim-ai-2020	103	3	]	]	-RRB-
kim-ai-2020	103	4	[	[	-LRB-
kim-ai-2020	103	5	10	10	CD
kim-ai-2020	103	6	:	:	:
kim-ai-2020	103	7	Note	note	VB
kim-ai-2020	103	8	that	that	IN
kim-ai-2020	103	9	this	this	DT
kim-ai-2020	103	10	moral	moral	JJ
kim-ai-2020	103	11	decision	decision	NN
kim-ai-2020	103	12	-	-	HYPH
kim-ai-2020	103	13	making	make	VBG
kim-ai-2020	103	14	process	process	NN
kim-ai-2020	103	15	can	can	MD
kim-ai-2020	103	16	be	be	VB
kim-ai-2020	103	17	modeled	model	VBN
kim-ai-2020	103	18	as	as	IN
kim-ai-2020	103	19	rule	rule	NN
kim-ai-2020	103	20	-	-	HYPH
kim-ai-2020	103	21	based	base	VBN
kim-ai-2020	103	22	,	,	,
kim-ai-2020	103	23	as	as	IN
kim-ai-2020	103	24	in	in	IN
kim-ai-2020	103	25	a	a	DT
kim-ai-2020	103	26	symbolic	symbolic	JJ
kim-ai-2020	103	27	AI	ai	NN
kim-ai-2020	103	28	system	system	NN
kim-ai-2020	103	29	,	,	,
kim-ai-2020	103	30	although	although	IN
kim-ai-2020	103	31	it	-PRON-	PRP
kim-ai-2020	103	32	does	do	VBZ
kim-ai-2020	103	33	not	not	RB
kim-ai-2020	103	34	have	have	VB
kim-ai-2020	103	35	to	to	TO
kim-ai-2020	103	36	be	be	VB
kim-ai-2020	103	37	.	.	.
kim-ai-2020	104	1	It	-PRON-	PRP
kim-ai-2020	104	2	can	can	MD
kim-ai-2020	104	3	be	be	VB
kim-ai-2020	104	4	based	base	VBN
kim-ai-2020	104	5	upon	upon	IN
kim-ai-2020	104	6	a	a	DT
kim-ai-2020	104	7	machine	machine	NN
kim-ai-2020	104	8	learning	learning	NN
kim-ai-2020	104	9	approach	approach	NN
kim-ai-2020	104	10	.	.	.
kim-ai-2020	105	1	It	-PRON-	PRP
kim-ai-2020	105	2	is	be	VBZ
kim-ai-2020	105	3	also	also	RB
kim-ai-2020	105	4	possible	possible	JJ
kim-ai-2020	105	5	to	to	TO
kim-ai-2020	105	6	combine	combine	VB
kim-ai-2020	105	7	both	both	DT
kim-ai-2020	105	8	.	.	.
kim-ai-2020	106	1	See	see	VB
kim-ai-2020	106	2	Vincent	Vincent	NNP
kim-ai-2020	106	3	Conitzer	Conitzer	NNP
kim-ai-2020	106	4	et	et	NNP
kim-ai-2020	106	5	al	al	NNP
kim-ai-2020	106	6	.	.	.
kim-ai-2020	107	1	2017	2017	CD
kim-ai-2020	107	2	.	.	.
kim-ai-2020	107	3	]	]	-RRB-
kim-ai-2020	108	1	On	on	IN
kim-ai-2020	108	2	the	the	DT
kim-ai-2020	108	3	other	other	JJ
kim-ai-2020	108	4	hand	hand	NN
kim-ai-2020	108	5	,	,	,
kim-ai-2020	108	6	military	military	JJ
kim-ai-2020	108	7	commanders	commander	NNS
kim-ai-2020	108	8	and	and	CC
kim-ai-2020	108	9	soldiers	soldier	NNS
kim-ai-2020	108	10	might	may	MD
kim-ai-2020	108	11	aim	aim	VB
kim-ai-2020	108	12	to	to	TO
kim-ai-2020	108	13	end	end	VB
kim-ai-2020	108	14	a	a	DT
kim-ai-2020	108	15	conflict	conflict	NN
kim-ai-2020	108	16	,	,	,
kim-ai-2020	108	17	bring	bring	VB
kim-ai-2020	108	18	peace	peace	NN
kim-ai-2020	108	19	,	,	,
kim-ai-2020	108	20	and	and	CC
kim-ai-2020	108	21	facilitate	facilitate	VB
kim-ai-2020	108	22	restoring	restore	VBG
kim-ai-2020	108	23	and	and	CC
kim-ai-2020	108	24	establishing	establish	VBG
kim-ai-2020	108	25	universally	universally	RB
kim-ai-2020	108	26	recognized	recognize	VBN
kim-ai-2020	108	27	human	human	JJ
kim-ai-2020	108	28	values	value	NNS
kim-ai-2020	108	29	such	such	JJ
kim-ai-2020	108	30	as	as	IN
kim-ai-2020	108	31	freedom	freedom	NN
kim-ai-2020	108	32	,	,	,
kim-ai-2020	108	33	equality	equality	NN
kim-ai-2020	108	34	,	,	,
kim-ai-2020	108	35	justice	justice	NN
kim-ai-2020	108	36	,	,	,
kim-ai-2020	108	37	and	and	CC
kim-ai-2020	108	38	self	self	NN
kim-ai-2020	108	39	-	-	HYPH
kim-ai-2020	108	40	determination	determination	NN
kim-ai-2020	108	41	.	.	.
kim-ai-2020	109	1	In	in	IN
kim-ai-2020	109	2	order	order	NN
kim-ai-2020	109	3	to	to	TO
kim-ai-2020	109	4	achieve	achieve	VB
kim-ai-2020	109	5	this	this	DT
kim-ai-2020	109	6	goal	goal	NN
kim-ai-2020	109	7	,	,	,
kim-ai-2020	109	8	they	-PRON-	PRP
kim-ai-2020	109	9	must	must	MD
kim-ai-2020	109	10	make	make	VB
kim-ai-2020	109	11	the	the	DT
kim-ai-2020	109	12	best	good	JJS
kim-ai-2020	109	13	strategic	strategic	JJ
kim-ai-2020	109	14	decisions	decision	NNS
kim-ai-2020	109	15	and	and	CC
kim-ai-2020	109	16	take	take	VB
kim-ai-2020	109	17	the	the	DT
kim-ai-2020	109	18	most	most	RBS
kim-ai-2020	109	19	appropriate	appropriate	JJ
kim-ai-2020	109	20	actions	action	NNS
kim-ai-2020	109	21	.	.	.
kim-ai-2020	110	1	In	in	IN
kim-ai-2020	110	2	deciding	decide	VBG
kim-ai-2020	110	3	on	on	IN
kim-ai-2020	110	4	those	those	DT
kim-ai-2020	110	5	actions	action	NNS
kim-ai-2020	110	6	,	,	,
kim-ai-2020	110	7	they	-PRON-	PRP
kim-ai-2020	110	8	are	be	VBP
kim-ai-2020	110	9	also	also	RB
kim-ai-2020	110	10	responsible	responsible	JJ
kim-ai-2020	110	11	for	for	IN
kim-ai-2020	110	12	abiding	abide	VBG
kim-ai-2020	110	13	by	by	IN
kim-ai-2020	110	14	the	the	DT
kim-ai-2020	110	15	principles	principle	NNS
kim-ai-2020	110	16	of	of	IN
kim-ai-2020	110	17	jus	jus	NN
kim-ai-2020	110	18	in	in	IN
kim-ai-2020	110	19	bello	bello	NN
kim-ai-2020	110	20	and	and	CC
kim-ai-2020	110	21	for	for	IN
kim-ai-2020	110	22	not	not	RB
kim-ai-2020	110	23	abdicating	abdicate	VBG
kim-ai-2020	110	24	their	-PRON-	PRP$
kim-ai-2020	110	25	moral	moral	JJ
kim-ai-2020	110	26	responsibility	responsibility	NN
kim-ai-2020	110	27	,	,	,
kim-ai-2020	110	28	protecting	protect	VBG
kim-ai-2020	110	29	civilians	civilian	NNS
kim-ai-2020	110	30	and	and	CC
kim-ai-2020	110	31	minimizing	minimize	VBG
kim-ai-2020	110	32	harm	harm	NN
kim-ai-2020	110	33	,	,	,
kim-ai-2020	110	34	violence	violence	NN
kim-ai-2020	110	35	,	,	,
kim-ai-2020	110	36	and	and	CC
kim-ai-2020	110	37	destruction	destruction	NN
kim-ai-2020	110	38	as	as	RB
kim-ai-2020	110	39	much	much	RB
kim-ai-2020	110	40	as	as	IN
kim-ai-2020	110	41	possible	possible	JJ
kim-ai-2020	110	42	.	.	.
kim-ai-2020	111	1	[	[	-LRB-
kim-ai-2020	111	2	footnoteRef:11	footnoteRef:11	NNP
kim-ai-2020	111	3	]	]	-RRB-
kim-ai-2020	111	4	The	the	DT
kim-ai-2020	111	5	goal	goal	NN
kim-ai-2020	111	6	of	of	IN
kim-ai-2020	111	7	military	military	JJ
kim-ai-2020	111	8	commanders	commander	NNS
kim-ai-2020	111	9	and	and	CC
kim-ai-2020	111	10	soldiers	soldier	NNS
kim-ai-2020	111	11	,	,	,
kim-ai-2020	111	12	therefore	therefore	RB
kim-ai-2020	111	13	,	,	,
kim-ai-2020	111	14	differs	differ	NNS
kim-ai-2020	111	15	from	from	IN
kim-ai-2020	111	16	those	those	DT
kim-ai-2020	111	17	of	of	IN
kim-ai-2020	111	18	moral	moral	JJ
kim-ai-2020	111	19	philosophers	philosopher	NNS
kim-ai-2020	111	20	or	or	CC
kim-ai-2020	111	21	of	of	IN
kim-ai-2020	111	22	the	the	DT
kim-ai-2020	111	23	engineers	engineer	NNS
kim-ai-2020	111	24	of	of	IN
kim-ai-2020	111	25	autonomous	autonomous	JJ
kim-ai-2020	111	26	weapons	weapon	NNS
kim-ai-2020	111	27	.	.	.
kim-ai-2020	112	1	They	-PRON-	PRP
kim-ai-2020	112	2	are	be	VBP
kim-ai-2020	112	3	obligated	obligate	VBN
kim-ai-2020	112	4	to	to	TO
kim-ai-2020	112	5	make	make	VB
kim-ai-2020	112	6	quick	quick	JJ
kim-ai-2020	112	7	decisions	decision	NNS
kim-ai-2020	112	8	in	in	IN
kim-ai-2020	112	9	a	a	DT
kim-ai-2020	112	10	life	life	NN
kim-ai-2020	112	11	-	-	HYPH
kim-ai-2020	112	12	or	or	CC
kim-ai-2020	112	13	-	-	HYPH
kim-ai-2020	112	14	death	death	NN
kim-ai-2020	112	15	situation	situation	NN
kim-ai-2020	112	16	while	while	IN
kim-ai-2020	112	17	working	work	VBG
kim-ai-2020	112	18	with	with	IN
kim-ai-2020	112	19	AI	AI	NNP
kim-ai-2020	112	20	-	-	HYPH
kim-ai-2020	112	21	powered	power	VBN
kim-ai-2020	112	22	military	military	JJ
kim-ai-2020	112	23	systems	system	NNS
kim-ai-2020	112	24	.	.	.
kim-ai-2020	113	1	[	[	-LRB-
kim-ai-2020	113	2	11	11	CD
kim-ai-2020	113	3	:	:	:
kim-ai-2020	113	4	For	for	IN
kim-ai-2020	113	5	the	the	DT
kim-ai-2020	113	6	principles	principle	NNS
kim-ai-2020	113	7	of	of	IN
kim-ai-2020	113	8	jus	jus	NN
kim-ai-2020	113	9	in	in	IN
kim-ai-2020	113	10	bello	bello	NN
kim-ai-2020	113	11	,	,	,
kim-ai-2020	113	12	see	see	VB
kim-ai-2020	113	13	International	International	NNP
kim-ai-2020	113	14	Committee	Committee	NNP
kim-ai-2020	113	15	of	of	IN
kim-ai-2020	113	16	the	the	DT
kim-ai-2020	113	17	Red	Red	NNP
kim-ai-2020	113	18	Cross	Cross	NNP
kim-ai-2020	113	19	2015	2015	CD
kim-ai-2020	113	20	.	.	.
kim-ai-2020	113	21	]	]	-RRB-
kim-ai-2020	114	1	These	these	DT
kim-ai-2020	114	2	different	different	JJ
kim-ai-2020	114	3	goals	goal	NNS
kim-ai-2020	114	4	and	and	CC
kim-ai-2020	114	5	interests	interest	NNS
kim-ai-2020	114	6	explain	explain	VBP
kim-ai-2020	114	7	why	why	WRB
kim-ai-2020	114	8	moral	moral	JJ
kim-ai-2020	114	9	philosophers	philosopher	NNS
kim-ai-2020	114	10	’	’	POS
kim-ai-2020	114	11	discussion	discussion	NN
kim-ai-2020	114	12	on	on	IN
kim-ai-2020	114	13	the	the	DT
kim-ai-2020	114	14	trolley	trolley	NN
kim-ai-2020	114	15	problem	problem	NN
kim-ai-2020	114	16	may	may	MD
kim-ai-2020	114	17	be	be	VB
kim-ai-2020	114	18	disappointing	disappointing	JJ
kim-ai-2020	114	19	to	to	IN
kim-ai-2020	114	20	AI	AI	NNP
kim-ai-2020	114	21	programmers	programmer	NNS
kim-ai-2020	114	22	or	or	CC
kim-ai-2020	114	23	military	military	JJ
kim-ai-2020	114	24	commanders	commander	NNS
kim-ai-2020	114	25	and	and	CC
kim-ai-2020	114	26	soldiers	soldier	NNS
kim-ai-2020	114	27	.	.	.
kim-ai-2020	115	1	Ethics	ethic	NNS
kim-ai-2020	115	2	does	do	VBZ
kim-ai-2020	115	3	not	not	RB
kim-ai-2020	115	4	provide	provide	VB
kim-ai-2020	115	5	an	an	DT
kim-ai-2020	115	6	easy	easy	JJ
kim-ai-2020	115	7	answer	answer	NN
kim-ai-2020	115	8	to	to	IN
kim-ai-2020	115	9	the	the	DT
kim-ai-2020	115	10	question	question	NN
kim-ai-2020	115	11	of	of	IN
kim-ai-2020	115	12	how	how	WRB
kim-ai-2020	115	13	one	one	PRP
kim-ai-2020	115	14	should	should	MD
kim-ai-2020	115	15	program	program	VB
kim-ai-2020	115	16	moral	moral	JJ
kim-ai-2020	115	17	decision	decision	NN
kim-ai-2020	115	18	-	-	HYPH
kim-ai-2020	115	19	making	making	NN
kim-ai-2020	115	20	into	into	IN
kim-ai-2020	115	21	intelligent	intelligent	JJ
kim-ai-2020	115	22	machines	machine	NNS
kim-ai-2020	115	23	.	.	.
kim-ai-2020	116	1	Nor	nor	CC
kim-ai-2020	116	2	does	do	VBZ
kim-ai-2020	116	3	it	-PRON-	PRP
kim-ai-2020	116	4	prescribe	prescribe	VB
kim-ai-2020	116	5	the	the	DT
kim-ai-2020	116	6	right	right	JJ
kim-ai-2020	116	7	moral	moral	JJ
kim-ai-2020	116	8	decision	decision	NN
kim-ai-2020	116	9	in	in	IN
kim-ai-2020	116	10	a	a	DT
kim-ai-2020	116	11	battlefield	battlefield	NN
kim-ai-2020	116	12	.	.	.
kim-ai-2020	117	1	But	but	CC
kim-ai-2020	117	2	taking	take	VBG
kim-ai-2020	117	3	this	this	DT
kim-ai-2020	117	4	as	as	IN
kim-ai-2020	117	5	a	a	DT
kim-ai-2020	117	6	shortcoming	shortcoming	NN
kim-ai-2020	117	7	of	of	IN
kim-ai-2020	117	8	ethics	ethic	NNS
kim-ai-2020	117	9	is	be	VBZ
kim-ai-2020	117	10	missing	miss	VBG
kim-ai-2020	117	11	the	the	DT
kim-ai-2020	117	12	point	point	NN
kim-ai-2020	117	13	.	.	.
kim-ai-2020	118	1	The	the	DT
kim-ai-2020	118	2	role	role	NN
kim-ai-2020	118	3	of	of	IN
kim-ai-2020	118	4	moral	moral	JJ
kim-ai-2020	118	5	philosophy	philosophy	NN
kim-ai-2020	118	6	is	be	VBZ
kim-ai-2020	118	7	not	not	RB
kim-ai-2020	118	8	to	to	TO
kim-ai-2020	118	9	make	make	VB
kim-ai-2020	118	10	decision	decision	NN
kim-ai-2020	118	11	-	-	HYPH
kim-ai-2020	118	12	making	make	VBG
kim-ai-2020	118	13	easier	easy	JJR
kim-ai-2020	118	14	but	but	CC
kim-ai-2020	118	15	to	to	TO
kim-ai-2020	118	16	highlight	highlight	VB
kim-ai-2020	118	17	and	and	CC
kim-ai-2020	118	18	articulate	articulate	VB
kim-ai-2020	118	19	the	the	DT
kim-ai-2020	118	20	difficulty	difficulty	NN
kim-ai-2020	118	21	and	and	CC
kim-ai-2020	118	22	complexity	complexity	NN
kim-ai-2020	118	23	involved	involve	VBN
kim-ai-2020	118	24	in	in	IN
kim-ai-2020	118	25	it	-PRON-	PRP
kim-ai-2020	118	26	.	.	.
kim-ai-2020	119	1	4	4	LS
kim-ai-2020	119	2	.	.	.
kim-ai-2020	120	1	Ethical	ethical	JJ
kim-ai-2020	120	2	Challenges	Challenges	NNPS
kim-ai-2020	120	3	from	from	IN
kim-ai-2020	120	4	Autonomous	Autonomous	NNP
kim-ai-2020	120	5	AI	AI	NNP
kim-ai-2020	120	6	Systems	Systems	NNP
kim-ai-2020	120	7	The	the	DT
kim-ai-2020	120	8	complexity	complexity	NN
kim-ai-2020	120	9	of	of	IN
kim-ai-2020	120	10	ethical	ethical	JJ
kim-ai-2020	120	11	questions	question	NNS
kim-ai-2020	120	12	means	mean	VBZ
kim-ai-2020	120	13	that	that	IN
kim-ai-2020	120	14	dealing	deal	VBG
kim-ai-2020	120	15	with	with	IN
kim-ai-2020	120	16	the	the	DT
kim-ai-2020	120	17	morality	morality	NN
kim-ai-2020	120	18	of	of	IN
kim-ai-2020	120	19	an	an	DT
kim-ai-2020	120	20	action	action	NN
kim-ai-2020	120	21	by	by	IN
kim-ai-2020	120	22	an	an	DT
kim-ai-2020	120	23	autonomous	autonomous	JJ
kim-ai-2020	120	24	AI	ai	NN
kim-ai-2020	120	25	system	system	NN
kim-ai-2020	120	26	will	will	MD
kim-ai-2020	120	27	require	require	VB
kim-ai-2020	120	28	more	more	JJR
kim-ai-2020	120	29	than	than	IN
kim-ai-2020	120	30	a	a	DT
kim-ai-2020	120	31	clever	clever	JJ
kim-ai-2020	120	32	engineering	engineering	NN
kim-ai-2020	120	33	or	or	CC
kim-ai-2020	120	34	programming	programming	NN
kim-ai-2020	120	35	solution	solution	NN
kim-ai-2020	120	36	.	.	.
kim-ai-2020	121	1	The	the	DT
kim-ai-2020	121	2	fact	fact	NN
kim-ai-2020	121	3	that	that	IN
kim-ai-2020	121	4	ethics	ethic	NNS
kim-ai-2020	121	5	does	do	VBZ
kim-ai-2020	121	6	not	not	RB
kim-ai-2020	121	7	eliminate	eliminate	VB
kim-ai-2020	121	8	the	the	DT
kim-ai-2020	121	9	inherent	inherent	JJ
kim-ai-2020	121	10	ambiguity	ambiguity	NN
kim-ai-2020	121	11	in	in	IN
kim-ai-2020	121	12	many	many	JJ
kim-ai-2020	121	13	moral	moral	JJ
kim-ai-2020	121	14	decisions	decision	NNS
kim-ai-2020	121	15	should	should	MD
kim-ai-2020	121	16	not	not	RB
kim-ai-2020	121	17	lead	lead	VB
kim-ai-2020	121	18	to	to	IN
kim-ai-2020	121	19	the	the	DT
kim-ai-2020	121	20	dismissal	dismissal	NN
kim-ai-2020	121	21	of	of	IN
kim-ai-2020	121	22	ethical	ethical	JJ
kim-ai-2020	121	23	challenges	challenge	NNS
kim-ai-2020	121	24	from	from	IN
kim-ai-2020	121	25	autonomous	autonomous	JJ
kim-ai-2020	121	26	AI	AI	NNP
kim-ai-2020	121	27	systems	system	NNS
kim-ai-2020	121	28	.	.	.
kim-ai-2020	122	1	By	by	IN
kim-ai-2020	122	2	injecting	inject	VBG
kim-ai-2020	122	3	the	the	DT
kim-ai-2020	122	4	capacity	capacity	NN
kim-ai-2020	122	5	for	for	IN
kim-ai-2020	122	6	autonomous	autonomous	JJ
kim-ai-2020	122	7	decision	decision	NN
kim-ai-2020	122	8	-	-	HYPH
kim-ai-2020	122	9	making	making	NN
kim-ai-2020	122	10	into	into	IN
kim-ai-2020	122	11	major	major	JJ
kim-ai-2020	122	12	systems	system	NNS
kim-ai-2020	122	13	and	and	CC
kim-ai-2020	122	14	tools	tool	NNS
kim-ai-2020	122	15	,	,	,
kim-ai-2020	122	16	AI	AI	NNP
kim-ai-2020	122	17	can	can	MD
kim-ai-2020	122	18	fundamentally	fundamentally	RB
kim-ai-2020	122	19	transform	transform	VB
kim-ai-2020	122	20	any	any	DT
kim-ai-2020	122	21	given	give	VBN
kim-ai-2020	122	22	field	field	NN
kim-ai-2020	122	23	.	.	.
kim-ai-2020	123	1	For	for	IN
kim-ai-2020	123	2	example	example	NN
kim-ai-2020	123	3	,	,	,
kim-ai-2020	123	4	AI	AI	NNP
kim-ai-2020	123	5	-	-	HYPH
kim-ai-2020	123	6	powered	power	VBN
kim-ai-2020	123	7	military	military	JJ
kim-ai-2020	123	8	robots	robot	NNS
kim-ai-2020	123	9	are	be	VBP
kim-ai-2020	123	10	not	not	RB
kim-ai-2020	123	11	just	just	RB
kim-ai-2020	123	12	another	another	DT
kim-ai-2020	123	13	kind	kind	NN
kim-ai-2020	123	14	of	of	IN
kim-ai-2020	123	15	weapon	weapon	NN
kim-ai-2020	123	16	.	.	.
kim-ai-2020	124	1	When	when	WRB
kim-ai-2020	124	2	widely	widely	RB
kim-ai-2020	124	3	deployed	deploy	VBN
kim-ai-2020	124	4	,	,	,
kim-ai-2020	124	5	they	-PRON-	PRP
kim-ai-2020	124	6	can	can	MD
kim-ai-2020	124	7	change	change	VB
kim-ai-2020	124	8	the	the	DT
kim-ai-2020	124	9	nature	nature	NN
kim-ai-2020	124	10	of	of	IN
kim-ai-2020	124	11	war	war	NN
kim-ai-2020	124	12	itself	-PRON-	PRP
kim-ai-2020	124	13	.	.	.
kim-ai-2020	125	1	Described	describe	VBN
kim-ai-2020	125	2	below	below	IN
kim-ai-2020	125	3	are	be	VBP
kim-ai-2020	125	4	some	some	DT
kim-ai-2020	125	5	of	of	IN
kim-ai-2020	125	6	the	the	DT
kim-ai-2020	125	7	significant	significant	JJ
kim-ai-2020	125	8	ethical	ethical	JJ
kim-ai-2020	125	9	challenges	challenge	NNS
kim-ai-2020	125	10	that	that	WDT
kim-ai-2020	125	11	autonomous	autonomous	JJ
kim-ai-2020	125	12	AI	AI	NNP
kim-ai-2020	125	13	systems	system	NNS
kim-ai-2020	125	14	such	such	JJ
kim-ai-2020	125	15	as	as	IN
kim-ai-2020	125	16	military	military	JJ
kim-ai-2020	125	17	robots	robot	NNS
kim-ai-2020	125	18	present	present	JJ
kim-ai-2020	125	19	.	.	.
kim-ai-2020	126	1	(	(	-LRB-
kim-ai-2020	126	2	a	a	LS
kim-ai-2020	126	3	)	)	-RRB-
kim-ai-2020	126	4	Moral	moral	JJ
kim-ai-2020	126	5	desensitization	desensitization	NN
kim-ai-2020	126	6	AI	ai	NN
kim-ai-2020	126	7	-	-	HYPH
kim-ai-2020	126	8	powered	power	VBN
kim-ai-2020	126	9	military	military	JJ
kim-ai-2020	126	10	robots	robot	NNS
kim-ai-2020	126	11	go	go	VBP
kim-ai-2020	126	12	further	further	RB
kim-ai-2020	126	13	than	than	IN
kim-ai-2020	126	14	any	any	DT
kim-ai-2020	126	15	remotely	remotely	RB
kim-ai-2020	126	16	-	-	HYPH
kim-ai-2020	126	17	operated	operate	VBN
kim-ai-2020	126	18	weapons	weapon	NNS
kim-ai-2020	126	19	.	.	.
kim-ai-2020	127	1	They	-PRON-	PRP
kim-ai-2020	127	2	can	can	MD
kim-ai-2020	127	3	identify	identify	VB
kim-ai-2020	127	4	a	a	DT
kim-ai-2020	127	5	target	target	NN
kim-ai-2020	127	6	and	and	CC
kim-ai-2020	127	7	initiate	initiate	VB
kim-ai-2020	127	8	an	an	DT
kim-ai-2020	127	9	attack	attack	NN
kim-ai-2020	127	10	on	on	IN
kim-ai-2020	127	11	their	-PRON-	PRP$
kim-ai-2020	127	12	own	own	JJ
kim-ai-2020	127	13	.	.	.
kim-ai-2020	128	1	Due	due	IN
kim-ai-2020	128	2	to	to	IN
kim-ai-2020	128	3	their	-PRON-	PRP$
kim-ai-2020	128	4	autonomy	autonomy	NN
kim-ai-2020	128	5	,	,	,
kim-ai-2020	128	6	military	military	JJ
kim-ai-2020	128	7	robots	robot	NNS
kim-ai-2020	128	8	can	can	MD
kim-ai-2020	128	9	significantly	significantly	RB
kim-ai-2020	128	10	increase	increase	VB
kim-ai-2020	128	11	the	the	DT
kim-ai-2020	128	12	distance	distance	NN
kim-ai-2020	128	13	between	between	IN
kim-ai-2020	128	14	the	the	DT
kim-ai-2020	128	15	party	party	NN
kim-ai-2020	128	16	that	that	WDT
kim-ai-2020	128	17	kills	kill	VBZ
kim-ai-2020	128	18	and	and	CC
kim-ai-2020	128	19	the	the	DT
kim-ai-2020	128	20	party	party	NN
kim-ai-2020	128	21	that	that	WDT
kim-ai-2020	128	22	gets	get	VBZ
kim-ai-2020	128	23	killed	kill	VBN
kim-ai-2020	128	24	(	(	-LRB-
kim-ai-2020	128	25	Sharkey	Sharkey	NNP
kim-ai-2020	128	26	2012	2012	CD
kim-ai-2020	128	27	)	)	-RRB-
kim-ai-2020	128	28	.	.	.
kim-ai-2020	129	1	This	this	DT
kim-ai-2020	129	2	increase	increase	NN
kim-ai-2020	129	3	,	,	,
kim-ai-2020	129	4	however	however	RB
kim-ai-2020	129	5	,	,	,
kim-ai-2020	129	6	may	may	MD
kim-ai-2020	129	7	lead	lead	VB
kim-ai-2020	129	8	people	people	NNS
kim-ai-2020	129	9	to	to	TO
kim-ai-2020	129	10	surrender	surrender	VB
kim-ai-2020	129	11	their	-PRON-	PRP$
kim-ai-2020	129	12	own	own	JJ
kim-ai-2020	129	13	moral	moral	JJ
kim-ai-2020	129	14	responsibility	responsibility	NN
kim-ai-2020	129	15	to	to	IN
kim-ai-2020	129	16	a	a	DT
kim-ai-2020	129	17	machine	machine	NN
kim-ai-2020	129	18	,	,	,
kim-ai-2020	129	19	thereby	thereby	RB
kim-ai-2020	129	20	resulting	result	VBG
kim-ai-2020	129	21	in	in	IN
kim-ai-2020	129	22	the	the	DT
kim-ai-2020	129	23	loss	loss	NN
kim-ai-2020	129	24	of	of	IN
kim-ai-2020	129	25	humanity	humanity	NN
kim-ai-2020	129	26	,	,	,
kim-ai-2020	129	27	which	which	WDT
kim-ai-2020	129	28	is	be	VBZ
kim-ai-2020	129	29	a	a	DT
kim-ai-2020	129	30	serious	serious	JJ
kim-ai-2020	129	31	moral	moral	JJ
kim-ai-2020	129	32	risk	risk	NN
kim-ai-2020	129	33	(	(	-LRB-
kim-ai-2020	129	34	Davis	Davis	NNP
kim-ai-2020	129	35	2007	2007	CD
kim-ai-2020	129	36	)	)	-RRB-
kim-ai-2020	129	37	.	.	.
kim-ai-2020	130	1	The	the	DT
kim-ai-2020	130	2	more	more	RBR
kim-ai-2020	130	3	autonomous	autonomous	JJ
kim-ai-2020	130	4	military	military	JJ
kim-ai-2020	130	5	robots	robot	NNS
kim-ai-2020	130	6	become	become	VBP
kim-ai-2020	130	7	,	,	,
kim-ai-2020	130	8	the	the	DT
kim-ai-2020	130	9	less	less	JJR
kim-ai-2020	130	10	responsibility	responsibility	NN
kim-ai-2020	130	11	for	for	IN
kim-ai-2020	130	12	life	life	NN
kim-ai-2020	130	13	-	-	HYPH
kim-ai-2020	130	14	or	or	CC
kim-ai-2020	130	15	-	-	HYPH
kim-ai-2020	130	16	death	death	NN
kim-ai-2020	130	17	decisions	decision	NNS
kim-ai-2020	130	18	may	may	MD
kim-ai-2020	130	19	come	come	VB
kim-ai-2020	130	20	to	to	TO
kim-ai-2020	130	21	rest	rest	VB
kim-ai-2020	130	22	with	with	IN
kim-ai-2020	130	23	humans	human	NNS
kim-ai-2020	130	24	.	.	.
kim-ai-2020	131	1	(	(	-LRB-
kim-ai-2020	131	2	b	b	NN
kim-ai-2020	131	3	)	)	-RRB-
kim-ai-2020	131	4	Unintended	Unintended	NNP
kim-ai-2020	131	5	outcomes	outcome	NNS
kim-ai-2020	131	6	As	as	IN
kim-ai-2020	131	7	military	military	JJ
kim-ai-2020	131	8	robots	robot	NNS
kim-ai-2020	131	9	make	make	VBP
kim-ai-2020	131	10	killing	kill	VBG
kim-ai-2020	131	11	easier	easy	JJR
kim-ai-2020	131	12	,	,	,
kim-ai-2020	131	13	small	small	JJ
kim-ai-2020	131	14	conflicts	conflict	NNS
kim-ai-2020	131	15	may	may	MD
kim-ai-2020	131	16	more	more	RBR
kim-ai-2020	131	17	quickly	quickly	RB
kim-ai-2020	131	18	escalate	escalate	VB
kim-ai-2020	131	19	to	to	IN
kim-ai-2020	131	20	a	a	DT
kim-ai-2020	131	21	war	war	NN
kim-ai-2020	131	22	.	.	.
kim-ai-2020	132	1	The	the	DT
kim-ai-2020	132	2	side	side	NN
kim-ai-2020	132	3	that	that	WDT
kim-ai-2020	132	4	deploys	deploy	VBZ
kim-ai-2020	132	5	AI	ai	NN
kim-ai-2020	132	6	-	-	HYPH
kim-ai-2020	132	7	powered	power	VBN
kim-ai-2020	132	8	military	military	JJ
kim-ai-2020	132	9	robots	robot	NNS
kim-ai-2020	132	10	is	be	VBZ
kim-ai-2020	132	11	likely	likely	JJ
kim-ai-2020	132	12	to	to	TO
kim-ai-2020	132	13	suffer	suffer	VB
kim-ai-2020	132	14	fewer	few	JJR
kim-ai-2020	132	15	casualties	casualty	NNS
kim-ai-2020	132	16	itself	-PRON-	PRP
kim-ai-2020	132	17	while	while	IN
kim-ai-2020	132	18	inflicting	inflict	VBG
kim-ai-2020	132	19	more	more	JJR
kim-ai-2020	132	20	casualties	casualty	NNS
kim-ai-2020	132	21	on	on	IN
kim-ai-2020	132	22	the	the	DT
kim-ai-2020	132	23	enemy	enemy	NN
kim-ai-2020	132	24	side	side	NN
kim-ai-2020	132	25	.	.	.
kim-ai-2020	133	1	This	this	DT
kim-ai-2020	133	2	may	may	MD
kim-ai-2020	133	3	make	make	VB
kim-ai-2020	133	4	the	the	DT
kim-ai-2020	133	5	military	military	NN
kim-ai-2020	133	6	more	more	RBR
kim-ai-2020	133	7	inclined	inclined	JJ
kim-ai-2020	133	8	to	to	TO
kim-ai-2020	133	9	start	start	VB
kim-ai-2020	133	10	a	a	DT
kim-ai-2020	133	11	war	war	NN
kim-ai-2020	133	12	.	.	.
kim-ai-2020	134	1	Ironically	ironically	RB
kim-ai-2020	134	2	,	,	,
kim-ai-2020	134	3	when	when	WRB
kim-ai-2020	134	4	everyone	everyone	NN
kim-ai-2020	134	5	thinks	think	VBZ
kim-ai-2020	134	6	and	and	CC
kim-ai-2020	134	7	acts	act	VBZ
kim-ai-2020	134	8	this	this	DT
kim-ai-2020	134	9	way	way	NN
kim-ai-2020	134	10	,	,	,
kim-ai-2020	134	11	the	the	DT
kim-ai-2020	134	12	number	number	NN
kim-ai-2020	134	13	of	of	IN
kim-ai-2020	134	14	wars	war	NNS
kim-ai-2020	134	15	and	and	CC
kim-ai-2020	134	16	the	the	DT
kim-ai-2020	134	17	overall	overall	JJ
kim-ai-2020	134	18	amount	amount	NN
kim-ai-2020	134	19	of	of	IN
kim-ai-2020	134	20	violence	violence	NN
kim-ai-2020	134	21	and	and	CC
kim-ai-2020	134	22	destruction	destruction	NN
kim-ai-2020	134	23	in	in	IN
kim-ai-2020	134	24	the	the	DT
kim-ai-2020	134	25	world	world	NN
kim-ai-2020	134	26	will	will	MD
kim-ai-2020	134	27	only	only	RB
kim-ai-2020	134	28	increase	increase	VB
kim-ai-2020	134	29	.	.	.
kim-ai-2020	135	1	[	[	-LRB-
kim-ai-2020	135	2	footnoteRef:12	footnoteref:12	NN
kim-ai-2020	135	3	]	]	-RRB-
kim-ai-2020	135	4	[	[	-LRB-
kim-ai-2020	135	5	12	12	CD
kim-ai-2020	135	6	:	:	:
kim-ai-2020	135	7	Kahn	Kahn	NNP
kim-ai-2020	135	8	(	(	-LRB-
kim-ai-2020	135	9	2012	2012	CD
kim-ai-2020	135	10	)	)	-RRB-
kim-ai-2020	135	11	also	also	RB
kim-ai-2020	135	12	argues	argue	VBZ
kim-ai-2020	135	13	that	that	IN
kim-ai-2020	135	14	the	the	DT
kim-ai-2020	135	15	resulting	result	VBG
kim-ai-2020	135	16	increase	increase	NN
kim-ai-2020	135	17	in	in	IN
kim-ai-2020	135	18	the	the	DT
kim-ai-2020	135	19	number	number	NN
kim-ai-2020	135	20	of	of	IN
kim-ai-2020	135	21	wars	war	NNS
kim-ai-2020	135	22	by	by	IN
kim-ai-2020	135	23	the	the	DT
kim-ai-2020	135	24	use	use	NN
kim-ai-2020	135	25	of	of	IN
kim-ai-2020	135	26	military	military	JJ
kim-ai-2020	135	27	robots	robot	NNS
kim-ai-2020	135	28	will	will	MD
kim-ai-2020	135	29	be	be	VB
kim-ai-2020	135	30	morally	morally	RB
kim-ai-2020	135	31	bad	bad	JJ
kim-ai-2020	135	32	.	.	.
kim-ai-2020	135	33	]	]	-RRB-
kim-ai-2020	136	1	(	(	-LRB-
kim-ai-2020	136	2	c	c	NN
kim-ai-2020	136	3	)	)	-RRB-
kim-ai-2020	136	4	Surrender	surrender	NN
kim-ai-2020	136	5	of	of	IN
kim-ai-2020	136	6	moral	moral	JJ
kim-ai-2020	136	7	agency	agency	NN
kim-ai-2020	136	8	Military	military	JJ
kim-ai-2020	136	9	robots	robot	NNS
kim-ai-2020	136	10	may	may	MD
kim-ai-2020	136	11	fail	fail	VB
kim-ai-2020	136	12	to	to	TO
kim-ai-2020	136	13	distinguish	distinguish	VB
kim-ai-2020	136	14	innocents	innocent	NNS
kim-ai-2020	136	15	from	from	IN
kim-ai-2020	136	16	combatants	combatant	NNS
kim-ai-2020	136	17	and	and	CC
kim-ai-2020	136	18	kill	kill	VB
kim-ai-2020	136	19	the	the	DT
kim-ai-2020	136	20	former	former	JJ
kim-ai-2020	136	21	.	.	.
kim-ai-2020	137	1	In	in	IN
kim-ai-2020	137	2	such	such	PDT
kim-ai-2020	137	3	a	a	DT
kim-ai-2020	137	4	case	case	NN
kim-ai-2020	137	5	,	,	,
kim-ai-2020	137	6	can	can	MD
kim-ai-2020	137	7	we	-PRON-	PRP
kim-ai-2020	137	8	be	be	VB
kim-ai-2020	137	9	justified	justify	VBN
kim-ai-2020	137	10	in	in	IN
kim-ai-2020	137	11	letting	let	VBG
kim-ai-2020	137	12	robots	robot	NNS
kim-ai-2020	137	13	take	take	VB
kim-ai-2020	137	14	the	the	DT
kim-ai-2020	137	15	lives	life	NNS
kim-ai-2020	137	16	of	of	IN
kim-ai-2020	137	17	other	other	JJ
kim-ai-2020	137	18	human	human	JJ
kim-ai-2020	137	19	beings	being	NNS
kim-ai-2020	137	20	?	?	.
kim-ai-2020	138	1	Some	some	DT
kim-ai-2020	138	2	may	may	MD
kim-ai-2020	138	3	argue	argue	VB
kim-ai-2020	138	4	that	that	IN
kim-ai-2020	138	5	only	only	RB
kim-ai-2020	138	6	humans	human	NNS
kim-ai-2020	138	7	should	should	MD
kim-ai-2020	138	8	decide	decide	VB
kim-ai-2020	138	9	to	to	TO
kim-ai-2020	138	10	kill	kill	VB
kim-ai-2020	138	11	other	other	JJ
kim-ai-2020	138	12	humans	human	NNS
kim-ai-2020	138	13	,	,	,
kim-ai-2020	138	14	not	not	RB
kim-ai-2020	138	15	machines	machine	NNS
kim-ai-2020	138	16	(	(	-LRB-
kim-ai-2020	138	17	Davis	Davis	NNP
kim-ai-2020	138	18	2007	2007	CD
kim-ai-2020	138	19	)	)	-RRB-
kim-ai-2020	138	20	.	.	.
kim-ai-2020	139	1	Is	be	VBZ
kim-ai-2020	139	2	it	-PRON-	PRP
kim-ai-2020	139	3	permissible	permissible	JJ
kim-ai-2020	139	4	for	for	IN
kim-ai-2020	139	5	people	people	NNS
kim-ai-2020	139	6	to	to	TO
kim-ai-2020	139	7	delegate	delegate	VB
kim-ai-2020	139	8	such	such	PDT
kim-ai-2020	139	9	a	a	DT
kim-ai-2020	139	10	decision	decision	NN
kim-ai-2020	139	11	to	to	IN
kim-ai-2020	139	12	robots	robot	NNS
kim-ai-2020	139	13	?	?	.
kim-ai-2020	140	1	(	(	-LRB-
kim-ai-2020	140	2	d	d	LS
kim-ai-2020	140	3	)	)	-RRB-
kim-ai-2020	140	4	Opacity	opacity	NN
kim-ai-2020	140	5	in	in	IN
kim-ai-2020	140	6	decision	decision	NN
kim-ai-2020	140	7	-	-	HYPH
kim-ai-2020	140	8	making	make	VBG
kim-ai-2020	140	9	Machine	Machine	NNP
kim-ai-2020	140	10	learning	learning	NN
kim-ai-2020	140	11	is	be	VBZ
kim-ai-2020	140	12	used	use	VBN
kim-ai-2020	140	13	to	to	TO
kim-ai-2020	140	14	build	build	VB
kim-ai-2020	140	15	many	many	JJ
kim-ai-2020	140	16	AI	AI	NNP
kim-ai-2020	140	17	systems	system	NNS
kim-ai-2020	140	18	today	today	NN
kim-ai-2020	140	19	.	.	.
kim-ai-2020	141	1	Instead	instead	RB
kim-ai-2020	141	2	of	of	IN
kim-ai-2020	141	3	prescribing	prescribe	VBG
kim-ai-2020	141	4	a	a	DT
kim-ai-2020	141	5	pre	pre	JJ
kim-ai-2020	141	6	-	-	JJ
kim-ai-2020	141	7	determined	determined	JJ
kim-ai-2020	141	8	algorithm	algorithm	NN
kim-ai-2020	141	9	,	,	,
kim-ai-2020	141	10	a	a	DT
kim-ai-2020	141	11	machine	machine	NN
kim-ai-2020	141	12	learning	learning	NN
kim-ai-2020	141	13	system	system	NN
kim-ai-2020	141	14	goes	go	VBZ
kim-ai-2020	141	15	through	through	IN
kim-ai-2020	141	16	a	a	DT
kim-ai-2020	141	17	so	so	RB
kim-ai-2020	141	18	-	-	HYPH
kim-ai-2020	141	19	called	call	VBN
kim-ai-2020	141	20	‘	'	``
kim-ai-2020	141	21	training	training	NN
kim-ai-2020	141	22	’	'	''
kim-ai-2020	141	23	process	process	NN
kim-ai-2020	141	24	in	in	IN
kim-ai-2020	141	25	order	order	NN
kim-ai-2020	141	26	to	to	TO
kim-ai-2020	141	27	produce	produce	VB
kim-ai-2020	141	28	the	the	DT
kim-ai-2020	141	29	final	final	JJ
kim-ai-2020	141	30	algorithm	algorithm	NN
kim-ai-2020	141	31	from	from	IN
kim-ai-2020	141	32	a	a	DT
kim-ai-2020	141	33	large	large	JJ
kim-ai-2020	141	34	amount	amount	NN
kim-ai-2020	141	35	of	of	IN
kim-ai-2020	141	36	data	datum	NNS
kim-ai-2020	141	37	.	.	.
kim-ai-2020	142	1	For	for	IN
kim-ai-2020	142	2	example	example	NN
kim-ai-2020	142	3	,	,	,
kim-ai-2020	142	4	a	a	DT
kim-ai-2020	142	5	machine	machine	NN
kim-ai-2020	142	6	learning	learning	NN
kim-ai-2020	142	7	system	system	NN
kim-ai-2020	142	8	may	may	MD
kim-ai-2020	142	9	generate	generate	VB
kim-ai-2020	142	10	an	an	DT
kim-ai-2020	142	11	algorithm	algorithm	NN
kim-ai-2020	142	12	that	that	WDT
kim-ai-2020	142	13	successfully	successfully	RB
kim-ai-2020	142	14	recognizes	recognize	VBZ
kim-ai-2020	142	15	cats	cat	NNS
kim-ai-2020	142	16	in	in	IN
kim-ai-2020	142	17	a	a	DT
kim-ai-2020	142	18	photo	photo	NN
kim-ai-2020	142	19	after	after	IN
kim-ai-2020	142	20	going	go	VBG
kim-ai-2020	142	21	through	through	IN
kim-ai-2020	142	22	millions	million	NNS
kim-ai-2020	142	23	of	of	IN
kim-ai-2020	142	24	photos	photo	NNS
kim-ai-2020	142	25	that	that	WDT
kim-ai-2020	142	26	show	show	VBP
kim-ai-2020	142	27	cats	cat	NNS
kim-ai-2020	142	28	in	in	IN
kim-ai-2020	142	29	many	many	JJ
kim-ai-2020	142	30	different	different	JJ
kim-ai-2020	142	31	postures	posture	NNS
kim-ai-2020	142	32	from	from	IN
kim-ai-2020	142	33	various	various	JJ
kim-ai-2020	142	34	angles	angle	NNS
kim-ai-2020	142	35	.	.	.
kim-ai-2020	143	1	[	[	-LRB-
kim-ai-2020	143	2	footnoteRef:13	footnoteref:13	NN
kim-ai-2020	143	3	]	]	-RRB-
kim-ai-2020	143	4	But	but	CC
kim-ai-2020	143	5	the	the	DT
kim-ai-2020	143	6	resulting	result	VBG
kim-ai-2020	143	7	algorithm	algorithm	NN
kim-ai-2020	143	8	is	be	VBZ
kim-ai-2020	143	9	a	a	DT
kim-ai-2020	143	10	complex	complex	JJ
kim-ai-2020	143	11	mathematical	mathematical	JJ
kim-ai-2020	143	12	formula	formula	NN
kim-ai-2020	143	13	and	and	CC
kim-ai-2020	143	14	not	not	RB
kim-ai-2020	143	15	something	something	NN
kim-ai-2020	143	16	that	that	WDT
kim-ai-2020	143	17	humans	human	NNS
kim-ai-2020	143	18	can	can	MD
kim-ai-2020	143	19	easily	easily	RB
kim-ai-2020	143	20	decipher	decipher	VB
kim-ai-2020	143	21	.	.	.
kim-ai-2020	144	1	This	this	DT
kim-ai-2020	144	2	means	mean	VBZ
kim-ai-2020	144	3	that	that	IN
kim-ai-2020	144	4	the	the	DT
kim-ai-2020	144	5	inner	inner	JJ
kim-ai-2020	144	6	workings	working	NNS
kim-ai-2020	144	7	of	of	IN
kim-ai-2020	144	8	a	a	DT
kim-ai-2020	144	9	machine	machine	NN
kim-ai-2020	144	10	learning	learn	VBG
kim-ai-2020	144	11	AI	ai	NN
kim-ai-2020	144	12	system	system	NN
kim-ai-2020	144	13	and	and	CC
kim-ai-2020	144	14	its	-PRON-	PRP$
kim-ai-2020	144	15	decision	decision	NN
kim-ai-2020	144	16	-	-	HYPH
kim-ai-2020	144	17	making	make	VBG
kim-ai-2020	144	18	process	process	NN
kim-ai-2020	144	19	is	be	VBZ
kim-ai-2020	144	20	opaque	opaque	JJ
kim-ai-2020	144	21	to	to	IN
kim-ai-2020	144	22	human	human	JJ
kim-ai-2020	144	23	understanding	understanding	NN
kim-ai-2020	144	24	,	,	,
kim-ai-2020	144	25	even	even	RB
kim-ai-2020	144	26	to	to	IN
kim-ai-2020	144	27	those	those	DT
kim-ai-2020	144	28	who	who	WP
kim-ai-2020	144	29	built	build	VBD
kim-ai-2020	144	30	the	the	DT
kim-ai-2020	144	31	system	system	NN
kim-ai-2020	144	32	itself	-PRON-	PRP
kim-ai-2020	144	33	(	(	-LRB-
kim-ai-2020	144	34	Knight	Knight	NNP
kim-ai-2020	144	35	2017	2017	CD
kim-ai-2020	144	36	)	)	-RRB-
kim-ai-2020	144	37	.	.	.
kim-ai-2020	145	1	In	in	IN
kim-ai-2020	145	2	cases	case	NNS
kim-ai-2020	145	3	where	where	WRB
kim-ai-2020	145	4	the	the	DT
kim-ai-2020	145	5	actions	action	NNS
kim-ai-2020	145	6	of	of	IN
kim-ai-2020	145	7	an	an	DT
kim-ai-2020	145	8	AI	AI	NNP
kim-ai-2020	145	9	system	system	NN
kim-ai-2020	145	10	can	can	MD
kim-ai-2020	145	11	have	have	VB
kim-ai-2020	145	12	grave	grave	JJ
kim-ai-2020	145	13	consequences	consequence	NNS
kim-ai-2020	145	14	such	such	JJ
kim-ai-2020	145	15	as	as	IN
kim-ai-2020	145	16	a	a	DT
kim-ai-2020	145	17	military	military	JJ
kim-ai-2020	145	18	robot	robot	NN
kim-ai-2020	145	19	,	,	,
kim-ai-2020	145	20	such	such	JJ
kim-ai-2020	145	21	opacity	opacity	NN
kim-ai-2020	145	22	becomes	become	VBZ
kim-ai-2020	145	23	a	a	DT
kim-ai-2020	145	24	serious	serious	JJ
kim-ai-2020	145	25	problem	problem	NN
kim-ai-2020	145	26	.	.	.
kim-ai-2020	146	1	[	[	-LRB-
kim-ai-2020	146	2	footnoteRef:14	footnoteRef:14	NNP
kim-ai-2020	146	3	]	]	-RRB-
kim-ai-2020	146	4	In	in	IN
kim-ai-2020	146	5	spite	spite	NN
kim-ai-2020	146	6	of	of	IN
kim-ai-2020	146	7	these	these	DT
kim-ai-2020	146	8	ethical	ethical	JJ
kim-ai-2020	146	9	concerns	concern	NNS
kim-ai-2020	146	10	,	,	,
kim-ai-2020	146	11	autonomous	autonomous	JJ
kim-ai-2020	146	12	AI	AI	NNP
kim-ai-2020	146	13	systems	system	NNS
kim-ai-2020	146	14	are	be	VBP
kim-ai-2020	146	15	likely	likely	JJ
kim-ai-2020	146	16	to	to	TO
kim-ai-2020	146	17	continue	continue	VB
kim-ai-2020	146	18	to	to	TO
kim-ai-2020	146	19	be	be	VB
kim-ai-2020	146	20	developed	develop	VBN
kim-ai-2020	146	21	and	and	CC
kim-ai-2020	146	22	adopted	adopt	VBN
kim-ai-2020	146	23	in	in	IN
kim-ai-2020	146	24	many	many	JJ
kim-ai-2020	146	25	areas	area	NNS
kim-ai-2020	146	26	as	as	IN
kim-ai-2020	146	27	a	a	DT
kim-ai-2020	146	28	way	way	NN
kim-ai-2020	146	29	to	to	TO
kim-ai-2020	146	30	increase	increase	VB
kim-ai-2020	146	31	efficiency	efficiency	NN
kim-ai-2020	146	32	and	and	CC
kim-ai-2020	146	33	lower	low	JJR
kim-ai-2020	146	34	cost	cost	NN
kim-ai-2020	146	35	.	.	.
kim-ai-2020	147	1	[	[	-LRB-
kim-ai-2020	147	2	13	13	CD
kim-ai-2020	147	3	:	:	:
kim-ai-2020	147	4	Google	Google	NNP
kim-ai-2020	147	5	’s	’s	POS
kim-ai-2020	147	6	research	research	NN
kim-ai-2020	147	7	team	team	NN
kim-ai-2020	147	8	created	create	VBD
kim-ai-2020	147	9	an	an	DT
kim-ai-2020	147	10	AI	AI	NNP
kim-ai-2020	147	11	algorithm	algorithm	NN
kim-ai-2020	147	12	that	that	WDT
kim-ai-2020	147	13	learned	learn	VBD
kim-ai-2020	147	14	how	how	WRB
kim-ai-2020	147	15	to	to	TO
kim-ai-2020	147	16	recognize	recognize	VB
kim-ai-2020	147	17	a	a	DT
kim-ai-2020	147	18	cat	cat	NN
kim-ai-2020	147	19	in	in	IN
kim-ai-2020	147	20	2012	2012	CD
kim-ai-2020	147	21	.	.	.
kim-ai-2020	148	1	The	the	DT
kim-ai-2020	148	2	neural	neural	JJ
kim-ai-2020	148	3	network	network	NN
kim-ai-2020	148	4	behind	behind	IN
kim-ai-2020	148	5	this	this	DT
kim-ai-2020	148	6	algorithm	algorithm	NN
kim-ai-2020	148	7	had	have	VBD
kim-ai-2020	148	8	an	an	DT
kim-ai-2020	148	9	array	array	NN
kim-ai-2020	148	10	of	of	IN
kim-ai-2020	148	11	16,000	16,000	CD
kim-ai-2020	148	12	processors	processor	NNS
kim-ai-2020	148	13	and	and	CC
kim-ai-2020	148	14	more	more	JJR
kim-ai-2020	148	15	than	than	IN
kim-ai-2020	148	16	one	one	CD
kim-ai-2020	148	17	billion	billion	CD
kim-ai-2020	148	18	connections	connection	NNS
kim-ai-2020	148	19	,	,	,
kim-ai-2020	148	20	and	and	CC
kim-ai-2020	148	21	unlabeled	unlabeled	JJ
kim-ai-2020	148	22	random	random	JJ
kim-ai-2020	148	23	thumbnail	thumbnail	NN
kim-ai-2020	148	24	images	image	NNS
kim-ai-2020	148	25	from	from	IN
kim-ai-2020	148	26	10	10	CD
kim-ai-2020	148	27	million	million	CD
kim-ai-2020	148	28	YouTube	YouTube	NNP
kim-ai-2020	148	29	videos	video	NNS
kim-ai-2020	148	30	allowed	allow	VBD
kim-ai-2020	148	31	this	this	DT
kim-ai-2020	148	32	algorithm	algorithm	NN
kim-ai-2020	148	33	to	to	TO
kim-ai-2020	148	34	learn	learn	VB
kim-ai-2020	148	35	to	to	TO
kim-ai-2020	148	36	identify	identify	VB
kim-ai-2020	148	37	cats	cat	NNS
kim-ai-2020	148	38	by	by	IN
kim-ai-2020	148	39	itself	-PRON-	PRP
kim-ai-2020	148	40	.	.	.
kim-ai-2020	149	1	See	see	VB
kim-ai-2020	149	2	Markoff	Markoff	NNP
kim-ai-2020	149	3	2012	2012	CD
kim-ai-2020	149	4	and	and	CC
kim-ai-2020	149	5	Liat	Liat	NNP
kim-ai-2020	149	6	2012	2012	CD
kim-ai-2020	149	7	.	.	.
kim-ai-2020	149	8	]	]	-RRB-
kim-ai-2020	150	1	[	[	-LRB-
kim-ai-2020	150	2	14	14	CD
kim-ai-2020	150	3	:	:	:
kim-ai-2020	150	4	This	this	DT
kim-ai-2020	150	5	black	black	JJ
kim-ai-2020	150	6	-	-	HYPH
kim-ai-2020	150	7	box	box	NN
kim-ai-2020	150	8	nature	nature	NN
kim-ai-2020	150	9	of	of	IN
kim-ai-2020	150	10	AI	AI	NNP
kim-ai-2020	150	11	systems	system	NNS
kim-ai-2020	150	12	powered	power	VBN
kim-ai-2020	150	13	by	by	IN
kim-ai-2020	150	14	machine	machine	NN
kim-ai-2020	150	15	learning	learning	NN
kim-ai-2020	150	16	has	have	VBZ
kim-ai-2020	150	17	raised	raise	VBN
kim-ai-2020	150	18	great	great	JJ
kim-ai-2020	150	19	concern	concern	NN
kim-ai-2020	150	20	among	among	IN
kim-ai-2020	150	21	many	many	JJ
kim-ai-2020	150	22	AI	AI	NNP
kim-ai-2020	150	23	researchers	researcher	NNS
kim-ai-2020	150	24	in	in	IN
kim-ai-2020	150	25	recent	recent	JJ
kim-ai-2020	150	26	years	year	NNS
kim-ai-2020	150	27	.	.	.
kim-ai-2020	151	1	This	this	DT
kim-ai-2020	151	2	is	be	VBZ
kim-ai-2020	151	3	problematic	problematic	JJ
kim-ai-2020	151	4	in	in	IN
kim-ai-2020	151	5	all	all	DT
kim-ai-2020	151	6	areas	area	NNS
kim-ai-2020	151	7	where	where	WRB
kim-ai-2020	151	8	these	these	DT
kim-ai-2020	151	9	AI	AI	NNP
kim-ai-2020	151	10	systems	system	NNS
kim-ai-2020	151	11	are	be	VBP
kim-ai-2020	151	12	used	use	VBN
kim-ai-2020	151	13	for	for	IN
kim-ai-2020	151	14	decision	decision	NN
kim-ai-2020	151	15	-	-	HYPH
kim-ai-2020	151	16	making	making	NN
kim-ai-2020	151	17	,	,	,
kim-ai-2020	151	18	not	not	RB
kim-ai-2020	151	19	just	just	RB
kim-ai-2020	151	20	in	in	IN
kim-ai-2020	151	21	military	military	JJ
kim-ai-2020	151	22	operations	operation	NNS
kim-ai-2020	151	23	.	.	.
kim-ai-2020	152	1	The	the	DT
kim-ai-2020	152	2	gravity	gravity	NN
kim-ai-2020	152	3	of	of	IN
kim-ai-2020	152	4	decisions	decision	NNS
kim-ai-2020	152	5	made	make	VBN
kim-ai-2020	152	6	in	in	IN
kim-ai-2020	152	7	a	a	DT
kim-ai-2020	152	8	military	military	JJ
kim-ai-2020	152	9	operation	operation	NN
kim-ai-2020	152	10	makes	make	VBZ
kim-ai-2020	152	11	this	this	DT
kim-ai-2020	152	12	problem	problem	NN
kim-ai-2020	152	13	even	even	RB
kim-ai-2020	152	14	more	more	RBR
kim-ai-2020	152	15	troublesome	troublesome	JJ
kim-ai-2020	152	16	.	.	.
kim-ai-2020	153	1	Fortunately	fortunately	RB
kim-ai-2020	153	2	,	,	,
kim-ai-2020	153	3	some	some	DT
kim-ai-2020	153	4	AI	AI	NNP
kim-ai-2020	153	5	researchers	researcher	NNS
kim-ai-2020	153	6	including	include	VBG
kim-ai-2020	153	7	those	those	DT
kim-ai-2020	153	8	in	in	IN
kim-ai-2020	153	9	the	the	DT
kim-ai-2020	153	10	US	US	NNP
kim-ai-2020	153	11	Department	Department	NNP
kim-ai-2020	153	12	of	of	IN
kim-ai-2020	153	13	Defense	Defense	NNP
kim-ai-2020	153	14	are	be	VBP
kim-ai-2020	153	15	actively	actively	RB
kim-ai-2020	153	16	working	work	VBG
kim-ai-2020	153	17	to	to	TO
kim-ai-2020	153	18	make	make	VB
kim-ai-2020	153	19	AI	AI	NNP
kim-ai-2020	153	20	systems	system	NNS
kim-ai-2020	153	21	explainable	explainable	JJ
kim-ai-2020	153	22	.	.	.
kim-ai-2020	154	1	But	but	CC
kim-ai-2020	154	2	until	until	IN
kim-ai-2020	154	3	such	such	JJ
kim-ai-2020	154	4	research	research	NN
kim-ai-2020	154	5	bears	bear	VBZ
kim-ai-2020	154	6	fruit	fruit	NN
kim-ai-2020	154	7	and	and	CC
kim-ai-2020	154	8	AI	AI	NNP
kim-ai-2020	154	9	systems	system	NNS
kim-ai-2020	154	10	become	become	VBP
kim-ai-2020	154	11	fully	fully	RB
kim-ai-2020	154	12	explainable	explainable	JJ
kim-ai-2020	154	13	,	,	,
kim-ai-2020	154	14	their	-PRON-	PRP$
kim-ai-2020	154	15	military	military	JJ
kim-ai-2020	154	16	use	use	NN
kim-ai-2020	154	17	means	mean	VBZ
kim-ai-2020	154	18	accepting	accept	VBG
kim-ai-2020	154	19	many	many	JJ
kim-ai-2020	154	20	unknown	unknown	JJ
kim-ai-2020	154	21	variables	variable	NNS
kim-ai-2020	154	22	and	and	CC
kim-ai-2020	154	23	unforeseeable	unforeseeable	JJ
kim-ai-2020	154	24	consequences	consequence	NNS
kim-ai-2020	154	25	.	.	.
kim-ai-2020	155	1	See	see	VB
kim-ai-2020	155	2	Turek	Turek	NNP
kim-ai-2020	155	3	n.d	n.d	NNP
kim-ai-2020	155	4	..	..	NFP
kim-ai-2020	155	5	]	]	-RRB-
kim-ai-2020	155	6	5	5	CD
kim-ai-2020	155	7	.	.	.
kim-ai-2020	156	1	AI	AI	NNP
kim-ai-2020	156	2	Applications	Applications	NNPS
kim-ai-2020	156	3	for	for	IN
kim-ai-2020	156	4	Libraries	library	NNS
kim-ai-2020	156	5	Do	do	VBP
kim-ai-2020	156	6	these	these	DT
kim-ai-2020	156	7	ethical	ethical	JJ
kim-ai-2020	156	8	concerns	concern	NNS
kim-ai-2020	156	9	outlined	outline	VBN
kim-ai-2020	156	10	above	above	IN
kim-ai-2020	156	11	apply	apply	VBP
kim-ai-2020	156	12	to	to	IN
kim-ai-2020	156	13	libraries	library	NNS
kim-ai-2020	156	14	?	?	.
kim-ai-2020	157	1	To	to	TO
kim-ai-2020	157	2	answer	answer	VB
kim-ai-2020	157	3	that	that	DT
kim-ai-2020	157	4	,	,	,
kim-ai-2020	157	5	let	let	VB
kim-ai-2020	157	6	us	-PRON-	PRP
kim-ai-2020	157	7	first	first	RB
kim-ai-2020	157	8	take	take	VB
kim-ai-2020	157	9	a	a	DT
kim-ai-2020	157	10	look	look	NN
kim-ai-2020	157	11	at	at	IN
kim-ai-2020	157	12	how	how	WRB
kim-ai-2020	157	13	AI	AI	NNP
kim-ai-2020	157	14	,	,	,
kim-ai-2020	157	15	particularly	particularly	RB
kim-ai-2020	157	16	machine	machine	NN
kim-ai-2020	157	17	learning	learning	NN
kim-ai-2020	157	18	,	,	,
kim-ai-2020	157	19	may	may	MD
kim-ai-2020	157	20	apply	apply	VB
kim-ai-2020	157	21	to	to	IN
kim-ai-2020	157	22	library	library	NN
kim-ai-2020	157	23	services	service	NNS
kim-ai-2020	157	24	and	and	CC
kim-ai-2020	157	25	operations	operation	NNS
kim-ai-2020	157	26	.	.	.
kim-ai-2020	158	1	AI	AI	NNP
kim-ai-2020	158	2	-	-	HYPH
kim-ai-2020	158	3	powered	power	VBN
kim-ai-2020	158	4	digital	digital	JJ
kim-ai-2020	158	5	assistants	assistant	NNS
kim-ai-2020	158	6	are	be	VBP
kim-ai-2020	158	7	likely	likely	JJ
kim-ai-2020	158	8	to	to	TO
kim-ai-2020	158	9	mediate	mediate	VB
kim-ai-2020	158	10	a	a	DT
kim-ai-2020	158	11	library	library	NN
kim-ai-2020	158	12	user	user	NN
kim-ai-2020	158	13	’s	’s	POS
kim-ai-2020	158	14	information	information	NN
kim-ai-2020	158	15	search	search	NN
kim-ai-2020	158	16	,	,	,
kim-ai-2020	158	17	discovery	discovery	NN
kim-ai-2020	158	18	,	,	,
kim-ai-2020	158	19	and	and	CC
kim-ai-2020	158	20	retrieval	retrieval	NN
kim-ai-2020	158	21	activities	activity	NNS
kim-ai-2020	158	22	in	in	IN
kim-ai-2020	158	23	the	the	DT
kim-ai-2020	158	24	near	near	JJ
kim-ai-2020	158	25	future	future	NN
kim-ai-2020	158	26	.	.	.
kim-ai-2020	159	1	In	in	IN
kim-ai-2020	159	2	recent	recent	JJ
kim-ai-2020	159	3	years	year	NNS
kim-ai-2020	159	4	,	,	,
kim-ai-2020	159	5	machine	machine	NN
kim-ai-2020	159	6	learning	learning	NN
kim-ai-2020	159	7	and	and	CC
kim-ai-2020	159	8	deep	deep	JJ
kim-ai-2020	159	9	learning	learning	NN
kim-ai-2020	159	10	have	have	VBP
kim-ai-2020	159	11	brought	bring	VBN
kim-ai-2020	159	12	significant	significant	JJ
kim-ai-2020	159	13	improvement	improvement	NN
kim-ai-2020	159	14	to	to	IN
kim-ai-2020	159	15	natural	natural	JJ
kim-ai-2020	159	16	language	language	NN
kim-ai-2020	159	17	processing	processing	NN
kim-ai-2020	159	18	(	(	-LRB-
kim-ai-2020	159	19	NLP	NLP	NNP
kim-ai-2020	159	20	)	)	-RRB-
kim-ai-2020	159	21	,	,	,
kim-ai-2020	159	22	which	which	WDT
kim-ai-2020	159	23	deals	deal	VBZ
kim-ai-2020	159	24	with	with	IN
kim-ai-2020	159	25	analyzing	analyze	VBG
kim-ai-2020	159	26	large	large	JJ
kim-ai-2020	159	27	amounts	amount	NNS
kim-ai-2020	159	28	of	of	IN
kim-ai-2020	159	29	natural	natural	JJ
kim-ai-2020	159	30	language	language	NN
kim-ai-2020	159	31	data	datum	NNS
kim-ai-2020	159	32	to	to	TO
kim-ai-2020	159	33	make	make	VB
kim-ai-2020	159	34	the	the	DT
kim-ai-2020	159	35	interaction	interaction	NN
kim-ai-2020	159	36	between	between	IN
kim-ai-2020	159	37	people	people	NNS
kim-ai-2020	159	38	and	and	CC
kim-ai-2020	159	39	machines	machine	NNS
kim-ai-2020	159	40	in	in	IN
kim-ai-2020	159	41	natural	natural	JJ
kim-ai-2020	159	42	languages	language	NNS
kim-ai-2020	159	43	possible	possible	JJ
kim-ai-2020	159	44	.	.	.
kim-ai-2020	160	1	For	for	IN
kim-ai-2020	160	2	instance	instance	NN
kim-ai-2020	160	3	,	,	,
kim-ai-2020	160	4	Google	Google	NNP
kim-ai-2020	160	5	Assistant	Assistant	NNP
kim-ai-2020	160	6	’s	’s	POS
kim-ai-2020	160	7	new	new	JJ
kim-ai-2020	160	8	feature	feature	NN
kim-ai-2020	160	9	‘	'	``
kim-ai-2020	160	10	duplex	duplex	NN
kim-ai-2020	160	11	’	'	''
kim-ai-2020	160	12	was	be	VBD
kim-ai-2020	160	13	shown	show	VBN
kim-ai-2020	160	14	to	to	TO
kim-ai-2020	160	15	successfully	successfully	RB
kim-ai-2020	160	16	make	make	VB
kim-ai-2020	160	17	a	a	DT
kim-ai-2020	160	18	phone	phone	NN
kim-ai-2020	160	19	reservation	reservation	NN
kim-ai-2020	160	20	with	with	IN
kim-ai-2020	160	21	restaurant	restaurant	NN
kim-ai-2020	160	22	staff	staff	NN
kim-ai-2020	160	23	in	in	IN
kim-ai-2020	160	24	2018	2018	CD
kim-ai-2020	160	25	(	(	-LRB-
kim-ai-2020	160	26	Welch	Welch	NNP
kim-ai-2020	160	27	2018	2018	CD
kim-ai-2020	160	28	)	)	-RRB-
kim-ai-2020	160	29	.	.	.
kim-ai-2020	161	1	Google	Google	NNP
kim-ai-2020	161	2	’s	’s	POS
kim-ai-2020	161	3	real	real	JJ
kim-ai-2020	161	4	-	-	HYPH
kim-ai-2020	161	5	time	time	NN
kim-ai-2020	161	6	translation	translation	NN
kim-ai-2020	161	7	capability	capability	NN
kim-ai-2020	161	8	for	for	IN
kim-ai-2020	161	9	40	40	CD
kim-ai-2020	161	10	different	different	JJ
kim-ai-2020	161	11	languages	language	NNS
kim-ai-2020	161	12	is	be	VBZ
kim-ai-2020	161	13	planned	plan	VBN
kim-ai-2020	161	14	to	to	TO
kim-ai-2020	161	15	become	become	VB
kim-ai-2020	161	16	part	part	NN
kim-ai-2020	161	17	of	of	IN
kim-ai-2020	161	18	the	the	DT
kim-ai-2020	161	19	features	feature	NNS
kim-ai-2020	161	20	for	for	IN
kim-ai-2020	161	21	all	all	DT
kim-ai-2020	161	22	headphones	headphone	NNS
kim-ai-2020	161	23	equipped	equip	VBN
kim-ai-2020	161	24	with	with	IN
kim-ai-2020	161	25	Google	Google	NNP
kim-ai-2020	161	26	Assistant	Assistant	NNP
kim-ai-2020	161	27	(	(	-LRB-
kim-ai-2020	161	28	Le	Le	NNP
kim-ai-2020	161	29	and	and	CC
kim-ai-2020	161	30	Schuster	Schuster	NNP
kim-ai-2020	161	31	2016	2016	CD
kim-ai-2020	161	32	,	,	,
kim-ai-2020	161	33	Porter	Porter	NNP
kim-ai-2020	161	34	2018	2018	CD
kim-ai-2020	161	35	)	)	-RRB-
kim-ai-2020	161	36	.	.	.
kim-ai-2020	162	1	As	as	IN
kim-ai-2020	162	2	digital	digital	JJ
kim-ai-2020	162	3	assistants	assistant	NNS
kim-ai-2020	162	4	become	become	VBP
kim-ai-2020	162	5	capable	capable	JJ
kim-ai-2020	162	6	of	of	IN
kim-ai-2020	162	7	handling	handle	VBG
kim-ai-2020	162	8	more	more	RBR
kim-ai-2020	162	9	sophisticated	sophisticated	JJ
kim-ai-2020	162	10	language	language	NN
kim-ai-2020	162	11	tasks	task	NNS
kim-ai-2020	162	12	,	,	,
kim-ai-2020	162	13	their	-PRON-	PRP$
kim-ai-2020	162	14	use	use	NN
kim-ai-2020	162	15	as	as	IN
kim-ai-2020	162	16	a	a	DT
kim-ai-2020	162	17	flexible	flexible	JJ
kim-ai-2020	162	18	voice	voice	NN
kim-ai-2020	162	19	user	user	NN
kim-ai-2020	162	20	interface	interface	NN
kim-ai-2020	162	21	will	will	MD
kim-ai-2020	162	22	only	only	RB
kim-ai-2020	162	23	increase	increase	VB
kim-ai-2020	162	24	.	.	.
kim-ai-2020	163	1	Such	such	JJ
kim-ai-2020	163	2	digital	digital	JJ
kim-ai-2020	163	3	assistants	assistant	NNS
kim-ai-2020	163	4	will	will	MD
kim-ai-2020	163	5	directly	directly	RB
kim-ai-2020	163	6	interact	interact	VB
kim-ai-2020	163	7	with	with	IN
kim-ai-2020	163	8	library	library	JJ
kim-ai-2020	163	9	systems	system	NNS
kim-ai-2020	163	10	and	and	CC
kim-ai-2020	163	11	applications	application	NNS
kim-ai-2020	163	12	,	,	,
kim-ai-2020	163	13	automatically	automatically	RB
kim-ai-2020	163	14	interpret	interpret	VB
kim-ai-2020	163	15	a	a	DT
kim-ai-2020	163	16	query	query	NN
kim-ai-2020	163	17	,	,	,
kim-ai-2020	163	18	and	and	CC
kim-ai-2020	163	19	return	return	VB
kim-ai-2020	163	20	results	result	NNS
kim-ai-2020	163	21	that	that	WDT
kim-ai-2020	163	22	they	-PRON-	PRP
kim-ai-2020	163	23	deem	deem	VBP
kim-ai-2020	163	24	to	to	TO
kim-ai-2020	163	25	be	be	VB
kim-ai-2020	163	26	most	most	RBS
kim-ai-2020	163	27	relevant	relevant	JJ
kim-ai-2020	163	28	.	.	.
kim-ai-2020	164	1	Those	those	DT
kim-ai-2020	164	2	digital	digital	JJ
kim-ai-2020	164	3	assistants	assistant	NNS
kim-ai-2020	164	4	can	can	MD
kim-ai-2020	164	5	also	also	RB
kim-ai-2020	164	6	be	be	VB
kim-ai-2020	164	7	equipped	equip	VBN
kim-ai-2020	164	8	to	to	TO
kim-ai-2020	164	9	handle	handle	VB
kim-ai-2020	164	10	the	the	DT
kim-ai-2020	164	11	library	library	NN
kim-ai-2020	164	12	’s	’s	,
kim-ai-2020	164	13	traditional	traditional	JJ
kim-ai-2020	164	14	reference	reference	NN
kim-ai-2020	164	15	or	or	CC
kim-ai-2020	164	16	readers	reader	NNS
kim-ai-2020	164	17	’	’	POS
kim-ai-2020	164	18	advisory	advisory	JJ
kim-ai-2020	164	19	service	service	NN
kim-ai-2020	164	20	.	.	.
kim-ai-2020	165	1	Integrated	integrate	VBN
kim-ai-2020	165	2	into	into	IN
kim-ai-2020	165	3	a	a	DT
kim-ai-2020	165	4	humanoid	humanoid	JJ
kim-ai-2020	165	5	robot	robot	NN
kim-ai-2020	165	6	body	body	NN
kim-ai-2020	165	7	,	,	,
kim-ai-2020	165	8	they	-PRON-	PRP
kim-ai-2020	165	9	may	may	MD
kim-ai-2020	165	10	even	even	RB
kim-ai-2020	165	11	greet	greet	VB
kim-ai-2020	165	12	library	library	JJ
kim-ai-2020	165	13	patrons	patron	NNS
kim-ai-2020	165	14	at	at	IN
kim-ai-2020	165	15	the	the	DT
kim-ai-2020	165	16	entrance	entrance	NN
kim-ai-2020	165	17	and	and	CC
kim-ai-2020	165	18	answer	answer	VB
kim-ai-2020	165	19	directional	directional	JJ
kim-ai-2020	165	20	questions	question	NNS
kim-ai-2020	165	21	about	about	IN
kim-ai-2020	165	22	the	the	DT
kim-ai-2020	165	23	library	library	NN
kim-ai-2020	165	24	building	building	NN
kim-ai-2020	165	25	.	.	.
kim-ai-2020	166	1	Cataloging	catalog	VBG
kim-ai-2020	166	2	,	,	,
kim-ai-2020	166	3	abstracting	abstracting	NN
kim-ai-2020	166	4	,	,	,
kim-ai-2020	166	5	and	and	CC
kim-ai-2020	166	6	indexing	indexing	NN
kim-ai-2020	166	7	are	be	VBP
kim-ai-2020	166	8	other	other	JJ
kim-ai-2020	166	9	areas	area	NNS
kim-ai-2020	166	10	where	where	WRB
kim-ai-2020	166	11	AI	AI	NNP
kim-ai-2020	166	12	will	will	MD
kim-ai-2020	166	13	be	be	VB
kim-ai-2020	166	14	actively	actively	RB
kim-ai-2020	166	15	utilized	utilize	VBN
kim-ai-2020	166	16	.	.	.
kim-ai-2020	167	1	Currently	currently	RB
kim-ai-2020	167	2	,	,	,
kim-ai-2020	167	3	those	those	DT
kim-ai-2020	167	4	tasks	task	NNS
kim-ai-2020	167	5	are	be	VBP
kim-ai-2020	167	6	performed	perform	VBN
kim-ai-2020	167	7	by	by	IN
kim-ai-2020	167	8	skilled	skilled	JJ
kim-ai-2020	167	9	professionals	professional	NNS
kim-ai-2020	167	10	.	.	.
kim-ai-2020	168	1	But	but	CC
kim-ai-2020	168	2	as	as	IN
kim-ai-2020	168	3	AI	AI	NNP
kim-ai-2020	168	4	applications	application	NNS
kim-ai-2020	168	5	become	become	VBP
kim-ai-2020	168	6	more	more	RBR
kim-ai-2020	168	7	sophisticated	sophisticated	JJ
kim-ai-2020	168	8	,	,	,
kim-ai-2020	168	9	we	-PRON-	PRP
kim-ai-2020	168	10	may	may	MD
kim-ai-2020	168	11	see	see	VB
kim-ai-2020	168	12	many	many	JJ
kim-ai-2020	168	13	of	of	IN
kim-ai-2020	168	14	those	those	DT
kim-ai-2020	168	15	tasks	task	NNS
kim-ai-2020	168	16	partially	partially	RB
kim-ai-2020	168	17	or	or	CC
kim-ai-2020	168	18	fully	fully	RB
kim-ai-2020	168	19	automated	automate	VBN
kim-ai-2020	168	20	and	and	CC
kim-ai-2020	168	21	handed	hand	VBD
kim-ai-2020	168	22	over	over	RP
kim-ai-2020	168	23	to	to	IN
kim-ai-2020	168	24	AI	AI	NNP
kim-ai-2020	168	25	systems	system	NNS
kim-ai-2020	168	26	.	.	.
kim-ai-2020	169	1	Machine	machine	NN
kim-ai-2020	169	2	learning	learning	NN
kim-ai-2020	169	3	and	and	CC
kim-ai-2020	169	4	deep	deep	JJ
kim-ai-2020	169	5	learning	learning	NN
kim-ai-2020	169	6	can	can	MD
kim-ai-2020	169	7	be	be	VB
kim-ai-2020	169	8	used	use	VBN
kim-ai-2020	169	9	to	to	TO
kim-ai-2020	169	10	extract	extract	VB
kim-ai-2020	169	11	key	key	JJ
kim-ai-2020	169	12	information	information	NN
kim-ai-2020	169	13	from	from	IN
kim-ai-2020	169	14	a	a	DT
kim-ai-2020	169	15	large	large	JJ
kim-ai-2020	169	16	number	number	NN
kim-ai-2020	169	17	of	of	IN
kim-ai-2020	169	18	documents	document	NNS
kim-ai-2020	169	19	or	or	CC
kim-ai-2020	169	20	from	from	IN
kim-ai-2020	169	21	information	information	NN
kim-ai-2020	169	22	-	-	HYPH
kim-ai-2020	169	23	rich	rich	JJ
kim-ai-2020	169	24	visual	visual	JJ
kim-ai-2020	169	25	materials	material	NNS
kim-ai-2020	169	26	such	such	JJ
kim-ai-2020	169	27	as	as	IN
kim-ai-2020	169	28	maps	map	NNS
kim-ai-2020	169	29	and	and	CC
kim-ai-2020	169	30	video	video	NN
kim-ai-2020	169	31	recordings	recording	NNS
kim-ai-2020	169	32	,	,	,
kim-ai-2020	169	33	and	and	CC
kim-ai-2020	169	34	generate	generate	VB
kim-ai-2020	169	35	metadata	metadata	NN
kim-ai-2020	169	36	or	or	CC
kim-ai-2020	169	37	a	a	DT
kim-ai-2020	169	38	summary	summary	NN
kim-ai-2020	169	39	.	.	.
kim-ai-2020	170	1	Since	since	IN
kim-ai-2020	170	2	machine	machine	NN
kim-ai-2020	170	3	learning	learning	NN
kim-ai-2020	170	4	is	be	VBZ
kim-ai-2020	170	5	new	new	JJ
kim-ai-2020	170	6	to	to	IN
kim-ai-2020	170	7	libraries	library	NNS
kim-ai-2020	170	8	,	,	,
kim-ai-2020	170	9	there	there	EX
kim-ai-2020	170	10	are	be	VBP
kim-ai-2020	170	11	a	a	DT
kim-ai-2020	170	12	relatively	relatively	RB
kim-ai-2020	170	13	small	small	JJ
kim-ai-2020	170	14	number	number	NN
kim-ai-2020	170	15	of	of	IN
kim-ai-2020	170	16	machine	machine	NN
kim-ai-2020	170	17	learning	learning	NN
kim-ai-2020	170	18	applications	application	NNS
kim-ai-2020	170	19	developed	develop	VBN
kim-ai-2020	170	20	for	for	IN
kim-ai-2020	170	21	libraries	library	NNS
kim-ai-2020	170	22	’	’	POS
kim-ai-2020	170	23	use	use	NN
kim-ai-2020	170	24	.	.	.
kim-ai-2020	171	1	They	-PRON-	PRP
kim-ai-2020	171	2	are	be	VBP
kim-ai-2020	171	3	likely	likely	JJ
kim-ai-2020	171	4	to	to	TO
kim-ai-2020	171	5	grow	grow	VB
kim-ai-2020	171	6	in	in	IN
kim-ai-2020	171	7	number	number	NN
kim-ai-2020	171	8	.	.	.
kim-ai-2020	172	1	Yewno	Yewno	NNP
kim-ai-2020	172	2	,	,	,
kim-ai-2020	172	3	Quartolio	Quartolio	NNP
kim-ai-2020	172	4	,	,	,
kim-ai-2020	172	5	and	and	CC
kim-ai-2020	172	6	Iris.ai	iris.ai	NN
kim-ai-2020	172	7	are	be	VBP
kim-ai-2020	172	8	examples	example	NNS
kim-ai-2020	172	9	of	of	IN
kim-ai-2020	172	10	the	the	DT
kim-ai-2020	172	11	vendor	vendor	NN
kim-ai-2020	172	12	products	product	NNS
kim-ai-2020	172	13	developed	develop	VBN
kim-ai-2020	172	14	with	with	IN
kim-ai-2020	172	15	machine	machine	NN
kim-ai-2020	172	16	learning	learning	NN
kim-ai-2020	172	17	and	and	CC
kim-ai-2020	172	18	deep	deep	JJ
kim-ai-2020	172	19	learning	learning	NN
kim-ai-2020	172	20	techniques	technique	NNS
kim-ai-2020	172	21	.	.	.
kim-ai-2020	173	1	Yewno	Yewno	NNP
kim-ai-2020	173	2	Discover	Discover	NNP
kim-ai-2020	173	3	displays	display	VBZ
kim-ai-2020	173	4	the	the	DT
kim-ai-2020	173	5	connections	connection	NNS
kim-ai-2020	173	6	between	between	IN
kim-ai-2020	173	7	different	different	JJ
kim-ai-2020	173	8	concepts	concept	NNS
kim-ai-2020	173	9	or	or	CC
kim-ai-2020	173	10	works	work	NNS
kim-ai-2020	173	11	in	in	IN
kim-ai-2020	173	12	library	library	NN
kim-ai-2020	173	13	materials	material	NNS
kim-ai-2020	173	14	.	.	.
kim-ai-2020	174	1	Quartolio	quartolio	JJ
kim-ai-2020	174	2	targets	target	NNS
kim-ai-2020	174	3	researchers	researcher	NNS
kim-ai-2020	174	4	looking	look	VBG
kim-ai-2020	174	5	to	to	TO
kim-ai-2020	174	6	discover	discover	VB
kim-ai-2020	174	7	untapped	untapped	JJ
kim-ai-2020	174	8	research	research	NN
kim-ai-2020	174	9	opportunities	opportunity	NNS
kim-ai-2020	174	10	based	base	VBN
kim-ai-2020	174	11	upon	upon	IN
kim-ai-2020	174	12	a	a	DT
kim-ai-2020	174	13	large	large	JJ
kim-ai-2020	174	14	amount	amount	NN
kim-ai-2020	174	15	of	of	IN
kim-ai-2020	174	16	data	datum	NNS
kim-ai-2020	174	17	that	that	WDT
kim-ai-2020	174	18	includes	include	VBZ
kim-ai-2020	174	19	articles	article	NNS
kim-ai-2020	174	20	,	,	,
kim-ai-2020	174	21	clinical	clinical	JJ
kim-ai-2020	174	22	trials	trial	NNS
kim-ai-2020	174	23	,	,	,
kim-ai-2020	174	24	patents	patent	NNS
kim-ai-2020	174	25	,	,	,
kim-ai-2020	174	26	and	and	CC
kim-ai-2020	174	27	notes	note	NNS
kim-ai-2020	174	28	.	.	.
kim-ai-2020	175	1	Similarly	similarly	RB
kim-ai-2020	175	2	,	,	,
kim-ai-2020	175	3	Iris.ai	Iris.ai	NNP
kim-ai-2020	175	4	helps	help	VBZ
kim-ai-2020	175	5	researchers	researcher	NNS
kim-ai-2020	175	6	identify	identify	VB
kim-ai-2020	175	7	and	and	CC
kim-ai-2020	175	8	review	review	VB
kim-ai-2020	175	9	a	a	DT
kim-ai-2020	175	10	large	large	JJ
kim-ai-2020	175	11	amount	amount	NN
kim-ai-2020	175	12	of	of	IN
kim-ai-2020	175	13	research	research	NN
kim-ai-2020	175	14	papers	paper	NNS
kim-ai-2020	175	15	and	and	CC
kim-ai-2020	175	16	patents	patent	NNS
kim-ai-2020	175	17	and	and	CC
kim-ai-2020	175	18	extracts	extract	VBZ
kim-ai-2020	175	19	key	key	JJ
kim-ai-2020	175	20	information	information	NN
kim-ai-2020	175	21	from	from	IN
kim-ai-2020	175	22	them	-PRON-	PRP
kim-ai-2020	175	23	.	.	.
kim-ai-2020	176	1	[	[	-LRB-
kim-ai-2020	176	2	footnoteRef:15	footnoteref:15	CD
kim-ai-2020	176	3	]	]	-RRB-
kim-ai-2020	176	4	Kira	Kira	NNP
kim-ai-2020	176	5	identifies	identifie	NNS
kim-ai-2020	176	6	,	,	,
kim-ai-2020	176	7	extracts	extract	VBZ
kim-ai-2020	176	8	,	,	,
kim-ai-2020	176	9	and	and	CC
kim-ai-2020	176	10	analyzes	analyze	VBZ
kim-ai-2020	176	11	text	text	NN
kim-ai-2020	176	12	in	in	IN
kim-ai-2020	176	13	contracts	contract	NNS
kim-ai-2020	176	14	and	and	CC
kim-ai-2020	176	15	other	other	JJ
kim-ai-2020	176	16	legal	legal	JJ
kim-ai-2020	176	17	documents	document	NNS
kim-ai-2020	176	18	.	.	.
kim-ai-2020	177	1	[	[	-LRB-
kim-ai-2020	177	2	footnoteRef:16	footnoteRef:16	NNP
kim-ai-2020	177	3	]	]	-RRB-
kim-ai-2020	177	4	None	none	NN
kim-ai-2020	177	5	of	of	IN
kim-ai-2020	177	6	these	these	DT
kim-ai-2020	177	7	applications	application	NNS
kim-ai-2020	177	8	performs	perform	VBZ
kim-ai-2020	177	9	fully	fully	RB
kim-ai-2020	177	10	automated	automate	VBN
kim-ai-2020	177	11	decision	decision	NN
kim-ai-2020	177	12	-	-	HYPH
kim-ai-2020	177	13	making	making	NN
kim-ai-2020	177	14	nor	nor	CC
kim-ai-2020	177	15	incorporates	incorporate	VBZ
kim-ai-2020	177	16	the	the	DT
kim-ai-2020	177	17	digital	digital	JJ
kim-ai-2020	177	18	assistant	assistant	NN
kim-ai-2020	177	19	feature	feature	NN
kim-ai-2020	177	20	.	.	.
kim-ai-2020	178	1	But	but	CC
kim-ai-2020	178	2	this	this	DT
kim-ai-2020	178	3	is	be	VBZ
kim-ai-2020	178	4	an	an	DT
kim-ai-2020	178	5	area	area	NN
kim-ai-2020	178	6	where	where	WRB
kim-ai-2020	178	7	information	information	NN
kim-ai-2020	178	8	systems	system	NNS
kim-ai-2020	178	9	vendors	vendor	NNS
kim-ai-2020	178	10	are	be	VBP
kim-ai-2020	178	11	increasingly	increasingly	RB
kim-ai-2020	178	12	focusing	focus	VBG
kim-ai-2020	178	13	.	.	.
kim-ai-2020	179	1	[	[	-LRB-
kim-ai-2020	179	2	15	15	CD
kim-ai-2020	179	3	:	:	:
kim-ai-2020	179	4	See	see	VB
kim-ai-2020	179	5	https://www.yewno.com/education	https://www.yewno.com/education	CD
kim-ai-2020	179	6	,	,	,
kim-ai-2020	179	7	https://quartolio.com/	https://quartolio.com/	CD
kim-ai-2020	179	8	,	,	,
kim-ai-2020	179	9	and	and	CC
kim-ai-2020	179	10	https://iris.ai/.	https://iris.ai/.	VBN
kim-ai-2020	179	11	]	]	-RRB-
kim-ai-2020	180	1	[	[	-LRB-
kim-ai-2020	180	2	16	16	CD
kim-ai-2020	180	3	:	:	:
kim-ai-2020	180	4	See	see	VB
kim-ai-2020	180	5	https://kirasystems.com/.	https://kirasystems.com/.	NNP
kim-ai-2020	181	1	Law	law	NN
kim-ai-2020	181	2	firms	firm	NNS
kim-ai-2020	181	3	are	be	VBP
kim-ai-2020	181	4	adopting	adopt	VBG
kim-ai-2020	181	5	similar	similar	JJ
kim-ai-2020	181	6	products	product	NNS
kim-ai-2020	181	7	to	to	TO
kim-ai-2020	181	8	automate	automate	VB
kim-ai-2020	181	9	and	and	CC
kim-ai-2020	181	10	expedite	expedite	VB
kim-ai-2020	181	11	their	-PRON-	PRP$
kim-ai-2020	181	12	legal	legal	JJ
kim-ai-2020	181	13	work	work	NN
kim-ai-2020	181	14	,	,	,
kim-ai-2020	181	15	and	and	CC
kim-ai-2020	181	16	law	law	NN
kim-ai-2020	181	17	librarians	librarian	NNS
kim-ai-2020	181	18	are	be	VBP
kim-ai-2020	181	19	discussing	discuss	VBG
kim-ai-2020	181	20	how	how	WRB
kim-ai-2020	181	21	the	the	DT
kim-ai-2020	181	22	use	use	NN
kim-ai-2020	181	23	of	of	IN
kim-ai-2020	181	24	AI	AI	NNP
kim-ai-2020	181	25	may	may	MD
kim-ai-2020	181	26	change	change	VB
kim-ai-2020	181	27	their	-PRON-	PRP$
kim-ai-2020	181	28	work	work	NN
kim-ai-2020	181	29	.	.	.
kim-ai-2020	182	1	See	see	VB
kim-ai-2020	182	2	Marr	Marr	NNP
kim-ai-2020	182	3	2018	2018	CD
kim-ai-2020	182	4	and	and	CC
kim-ai-2020	182	5	Talley	Talley	NNP
kim-ai-2020	182	6	2016	2016	CD
kim-ai-2020	182	7	.	.	.
kim-ai-2020	182	8	]	]	-RRB-
kim-ai-2020	183	1	Libraries	library	NNS
kim-ai-2020	183	2	themselves	-PRON-	PRP
kim-ai-2020	183	3	are	be	VBP
kim-ai-2020	183	4	also	also	RB
kim-ai-2020	183	5	experimenting	experiment	VBG
kim-ai-2020	183	6	with	with	IN
kim-ai-2020	183	7	AI	AI	NNP
kim-ai-2020	183	8	to	to	TO
kim-ai-2020	183	9	test	test	VB
kim-ai-2020	183	10	its	-PRON-	PRP$
kim-ai-2020	183	11	potential	potential	NN
kim-ai-2020	183	12	for	for	IN
kim-ai-2020	183	13	library	library	NN
kim-ai-2020	183	14	services	service	NNS
kim-ai-2020	183	15	and	and	CC
kim-ai-2020	183	16	operations	operation	NNS
kim-ai-2020	183	17	.	.	.
kim-ai-2020	184	1	Some	some	DT
kim-ai-2020	184	2	are	be	VBP
kim-ai-2020	184	3	focusing	focus	VBG
kim-ai-2020	184	4	on	on	IN
kim-ai-2020	184	5	using	use	VBG
kim-ai-2020	184	6	AI	AI	NNP
kim-ai-2020	184	7	,	,	,
kim-ai-2020	184	8	particularly	particularly	RB
kim-ai-2020	184	9	the	the	DT
kim-ai-2020	184	10	voice	voice	NN
kim-ai-2020	184	11	user	user	NN
kim-ai-2020	184	12	interface	interface	NN
kim-ai-2020	184	13	aspect	aspect	NN
kim-ai-2020	184	14	of	of	IN
kim-ai-2020	184	15	the	the	DT
kim-ai-2020	184	16	digital	digital	JJ
kim-ai-2020	184	17	assistant	assistant	NN
kim-ai-2020	184	18	,	,	,
kim-ai-2020	184	19	in	in	IN
kim-ai-2020	184	20	order	order	NN
kim-ai-2020	184	21	to	to	TO
kim-ai-2020	184	22	improve	improve	VB
kim-ai-2020	184	23	existing	exist	VBG
kim-ai-2020	184	24	services	service	NNS
kim-ai-2020	184	25	.	.	.
kim-ai-2020	185	1	The	the	DT
kim-ai-2020	185	2	University	University	NNP
kim-ai-2020	185	3	of	of	IN
kim-ai-2020	185	4	Oklahoma	Oklahoma	NNP
kim-ai-2020	185	5	Libraries	library	NNS
kim-ai-2020	185	6	have	have	VBP
kim-ai-2020	185	7	been	be	VBN
kim-ai-2020	185	8	building	build	VBG
kim-ai-2020	185	9	an	an	DT
kim-ai-2020	185	10	Alexa	alexa	JJ
kim-ai-2020	185	11	application	application	NN
kim-ai-2020	185	12	to	to	TO
kim-ai-2020	185	13	provide	provide	VB
kim-ai-2020	185	14	basic	basic	JJ
kim-ai-2020	185	15	reference	reference	NN
kim-ai-2020	185	16	service	service	NN
kim-ai-2020	185	17	to	to	IN
kim-ai-2020	185	18	their	-PRON-	PRP$
kim-ai-2020	185	19	students	student	NNS
kim-ai-2020	185	20	.	.	.
kim-ai-2020	186	1	[	[	-LRB-
kim-ai-2020	186	2	footnoteRef:17	footnoteRef:17	NNP
kim-ai-2020	186	3	]	]	-RRB-
kim-ai-2020	186	4	At	at	IN
kim-ai-2020	186	5	the	the	DT
kim-ai-2020	186	6	University	University	NNP
kim-ai-2020	186	7	of	of	IN
kim-ai-2020	186	8	Pretoria	Pretoria	NNP
kim-ai-2020	186	9	Library	Library	NNP
kim-ai-2020	186	10	in	in	IN
kim-ai-2020	186	11	South	South	NNP
kim-ai-2020	186	12	Africa	Africa	NNP
kim-ai-2020	186	13	,	,	,
kim-ai-2020	186	14	a	a	DT
kim-ai-2020	186	15	robot	robot	NN
kim-ai-2020	186	16	named	name	VBN
kim-ai-2020	186	17	‘	'	``
kim-ai-2020	186	18	Libby	Libby	NNP
kim-ai-2020	186	19	’	'	''
kim-ai-2020	186	20	already	already	RB
kim-ai-2020	186	21	interacts	interact	VBZ
kim-ai-2020	186	22	with	with	IN
kim-ai-2020	186	23	patrons	patron	NNS
kim-ai-2020	186	24	by	by	IN
kim-ai-2020	186	25	providing	provide	VBG
kim-ai-2020	186	26	guidance	guidance	NN
kim-ai-2020	186	27	,	,	,
kim-ai-2020	186	28	answering	answer	VBG
kim-ai-2020	186	29	questions	question	NNS
kim-ai-2020	186	30	,	,	,
kim-ai-2020	186	31	conducting	conduct	VBG
kim-ai-2020	186	32	surveys	survey	NNS
kim-ai-2020	186	33	,	,	,
kim-ai-2020	186	34	and	and	CC
kim-ai-2020	186	35	displaying	display	VBG
kim-ai-2020	186	36	marketing	marketing	NN
kim-ai-2020	186	37	videos	video	NNS
kim-ai-2020	186	38	(	(	-LRB-
kim-ai-2020	186	39	Mahlangu	Mahlangu	NNP
kim-ai-2020	186	40	2019	2019	CD
kim-ai-2020	186	41	)	)	-RRB-
kim-ai-2020	186	42	.	.	.
kim-ai-2020	187	1	[	[	-LRB-
kim-ai-2020	187	2	17	17	CD
kim-ai-2020	187	3	:	:	:
kim-ai-2020	187	4	University	University	NNP
kim-ai-2020	187	5	of	of	IN
kim-ai-2020	187	6	Oklahoma	Oklahoma	NNP
kim-ai-2020	187	7	Libraries	Libraries	NNPS
kim-ai-2020	187	8	are	be	VBP
kim-ai-2020	187	9	building	build	VBG
kim-ai-2020	187	10	an	an	DT
kim-ai-2020	187	11	Alexa	alexa	JJ
kim-ai-2020	187	12	application	application	NN
kim-ai-2020	187	13	that	that	WDT
kim-ai-2020	187	14	will	will	MD
kim-ai-2020	187	15	provide	provide	VB
kim-ai-2020	187	16	some	some	DT
kim-ai-2020	187	17	basic	basic	JJ
kim-ai-2020	187	18	reference	reference	NN
kim-ai-2020	187	19	service	service	NN
kim-ai-2020	187	20	to	to	IN
kim-ai-2020	187	21	their	-PRON-	PRP$
kim-ai-2020	187	22	students	student	NNS
kim-ai-2020	187	23	.	.	.
kim-ai-2020	188	1	Also	also	RB
kim-ai-2020	188	2	,	,	,
kim-ai-2020	188	3	their	-PRON-	PRP$
kim-ai-2020	188	4	PAIR	PAIR	NNP
kim-ai-2020	188	5	registry	registry	NN
kim-ai-2020	188	6	attempts	attempt	VBZ
kim-ai-2020	188	7	to	to	TO
kim-ai-2020	188	8	compile	compile	VB
kim-ai-2020	188	9	all	all	DT
kim-ai-2020	188	10	AI	ai	NN
kim-ai-2020	188	11	-	-	HYPH
kim-ai-2020	188	12	related	relate	VBN
kim-ai-2020	188	13	projects	project	NNS
kim-ai-2020	188	14	at	at	IN
kim-ai-2020	188	15	libraries	library	NNS
kim-ai-2020	188	16	.	.	.
kim-ai-2020	189	1	See	see	VB
kim-ai-2020	189	2	https://pair.libraries.ou.edu	https://pair.libraries.ou.edu	NN
kim-ai-2020	189	3	.	.	.
kim-ai-2020	189	4	]	]	-RRB-
kim-ai-2020	190	1	Other	other	JJ
kim-ai-2020	190	2	libraries	library	NNS
kim-ai-2020	190	3	are	be	VBP
kim-ai-2020	190	4	applying	apply	VBG
kim-ai-2020	190	5	AI	AI	NNP
kim-ai-2020	190	6	to	to	IN
kim-ai-2020	190	7	extracting	extract	VBG
kim-ai-2020	190	8	information	information	NN
kim-ai-2020	190	9	and	and	CC
kim-ai-2020	190	10	automating	automate	VBG
kim-ai-2020	190	11	metadata	metadata	NN
kim-ai-2020	190	12	generation	generation	NN
kim-ai-2020	190	13	for	for	IN
kim-ai-2020	190	14	digital	digital	JJ
kim-ai-2020	190	15	materials	material	NNS
kim-ai-2020	190	16	to	to	TO
kim-ai-2020	190	17	enhance	enhance	VB
kim-ai-2020	190	18	their	-PRON-	PRP$
kim-ai-2020	190	19	discovery	discovery	NN
kim-ai-2020	190	20	and	and	CC
kim-ai-2020	190	21	use	use	NN
kim-ai-2020	190	22	.	.	.
kim-ai-2020	191	1	The	the	DT
kim-ai-2020	191	2	Library	Library	NNP
kim-ai-2020	191	3	of	of	IN
kim-ai-2020	191	4	Congress	Congress	NNP
kim-ai-2020	191	5	has	have	VBZ
kim-ai-2020	191	6	worked	work	VBN
kim-ai-2020	191	7	on	on	IN
kim-ai-2020	191	8	detecting	detect	VBG
kim-ai-2020	191	9	features	feature	NNS
kim-ai-2020	191	10	,	,	,
kim-ai-2020	191	11	such	such	JJ
kim-ai-2020	191	12	as	as	IN
kim-ai-2020	191	13	railroads	railroad	NNS
kim-ai-2020	191	14	in	in	IN
kim-ai-2020	191	15	maps	map	NNS
kim-ai-2020	191	16	,	,	,
kim-ai-2020	191	17	using	use	VBG
kim-ai-2020	191	18	the	the	DT
kim-ai-2020	191	19	convolutional	convolutional	JJ
kim-ai-2020	191	20	neural	neural	JJ
kim-ai-2020	191	21	network	network	NN
kim-ai-2020	191	22	model	model	NN
kim-ai-2020	191	23	,	,	,
kim-ai-2020	191	24	and	and	CC
kim-ai-2020	191	25	issued	issue	VBD
kim-ai-2020	191	26	a	a	DT
kim-ai-2020	191	27	solicitation	solicitation	NN
kim-ai-2020	191	28	for	for	IN
kim-ai-2020	191	29	a	a	DT
kim-ai-2020	191	30	machine	machine	NN
kim-ai-2020	191	31	learning	learning	NN
kim-ai-2020	191	32	and	and	CC
kim-ai-2020	191	33	deep	deep	JJ
kim-ai-2020	191	34	learning	learning	NN
kim-ai-2020	191	35	pilot	pilot	NN
kim-ai-2020	191	36	program	program	NN
kim-ai-2020	191	37	that	that	WDT
kim-ai-2020	191	38	will	will	MD
kim-ai-2020	191	39	maximize	maximize	VB
kim-ai-2020	191	40	the	the	DT
kim-ai-2020	191	41	use	use	NN
kim-ai-2020	191	42	of	of	IN
kim-ai-2020	191	43	its	-PRON-	PRP$
kim-ai-2020	191	44	digital	digital	JJ
kim-ai-2020	191	45	collections	collection	NNS
kim-ai-2020	191	46	in	in	IN
kim-ai-2020	191	47	2019	2019	CD
kim-ai-2020	191	48	.	.	.
kim-ai-2020	192	1	[	[	-LRB-
kim-ai-2020	192	2	footnoteRef:18	footnoteRef:18	NNP
kim-ai-2020	192	3	]	]	-RRB-
kim-ai-2020	192	4	Indiana	Indiana	NNP
kim-ai-2020	192	5	University	University	NNP
kim-ai-2020	192	6	Libraries	Libraries	NNPS
kim-ai-2020	192	7	,	,	,
kim-ai-2020	192	8	AVP	AVP	NNP
kim-ai-2020	192	9	,	,	,
kim-ai-2020	192	10	University	University	NNP
kim-ai-2020	192	11	of	of	IN
kim-ai-2020	192	12	Texas	Texas	NNP
kim-ai-2020	192	13	Austin	Austin	NNP
kim-ai-2020	192	14	School	School	NNP
kim-ai-2020	192	15	of	of	IN
kim-ai-2020	192	16	Information	Information	NNP
kim-ai-2020	192	17	,	,	,
kim-ai-2020	192	18	and	and	CC
kim-ai-2020	192	19	the	the	DT
kim-ai-2020	192	20	New	New	NNP
kim-ai-2020	192	21	York	York	NNP
kim-ai-2020	192	22	Public	Public	NNP
kim-ai-2020	192	23	Library	Library	NNP
kim-ai-2020	192	24	are	be	VBP
kim-ai-2020	192	25	jointly	jointly	RB
kim-ai-2020	192	26	developing	develop	VBG
kim-ai-2020	192	27	the	the	DT
kim-ai-2020	192	28	Audiovisual	Audiovisual	NNP
kim-ai-2020	192	29	Metadata	Metadata	NNP
kim-ai-2020	192	30	Platform	Platform	NNP
kim-ai-2020	192	31	(	(	-LRB-
kim-ai-2020	192	32	AMP	AMP	NNP
kim-ai-2020	192	33	)	)	-RRB-
kim-ai-2020	192	34	,	,	,
kim-ai-2020	192	35	using	use	VBG
kim-ai-2020	192	36	many	many	JJ
kim-ai-2020	192	37	AI	ai	NN
kim-ai-2020	192	38	tools	tool	NNS
kim-ai-2020	192	39	in	in	IN
kim-ai-2020	192	40	order	order	NN
kim-ai-2020	192	41	to	to	TO
kim-ai-2020	192	42	automatically	automatically	RB
kim-ai-2020	192	43	generate	generate	VB
kim-ai-2020	192	44	metadata	metadata	NN
kim-ai-2020	192	45	for	for	IN
kim-ai-2020	192	46	audiovisual	audiovisual	JJ
kim-ai-2020	192	47	materials	material	NNS
kim-ai-2020	192	48	,	,	,
kim-ai-2020	192	49	which	which	WDT
kim-ai-2020	192	50	collection	collection	NN
kim-ai-2020	192	51	managers	manager	NNS
kim-ai-2020	192	52	can	can	MD
kim-ai-2020	192	53	use	use	VB
kim-ai-2020	192	54	to	to	TO
kim-ai-2020	192	55	supplement	supplement	VB
kim-ai-2020	192	56	their	-PRON-	PRP$
kim-ai-2020	192	57	archival	archival	NN
kim-ai-2020	192	58	description	description	NN
kim-ai-2020	192	59	and	and	CC
kim-ai-2020	192	60	processing	processing	NN
kim-ai-2020	192	61	workflows	workflow	NNS
kim-ai-2020	192	62	.	.	.
kim-ai-2020	193	1	[	[	-LRB-
kim-ai-2020	193	2	footnoteRef:19	footnoteref:19	NN
kim-ai-2020	193	3	]	]	-RRB-
kim-ai-2020	193	4	[	[	-LRB-
kim-ai-2020	193	5	18	18	CD
kim-ai-2020	193	6	:	:	:
kim-ai-2020	193	7	See	see	VB
kim-ai-2020	193	8	Blewer	Blewer	NNP
kim-ai-2020	193	9	,	,	,
kim-ai-2020	193	10	Kim	Kim	NNP
kim-ai-2020	193	11	,	,	,
kim-ai-2020	193	12	and	and	CC
kim-ai-2020	193	13	Phetteplace	Phetteplace	NNP
kim-ai-2020	193	14	2018	2018	CD
kim-ai-2020	193	15	and	and	CC
kim-ai-2020	193	16	Price	Price	NNP
kim-ai-2020	193	17	2019	2019	CD
kim-ai-2020	193	18	.	.	.
kim-ai-2020	193	19	]	]	-RRB-
kim-ai-2020	194	1	[	[	-LRB-
kim-ai-2020	194	2	19	19	CD
kim-ai-2020	194	3	:	:	:
kim-ai-2020	194	4	The	the	DT
kim-ai-2020	194	5	AMP	AMP	NNP
kim-ai-2020	194	6	wiki	wiki	NN
kim-ai-2020	194	7	is	be	VBZ
kim-ai-2020	194	8	https://wiki.dlib.indiana.edu/pages/viewpage.action?pageId=531699941	https://wiki.dlib.indiana.edu/pages/viewpage.action?pageid=531699941	NN
kim-ai-2020	194	9	.	.	.
kim-ai-2020	195	1	The	the	DT
kim-ai-2020	195	2	Audiovisual	audiovisual	JJ
kim-ai-2020	195	3	Metadata	metadata	NN
kim-ai-2020	195	4	Platform	platform	NN
kim-ai-2020	195	5	Pilot	pilot	NN
kim-ai-2020	195	6	Development	Development	NNP
kim-ai-2020	195	7	(	(	-LRB-
kim-ai-2020	195	8	AMPPD	AMPPD	NNP
kim-ai-2020	195	9	)	)	-RRB-
kim-ai-2020	195	10	project	project	NN
kim-ai-2020	195	11	was	be	VBD
kim-ai-2020	195	12	presented	present	VBN
kim-ai-2020	195	13	at	at	IN
kim-ai-2020	195	14	Code4Lib	Code4Lib	NNP
kim-ai-2020	195	15	2020	2020	CD
kim-ai-2020	195	16	(	(	-LRB-
kim-ai-2020	195	17	Averkamp	Averkamp	NNP
kim-ai-2020	195	18	and	and	CC
kim-ai-2020	195	19	Hardesty	Hardesty	NNP
kim-ai-2020	195	20	2020	2020	CD
kim-ai-2020	195	21	)	)	-RRB-
kim-ai-2020	195	22	.	.	.
kim-ai-2020	195	23	]	]	-RRB-
kim-ai-2020	196	1	Some	some	DT
kim-ai-2020	196	2	libraries	library	NNS
kim-ai-2020	196	3	are	be	VBP
kim-ai-2020	196	4	also	also	RB
kim-ai-2020	196	5	testing	test	VBG
kim-ai-2020	196	6	out	out	RP
kim-ai-2020	196	7	AI	AI	NNP
kim-ai-2020	196	8	as	as	IN
kim-ai-2020	196	9	a	a	DT
kim-ai-2020	196	10	tool	tool	NN
kim-ai-2020	196	11	for	for	IN
kim-ai-2020	196	12	evaluating	evaluate	VBG
kim-ai-2020	196	13	services	service	NNS
kim-ai-2020	196	14	and	and	CC
kim-ai-2020	196	15	operations	operation	NNS
kim-ai-2020	196	16	.	.	.
kim-ai-2020	197	1	The	the	DT
kim-ai-2020	197	2	University	University	NNP
kim-ai-2020	197	3	of	of	IN
kim-ai-2020	197	4	Rochester	Rochester	NNP
kim-ai-2020	197	5	Libraries	Libraries	NNP
kim-ai-2020	197	6	applied	apply	VBD
kim-ai-2020	197	7	deep	deep	JJ
kim-ai-2020	197	8	learning	learning	NN
kim-ai-2020	197	9	to	to	IN
kim-ai-2020	197	10	the	the	DT
kim-ai-2020	197	11	library	library	NN
kim-ai-2020	197	12	’s	’s	POS
kim-ai-2020	197	13	space	space	NN
kim-ai-2020	197	14	assessment	assessment	NN
kim-ai-2020	197	15	to	to	TO
kim-ai-2020	197	16	determine	determine	VB
kim-ai-2020	197	17	the	the	DT
kim-ai-2020	197	18	optimal	optimal	JJ
kim-ai-2020	197	19	staffing	staffing	NN
kim-ai-2020	197	20	level	level	NN
kim-ai-2020	197	21	and	and	CC
kim-ai-2020	197	22	building	building	NN
kim-ai-2020	197	23	hours	hour	NNS
kim-ai-2020	197	24	.	.	.
kim-ai-2020	198	1	The	the	DT
kim-ai-2020	198	2	University	University	NNP
kim-ai-2020	198	3	of	of	IN
kim-ai-2020	198	4	Illinois	Illinois	NNP
kim-ai-2020	198	5	Urbana	Urbana	NNP
kim-ai-2020	198	6	-	-	HYPH
kim-ai-2020	198	7	Champaign	Champaign	NNP
kim-ai-2020	198	8	Libraries	Libraries	NNPS
kim-ai-2020	198	9	used	use	VBD
kim-ai-2020	198	10	machine	machine	NN
kim-ai-2020	198	11	learning	learn	VBG
kim-ai-2020	198	12	to	to	TO
kim-ai-2020	198	13	conduct	conduct	VB
kim-ai-2020	198	14	sentiment	sentiment	NN
kim-ai-2020	198	15	analysis	analysis	NN
kim-ai-2020	198	16	on	on	IN
kim-ai-2020	198	17	their	-PRON-	PRP$
kim-ai-2020	198	18	reference	reference	NN
kim-ai-2020	198	19	chat	chat	NN
kim-ai-2020	198	20	log	log	NN
kim-ai-2020	198	21	(	(	-LRB-
kim-ai-2020	198	22	Blewer	Blewer	NNP
kim-ai-2020	198	23	,	,	,
kim-ai-2020	198	24	Kim	Kim	NNP
kim-ai-2020	198	25	,	,	,
kim-ai-2020	198	26	and	and	CC
kim-ai-2020	198	27	Phetteplace	Phetteplace	NNP
kim-ai-2020	198	28	2018	2018	CD
kim-ai-2020	198	29	)	)	-RRB-
kim-ai-2020	198	30	.	.	.
kim-ai-2020	199	1	6	6	CD
kim-ai-2020	199	2	.	.	.
kim-ai-2020	200	1	Ethical	ethical	JJ
kim-ai-2020	200	2	Challenges	Challenges	NNPS
kim-ai-2020	200	3	from	from	IN
kim-ai-2020	200	4	the	the	DT
kim-ai-2020	200	5	Personalized	Personalized	NNP
kim-ai-2020	200	6	and	and	CC
kim-ai-2020	200	7	Automated	Automated	NNP
kim-ai-2020	200	8	Information	Information	NNP
kim-ai-2020	200	9	Environment	Environment	NNP
kim-ai-2020	200	10	Do	do	VBP
kim-ai-2020	200	11	these	these	DT
kim-ai-2020	200	12	current	current	JJ
kim-ai-2020	200	13	and	and	CC
kim-ai-2020	200	14	future	future	JJ
kim-ai-2020	200	15	AI	AI	NNP
kim-ai-2020	200	16	applications	application	NNS
kim-ai-2020	200	17	for	for	IN
kim-ai-2020	200	18	libraries	library	NNS
kim-ai-2020	200	19	pose	pose	VBP
kim-ai-2020	200	20	ethical	ethical	JJ
kim-ai-2020	200	21	challenges	challenge	NNS
kim-ai-2020	200	22	similar	similar	JJ
kim-ai-2020	200	23	to	to	IN
kim-ai-2020	200	24	those	those	DT
kim-ai-2020	200	25	that	that	WDT
kim-ai-2020	200	26	we	-PRON-	PRP
kim-ai-2020	200	27	discussed	discuss	VBD
kim-ai-2020	200	28	earlier	early	RBR
kim-ai-2020	200	29	?	?	.
kim-ai-2020	201	1	Since	since	IN
kim-ai-2020	201	2	information	information	NN
kim-ai-2020	201	3	query	query	NN
kim-ai-2020	201	4	,	,	,
kim-ai-2020	201	5	discovery	discovery	NN
kim-ai-2020	201	6	,	,	,
kim-ai-2020	201	7	and	and	CC
kim-ai-2020	201	8	retrieval	retrieval	NN
kim-ai-2020	201	9	rarely	rarely	RB
kim-ai-2020	201	10	involve	involve	VBP
kim-ai-2020	201	11	life	life	NN
kim-ai-2020	201	12	-	-	HYPH
kim-ai-2020	201	13	or	or	CC
kim-ai-2020	201	14	-	-	HYPH
kim-ai-2020	201	15	death	death	NN
kim-ai-2020	201	16	situations	situation	NNS
kim-ai-2020	201	17	,	,	,
kim-ai-2020	201	18	stakes	stake	NNS
kim-ai-2020	201	19	seem	seem	VBP
kim-ai-2020	201	20	to	to	TO
kim-ai-2020	201	21	be	be	VB
kim-ai-2020	201	22	certainly	certainly	RB
kim-ai-2020	201	23	lower	low	JJR
kim-ai-2020	201	24	.	.	.
kim-ai-2020	202	1	But	but	CC
kim-ai-2020	202	2	an	an	DT
kim-ai-2020	202	3	AI	ai	NN
kim-ai-2020	202	4	-	-	HYPH
kim-ai-2020	202	5	driven	drive	VBN
kim-ai-2020	202	6	automated	automated	JJ
kim-ai-2020	202	7	information	information	NN
kim-ai-2020	202	8	environment	environment	NN
kim-ai-2020	202	9	does	do	VBZ
kim-ai-2020	202	10	raise	raise	VB
kim-ai-2020	202	11	its	-PRON-	PRP$
kim-ai-2020	202	12	own	own	JJ
kim-ai-2020	202	13	distinct	distinct	JJ
kim-ai-2020	202	14	ethical	ethical	JJ
kim-ai-2020	202	15	challenges	challenge	NNS
kim-ai-2020	202	16	.	.	.
kim-ai-2020	203	1	(	(	-LRB-
kim-ai-2020	203	2	i	i	NN
kim-ai-2020	203	3	)	)	-RRB-
kim-ai-2020	203	4	Intellectual	intellectual	JJ
kim-ai-2020	203	5	isolation	isolation	NN
kim-ai-2020	203	6	and	and	CC
kim-ai-2020	203	7	bigotry	bigotry	NN
kim-ai-2020	203	8	hampering	hamper	VBG
kim-ai-2020	203	9	civic	civic	JJ
kim-ai-2020	203	10	discourse	discourse	NN
kim-ai-2020	203	11	Many	many	JJ
kim-ai-2020	203	12	AI	AI	NNP
kim-ai-2020	203	13	applications	application	NNS
kim-ai-2020	203	14	that	that	WDT
kim-ai-2020	203	15	assist	assist	VBP
kim-ai-2020	203	16	with	with	IN
kim-ai-2020	203	17	information	information	NN
kim-ai-2020	203	18	seeking	seek	VBG
kim-ai-2020	203	19	activities	activity	NNS
kim-ai-2020	203	20	promise	promise	VBP
kim-ai-2020	203	21	a	a	DT
kim-ai-2020	203	22	higher	high	JJR
kim-ai-2020	203	23	level	level	NN
kim-ai-2020	203	24	of	of	IN
kim-ai-2020	203	25	personalization	personalization	NN
kim-ai-2020	203	26	.	.	.
kim-ai-2020	204	1	But	but	CC
kim-ai-2020	204	2	a	a	DT
kim-ai-2020	204	3	highly	highly	RB
kim-ai-2020	204	4	personalized	personalized	JJ
kim-ai-2020	204	5	information	information	NN
kim-ai-2020	204	6	environment	environment	NN
kim-ai-2020	204	7	often	often	RB
kim-ai-2020	204	8	traps	trap	VBZ
kim-ai-2020	204	9	people	people	NNS
kim-ai-2020	204	10	in	in	IN
kim-ai-2020	204	11	their	-PRON-	PRP$
kim-ai-2020	204	12	own	own	JJ
kim-ai-2020	204	13	so	so	RB
kim-ai-2020	204	14	-	-	HYPH
kim-ai-2020	204	15	called	call	VBN
kim-ai-2020	204	16	‘	'	``
kim-ai-2020	204	17	filter	filter	NN
kim-ai-2020	204	18	bubble	bubble	NN
kim-ai-2020	204	19	,	,	,
kim-ai-2020	204	20	’	'	''
kim-ai-2020	204	21	as	as	IN
kim-ai-2020	204	22	we	-PRON-	PRP
kim-ai-2020	204	23	have	have	VBP
kim-ai-2020	204	24	been	be	VBN
kim-ai-2020	204	25	increasingly	increasingly	RB
kim-ai-2020	204	26	seeing	see	VBG
kim-ai-2020	204	27	in	in	IN
kim-ai-2020	204	28	today	today	NN
kim-ai-2020	204	29	’s	’s	POS
kim-ai-2020	204	30	social	social	JJ
kim-ai-2020	204	31	media	medium	NNS
kim-ai-2020	204	32	channels	channel	NNS
kim-ai-2020	204	33	,	,	,
kim-ai-2020	204	34	news	news	NN
kim-ai-2020	204	35	websites	website	NNS
kim-ai-2020	204	36	,	,	,
kim-ai-2020	204	37	and	and	CC
kim-ai-2020	204	38	commercial	commercial	JJ
kim-ai-2020	204	39	search	search	NN
kim-ai-2020	204	40	engines	engine	NNS
kim-ai-2020	204	41	,	,	,
kim-ai-2020	204	42	where	where	WRB
kim-ai-2020	204	43	such	such	JJ
kim-ai-2020	204	44	personalization	personalization	NN
kim-ai-2020	204	45	is	be	VBZ
kim-ai-2020	204	46	provided	provide	VBN
kim-ai-2020	204	47	by	by	IN
kim-ai-2020	204	48	machine	machine	NN
kim-ai-2020	204	49	learning	learning	NN
kim-ai-2020	204	50	and	and	CC
kim-ai-2020	204	51	deep	deep	JJ
kim-ai-2020	204	52	learning	learning	NN
kim-ai-2020	204	53	.	.	.
kim-ai-2020	205	1	[	[	-LRB-
kim-ai-2020	205	2	footnoteRef:20	footnoteRef:20	NNP
kim-ai-2020	205	3	]	]	-RRB-
kim-ai-2020	205	4	Sophisticated	Sophisticated	NNP
kim-ai-2020	205	5	AI	AI	NNP
kim-ai-2020	205	6	algorithms	algorithm	NNS
kim-ai-2020	205	7	are	be	VBP
kim-ai-2020	205	8	already	already	RB
kim-ai-2020	205	9	curating	curate	VBG
kim-ai-2020	205	10	and	and	CC
kim-ai-2020	205	11	pushing	push	VBG
kim-ai-2020	205	12	information	information	NN
kim-ai-2020	205	13	feeds	feed	VBZ
kim-ai-2020	205	14	based	base	VBN
kim-ai-2020	205	15	upon	upon	IN
kim-ai-2020	205	16	the	the	DT
kim-ai-2020	205	17	person	person	NN
kim-ai-2020	205	18	’s	’s	POS
kim-ai-2020	205	19	past	past	JJ
kim-ai-2020	205	20	search	search	NN
kim-ai-2020	205	21	and	and	CC
kim-ai-2020	205	22	click	click	VB
kim-ai-2020	205	23	behavior	behavior	NN
kim-ai-2020	205	24	.	.	.
kim-ai-2020	206	1	The	the	DT
kim-ai-2020	206	2	result	result	NN
kim-ai-2020	206	3	is	be	VBZ
kim-ai-2020	206	4	that	that	IN
kim-ai-2020	206	5	information	information	NN
kim-ai-2020	206	6	seekers	seeker	NNS
kim-ai-2020	206	7	are	be	VBP
kim-ai-2020	206	8	provided	provide	VBN
kim-ai-2020	206	9	with	with	IN
kim-ai-2020	206	10	information	information	NN
kim-ai-2020	206	11	that	that	WDT
kim-ai-2020	206	12	conforms	conform	VBZ
kim-ai-2020	206	13	and	and	CC
kim-ai-2020	206	14	reinforces	reinforce	VBZ
kim-ai-2020	206	15	their	-PRON-	PRP$
kim-ai-2020	206	16	existing	exist	VBG
kim-ai-2020	206	17	beliefs	belief	NNS
kim-ai-2020	206	18	and	and	CC
kim-ai-2020	206	19	interests	interest	NNS
kim-ai-2020	206	20	.	.	.
kim-ai-2020	207	1	Views	view	NNS
kim-ai-2020	207	2	that	that	WDT
kim-ai-2020	207	3	are	be	VBP
kim-ai-2020	207	4	novel	novel	JJ
kim-ai-2020	207	5	or	or	CC
kim-ai-2020	207	6	contrast	contrast	NN
kim-ai-2020	207	7	with	with	IN
kim-ai-2020	207	8	their	-PRON-	PRP$
kim-ai-2020	207	9	existing	exist	VBG
kim-ai-2020	207	10	beliefs	belief	NNS
kim-ai-2020	207	11	are	be	VBP
kim-ai-2020	207	12	increasingly	increasingly	RB
kim-ai-2020	207	13	suppressed	suppress	VBN
kim-ai-2020	207	14	and	and	CC
kim-ai-2020	207	15	become	become	VBN
kim-ai-2020	207	16	invisible	invisible	JJ
kim-ai-2020	207	17	without	without	IN
kim-ai-2020	207	18	them	-PRON-	PRP
kim-ai-2020	207	19	even	even	RB
kim-ai-2020	207	20	realizing	realize	VBG
kim-ai-2020	207	21	.	.	.
kim-ai-2020	208	1	[	[	-LRB-
kim-ai-2020	208	2	20	20	CD
kim-ai-2020	208	3	:	:	:
kim-ai-2020	208	4	See	see	VB
kim-ai-2020	208	5	Pariser	Pariser	NNP
kim-ai-2020	208	6	2012	2012	CD
kim-ai-2020	208	7	.	.	.
kim-ai-2020	208	8	]	]	-RRB-
kim-ai-2020	209	1	Such	such	JJ
kim-ai-2020	209	2	lack	lack	NN
kim-ai-2020	209	3	of	of	IN
kim-ai-2020	209	4	exposure	exposure	NN
kim-ai-2020	209	5	to	to	IN
kim-ai-2020	209	6	opposing	oppose	VBG
kim-ai-2020	209	7	views	view	NNS
kim-ai-2020	209	8	leads	lead	VBZ
kim-ai-2020	209	9	information	information	NN
kim-ai-2020	209	10	users	user	NNS
kim-ai-2020	209	11	to	to	IN
kim-ai-2020	209	12	intellectual	intellectual	JJ
kim-ai-2020	209	13	isolation	isolation	NN
kim-ai-2020	209	14	and	and	CC
kim-ai-2020	209	15	even	even	RB
kim-ai-2020	209	16	bigotry	bigotry	NN
kim-ai-2020	209	17	.	.	.
kim-ai-2020	210	1	Highly	highly	RB
kim-ai-2020	210	2	personalized	personalized	JJ
kim-ai-2020	210	3	information	information	NN
kim-ai-2020	210	4	environments	environment	NNS
kim-ai-2020	210	5	powered	power	VBN
kim-ai-2020	210	6	by	by	IN
kim-ai-2020	210	7	AI	AI	NNP
kim-ai-2020	210	8	can	can	MD
kim-ai-2020	210	9	actively	actively	RB
kim-ai-2020	210	10	restrict	restrict	VB
kim-ai-2020	210	11	ways	way	NNS
kim-ai-2020	210	12	in	in	IN
kim-ai-2020	210	13	which	which	WDT
kim-ai-2020	210	14	people	people	NNS
kim-ai-2020	210	15	develop	develop	VBP
kim-ai-2020	210	16	balanced	balanced	JJ
kim-ai-2020	210	17	and	and	CC
kim-ai-2020	210	18	informed	informed	JJ
kim-ai-2020	210	19	opinions	opinion	NNS
kim-ai-2020	210	20	,	,	,
kim-ai-2020	210	21	thereby	thereby	RB
kim-ai-2020	210	22	intensifying	intensify	VBG
kim-ai-2020	210	23	and	and	CC
kim-ai-2020	210	24	perpetuating	perpetuate	VBG
kim-ai-2020	210	25	social	social	JJ
kim-ai-2020	210	26	discord	discord	NN
kim-ai-2020	210	27	and	and	CC
kim-ai-2020	210	28	disrupting	disrupt	VBG
kim-ai-2020	210	29	civic	civic	JJ
kim-ai-2020	210	30	discourse	discourse	NN
kim-ai-2020	210	31	.	.	.
kim-ai-2020	211	1	Such	such	JJ
kim-ai-2020	211	2	social	social	JJ
kim-ai-2020	211	3	discord	discord	NN
kim-ai-2020	211	4	and	and	CC
kim-ai-2020	211	5	disruption	disruption	NN
kim-ai-2020	211	6	in	in	IN
kim-ai-2020	211	7	civil	civil	JJ
kim-ai-2020	211	8	discourse	discourse	NN
kim-ai-2020	211	9	is	be	VBZ
kim-ai-2020	211	10	likely	likely	JJ
kim-ai-2020	211	11	to	to	TO
kim-ai-2020	211	12	disproportionately	disproportionately	RB
kim-ai-2020	211	13	affect	affect	VB
kim-ai-2020	211	14	those	those	DT
kim-ai-2020	211	15	with	with	IN
kim-ai-2020	211	16	least	least	JJS
kim-ai-2020	211	17	privileges	privilege	NNS
kim-ai-2020	211	18	.	.	.
kim-ai-2020	212	1	In	in	IN
kim-ai-2020	212	2	this	this	DT
kim-ai-2020	212	3	sense	sense	NN
kim-ai-2020	212	4	,	,	,
kim-ai-2020	212	5	intellectual	intellectual	JJ
kim-ai-2020	212	6	isolation	isolation	NN
kim-ai-2020	212	7	and	and	CC
kim-ai-2020	212	8	bigotry	bigotry	NN
kim-ai-2020	212	9	has	have	VBZ
kim-ai-2020	212	10	a	a	DT
kim-ai-2020	212	11	distinctly	distinctly	RB
kim-ai-2020	212	12	moral	moral	JJ
kim-ai-2020	212	13	impact	impact	NN
kim-ai-2020	212	14	on	on	IN
kim-ai-2020	212	15	society	society	NN
kim-ai-2020	212	16	.	.	.
kim-ai-2020	213	1	(	(	-LRB-
kim-ai-2020	213	2	ii	ii	NNP
kim-ai-2020	213	3	)	)	-RRB-
kim-ai-2020	213	4	Weakening	Weakening	NNP
kim-ai-2020	213	5	of	of	IN
kim-ai-2020	213	6	cognitive	cognitive	JJ
kim-ai-2020	213	7	agency	agency	NN
kim-ai-2020	213	8	and	and	CC
kim-ai-2020	213	9	autonomy	autonomy	NN
kim-ai-2020	213	10	We	-PRON-	PRP
kim-ai-2020	213	11	have	have	VBP
kim-ai-2020	213	12	seen	see	VBN
kim-ai-2020	213	13	earlier	early	RBR
kim-ai-2020	213	14	that	that	IN
kim-ai-2020	213	15	AI	AI	NNP
kim-ai-2020	213	16	-	-	HYPH
kim-ai-2020	213	17	powered	power	VBN
kim-ai-2020	213	18	digital	digital	JJ
kim-ai-2020	213	19	assistants	assistant	NNS
kim-ai-2020	213	20	are	be	VBP
kim-ai-2020	213	21	likely	likely	JJ
kim-ai-2020	213	22	to	to	TO
kim-ai-2020	213	23	mediate	mediate	VB
kim-ai-2020	213	24	people	people	NNS
kim-ai-2020	213	25	’s	’s	POS
kim-ai-2020	213	26	information	information	NN
kim-ai-2020	213	27	search	search	NN
kim-ai-2020	213	28	,	,	,
kim-ai-2020	213	29	discovery	discovery	NN
kim-ai-2020	213	30	,	,	,
kim-ai-2020	213	31	and	and	CC
kim-ai-2020	213	32	retrieval	retrieval	NN
kim-ai-2020	213	33	activities	activity	NNS
kim-ai-2020	213	34	in	in	IN
kim-ai-2020	213	35	the	the	DT
kim-ai-2020	213	36	near	near	JJ
kim-ai-2020	213	37	future	future	NN
kim-ai-2020	213	38	.	.	.
kim-ai-2020	214	1	As	as	IN
kim-ai-2020	214	2	those	those	DT
kim-ai-2020	214	3	digital	digital	JJ
kim-ai-2020	214	4	assistants	assistant	NNS
kim-ai-2020	214	5	become	become	VBP
kim-ai-2020	214	6	more	more	RBR
kim-ai-2020	214	7	capable	capable	JJ
kim-ai-2020	214	8	,	,	,
kim-ai-2020	214	9	they	-PRON-	PRP
kim-ai-2020	214	10	will	will	MD
kim-ai-2020	214	11	go	go	VB
kim-ai-2020	214	12	beyond	beyond	IN
kim-ai-2020	214	13	listing	list	VBG
kim-ai-2020	214	14	available	available	JJ
kim-ai-2020	214	15	information	information	NN
kim-ai-2020	214	16	.	.	.
kim-ai-2020	215	1	They	-PRON-	PRP
kim-ai-2020	215	2	will	will	MD
kim-ai-2020	215	3	further	further	RB
kim-ai-2020	215	4	choose	choose	VB
kim-ai-2020	215	5	what	what	WP
kim-ai-2020	215	6	they	-PRON-	PRP
kim-ai-2020	215	7	deem	deem	VBP
kim-ai-2020	215	8	to	to	TO
kim-ai-2020	215	9	be	be	VB
kim-ai-2020	215	10	most	most	RBS
kim-ai-2020	215	11	relevant	relevant	JJ
kim-ai-2020	215	12	to	to	IN
kim-ai-2020	215	13	their	-PRON-	PRP$
kim-ai-2020	215	14	users	user	NNS
kim-ai-2020	215	15	and	and	CC
kim-ai-2020	215	16	proceed	proceed	VB
kim-ai-2020	215	17	to	to	TO
kim-ai-2020	215	18	recommend	recommend	VB
kim-ai-2020	215	19	or	or	CC
kim-ai-2020	215	20	autonomously	autonomously	RB
kim-ai-2020	215	21	execute	execute	VB
kim-ai-2020	215	22	the	the	DT
kim-ai-2020	215	23	best	good	JJS
kim-ai-2020	215	24	course	course	NN
kim-ai-2020	215	25	of	of	IN
kim-ai-2020	215	26	action	action	NN
kim-ai-2020	215	27	.	.	.
kim-ai-2020	216	1	[	[	-LRB-
kim-ai-2020	216	2	footnoteRef:21	footnoteRef:21	NNP
kim-ai-2020	216	3	]	]	-RRB-
kim-ai-2020	216	4	Other	other	JJ
kim-ai-2020	216	5	AI	ai	NN
kim-ai-2020	216	6	-	-	HYPH
kim-ai-2020	216	7	driven	drive	VBN
kim-ai-2020	216	8	features	feature	NNS
kim-ai-2020	216	9	,	,	,
kim-ai-2020	216	10	such	such	JJ
kim-ai-2020	216	11	as	as	IN
kim-ai-2020	216	12	extracting	extract	VBG
kim-ai-2020	216	13	key	key	JJ
kim-ai-2020	216	14	information	information	NN
kim-ai-2020	216	15	or	or	CC
kim-ai-2020	216	16	generating	generate	VBG
kim-ai-2020	216	17	a	a	DT
kim-ai-2020	216	18	summary	summary	NN
kim-ai-2020	216	19	of	of	IN
kim-ai-2020	216	20	a	a	DT
kim-ai-2020	216	21	large	large	JJ
kim-ai-2020	216	22	amount	amount	NN
kim-ai-2020	216	23	of	of	IN
kim-ai-2020	216	24	information	information	NN
kim-ai-2020	216	25	,	,	,
kim-ai-2020	216	26	are	be	VBP
kim-ai-2020	216	27	also	also	RB
kim-ai-2020	216	28	likely	likely	JJ
kim-ai-2020	216	29	to	to	TO
kim-ai-2020	216	30	be	be	VB
kim-ai-2020	216	31	included	include	VBN
kim-ai-2020	216	32	in	in	IN
kim-ai-2020	216	33	future	future	JJ
kim-ai-2020	216	34	information	information	NN
kim-ai-2020	216	35	systems	system	NNS
kim-ai-2020	216	36	,	,	,
kim-ai-2020	216	37	and	and	CC
kim-ai-2020	216	38	they	-PRON-	PRP
kim-ai-2020	216	39	may	may	MD
kim-ai-2020	216	40	deliver	deliver	VB
kim-ai-2020	216	41	key	key	JJ
kim-ai-2020	216	42	information	information	NN
kim-ai-2020	216	43	or	or	CC
kim-ai-2020	216	44	summaries	summary	NNS
kim-ai-2020	216	45	even	even	RB
kim-ai-2020	216	46	before	before	IN
kim-ai-2020	216	47	the	the	DT
kim-ai-2020	216	48	request	request	NN
kim-ai-2020	216	49	is	be	VBZ
kim-ai-2020	216	50	made	make	VBN
kim-ai-2020	216	51	,	,	,
kim-ai-2020	216	52	based	base	VBN
kim-ai-2020	216	53	upon	upon	IN
kim-ai-2020	216	54	constant	constant	JJ
kim-ai-2020	216	55	monitoring	monitoring	NN
kim-ai-2020	216	56	of	of	IN
kim-ai-2020	216	57	the	the	DT
kim-ai-2020	216	58	user	user	NN
kim-ai-2020	216	59	’s	’s	POS
kim-ai-2020	216	60	activities	activity	NNS
kim-ai-2020	216	61	.	.	.
kim-ai-2020	217	1	[	[	-LRB-
kim-ai-2020	217	2	21	21	CD
kim-ai-2020	217	3	:	:	:
kim-ai-2020	217	4	Please	please	UH
kim-ai-2020	217	5	note	note	VB
kim-ai-2020	217	6	that	that	IN
kim-ai-2020	217	7	this	this	DT
kim-ai-2020	217	8	is	be	VBZ
kim-ai-2020	217	9	a	a	DT
kim-ai-2020	217	10	simplified	simplified	JJ
kim-ai-2020	217	11	scenario	scenario	NN
kim-ai-2020	217	12	.	.	.
kim-ai-2020	218	1	The	the	DT
kim-ai-2020	218	2	features	feature	NNS
kim-ai-2020	218	3	can	can	MD
kim-ai-2020	218	4	also	also	RB
kim-ai-2020	218	5	be	be	VB
kim-ai-2020	218	6	built	build	VBN
kim-ai-2020	218	7	in	in	IN
kim-ai-2020	218	8	the	the	DT
kim-ai-2020	218	9	information	information	NN
kim-ai-2020	218	10	system	system	NN
kim-ai-2020	218	11	itself	-PRON-	PRP
kim-ai-2020	218	12	rather	rather	RB
kim-ai-2020	218	13	than	than	IN
kim-ai-2020	218	14	the	the	DT
kim-ai-2020	218	15	agent	agent	NN
kim-ai-2020	218	16	.	.	.
kim-ai-2020	218	17	]	]	-RRB-
kim-ai-2020	219	1	In	in	IN
kim-ai-2020	219	2	such	such	PDT
kim-ai-2020	219	3	a	a	DT
kim-ai-2020	219	4	scenario	scenario	NN
kim-ai-2020	219	5	,	,	,
kim-ai-2020	219	6	an	an	DT
kim-ai-2020	219	7	information	information	NN
kim-ai-2020	219	8	seeker	seeker	NN
kim-ai-2020	219	9	’s	’s	POS
kim-ai-2020	219	10	cognitive	cognitive	JJ
kim-ai-2020	219	11	agency	agency	NN
kim-ai-2020	219	12	is	be	VBZ
kim-ai-2020	219	13	likely	likely	JJ
kim-ai-2020	219	14	to	to	TO
kim-ai-2020	219	15	suffer	suffer	VB
kim-ai-2020	219	16	a	a	DT
kim-ai-2020	219	17	loss	loss	NN
kim-ai-2020	219	18	.	.	.
kim-ai-2020	220	1	Crucial	crucial	JJ
kim-ai-2020	220	2	to	to	IN
kim-ai-2020	220	3	cognitive	cognitive	JJ
kim-ai-2020	220	4	agency	agency	NN
kim-ai-2020	220	5	is	be	VBZ
kim-ai-2020	220	6	the	the	DT
kim-ai-2020	220	7	mental	mental	JJ
kim-ai-2020	220	8	capacity	capacity	NN
kim-ai-2020	220	9	to	to	TO
kim-ai-2020	220	10	critically	critically	RB
kim-ai-2020	220	11	review	review	VB
kim-ai-2020	220	12	a	a	DT
kim-ai-2020	220	13	variety	variety	NN
kim-ai-2020	220	14	of	of	IN
kim-ai-2020	220	15	information	information	NN
kim-ai-2020	220	16	,	,	,
kim-ai-2020	220	17	judge	judge	VBP
kim-ai-2020	220	18	what	what	WP
kim-ai-2020	220	19	is	be	VBZ
kim-ai-2020	220	20	and	and	CC
kim-ai-2020	220	21	is	be	VBZ
kim-ai-2020	220	22	not	not	RB
kim-ai-2020	220	23	relevant	relevant	JJ
kim-ai-2020	220	24	,	,	,
kim-ai-2020	220	25	and	and	CC
kim-ai-2020	220	26	interpret	interpret	VB
kim-ai-2020	220	27	how	how	WRB
kim-ai-2020	220	28	they	-PRON-	PRP
kim-ai-2020	220	29	relate	relate	VBP
kim-ai-2020	220	30	to	to	IN
kim-ai-2020	220	31	other	other	JJ
kim-ai-2020	220	32	existing	exist	VBG
kim-ai-2020	220	33	beliefs	belief	NNS
kim-ai-2020	220	34	and	and	CC
kim-ai-2020	220	35	opinions	opinion	NNS
kim-ai-2020	220	36	.	.	.
kim-ai-2020	221	1	If	if	IN
kim-ai-2020	221	2	AI	AI	NNP
kim-ai-2020	221	3	assumes	assume	VBZ
kim-ai-2020	221	4	those	those	DT
kim-ai-2020	221	5	tasks	task	NNS
kim-ai-2020	221	6	,	,	,
kim-ai-2020	221	7	the	the	DT
kim-ai-2020	221	8	opportunities	opportunity	NNS
kim-ai-2020	221	9	for	for	IN
kim-ai-2020	221	10	information	information	NN
kim-ai-2020	221	11	seekers	seeker	NNS
kim-ai-2020	221	12	to	to	TO
kim-ai-2020	221	13	exercise	exercise	VB
kim-ai-2020	221	14	their	-PRON-	PRP$
kim-ai-2020	221	15	own	own	JJ
kim-ai-2020	221	16	cognitive	cognitive	JJ
kim-ai-2020	221	17	agency	agency	NN
kim-ai-2020	221	18	will	will	MD
kim-ai-2020	221	19	surely	surely	RB
kim-ai-2020	221	20	decrease	decrease	VB
kim-ai-2020	221	21	.	.	.
kim-ai-2020	222	1	Cognitive	cognitive	JJ
kim-ai-2020	222	2	deskilling	deskilling	NN
kim-ai-2020	222	3	and	and	CC
kim-ai-2020	222	4	the	the	DT
kim-ai-2020	222	5	subsequent	subsequent	JJ
kim-ai-2020	222	6	weakening	weakening	NN
kim-ai-2020	222	7	of	of	IN
kim-ai-2020	222	8	people	people	NNS
kim-ai-2020	222	9	’s	’s	POS
kim-ai-2020	222	10	agency	agency	NN
kim-ai-2020	222	11	in	in	IN
kim-ai-2020	222	12	the	the	DT
kim-ai-2020	222	13	AI	AI	NNP
kim-ai-2020	222	14	-powered	-powere	VBN
kim-ai-2020	222	15	automated	automate	VBN
kim-ai-2020	222	16	information	information	NN
kim-ai-2020	222	17	environment	environment	NN
kim-ai-2020	222	18	presents	present	VBZ
kim-ai-2020	222	19	an	an	DT
kim-ai-2020	222	20	ethical	ethical	JJ
kim-ai-2020	222	21	challenge	challenge	NN
kim-ai-2020	222	22	because	because	IN
kim-ai-2020	222	23	such	such	JJ
kim-ai-2020	222	24	agency	agency	NN
kim-ai-2020	222	25	is	be	VBZ
kim-ai-2020	222	26	necessary	necessary	JJ
kim-ai-2020	222	27	for	for	IN
kim-ai-2020	222	28	a	a	DT
kim-ai-2020	222	29	person	person	NN
kim-ai-2020	222	30	to	to	TO
kim-ai-2020	222	31	be	be	VB
kim-ai-2020	222	32	a	a	DT
kim-ai-2020	222	33	fully	fully	RB
kim-ai-2020	222	34	functioning	function	VBG
kim-ai-2020	222	35	moral	moral	JJ
kim-ai-2020	222	36	agent	agent	NN
kim-ai-2020	222	37	in	in	IN
kim-ai-2020	222	38	society	society	NN
kim-ai-2020	222	39	.	.	.
kim-ai-2020	223	1	[	[	-LRB-
kim-ai-2020	223	2	footnoteRef:22	footnoteref:22	NN
kim-ai-2020	223	3	]	]	-RRB-
kim-ai-2020	223	4	[	[	-LRB-
kim-ai-2020	223	5	22	22	CD
kim-ai-2020	223	6	:	:	:
kim-ai-2020	223	7	Outside	outside	IN
kim-ai-2020	223	8	of	of	IN
kim-ai-2020	223	9	the	the	DT
kim-ai-2020	223	10	automated	automated	JJ
kim-ai-2020	223	11	information	information	NN
kim-ai-2020	223	12	environment	environment	NN
kim-ai-2020	223	13	,	,	,
kim-ai-2020	223	14	AI	AI	NNP
kim-ai-2020	223	15	has	have	VBZ
kim-ai-2020	223	16	a	a	DT
kim-ai-2020	223	17	strong	strong	JJ
kim-ai-2020	223	18	potential	potential	NN
kim-ai-2020	223	19	to	to	TO
kim-ai-2020	223	20	engender	engender	VB
kim-ai-2020	223	21	moral	moral	JJ
kim-ai-2020	223	22	deskilling	deskilling	NN
kim-ai-2020	223	23	.	.	.
kim-ai-2020	224	1	Vallor	Vallor	NNP
kim-ai-2020	224	2	(	(	-LRB-
kim-ai-2020	224	3	2015	2015	CD
kim-ai-2020	224	4	)	)	-RRB-
kim-ai-2020	224	5	points	point	VBZ
kim-ai-2020	224	6	out	out	RP
kim-ai-2020	224	7	that	that	IN
kim-ai-2020	224	8	automated	automate	VBN
kim-ai-2020	224	9	weapons	weapon	NNS
kim-ai-2020	224	10	will	will	MD
kim-ai-2020	224	11	lead	lead	VB
kim-ai-2020	224	12	to	to	IN
kim-ai-2020	224	13	soldiers	soldier	NNS
kim-ai-2020	224	14	’	’	POS
kim-ai-2020	224	15	moral	moral	JJ
kim-ai-2020	224	16	deskilling	deskilling	NN
kim-ai-2020	224	17	in	in	IN
kim-ai-2020	224	18	the	the	DT
kim-ai-2020	224	19	use	use	NN
kim-ai-2020	224	20	of	of	IN
kim-ai-2020	224	21	military	military	JJ
kim-ai-2020	224	22	force	force	NN
kim-ai-2020	224	23	;	;	:
kim-ai-2020	224	24	new	new	JJ
kim-ai-2020	224	25	media	medium	NNS
kim-ai-2020	224	26	practices	practice	NNS
kim-ai-2020	224	27	of	of	IN
kim-ai-2020	224	28	multitasking	multitasking	NN
kim-ai-2020	224	29	may	may	MD
kim-ai-2020	224	30	result	result	VB
kim-ai-2020	224	31	in	in	IN
kim-ai-2020	224	32	deskilling	deskille	VBG
kim-ai-2020	224	33	in	in	IN
kim-ai-2020	224	34	moral	moral	JJ
kim-ai-2020	224	35	attention	attention	NN
kim-ai-2020	224	36	;	;	:
kim-ai-2020	224	37	and	and	CC
kim-ai-2020	224	38	social	social	JJ
kim-ai-2020	224	39	robots	robot	NNS
kim-ai-2020	224	40	can	can	MD
kim-ai-2020	224	41	cause	cause	VB
kim-ai-2020	224	42	moral	moral	JJ
kim-ai-2020	224	43	deskilling	deskilling	NN
kim-ai-2020	224	44	in	in	IN
kim-ai-2020	224	45	practices	practice	NNS
kim-ai-2020	224	46	of	of	IN
kim-ai-2020	224	47	human	human	JJ
kim-ai-2020	224	48	caregiving	caregiving	NN
kim-ai-2020	224	49	.	.	.
kim-ai-2020	224	50	]	]	-RRB-
kim-ai-2020	225	1	(	(	-LRB-
kim-ai-2020	225	2	iii	iii	NNP
kim-ai-2020	225	3	)	)	-RRB-
kim-ai-2020	225	4	Social	social	JJ
kim-ai-2020	225	5	impact	impact	NN
kim-ai-2020	225	6	of	of	IN
kim-ai-2020	225	7	scholarship	scholarship	NN
kim-ai-2020	225	8	and	and	CC
kim-ai-2020	225	9	research	research	NN
kim-ai-2020	225	10	from	from	IN
kim-ai-2020	225	11	flawed	flawed	JJ
kim-ai-2020	225	12	AI	AI	NNP
kim-ai-2020	225	13	algorithms	algorithm	NNS
kim-ai-2020	225	14	Previously	previously	RB
kim-ai-2020	225	15	,	,	,
kim-ai-2020	225	16	we	-PRON-	PRP
kim-ai-2020	225	17	have	have	VBP
kim-ai-2020	225	18	seen	see	VBN
kim-ai-2020	225	19	that	that	IN
kim-ai-2020	225	20	deep	deep	JJ
kim-ai-2020	225	21	learning	learning	NN
kim-ai-2020	225	22	applications	application	NNS
kim-ai-2020	225	23	are	be	VBP
kim-ai-2020	225	24	opaque	opaque	JJ
kim-ai-2020	225	25	to	to	IN
kim-ai-2020	225	26	human	human	JJ
kim-ai-2020	225	27	understanding	understanding	NN
kim-ai-2020	225	28	.	.	.
kim-ai-2020	226	1	This	this	DT
kim-ai-2020	226	2	lack	lack	NN
kim-ai-2020	226	3	of	of	IN
kim-ai-2020	226	4	transparency	transparency	NN
kim-ai-2020	226	5	and	and	CC
kim-ai-2020	226	6	explainability	explainability	NN
kim-ai-2020	226	7	raises	raise	VBZ
kim-ai-2020	226	8	the	the	DT
kim-ai-2020	226	9	question	question	NN
kim-ai-2020	226	10	of	of	IN
kim-ai-2020	226	11	whether	whether	IN
kim-ai-2020	226	12	it	-PRON-	PRP
kim-ai-2020	226	13	is	be	VBZ
kim-ai-2020	226	14	moral	moral	JJ
kim-ai-2020	226	15	to	to	TO
kim-ai-2020	226	16	rely	rely	VB
kim-ai-2020	226	17	on	on	IN
kim-ai-2020	226	18	AI	AI	NNP
kim-ai-2020	226	19	-	-	HYPH
kim-ai-2020	226	20	powered	power	VBN
kim-ai-2020	226	21	military	military	JJ
kim-ai-2020	226	22	robots	robot	NNS
kim-ai-2020	226	23	for	for	IN
kim-ai-2020	226	24	life	life	NN
kim-ai-2020	226	25	-	-	HYPH
kim-ai-2020	226	26	or	or	CC
kim-ai-2020	226	27	-	-	HYPH
kim-ai-2020	226	28	death	death	NN
kim-ai-2020	226	29	decisions	decision	NNS
kim-ai-2020	226	30	.	.	.
kim-ai-2020	227	1	Does	do	VBZ
kim-ai-2020	227	2	the	the	DT
kim-ai-2020	227	3	AI	ai	NN
kim-ai-2020	227	4	-	-	HYPH
kim-ai-2020	227	5	powered	power	VBN
kim-ai-2020	227	6	information	information	NN
kim-ai-2020	227	7	environment	environment	NN
kim-ai-2020	227	8	have	have	VBP
kim-ai-2020	227	9	a	a	DT
kim-ai-2020	227	10	similar	similar	JJ
kim-ai-2020	227	11	problem	problem	NN
kim-ai-2020	227	12	?	?	.
kim-ai-2020	228	1	As	as	IN
kim-ai-2020	228	2	explained	explain	VBN
kim-ai-2020	228	3	in	in	IN
kim-ai-2020	228	4	the	the	DT
kim-ai-2020	228	5	beginning	beginning	NN
kim-ai-2020	228	6	of	of	IN
kim-ai-2020	228	7	this	this	DT
kim-ai-2020	228	8	chapter	chapter	NN
kim-ai-2020	228	9	,	,	,
kim-ai-2020	228	10	machine	machine	NN
kim-ai-2020	228	11	learning	learning	NN
kim-ai-2020	228	12	applications	application	NNS
kim-ai-2020	228	13	base	base	VBP
kim-ai-2020	228	14	their	-PRON-	PRP$
kim-ai-2020	228	15	recommendations	recommendation	NNS
kim-ai-2020	228	16	and	and	CC
kim-ai-2020	228	17	predictions	prediction	NNS
kim-ai-2020	228	18	upon	upon	IN
kim-ai-2020	228	19	the	the	DT
kim-ai-2020	228	20	patterns	pattern	NNS
kim-ai-2020	228	21	in	in	IN
kim-ai-2020	228	22	past	past	JJ
kim-ai-2020	228	23	data	datum	NNS
kim-ai-2020	228	24	.	.	.
kim-ai-2020	229	1	Their	-PRON-	PRP$
kim-ai-2020	229	2	predictions	prediction	NNS
kim-ai-2020	229	3	and	and	CC
kim-ai-2020	229	4	recommendations	recommendation	NNS
kim-ai-2020	229	5	are	be	VBP
kim-ai-2020	229	6	in	in	IN
kim-ai-2020	229	7	this	this	DT
kim-ai-2020	229	8	sense	sense	NN
kim-ai-2020	229	9	inherently	inherently	RB
kim-ai-2020	229	10	conservative	conservative	JJ
kim-ai-2020	229	11	and	and	CC
kim-ai-2020	229	12	become	become	VBP
kim-ai-2020	229	13	outdated	outdated	JJ
kim-ai-2020	229	14	when	when	WRB
kim-ai-2020	229	15	they	-PRON-	PRP
kim-ai-2020	229	16	fail	fail	VBP
kim-ai-2020	229	17	to	to	TO
kim-ai-2020	229	18	reflect	reflect	VB
kim-ai-2020	229	19	the	the	DT
kim-ai-2020	229	20	new	new	JJ
kim-ai-2020	229	21	social	social	JJ
kim-ai-2020	229	22	views	view	NNS
kim-ai-2020	229	23	and	and	CC
kim-ai-2020	229	24	material	material	NN
kim-ai-2020	229	25	conditions	condition	NNS
kim-ai-2020	229	26	that	that	WDT
kim-ai-2020	229	27	no	no	RB
kim-ai-2020	229	28	longer	long	RBR
kim-ai-2020	229	29	fit	fit	VBP
kim-ai-2020	229	30	the	the	DT
kim-ai-2020	229	31	past	past	JJ
kim-ai-2020	229	32	patterns	pattern	NNS
kim-ai-2020	229	33	.	.	.
kim-ai-2020	230	1	Furthermore	furthermore	RB
kim-ai-2020	230	2	,	,	,
kim-ai-2020	230	3	each	each	DT
kim-ai-2020	230	4	data	data	NN
kim-ai-2020	230	5	set	set	NN
kim-ai-2020	230	6	is	be	VBZ
kim-ai-2020	230	7	a	a	DT
kim-ai-2020	230	8	social	social	JJ
kim-ai-2020	230	9	construct	construct	NN
kim-ai-2020	230	10	that	that	WDT
kim-ai-2020	230	11	reflects	reflect	VBZ
kim-ai-2020	230	12	particular	particular	JJ
kim-ai-2020	230	13	values	value	NNS
kim-ai-2020	230	14	and	and	CC
kim-ai-2020	230	15	choices	choice	NNS
kim-ai-2020	230	16	such	such	JJ
kim-ai-2020	230	17	as	as	IN
kim-ai-2020	230	18	who	who	WP
kim-ai-2020	230	19	decided	decide	VBD
kim-ai-2020	230	20	to	to	TO
kim-ai-2020	230	21	collect	collect	VB
kim-ai-2020	230	22	the	the	DT
kim-ai-2020	230	23	data	datum	NNS
kim-ai-2020	230	24	and	and	CC
kim-ai-2020	230	25	for	for	IN
kim-ai-2020	230	26	what	what	WDT
kim-ai-2020	230	27	purpose	purpose	NN
kim-ai-2020	230	28	;	;	:
kim-ai-2020	230	29	who	who	WP
kim-ai-2020	230	30	labeled	label	VBD
kim-ai-2020	230	31	data	datum	NNS
kim-ai-2020	230	32	;	;	:
kim-ai-2020	230	33	what	what	WDT
kim-ai-2020	230	34	criteria	criterion	NNS
kim-ai-2020	230	35	or	or	CC
kim-ai-2020	230	36	beliefs	belief	NNS
kim-ai-2020	230	37	guided	guide	VBD
kim-ai-2020	230	38	such	such	JJ
kim-ai-2020	230	39	labeling	labeling	NN
kim-ai-2020	230	40	;	;	:
kim-ai-2020	230	41	what	what	WP
kim-ai-2020	230	42	taxonomies	taxonomy	NNS
kim-ai-2020	230	43	were	be	VBD
kim-ai-2020	230	44	used	use	VBN
kim-ai-2020	230	45	and	and	CC
kim-ai-2020	230	46	why	why	WRB
kim-ai-2020	230	47	(	(	-LRB-
kim-ai-2020	230	48	Davis	Davis	NNP
kim-ai-2020	230	49	2020	2020	CD
kim-ai-2020	230	50	)	)	-RRB-
kim-ai-2020	230	51	.	.	.
kim-ai-2020	231	1	No	no	DT
kim-ai-2020	231	2	data	data	NN
kim-ai-2020	231	3	set	set	NN
kim-ai-2020	231	4	can	can	MD
kim-ai-2020	231	5	capture	capture	VB
kim-ai-2020	231	6	all	all	PDT
kim-ai-2020	231	7	the	the	DT
kim-ai-2020	231	8	variables	variable	NNS
kim-ai-2020	231	9	and	and	CC
kim-ai-2020	231	10	elements	element	NNS
kim-ai-2020	231	11	of	of	IN
kim-ai-2020	231	12	the	the	DT
kim-ai-2020	231	13	phenomenon	phenomenon	NN
kim-ai-2020	231	14	that	that	WDT
kim-ai-2020	231	15	it	-PRON-	PRP
kim-ai-2020	231	16	describes	describe	VBZ
kim-ai-2020	231	17	.	.	.
kim-ai-2020	232	1	Furthermore	furthermore	RB
kim-ai-2020	232	2	,	,	,
kim-ai-2020	232	3	data	data	NN
kim-ai-2020	232	4	sets	set	NNS
kim-ai-2020	232	5	used	use	VBN
kim-ai-2020	232	6	for	for	IN
kim-ai-2020	232	7	training	train	VBG
kim-ai-2020	232	8	machine	machine	NN
kim-ai-2020	232	9	learning	learning	NN
kim-ai-2020	232	10	and	and	CC
kim-ai-2020	232	11	deep	deep	JJ
kim-ai-2020	232	12	learning	learning	NN
kim-ai-2020	232	13	algorithms	algorithm	NNS
kim-ai-2020	232	14	may	may	MD
kim-ai-2020	232	15	not	not	RB
kim-ai-2020	232	16	be	be	VB
kim-ai-2020	232	17	representational	representational	JJ
kim-ai-2020	232	18	samples	sample	NNS
kim-ai-2020	232	19	for	for	IN
kim-ai-2020	232	20	all	all	DT
kim-ai-2020	232	21	relevant	relevant	JJ
kim-ai-2020	232	22	subgroups	subgroup	NNS
kim-ai-2020	232	23	.	.	.
kim-ai-2020	233	1	In	in	IN
kim-ai-2020	233	2	such	such	PDT
kim-ai-2020	233	3	a	a	DT
kim-ai-2020	233	4	case	case	NN
kim-ai-2020	233	5	,	,	,
kim-ai-2020	233	6	an	an	DT
kim-ai-2020	233	7	algorithm	algorithm	NN
kim-ai-2020	233	8	will	will	MD
kim-ai-2020	233	9	not	not	RB
kim-ai-2020	233	10	be	be	VB
kim-ai-2020	233	11	able	able	JJ
kim-ai-2020	233	12	to	to	TO
kim-ai-2020	233	13	perform	perform	VB
kim-ai-2020	233	14	equally	equally	RB
kim-ai-2020	233	15	well	well	RB
kim-ai-2020	233	16	in	in	IN
kim-ai-2020	233	17	all	all	DT
kim-ai-2020	233	18	subgroups	subgroup	NNS
kim-ai-2020	233	19	.	.	.
kim-ai-2020	234	1	Creating	create	VBG
kim-ai-2020	234	2	a	a	DT
kim-ai-2020	234	3	large	large	JJ
kim-ai-2020	234	4	data	data	NN
kim-ai-2020	234	5	set	set	NN
kim-ai-2020	234	6	is	be	VBZ
kim-ai-2020	234	7	also	also	RB
kim-ai-2020	234	8	costly	costly	JJ
kim-ai-2020	234	9	.	.	.
kim-ai-2020	235	1	Consequently	consequently	RB
kim-ai-2020	235	2	,	,	,
kim-ai-2020	235	3	developers	developer	NNS
kim-ai-2020	235	4	often	often	RB
kim-ai-2020	235	5	simply	simply	RB
kim-ai-2020	235	6	take	take	VBP
kim-ai-2020	235	7	data	data	NN
kim-ai-2020	235	8	sets	set	NNS
kim-ai-2020	235	9	that	that	WDT
kim-ai-2020	235	10	are	be	VBP
kim-ai-2020	235	11	available	available	JJ
kim-ai-2020	235	12	to	to	IN
kim-ai-2020	235	13	them	-PRON-	PRP
kim-ai-2020	235	14	.	.	.
kim-ai-2020	236	1	Those	those	DT
kim-ai-2020	236	2	data	data	NN
kim-ai-2020	236	3	sets	set	NNS
kim-ai-2020	236	4	are	be	VBP
kim-ai-2020	236	5	likely	likely	JJ
kim-ai-2020	236	6	to	to	TO
kim-ai-2020	236	7	come	come	VB
kim-ai-2020	236	8	with	with	IN
kim-ai-2020	236	9	inherent	inherent	JJ
kim-ai-2020	236	10	limitations	limitation	NNS
kim-ai-2020	236	11	such	such	JJ
kim-ai-2020	236	12	as	as	IN
kim-ai-2020	236	13	omissions	omission	NNS
kim-ai-2020	236	14	,	,	,
kim-ai-2020	236	15	inaccuracies	inaccuracy	NNS
kim-ai-2020	236	16	,	,	,
kim-ai-2020	236	17	errors	error	NNS
kim-ai-2020	236	18	,	,	,
kim-ai-2020	236	19	and	and	CC
kim-ai-2020	236	20	hidden	hide	VBN
kim-ai-2020	236	21	biases	bias	NNS
kim-ai-2020	236	22	.	.	.
kim-ai-2020	237	1	AI	AI	NNP
kim-ai-2020	237	2	algorithms	algorithm	NNS
kim-ai-2020	237	3	trained	train	VBD
kim-ai-2020	237	4	with	with	IN
kim-ai-2020	237	5	these	these	DT
kim-ai-2020	237	6	flawed	flawed	JJ
kim-ai-2020	237	7	data	data	NN
kim-ai-2020	237	8	sets	set	NNS
kim-ai-2020	237	9	can	can	MD
kim-ai-2020	237	10	fail	fail	VB
kim-ai-2020	237	11	unexpectedly	unexpectedly	RB
kim-ai-2020	237	12	,	,	,
kim-ai-2020	237	13	revealing	reveal	VBG
kim-ai-2020	237	14	those	those	DT
kim-ai-2020	237	15	limitations	limitation	NNS
kim-ai-2020	237	16	in	in	IN
kim-ai-2020	237	17	the	the	DT
kim-ai-2020	237	18	data	data	NN
kim-ai-2020	237	19	sets	set	NNS
kim-ai-2020	237	20	.	.	.
kim-ai-2020	238	1	For	for	IN
kim-ai-2020	238	2	example	example	NN
kim-ai-2020	238	3	,	,	,
kim-ai-2020	238	4	it	-PRON-	PRP
kim-ai-2020	238	5	has	have	VBZ
kim-ai-2020	238	6	been	be	VBN
kim-ai-2020	238	7	reported	report	VBN
kim-ai-2020	238	8	that	that	IN
kim-ai-2020	238	9	the	the	DT
kim-ai-2020	238	10	success	success	NN
kim-ai-2020	238	11	rate	rate	NN
kim-ai-2020	238	12	of	of	IN
kim-ai-2020	238	13	a	a	DT
kim-ai-2020	238	14	facial	facial	JJ
kim-ai-2020	238	15	recognition	recognition	NN
kim-ai-2020	238	16	algorithm	algorithm	NN
kim-ai-2020	238	17	plunges	plunge	NNS
kim-ai-2020	238	18	from	from	IN
kim-ai-2020	238	19	99	99	CD
kim-ai-2020	238	20	%	%	NN
kim-ai-2020	238	21	to	to	TO
kim-ai-2020	238	22	35	35	CD
kim-ai-2020	238	23	%	%	NN
kim-ai-2020	238	24	when	when	WRB
kim-ai-2020	238	25	the	the	DT
kim-ai-2020	238	26	group	group	NN
kim-ai-2020	238	27	of	of	IN
kim-ai-2020	238	28	subjects	subject	NNS
kim-ai-2020	238	29	changes	change	NNS
kim-ai-2020	238	30	from	from	IN
kim-ai-2020	238	31	white	white	JJ
kim-ai-2020	238	32	men	man	NNS
kim-ai-2020	238	33	to	to	IN
kim-ai-2020	238	34	dark	dark	JJ
kim-ai-2020	238	35	-	-	HYPH
kim-ai-2020	238	36	skinned	skinned	JJ
kim-ai-2020	238	37	women	woman	NNS
kim-ai-2020	238	38	because	because	IN
kim-ai-2020	238	39	it	-PRON-	PRP
kim-ai-2020	238	40	was	be	VBD
kim-ai-2020	238	41	trained	train	VBN
kim-ai-2020	238	42	mostly	mostly	RB
kim-ai-2020	238	43	with	with	IN
kim-ai-2020	238	44	the	the	DT
kim-ai-2020	238	45	photographs	photograph	NNS
kim-ai-2020	238	46	of	of	IN
kim-ai-2020	238	47	white	white	JJ
kim-ai-2020	238	48	men	man	NNS
kim-ai-2020	238	49	(	(	-LRB-
kim-ai-2020	238	50	Lohr	Lohr	NNP
kim-ai-2020	238	51	2018	2018	CD
kim-ai-2020	238	52	)	)	-RRB-
kim-ai-2020	238	53	.	.	.
kim-ai-2020	239	1	Adopting	adopt	VBG
kim-ai-2020	239	2	such	such	PDT
kim-ai-2020	239	3	a	a	DT
kim-ai-2020	239	4	faulty	faulty	JJ
kim-ai-2020	239	5	algorithm	algorithm	NN
kim-ai-2020	239	6	for	for	IN
kim-ai-2020	239	7	any	any	DT
kim-ai-2020	239	8	real	real	JJ
kim-ai-2020	239	9	-	-	HYPH
kim-ai-2020	239	10	life	life	NN
kim-ai-2020	239	11	use	use	NN
kim-ai-2020	239	12	at	at	IN
kim-ai-2020	239	13	a	a	DT
kim-ai-2020	239	14	large	large	JJ
kim-ai-2020	239	15	scale	scale	NN
kim-ai-2020	239	16	would	would	MD
kim-ai-2020	239	17	be	be	VB
kim-ai-2020	239	18	entirely	entirely	RB
kim-ai-2020	239	19	unethical	unethical	JJ
kim-ai-2020	239	20	.	.	.
kim-ai-2020	240	1	For	for	IN
kim-ai-2020	240	2	the	the	DT
kim-ai-2020	240	3	context	context	NN
kim-ai-2020	240	4	of	of	IN
kim-ai-2020	240	5	libraries	library	NNS
kim-ai-2020	240	6	,	,	,
kim-ai-2020	240	7	imagine	imagine	VB
kim-ai-2020	240	8	using	use	VBG
kim-ai-2020	240	9	such	such	PDT
kim-ai-2020	240	10	a	a	DT
kim-ai-2020	240	11	face	face	NN
kim-ai-2020	240	12	-	-	HYPH
kim-ai-2020	240	13	recognition	recognition	NN
kim-ai-2020	240	14	algorithm	algorithm	NN
kim-ai-2020	240	15	to	to	TO
kim-ai-2020	240	16	generate	generate	VB
kim-ai-2020	240	17	metadata	metadata	NN
kim-ai-2020	240	18	for	for	IN
kim-ai-2020	240	19	digitized	digitized	JJ
kim-ai-2020	240	20	historical	historical	JJ
kim-ai-2020	240	21	photographs	photograph	NNS
kim-ai-2020	240	22	,	,	,
kim-ai-2020	240	23	or	or	CC
kim-ai-2020	240	24	a	a	DT
kim-ai-2020	240	25	similarly	similarly	RB
kim-ai-2020	240	26	flawed	flawed	JJ
kim-ai-2020	240	27	audio	audio	JJ
kim-ai-2020	240	28	transcription	transcription	NN
kim-ai-2020	240	29	algorithm	algorithm	NNP
kim-ai-2020	240	30	to	to	TO
kim-ai-2020	240	31	transcribe	transcribe	VB
kim-ai-2020	240	32	archival	archival	NNP
kim-ai-2020	240	33	audio	audio	NNP
kim-ai-2020	240	34	recordings	recording	NNS
kim-ai-2020	240	35	.	.	.
kim-ai-2020	241	1	Just	just	RB
kim-ai-2020	241	2	like	like	IN
kim-ai-2020	241	3	those	those	DT
kim-ai-2020	241	4	faulty	faulty	JJ
kim-ai-2020	241	5	algorithms	algorithm	NNS
kim-ai-2020	241	6	,	,	,
kim-ai-2020	241	7	an	an	DT
kim-ai-2020	241	8	AI	AI	NNP
kim-ai-2020	241	9	-	-	HYPH
kim-ai-2020	241	10	powered	power	VBN
kim-ai-2020	241	11	automated	automated	JJ
kim-ai-2020	241	12	information	information	NN
kim-ai-2020	241	13	environment	environment	NN
kim-ai-2020	241	14	can	can	MD
kim-ai-2020	241	15	provide	provide	VB
kim-ai-2020	241	16	information	information	NN
kim-ai-2020	241	17	,	,	,
kim-ai-2020	241	18	recommendations	recommendation	NNS
kim-ai-2020	241	19	,	,	,
kim-ai-2020	241	20	and	and	CC
kim-ai-2020	241	21	predictions	prediction	NNS
kim-ai-2020	241	22	affected	affect	VBN
kim-ai-2020	241	23	by	by	IN
kim-ai-2020	241	24	similar	similar	JJ
kim-ai-2020	241	25	limitations	limitation	NNS
kim-ai-2020	241	26	existing	exist	VBG
kim-ai-2020	241	27	in	in	IN
kim-ai-2020	241	28	many	many	JJ
kim-ai-2020	241	29	data	datum	NNS
kim-ai-2020	241	30	sets	set	NNS
kim-ai-2020	241	31	.	.	.
kim-ai-2020	242	1	The	the	DT
kim-ai-2020	242	2	more	more	RBR
kim-ai-2020	242	3	seamless	seamless	NN
kim-ai-2020	242	4	such	such	PDT
kim-ai-2020	242	5	an	an	DT
kim-ai-2020	242	6	information	information	NN
kim-ai-2020	242	7	environment	environment	NN
kim-ai-2020	242	8	is	be	VBZ
kim-ai-2020	242	9	,	,	,
kim-ai-2020	242	10	the	the	DT
kim-ai-2020	242	11	more	more	RBR
kim-ai-2020	242	12	invisible	invisible	JJ
kim-ai-2020	242	13	those	those	DT
kim-ai-2020	242	14	limitations	limitation	NNS
kim-ai-2020	242	15	become	become	VBP
kim-ai-2020	242	16	.	.	.
kim-ai-2020	243	1	Automated	automate	VBN
kim-ai-2020	243	2	information	information	NN
kim-ai-2020	243	3	systems	system	NNS
kim-ai-2020	243	4	from	from	IN
kim-ai-2020	243	5	libraries	library	NNS
kim-ai-2020	243	6	may	may	MD
kim-ai-2020	243	7	not	not	RB
kim-ai-2020	243	8	be	be	VB
kim-ai-2020	243	9	involved	involve	VBN
kim-ai-2020	243	10	in	in	IN
kim-ai-2020	243	11	decisions	decision	NNS
kim-ai-2020	243	12	that	that	WDT
kim-ai-2020	243	13	have	have	VBP
kim-ai-2020	243	14	a	a	DT
kim-ai-2020	243	15	direct	direct	JJ
kim-ai-2020	243	16	and	and	CC
kim-ai-2020	243	17	immediate	immediate	JJ
kim-ai-2020	243	18	impact	impact	NN
kim-ai-2020	243	19	on	on	IN
kim-ai-2020	243	20	people	people	NNS
kim-ai-2020	243	21	’s	’s	POS
kim-ai-2020	243	22	lives	life	NNS
kim-ai-2020	243	23	,	,	,
kim-ai-2020	243	24	such	such	JJ
kim-ai-2020	243	25	as	as	IN
kim-ai-2020	243	26	setting	set	VBG
kim-ai-2020	243	27	a	a	DT
kim-ai-2020	243	28	bail	bail	NN
kim-ai-2020	243	29	amount	amount	NN
kim-ai-2020	243	30	or	or	CC
kim-ai-2020	243	31	determining	determine	VBG
kim-ai-2020	243	32	the	the	DT
kim-ai-2020	243	33	Medicaid	Medicaid	NNP
kim-ai-2020	243	34	payment	payment	NN
kim-ai-2020	243	35	to	to	TO
kim-ai-2020	243	36	be	be	VB
kim-ai-2020	243	37	paid,[footnoteRef:23	paid,[footnoteref:23	CD
kim-ai-2020	243	38	]	]	-RRB-
kim-ai-2020	243	39	but	but	CC
kim-ai-2020	243	40	automated	automate	VBD
kim-ai-2020	243	41	information	information	NN
kim-ai-2020	243	42	systems	system	NNS
kim-ai-2020	243	43	that	that	WDT
kim-ai-2020	243	44	are	be	VBP
kim-ai-2020	243	45	widely	widely	RB
kim-ai-2020	243	46	adopted	adopt	VBN
kim-ai-2020	243	47	and	and	CC
kim-ai-2020	243	48	used	use	VBN
kim-ai-2020	243	49	for	for	IN
kim-ai-2020	243	50	research	research	NN
kim-ai-2020	243	51	and	and	CC
kim-ai-2020	243	52	scholarship	scholarship	NN
kim-ai-2020	243	53	will	will	MD
kim-ai-2020	243	54	impact	impact	VB
kim-ai-2020	243	55	real	real	JJ
kim-ai-2020	243	56	-	-	HYPH
kim-ai-2020	243	57	life	life	NN
kim-ai-2020	243	58	policies	policy	NNS
kim-ai-2020	243	59	and	and	CC
kim-ai-2020	243	60	regulations	regulation	NNS
kim-ai-2020	243	61	in	in	IN
kim-ai-2020	243	62	areas	area	NNS
kim-ai-2020	243	63	such	such	JJ
kim-ai-2020	243	64	as	as	IN
kim-ai-2020	243	65	healthcare	healthcare	NN
kim-ai-2020	243	66	and	and	CC
kim-ai-2020	243	67	the	the	DT
kim-ai-2020	243	68	economy	economy	NN
kim-ai-2020	243	69	.	.	.
kim-ai-2020	244	1	Undiscovered	undiscovered	JJ
kim-ai-2020	244	2	flaws	flaw	NNS
kim-ai-2020	244	3	will	will	MD
kim-ai-2020	244	4	undermine	undermine	VB
kim-ai-2020	244	5	the	the	DT
kim-ai-2020	244	6	validity	validity	NN
kim-ai-2020	244	7	of	of	IN
kim-ai-2020	244	8	the	the	DT
kim-ai-2020	244	9	scholarly	scholarly	JJ
kim-ai-2020	244	10	output	output	NN
kim-ai-2020	244	11	that	that	WDT
kim-ai-2020	244	12	utilized	utilize	VBD
kim-ai-2020	244	13	those	those	DT
kim-ai-2020	244	14	automated	automate	VBN
kim-ai-2020	244	15	information	information	NN
kim-ai-2020	244	16	systems	system	NNS
kim-ai-2020	244	17	and	and	CC
kim-ai-2020	244	18	can	can	MD
kim-ai-2020	244	19	further	far	RBR
kim-ai-2020	244	20	inflict	inflict	VB
kim-ai-2020	244	21	serious	serious	JJ
kim-ai-2020	244	22	harm	harm	NN
kim-ai-2020	244	23	on	on	IN
kim-ai-2020	244	24	certain	certain	JJ
kim-ai-2020	244	25	groups	group	NNS
kim-ai-2020	244	26	of	of	IN
kim-ai-2020	244	27	people	people	NNS
kim-ai-2020	244	28	through	through	IN
kim-ai-2020	244	29	those	those	DT
kim-ai-2020	244	30	policies	policy	NNS
kim-ai-2020	244	31	and	and	CC
kim-ai-2020	244	32	regulations	regulation	NNS
kim-ai-2020	244	33	.	.	.
kim-ai-2020	245	1	[	[	-LRB-
kim-ai-2020	245	2	23	23	CD
kim-ai-2020	245	3	:	:	:
kim-ai-2020	245	4	See	see	VB
kim-ai-2020	245	5	Tashea	Tashea	NNP
kim-ai-2020	245	6	2017	2017	CD
kim-ai-2020	245	7	and	and	CC
kim-ai-2020	245	8	Stanley	Stanley	NNP
kim-ai-2020	245	9	2017	2017	CD
kim-ai-2020	245	10	.	.	.
kim-ai-2020	245	11	]	]	-RRB-
kim-ai-2020	246	1	7	7	LS
kim-ai-2020	246	2	.	.	.
kim-ai-2020	247	1	Moral	Moral	NNP
kim-ai-2020	247	2	Intelligence	Intelligence	NNP
kim-ai-2020	247	3	and	and	CC
kim-ai-2020	247	4	Rethinking	rethink	VBG
kim-ai-2020	247	5	the	the	DT
kim-ai-2020	247	6	Role	Role	NNP
kim-ai-2020	247	7	of	of	IN
kim-ai-2020	247	8	AI	AI	NNP
kim-ai-2020	247	9	In	in	IN
kim-ai-2020	247	10	this	this	DT
kim-ai-2020	247	11	chapter	chapter	NN
kim-ai-2020	247	12	,	,	,
kim-ai-2020	247	13	I	-PRON-	PRP
kim-ai-2020	247	14	discussed	discuss	VBD
kim-ai-2020	247	15	four	four	CD
kim-ai-2020	247	16	significant	significant	JJ
kim-ai-2020	247	17	ethical	ethical	JJ
kim-ai-2020	247	18	challenges	challenge	NNS
kim-ai-2020	247	19	that	that	IN
kim-ai-2020	247	20	automating	automate	VBG
kim-ai-2020	247	21	decisions	decision	NNS
kim-ai-2020	247	22	and	and	CC
kim-ai-2020	247	23	actions	action	NNS
kim-ai-2020	247	24	with	with	IN
kim-ai-2020	247	25	AI	AI	NNP
kim-ai-2020	247	26	presents	present	NNS
kim-ai-2020	247	27	:	:	:
kim-ai-2020	247	28	(	(	-LRB-
kim-ai-2020	247	29	a	a	DT
kim-ai-2020	247	30	)	)	-RRB-
kim-ai-2020	247	31	moral	moral	JJ
kim-ai-2020	247	32	desensitization	desensitization	NN
kim-ai-2020	247	33	;	;	,
kim-ai-2020	247	34	(	(	-LRB-
kim-ai-2020	247	35	b	b	NN
kim-ai-2020	247	36	)	)	-RRB-
kim-ai-2020	247	37	unintended	unintended	JJ
kim-ai-2020	247	38	outcomes	outcome	NNS
kim-ai-2020	247	39	;	;	,
kim-ai-2020	247	40	(	(	-LRB-
kim-ai-2020	247	41	c	c	NN
kim-ai-2020	247	42	)	)	-RRB-
kim-ai-2020	247	43	surrender	surrender	NN
kim-ai-2020	247	44	of	of	IN
kim-ai-2020	247	45	moral	moral	JJ
kim-ai-2020	247	46	agency	agency	NN
kim-ai-2020	247	47	;	;	:
kim-ai-2020	247	48	(	(	-LRB-
kim-ai-2020	247	49	d	d	NN
kim-ai-2020	247	50	)	)	-RRB-
kim-ai-2020	247	51	opacity	opacity	NN
kim-ai-2020	247	52	in	in	IN
kim-ai-2020	247	53	decision	decision	NN
kim-ai-2020	247	54	-	-	HYPH
kim-ai-2020	247	55	making	making	NN
kim-ai-2020	247	56	.	.	.
kim-ai-2020	248	1	[	[	-LRB-
kim-ai-2020	248	2	footnoteRef:24	footnoteRef:24	NNP
kim-ai-2020	248	3	]	]	-RRB-
kim-ai-2020	248	4	I	-PRON-	PRP
kim-ai-2020	248	5	also	also	RB
kim-ai-2020	248	6	examined	examine	VBD
kim-ai-2020	248	7	somewhat	somewhat	RB
kim-ai-2020	248	8	different	different	JJ
kim-ai-2020	248	9	but	but	CC
kim-ai-2020	248	10	equally	equally	RB
kim-ai-2020	248	11	significant	significant	JJ
kim-ai-2020	248	12	ethical	ethical	JJ
kim-ai-2020	248	13	challenges	challenge	NNS
kim-ai-2020	248	14	in	in	IN
kim-ai-2020	248	15	relation	relation	NN
kim-ai-2020	248	16	to	to	IN
kim-ai-2020	248	17	the	the	DT
kim-ai-2020	248	18	AI	AI	NNP
kim-ai-2020	248	19	-	-	HYPH
kim-ai-2020	248	20	powered	power	VBN
kim-ai-2020	248	21	automated	automated	JJ
kim-ai-2020	248	22	information	information	NN
kim-ai-2020	248	23	environment	environment	NN
kim-ai-2020	248	24	,	,	,
kim-ai-2020	248	25	which	which	WDT
kim-ai-2020	248	26	is	be	VBZ
kim-ai-2020	248	27	likely	likely	JJ
kim-ai-2020	248	28	to	to	TO
kim-ai-2020	248	29	surround	surround	VB
kim-ai-2020	248	30	us	-PRON-	PRP
kim-ai-2020	248	31	in	in	IN
kim-ai-2020	248	32	the	the	DT
kim-ai-2020	248	33	future	future	NN
kim-ai-2020	248	34	:	:	:
kim-ai-2020	248	35	(	(	-LRB-
kim-ai-2020	248	36	i	i	NN
kim-ai-2020	248	37	)	)	-RRB-
kim-ai-2020	248	38	intellectual	intellectual	JJ
kim-ai-2020	248	39	isolation	isolation	NN
kim-ai-2020	248	40	and	and	CC
kim-ai-2020	248	41	bigotry	bigotry	NN
kim-ai-2020	248	42	hampering	hamper	VBG
kim-ai-2020	248	43	civic	civic	JJ
kim-ai-2020	248	44	discourse	discourse	NN
kim-ai-2020	248	45	;	;	:
kim-ai-2020	248	46	(	(	-LRB-
kim-ai-2020	248	47	ii	ii	NN
kim-ai-2020	248	48	)	)	-RRB-
kim-ai-2020	248	49	weakening	weaken	VBG
kim-ai-2020	248	50	of	of	IN
kim-ai-2020	248	51	cognitive	cognitive	JJ
kim-ai-2020	248	52	agency	agency	NN
kim-ai-2020	248	53	and	and	CC
kim-ai-2020	248	54	autonomy	autonomy	NN
kim-ai-2020	248	55	;	;	,
kim-ai-2020	248	56	(	(	-LRB-
kim-ai-2020	248	57	iii	iii	NNP
kim-ai-2020	248	58	)	)	-RRB-
kim-ai-2020	248	59	social	social	JJ
kim-ai-2020	248	60	impact	impact	NN
kim-ai-2020	248	61	of	of	IN
kim-ai-2020	248	62	scholarship	scholarship	NN
kim-ai-2020	248	63	and	and	CC
kim-ai-2020	248	64	research	research	NN
kim-ai-2020	248	65	based	base	VBN
kim-ai-2020	248	66	upon	upon	IN
kim-ai-2020	248	67	flawed	flawed	JJ
kim-ai-2020	248	68	AI	AI	NNP
kim-ai-2020	248	69	algorithms	algorithm	NNS
kim-ai-2020	248	70	.	.	.
kim-ai-2020	249	1	[	[	-LRB-
kim-ai-2020	249	2	24	24	CD
kim-ai-2020	249	3	:	:	:
kim-ai-2020	249	4	This	this	DT
kim-ai-2020	249	5	is	be	VBZ
kim-ai-2020	249	6	by	by	IN
kim-ai-2020	249	7	no	no	DT
kim-ai-2020	249	8	means	means	NN
kim-ai-2020	249	9	an	an	DT
kim-ai-2020	249	10	exhaustive	exhaustive	JJ
kim-ai-2020	249	11	list	list	NN
kim-ai-2020	249	12	.	.	.
kim-ai-2020	250	1	User	user	NN
kim-ai-2020	250	2	privacy	privacy	NN
kim-ai-2020	250	3	and	and	CC
kim-ai-2020	250	4	potential	potential	JJ
kim-ai-2020	250	5	surveillance	surveillance	NN
kim-ai-2020	250	6	are	be	VBP
kim-ai-2020	250	7	examples	example	NNS
kim-ai-2020	250	8	of	of	IN
kim-ai-2020	250	9	other	other	JJ
kim-ai-2020	250	10	important	important	JJ
kim-ai-2020	250	11	ethical	ethical	JJ
kim-ai-2020	250	12	challenges	challenge	NNS
kim-ai-2020	250	13	,	,	,
kim-ai-2020	250	14	which	which	WDT
kim-ai-2020	250	15	I	-PRON-	PRP
kim-ai-2020	250	16	do	do	VBP
kim-ai-2020	250	17	not	not	RB
kim-ai-2020	250	18	discuss	discuss	VB
kim-ai-2020	250	19	here	here	RB
kim-ai-2020	250	20	.	.	.
kim-ai-2020	250	21	]	]	-RRB-
kim-ai-2020	251	1	The	the	DT
kim-ai-2020	251	2	challenges	challenge	NNS
kim-ai-2020	251	3	related	relate	VBN
kim-ai-2020	251	4	to	to	IN
kim-ai-2020	251	5	the	the	DT
kim-ai-2020	251	6	AI	AI	NNP
kim-ai-2020	251	7	-	-	HYPH
kim-ai-2020	251	8	powered	power	VBN
kim-ai-2020	251	9	automated	automated	JJ
kim-ai-2020	251	10	information	information	NN
kim-ai-2020	251	11	environment	environment	NN
kim-ai-2020	251	12	are	be	VBP
kim-ai-2020	251	13	relevant	relevant	JJ
kim-ai-2020	251	14	to	to	IN
kim-ai-2020	251	15	libraries	library	NNS
kim-ai-2020	251	16	,	,	,
kim-ai-2020	251	17	which	which	WDT
kim-ai-2020	251	18	will	will	MD
kim-ai-2020	251	19	soon	soon	RB
kim-ai-2020	251	20	be	be	VB
kim-ai-2020	251	21	acquiring	acquire	VBG
kim-ai-2020	251	22	,	,	,
kim-ai-2020	251	23	building	build	VBG
kim-ai-2020	251	24	,	,	,
kim-ai-2020	251	25	customizing	customizing	NN
kim-ai-2020	251	26	,	,	,
kim-ai-2020	251	27	and	and	CC
kim-ai-2020	251	28	implementing	implement	VBG
kim-ai-2020	251	29	many	many	JJ
kim-ai-2020	251	30	personalized	personalize	VBN
kim-ai-2020	251	31	and	and	CC
kim-ai-2020	251	32	automated	automate	VBD
kim-ai-2020	251	33	information	information	NN
kim-ai-2020	251	34	systems	system	NNS
kim-ai-2020	251	35	for	for	IN
kim-ai-2020	251	36	their	-PRON-	PRP$
kim-ai-2020	251	37	users	user	NNS
kim-ai-2020	251	38	.	.	.
kim-ai-2020	252	1	At	at	IN
kim-ai-2020	252	2	present	present	NN
kim-ai-2020	252	3	,	,	,
kim-ai-2020	252	4	libraries	library	NNS
kim-ai-2020	252	5	are	be	VBP
kim-ai-2020	252	6	at	at	IN
kim-ai-2020	252	7	an	an	DT
kim-ai-2020	252	8	early	early	JJ
kim-ai-2020	252	9	stage	stage	NN
kim-ai-2020	252	10	in	in	IN
kim-ai-2020	252	11	developing	develop	VBG
kim-ai-2020	252	12	AI	ai	NN
kim-ai-2020	252	13	applications	application	NNS
kim-ai-2020	252	14	and	and	CC
kim-ai-2020	252	15	applying	apply	VBG
kim-ai-2020	252	16	machine	machine	NN
kim-ai-2020	252	17	learning	learning	NN
kim-ai-2020	252	18	and	and	CC
kim-ai-2020	252	19	deep	deep	JJ
kim-ai-2020	252	20	learning	learning	NN
kim-ai-2020	252	21	techniques	technique	NNS
kim-ai-2020	252	22	to	to	TO
kim-ai-2020	252	23	improve	improve	VB
kim-ai-2020	252	24	library	library	JJ
kim-ai-2020	252	25	services	service	NNS
kim-ai-2020	252	26	,	,	,
kim-ai-2020	252	27	systems	system	NNS
kim-ai-2020	252	28	,	,	,
kim-ai-2020	252	29	and	and	CC
kim-ai-2020	252	30	operations	operation	NNS
kim-ai-2020	252	31	.	.	.
kim-ai-2020	253	1	But	but	CC
kim-ai-2020	253	2	the	the	DT
kim-ai-2020	253	3	general	general	JJ
kim-ai-2020	253	4	issues	issue	NNS
kim-ai-2020	253	5	of	of	IN
kim-ai-2020	253	6	hidden	hide	VBN
kim-ai-2020	253	7	biases	bias	NNS
kim-ai-2020	253	8	and	and	CC
kim-ai-2020	253	9	the	the	DT
kim-ai-2020	253	10	lack	lack	NN
kim-ai-2020	253	11	of	of	IN
kim-ai-2020	253	12	explainability	explainability	NN
kim-ai-2020	253	13	in	in	IN
kim-ai-2020	253	14	machine	machine	NN
kim-ai-2020	253	15	learning	learning	NN
kim-ai-2020	253	16	and	and	CC
kim-ai-2020	253	17	deep	deep	JJ
kim-ai-2020	253	18	learning	learning	NN
kim-ai-2020	253	19	are	be	VBP
kim-ai-2020	253	20	quickly	quickly	RB
kim-ai-2020	253	21	gaining	gain	VBG
kim-ai-2020	253	22	awareness	awareness	NN
kim-ai-2020	253	23	in	in	IN
kim-ai-2020	253	24	the	the	DT
kim-ai-2020	253	25	library	library	NN
kim-ai-2020	253	26	community	community	NN
kim-ai-2020	253	27	.	.	.
kim-ai-2020	254	1	As	as	IN
kim-ai-2020	254	2	we	-PRON-	PRP
kim-ai-2020	254	3	have	have	VBP
kim-ai-2020	254	4	seen	see	VBN
kim-ai-2020	254	5	in	in	IN
kim-ai-2020	254	6	the	the	DT
kim-ai-2020	254	7	trolley	trolley	NN
kim-ai-2020	254	8	problem	problem	NN
kim-ai-2020	254	9	,	,	,
kim-ai-2020	254	10	whether	whether	IN
kim-ai-2020	254	11	a	a	DT
kim-ai-2020	254	12	certain	certain	JJ
kim-ai-2020	254	13	action	action	NN
kim-ai-2020	254	14	is	be	VBZ
kim-ai-2020	254	15	moral	moral	JJ
kim-ai-2020	254	16	or	or	CC
kim-ai-2020	254	17	not	not	RB
kim-ai-2020	254	18	is	be	VBZ
kim-ai-2020	254	19	a	a	DT
kim-ai-2020	254	20	line	line	NN
kim-ai-2020	254	21	that	that	WDT
kim-ai-2020	254	22	can	can	MD
kim-ai-2020	254	23	not	not	RB
kim-ai-2020	254	24	be	be	VB
kim-ai-2020	254	25	drawn	draw	VBN
kim-ai-2020	254	26	with	with	IN
kim-ai-2020	254	27	absolute	absolute	JJ
kim-ai-2020	254	28	clarity	clarity	NN
kim-ai-2020	254	29	.	.	.
kim-ai-2020	255	1	It	-PRON-	PRP
kim-ai-2020	255	2	is	be	VBZ
kim-ai-2020	255	3	entirely	entirely	RB
kim-ai-2020	255	4	possible	possible	JJ
kim-ai-2020	255	5	for	for	IN
kim-ai-2020	255	6	fully	fully	RB
kim-ai-2020	255	7	-	-	HYPH
kim-ai-2020	255	8	functioning	function	VBG
kim-ai-2020	255	9	moral	moral	JJ
kim-ai-2020	255	10	agents	agent	NNS
kim-ai-2020	255	11	to	to	TO
kim-ai-2020	255	12	make	make	VB
kim-ai-2020	255	13	different	different	JJ
kim-ai-2020	255	14	judgements	judgement	NNS
kim-ai-2020	255	15	in	in	IN
kim-ai-2020	255	16	a	a	DT
kim-ai-2020	255	17	particular	particular	JJ
kim-ai-2020	255	18	situation	situation	NN
kim-ai-2020	255	19	.	.	.
kim-ai-2020	256	1	Additionally	additionally	RB
kim-ai-2020	256	2	,	,	,
kim-ai-2020	256	3	there	there	EX
kim-ai-2020	256	4	is	be	VBZ
kim-ai-2020	256	5	also	also	RB
kim-ai-2020	256	6	the	the	DT
kim-ai-2020	256	7	matter	matter	NN
kim-ai-2020	256	8	of	of	IN
kim-ai-2020	256	9	morality	morality	NN
kim-ai-2020	256	10	that	that	IN
kim-ai-2020	256	11	our	-PRON-	PRP$
kim-ai-2020	256	12	tools	tool	NNS
kim-ai-2020	256	13	and	and	CC
kim-ai-2020	256	14	systems	system	NNS
kim-ai-2020	256	15	display	display	NN
kim-ai-2020	256	16	.	.	.
kim-ai-2020	257	1	This	this	DT
kim-ai-2020	257	2	is	be	VBZ
kim-ai-2020	257	3	called	call	VBN
kim-ai-2020	257	4	machine	machine	NN
kim-ai-2020	257	5	morality	morality	NN
kim-ai-2020	257	6	in	in	IN
kim-ai-2020	257	7	relation	relation	NN
kim-ai-2020	257	8	to	to	IN
kim-ai-2020	257	9	AI	AI	NNP
kim-ai-2020	257	10	systems	system	NNS
kim-ai-2020	257	11	,	,	,
kim-ai-2020	257	12	since	since	IN
kim-ai-2020	257	13	AI	AI	NNP
kim-ai-2020	257	14	systems	system	NNS
kim-ai-2020	257	15	are	be	VBP
kim-ai-2020	257	16	intelligent	intelligent	JJ
kim-ai-2020	257	17	machines	machine	NNS
kim-ai-2020	257	18	.	.	.
kim-ai-2020	258	1	Wallach	Wallach	NNP
kim-ai-2020	258	2	and	and	CC
kim-ai-2020	258	3	Allen	Allen	NNP
kim-ai-2020	258	4	(	(	-LRB-
kim-ai-2020	258	5	2009	2009	CD
kim-ai-2020	258	6	)	)	-RRB-
kim-ai-2020	258	7	argue	argue	VBP
kim-ai-2020	258	8	that	that	IN
kim-ai-2020	258	9	there	there	EX
kim-ai-2020	258	10	are	be	VBP
kim-ai-2020	258	11	three	three	CD
kim-ai-2020	258	12	distinct	distinct	JJ
kim-ai-2020	258	13	levels	level	NNS
kim-ai-2020	258	14	of	of	IN
kim-ai-2020	258	15	machine	machine	NN
kim-ai-2020	258	16	morality	morality	NN
kim-ai-2020	258	17	:	:	:
kim-ai-2020	258	18	operational	operational	JJ
kim-ai-2020	258	19	morality	morality	NN
kim-ai-2020	258	20	,	,	,
kim-ai-2020	258	21	functional	functional	JJ
kim-ai-2020	258	22	morality	morality	NN
kim-ai-2020	258	23	,	,	,
kim-ai-2020	258	24	and	and	CC
kim-ai-2020	258	25	full	full	JJ
kim-ai-2020	258	26	moral	moral	JJ
kim-ai-2020	258	27	agency	agency	NN
kim-ai-2020	258	28	(	(	-LRB-
kim-ai-2020	258	29	26	26	CD
kim-ai-2020	258	30	)	)	-RRB-
kim-ai-2020	258	31	.	.	.
kim-ai-2020	259	1	Operational	operational	JJ
kim-ai-2020	259	2	morality	morality	NN
kim-ai-2020	259	3	is	be	VBZ
kim-ai-2020	259	4	found	find	VBN
kim-ai-2020	259	5	in	in	IN
kim-ai-2020	259	6	systems	system	NNS
kim-ai-2020	259	7	that	that	WDT
kim-ai-2020	259	8	are	be	VBP
kim-ai-2020	259	9	low	low	JJ
kim-ai-2020	259	10	in	in	IN
kim-ai-2020	259	11	both	both	CC
kim-ai-2020	259	12	autonomy	autonomy	NN
kim-ai-2020	259	13	and	and	CC
kim-ai-2020	259	14	ethical	ethical	JJ
kim-ai-2020	259	15	sensitivity	sensitivity	NN
kim-ai-2020	259	16	.	.	.
kim-ai-2020	260	1	At	at	IN
kim-ai-2020	260	2	this	this	DT
kim-ai-2020	260	3	level	level	NN
kim-ai-2020	260	4	of	of	IN
kim-ai-2020	260	5	machine	machine	NN
kim-ai-2020	260	6	morality	morality	NN
kim-ai-2020	260	7	,	,	,
kim-ai-2020	260	8	a	a	DT
kim-ai-2020	260	9	machine	machine	NN
kim-ai-2020	260	10	or	or	CC
kim-ai-2020	260	11	a	a	DT
kim-ai-2020	260	12	tool	tool	NN
kim-ai-2020	260	13	is	be	VBZ
kim-ai-2020	260	14	given	give	VBN
kim-ai-2020	260	15	a	a	DT
kim-ai-2020	260	16	mechanism	mechanism	NN
kim-ai-2020	260	17	that	that	WDT
kim-ai-2020	260	18	prevents	prevent	VBZ
kim-ai-2020	260	19	its	-PRON-	PRP$
kim-ai-2020	260	20	immoral	immoral	JJ
kim-ai-2020	260	21	use	use	NN
kim-ai-2020	260	22	,	,	,
kim-ai-2020	260	23	but	but	CC
kim-ai-2020	260	24	the	the	DT
kim-ai-2020	260	25	mechanism	mechanism	NN
kim-ai-2020	260	26	is	be	VBZ
kim-ai-2020	260	27	within	within	IN
kim-ai-2020	260	28	the	the	DT
kim-ai-2020	260	29	full	full	JJ
kim-ai-2020	260	30	control	control	NN
kim-ai-2020	260	31	of	of	IN
kim-ai-2020	260	32	the	the	DT
kim-ai-2020	260	33	machine	machine	NN
kim-ai-2020	260	34	or	or	CC
kim-ai-2020	260	35	the	the	DT
kim-ai-2020	260	36	tool	tool	NN
kim-ai-2020	260	37	’s	’s	POS
kim-ai-2020	260	38	designers	designer	NNS
kim-ai-2020	260	39	and	and	CC
kim-ai-2020	260	40	users	user	NNS
kim-ai-2020	260	41	.	.	.
kim-ai-2020	261	1	Such	such	JJ
kim-ai-2020	261	2	operational	operational	JJ
kim-ai-2020	261	3	morality	morality	NN
kim-ai-2020	261	4	exists	exist	VBZ
kim-ai-2020	261	5	in	in	IN
kim-ai-2020	261	6	a	a	DT
kim-ai-2020	261	7	gun	gun	NN
kim-ai-2020	261	8	with	with	IN
kim-ai-2020	261	9	a	a	DT
kim-ai-2020	261	10	childproof	childproof	JJ
kim-ai-2020	261	11	safety	safety	NN
kim-ai-2020	261	12	mechanism	mechanism	NN
kim-ai-2020	261	13	,	,	,
kim-ai-2020	261	14	for	for	IN
kim-ai-2020	261	15	example	example	NN
kim-ai-2020	261	16	.	.	.
kim-ai-2020	262	1	A	a	DT
kim-ai-2020	262	2	gun	gun	NN
kim-ai-2020	262	3	with	with	IN
kim-ai-2020	262	4	a	a	DT
kim-ai-2020	262	5	safety	safety	NN
kim-ai-2020	262	6	mechanism	mechanism	NN
kim-ai-2020	262	7	is	be	VBZ
kim-ai-2020	262	8	hardly	hardly	RB
kim-ai-2020	262	9	autonomous	autonomous	JJ
kim-ai-2020	262	10	nor	nor	CC
kim-ai-2020	262	11	sensitive	sensitive	JJ
kim-ai-2020	262	12	to	to	IN
kim-ai-2020	262	13	any	any	DT
kim-ai-2020	262	14	ethical	ethical	JJ
kim-ai-2020	262	15	concerns	concern	NNS
kim-ai-2020	262	16	related	relate	VBN
kim-ai-2020	262	17	to	to	IN
kim-ai-2020	262	18	its	-PRON-	PRP$
kim-ai-2020	262	19	use	use	NN
kim-ai-2020	262	20	.	.	.
kim-ai-2020	263	1	By	by	IN
kim-ai-2020	263	2	contrast	contrast	NN
kim-ai-2020	263	3	,	,	,
kim-ai-2020	263	4	machines	machine	NNS
kim-ai-2020	263	5	with	with	IN
kim-ai-2020	263	6	functional	functional	JJ
kim-ai-2020	263	7	morality	morality	NN
kim-ai-2020	263	8	do	do	VBP
kim-ai-2020	263	9	possess	possess	VB
kim-ai-2020	263	10	a	a	DT
kim-ai-2020	263	11	certain	certain	JJ
kim-ai-2020	263	12	level	level	NN
kim-ai-2020	263	13	of	of	IN
kim-ai-2020	263	14	autonomy	autonomy	NN
kim-ai-2020	263	15	and	and	CC
kim-ai-2020	263	16	ethical	ethical	JJ
kim-ai-2020	263	17	sensitivity	sensitivity	NN
kim-ai-2020	263	18	.	.	.
kim-ai-2020	264	1	They	-PRON-	PRP
kim-ai-2020	264	2	may	may	MD
kim-ai-2020	264	3	include	include	VB
kim-ai-2020	264	4	an	an	DT
kim-ai-2020	264	5	AI	AI	NNP
kim-ai-2020	264	6	system	system	NN
kim-ai-2020	264	7	with	with	IN
kim-ai-2020	264	8	significant	significant	JJ
kim-ai-2020	264	9	autonomy	autonomy	NN
kim-ai-2020	264	10	but	but	CC
kim-ai-2020	264	11	little	little	JJ
kim-ai-2020	264	12	ethical	ethical	JJ
kim-ai-2020	264	13	sensitivity	sensitivity	NN
kim-ai-2020	264	14	or	or	CC
kim-ai-2020	264	15	an	an	DT
kim-ai-2020	264	16	AI	AI	NNP
kim-ai-2020	264	17	system	system	NN
kim-ai-2020	264	18	with	with	IN
kim-ai-2020	264	19	little	little	JJ
kim-ai-2020	264	20	autonomy	autonomy	NN
kim-ai-2020	264	21	but	but	CC
kim-ai-2020	264	22	high	high	JJ
kim-ai-2020	264	23	ethical	ethical	JJ
kim-ai-2020	264	24	sensitivity	sensitivity	NN
kim-ai-2020	264	25	.	.	.
kim-ai-2020	265	1	An	an	DT
kim-ai-2020	265	2	autonomous	autonomous	JJ
kim-ai-2020	265	3	drone	drone	NN
kim-ai-2020	265	4	would	would	MD
kim-ai-2020	265	5	fall	fall	VB
kim-ai-2020	265	6	under	under	IN
kim-ai-2020	265	7	the	the	DT
kim-ai-2020	265	8	former	former	JJ
kim-ai-2020	265	9	category	category	NN
kim-ai-2020	265	10	,	,	,
kim-ai-2020	265	11	while	while	IN
kim-ai-2020	265	12	MedEthEx	medethex	NN
kim-ai-2020	265	13	,	,	,
kim-ai-2020	265	14	an	an	DT
kim-ai-2020	265	15	ethical	ethical	JJ
kim-ai-2020	265	16	decision	decision	NN
kim-ai-2020	265	17	-	-	HYPH
kim-ai-2020	265	18	support	support	NN
kim-ai-2020	265	19	AI	AI	NNP
kim-ai-2020	265	20	recommendation	recommendation	NN
kim-ai-2020	265	21	system	system	NN
kim-ai-2020	265	22	for	for	IN
kim-ai-2020	265	23	clinicians	clinician	NNS
kim-ai-2020	265	24	,	,	,
kim-ai-2020	265	25	would	would	MD
kim-ai-2020	265	26	be	be	VB
kim-ai-2020	265	27	of	of	IN
kim-ai-2020	265	28	the	the	DT
kim-ai-2020	265	29	latter	latter	JJ
kim-ai-2020	265	30	.	.	.
kim-ai-2020	266	1	Lastly	lastly	RB
kim-ai-2020	266	2	,	,	,
kim-ai-2020	266	3	systems	system	NNS
kim-ai-2020	266	4	with	with	IN
kim-ai-2020	266	5	high	high	JJ
kim-ai-2020	266	6	autonomy	autonomy	NN
kim-ai-2020	266	7	and	and	CC
kim-ai-2020	266	8	high	high	JJ
kim-ai-2020	266	9	ethical	ethical	JJ
kim-ai-2020	266	10	sensitivity	sensitivity	NN
kim-ai-2020	266	11	can	can	MD
kim-ai-2020	266	12	be	be	VB
kim-ai-2020	266	13	regarded	regard	VBN
kim-ai-2020	266	14	as	as	IN
kim-ai-2020	266	15	having	have	VBG
kim-ai-2020	266	16	full	full	JJ
kim-ai-2020	266	17	moral	moral	JJ
kim-ai-2020	266	18	agency	agency	NN
kim-ai-2020	266	19	,	,	,
kim-ai-2020	266	20	as	as	RB
kim-ai-2020	266	21	much	much	RB
kim-ai-2020	266	22	as	as	IN
kim-ai-2020	266	23	humans	human	NNS
kim-ai-2020	266	24	do	do	VBP
kim-ai-2020	266	25	.	.	.
kim-ai-2020	267	1	This	this	DT
kim-ai-2020	267	2	means	mean	VBZ
kim-ai-2020	267	3	that	that	IN
kim-ai-2020	267	4	those	those	DT
kim-ai-2020	267	5	systems	system	NNS
kim-ai-2020	267	6	would	would	MD
kim-ai-2020	267	7	have	have	VB
kim-ai-2020	267	8	a	a	DT
kim-ai-2020	267	9	mental	mental	JJ
kim-ai-2020	267	10	representation	representation	NN
kim-ai-2020	267	11	of	of	IN
kim-ai-2020	267	12	values	value	NNS
kim-ai-2020	267	13	and	and	CC
kim-ai-2020	267	14	the	the	DT
kim-ai-2020	267	15	capacity	capacity	NN
kim-ai-2020	267	16	for	for	IN
kim-ai-2020	267	17	moral	moral	JJ
kim-ai-2020	267	18	reasoning	reasoning	NN
kim-ai-2020	267	19	.	.	.
kim-ai-2020	268	1	Such	such	JJ
kim-ai-2020	268	2	machines	machine	NNS
kim-ai-2020	268	3	can	can	MD
kim-ai-2020	268	4	be	be	VB
kim-ai-2020	268	5	held	hold	VBN
kim-ai-2020	268	6	morally	morally	RB
kim-ai-2020	268	7	responsible	responsible	JJ
kim-ai-2020	268	8	for	for	IN
kim-ai-2020	268	9	their	-PRON-	PRP$
kim-ai-2020	268	10	actions	action	NNS
kim-ai-2020	268	11	.	.	.
kim-ai-2020	269	1	We	-PRON-	PRP
kim-ai-2020	269	2	do	do	VBP
kim-ai-2020	269	3	not	not	RB
kim-ai-2020	269	4	know	know	VB
kim-ai-2020	269	5	whether	whether	IN
kim-ai-2020	269	6	AI	AI	NNP
kim-ai-2020	269	7	will	will	MD
kim-ai-2020	269	8	be	be	VB
kim-ai-2020	269	9	able	able	JJ
kim-ai-2020	269	10	to	to	TO
kim-ai-2020	269	11	produce	produce	VB
kim-ai-2020	269	12	such	such	PDT
kim-ai-2020	269	13	a	a	DT
kim-ai-2020	269	14	machine	machine	NN
kim-ai-2020	269	15	with	with	IN
kim-ai-2020	269	16	full	full	JJ
kim-ai-2020	269	17	moral	moral	JJ
kim-ai-2020	269	18	agency	agency	NN
kim-ai-2020	269	19	.	.	.
kim-ai-2020	270	1	If	if	IN
kim-ai-2020	270	2	the	the	DT
kim-ai-2020	270	3	current	current	JJ
kim-ai-2020	270	4	direction	direction	NN
kim-ai-2020	270	5	to	to	TO
kim-ai-2020	270	6	automate	automate	VB
kim-ai-2020	270	7	more	more	JJR
kim-ai-2020	270	8	and	and	CC
kim-ai-2020	270	9	more	more	JJR
kim-ai-2020	270	10	human	human	JJ
kim-ai-2020	270	11	tasks	task	NNS
kim-ai-2020	270	12	for	for	IN
kim-ai-2020	270	13	cost	cost	NN
kim-ai-2020	270	14	savings	saving	NNS
kim-ai-2020	270	15	and	and	CC
kim-ai-2020	270	16	efficiency	efficiency	NN
kim-ai-2020	270	17	at	at	IN
kim-ai-2020	270	18	scale	scale	NN
kim-ai-2020	270	19	continues	continue	VBZ
kim-ai-2020	270	20	,	,	,
kim-ai-2020	270	21	however	however	RB
kim-ai-2020	270	22	,	,	,
kim-ai-2020	270	23	most	most	JJS
kim-ai-2020	270	24	of	of	IN
kim-ai-2020	270	25	the	the	DT
kim-ai-2020	270	26	more	more	RBR
kim-ai-2020	270	27	sophisticated	sophisticated	JJ
kim-ai-2020	270	28	AI	AI	NNP
kim-ai-2020	270	29	applications	application	NNS
kim-ai-2020	270	30	to	to	TO
kim-ai-2020	270	31	come	come	VB
kim-ai-2020	270	32	will	will	MD
kim-ai-2020	270	33	be	be	VB
kim-ai-2020	270	34	of	of	IN
kim-ai-2020	270	35	the	the	DT
kim-ai-2020	270	36	kind	kind	NN
kim-ai-2020	270	37	with	with	IN
kim-ai-2020	270	38	functional	functional	JJ
kim-ai-2020	270	39	morality	morality	NN
kim-ai-2020	270	40	,	,	,
kim-ai-2020	270	41	particularly	particularly	RB
kim-ai-2020	270	42	the	the	DT
kim-ai-2020	270	43	kind	kind	NN
kim-ai-2020	270	44	that	that	WDT
kim-ai-2020	270	45	combines	combine	VBZ
kim-ai-2020	270	46	a	a	DT
kim-ai-2020	270	47	relatively	relatively	RB
kim-ai-2020	270	48	high	high	JJ
kim-ai-2020	270	49	level	level	NN
kim-ai-2020	270	50	of	of	IN
kim-ai-2020	270	51	autonomy	autonomy	NN
kim-ai-2020	270	52	and	and	CC
kim-ai-2020	270	53	a	a	DT
kim-ai-2020	270	54	lower	low	JJR
kim-ai-2020	270	55	level	level	NN
kim-ai-2020	270	56	of	of	IN
kim-ai-2020	270	57	ethical	ethical	JJ
kim-ai-2020	270	58	sensitivity	sensitivity	NN
kim-ai-2020	270	59	.	.	.
kim-ai-2020	271	1	In	in	IN
kim-ai-2020	271	2	the	the	DT
kim-ai-2020	271	3	beginning	beginning	NN
kim-ai-2020	271	4	of	of	IN
kim-ai-2020	271	5	this	this	DT
kim-ai-2020	271	6	chapter	chapter	NN
kim-ai-2020	271	7	,	,	,
kim-ai-2020	271	8	I	-PRON-	PRP
kim-ai-2020	271	9	mentioned	mention	VBD
kim-ai-2020	271	10	that	that	IN
kim-ai-2020	271	11	the	the	DT
kim-ai-2020	271	12	goal	goal	NN
kim-ai-2020	271	13	of	of	IN
kim-ai-2020	271	14	AI	AI	NNP
kim-ai-2020	271	15	is	be	VBZ
kim-ai-2020	271	16	to	to	TO
kim-ai-2020	271	17	create	create	VB
kim-ai-2020	271	18	an	an	DT
kim-ai-2020	271	19	artificial	artificial	JJ
kim-ai-2020	271	20	system	system	NN
kim-ai-2020	271	21	–	–	:
kim-ai-2020	271	22	whether	whether	IN
kim-ai-2020	271	23	it	-PRON-	PRP
kim-ai-2020	271	24	be	be	VB
kim-ai-2020	271	25	a	a	DT
kim-ai-2020	271	26	piece	piece	NN
kim-ai-2020	271	27	of	of	IN
kim-ai-2020	271	28	software	software	NN
kim-ai-2020	271	29	or	or	CC
kim-ai-2020	271	30	a	a	DT
kim-ai-2020	271	31	machine	machine	NN
kim-ai-2020	271	32	with	with	IN
kim-ai-2020	271	33	a	a	DT
kim-ai-2020	271	34	physical	physical	JJ
kim-ai-2020	271	35	body	body	NN
kim-ai-2020	271	36	–	–	:
kim-ai-2020	271	37	that	that	DT
kim-ai-2020	271	38	is	be	VBZ
kim-ai-2020	271	39	as	as	RB
kim-ai-2020	271	40	intelligent	intelligent	JJ
kim-ai-2020	271	41	as	as	IN
kim-ai-2020	271	42	a	a	DT
kim-ai-2020	271	43	human	human	NN
kim-ai-2020	271	44	in	in	IN
kim-ai-2020	271	45	its	-PRON-	PRP$
kim-ai-2020	271	46	performance	performance	NN
kim-ai-2020	271	47	,	,	,
kim-ai-2020	271	48	either	either	CC
kim-ai-2020	271	49	broadly	broadly	RB
kim-ai-2020	271	50	in	in	IN
kim-ai-2020	271	51	all	all	DT
kim-ai-2020	271	52	areas	area	NNS
kim-ai-2020	271	53	of	of	IN
kim-ai-2020	271	54	human	human	JJ
kim-ai-2020	271	55	activities	activity	NNS
kim-ai-2020	271	56	or	or	CC
kim-ai-2020	271	57	narrowly	narrowly	RB
kim-ai-2020	271	58	in	in	IN
kim-ai-2020	271	59	a	a	DT
kim-ai-2020	271	60	specific	specific	JJ
kim-ai-2020	271	61	activity	activity	NN
kim-ai-2020	271	62	,	,	,
kim-ai-2020	271	63	such	such	JJ
kim-ai-2020	271	64	as	as	IN
kim-ai-2020	271	65	playing	play	VBG
kim-ai-2020	271	66	chess	chess	NN
kim-ai-2020	271	67	or	or	CC
kim-ai-2020	271	68	driving	driving	NN
kim-ai-2020	271	69	.	.	.
kim-ai-2020	272	1	But	but	CC
kim-ai-2020	272	2	what	what	WP
kim-ai-2020	272	3	does	do	VBZ
kim-ai-2020	272	4	“	"	``
kim-ai-2020	272	5	as	as	RB
kim-ai-2020	272	6	intelligent	intelligent	JJ
kim-ai-2020	272	7	as	as	IN
kim-ai-2020	272	8	a	a	DT
kim-ai-2020	272	9	human	human	NN
kim-ai-2020	272	10	”	"	''
kim-ai-2020	272	11	exactly	exactly	RB
kim-ai-2020	272	12	mean	mean	VB
kim-ai-2020	272	13	?	?	.
kim-ai-2020	273	1	If	if	IN
kim-ai-2020	273	2	morality	morality	NN
kim-ai-2020	273	3	is	be	VBZ
kim-ai-2020	273	4	an	an	DT
kim-ai-2020	273	5	integral	integral	JJ
kim-ai-2020	273	6	component	component	NN
kim-ai-2020	273	7	of	of	IN
kim-ai-2020	273	8	human	human	JJ
kim-ai-2020	273	9	-	-	HYPH
kim-ai-2020	273	10	level	level	NN
kim-ai-2020	273	11	intelligence	intelligence	NN
kim-ai-2020	273	12	,	,	,
kim-ai-2020	273	13	AI	AI	NNP
kim-ai-2020	273	14	research	research	NN
kim-ai-2020	273	15	needs	need	VBZ
kim-ai-2020	273	16	to	to	TO
kim-ai-2020	273	17	pay	pay	VB
kim-ai-2020	273	18	more	more	JJR
kim-ai-2020	273	19	attention	attention	NN
kim-ai-2020	273	20	to	to	IN
kim-ai-2020	273	21	intelligence	intelligence	NN
kim-ai-2020	273	22	,	,	,
kim-ai-2020	273	23	not	not	RB
kim-ai-2020	273	24	just	just	RB
kim-ai-2020	273	25	in	in	IN
kim-ai-2020	273	26	accomplishing	accomplish	VBG
kim-ai-2020	273	27	a	a	DT
kim-ai-2020	273	28	goal	goal	NN
kim-ai-2020	273	29	but	but	CC
kim-ai-2020	273	30	also	also	RB
kim-ai-2020	273	31	in	in	IN
kim-ai-2020	273	32	doing	do	VBG
kim-ai-2020	273	33	so	so	RB
kim-ai-2020	273	34	ethically	ethically	RB
kim-ai-2020	273	35	.	.	.
kim-ai-2020	274	1	[	[	-LRB-
kim-ai-2020	274	2	footnoteRef:25	footnoteRef:25	NNP
kim-ai-2020	274	3	]	]	-RRB-
kim-ai-2020	274	4	In	in	IN
kim-ai-2020	274	5	that	that	DT
kim-ai-2020	274	6	light	light	NN
kim-ai-2020	274	7	,	,	,
kim-ai-2020	274	8	it	-PRON-	PRP
kim-ai-2020	274	9	is	be	VBZ
kim-ai-2020	274	10	meaningful	meaningful	JJ
kim-ai-2020	274	11	to	to	TO
kim-ai-2020	274	12	ask	ask	VB
kim-ai-2020	274	13	what	what	WDT
kim-ai-2020	274	14	level	level	NN
kim-ai-2020	274	15	of	of	IN
kim-ai-2020	274	16	autonomy	autonomy	NN
kim-ai-2020	274	17	and	and	CC
kim-ai-2020	274	18	ethical	ethical	JJ
kim-ai-2020	274	19	sensitivity	sensitivity	NN
kim-ai-2020	274	20	a	a	DT
kim-ai-2020	274	21	given	give	VBN
kim-ai-2020	274	22	AI	ai	NN
kim-ai-2020	274	23	system	system	NN
kim-ai-2020	274	24	is	be	VBZ
kim-ai-2020	274	25	equipped	equip	VBN
kim-ai-2020	274	26	with	with	IN
kim-ai-2020	274	27	,	,	,
kim-ai-2020	274	28	and	and	CC
kim-ai-2020	274	29	what	what	WDT
kim-ai-2020	274	30	level	level	NN
kim-ai-2020	274	31	of	of	IN
kim-ai-2020	274	32	machine	machine	NN
kim-ai-2020	274	33	morality	morality	NN
kim-ai-2020	274	34	is	be	VBZ
kim-ai-2020	274	35	appropriate	appropriate	JJ
kim-ai-2020	274	36	for	for	IN
kim-ai-2020	274	37	its	-PRON-	PRP$
kim-ai-2020	274	38	purpose	purpose	NN
kim-ai-2020	274	39	.	.	.
kim-ai-2020	275	1	[	[	-LRB-
kim-ai-2020	275	2	25	25	CD
kim-ai-2020	275	3	:	:	:
kim-ai-2020	275	4	Here	here	RB
kim-ai-2020	275	5	,	,	,
kim-ai-2020	275	6	I	-PRON-	PRP
kim-ai-2020	275	7	regard	regard	VBP
kim-ai-2020	275	8	intelligence	intelligence	NN
kim-ai-2020	275	9	as	as	IN
kim-ai-2020	275	10	the	the	DT
kim-ai-2020	275	11	ability	ability	NN
kim-ai-2020	275	12	to	to	TO
kim-ai-2020	275	13	accomplish	accomplish	VB
kim-ai-2020	275	14	complex	complex	JJ
kim-ai-2020	275	15	goals	goal	NNS
kim-ai-2020	275	16	following	follow	VBG
kim-ai-2020	275	17	Tegmark	Tegmark	NNP
kim-ai-2020	275	18	2017	2017	CD
kim-ai-2020	275	19	.	.	.
kim-ai-2020	276	1	For	for	IN
kim-ai-2020	276	2	more	more	JJR
kim-ai-2020	276	3	discussion	discussion	NN
kim-ai-2020	276	4	on	on	IN
kim-ai-2020	276	5	intelligence	intelligence	NN
kim-ai-2020	276	6	and	and	CC
kim-ai-2020	276	7	goals	goal	NNS
kim-ai-2020	276	8	,	,	,
kim-ai-2020	276	9	see	see	VB
kim-ai-2020	276	10	Chapter	chapter	NN
kim-ai-2020	276	11	2	2	CD
kim-ai-2020	276	12	and	and	CC
kim-ai-2020	276	13	Chapter	Chapter	NNP
kim-ai-2020	276	14	7	7	CD
kim-ai-2020	276	15	.	.	.
kim-ai-2020	276	16	]	]	-RRB-
kim-ai-2020	277	1	In	in	IN
kim-ai-2020	277	2	designing	design	VBG
kim-ai-2020	277	3	an	an	DT
kim-ai-2020	277	4	AI	AI	NNP
kim-ai-2020	277	5	system	system	NN
kim-ai-2020	277	6	,	,	,
kim-ai-2020	277	7	it	-PRON-	PRP
kim-ai-2020	277	8	would	would	MD
kim-ai-2020	277	9	be	be	VB
kim-ai-2020	277	10	also	also	RB
kim-ai-2020	277	11	helpful	helpful	JJ
kim-ai-2020	277	12	to	to	TO
kim-ai-2020	277	13	determine	determine	VB
kim-ai-2020	277	14	what	what	WDT
kim-ai-2020	277	15	level	level	NN
kim-ai-2020	277	16	of	of	IN
kim-ai-2020	277	17	autonomy	autonomy	NN
kim-ai-2020	277	18	and	and	CC
kim-ai-2020	277	19	ethical	ethical	JJ
kim-ai-2020	277	20	sensitivity	sensitivity	NN
kim-ai-2020	277	21	would	would	MD
kim-ai-2020	277	22	be	be	VB
kim-ai-2020	277	23	best	well	RBS
kim-ai-2020	277	24	suited	suit	VBN
kim-ai-2020	277	25	for	for	IN
kim-ai-2020	277	26	the	the	DT
kim-ai-2020	277	27	purpose	purpose	NN
kim-ai-2020	277	28	of	of	IN
kim-ai-2020	277	29	the	the	DT
kim-ai-2020	277	30	system	system	NN
kim-ai-2020	277	31	and	and	CC
kim-ai-2020	277	32	whether	whether	IN
kim-ai-2020	277	33	it	-PRON-	PRP
kim-ai-2020	277	34	is	be	VBZ
kim-ai-2020	277	35	feasible	feasible	JJ
kim-ai-2020	277	36	to	to	TO
kim-ai-2020	277	37	provide	provide	VB
kim-ai-2020	277	38	that	that	DT
kim-ai-2020	277	39	level	level	NN
kim-ai-2020	277	40	of	of	IN
kim-ai-2020	277	41	machine	machine	NN
kim-ai-2020	277	42	morality	morality	NN
kim-ai-2020	277	43	for	for	IN
kim-ai-2020	277	44	the	the	DT
kim-ai-2020	277	45	system	system	NN
kim-ai-2020	277	46	in	in	IN
kim-ai-2020	277	47	question	question	NN
kim-ai-2020	277	48	.	.	.
kim-ai-2020	278	1	The	the	DT
kim-ai-2020	278	2	narrower	narrow	JJR
kim-ai-2020	278	3	the	the	DT
kim-ai-2020	278	4	function	function	NN
kim-ai-2020	278	5	or	or	CC
kim-ai-2020	278	6	the	the	DT
kim-ai-2020	278	7	domain	domain	NN
kim-ai-2020	278	8	of	of	IN
kim-ai-2020	278	9	an	an	DT
kim-ai-2020	278	10	AI	AI	NNP
kim-ai-2020	278	11	system	system	NN
kim-ai-2020	278	12	is	be	VBZ
kim-ai-2020	278	13	,	,	,
kim-ai-2020	278	14	the	the	DT
kim-ai-2020	278	15	easier	easy	JJR
kim-ai-2020	278	16	it	-PRON-	PRP
kim-ai-2020	278	17	will	will	MD
kim-ai-2020	278	18	be	be	VB
kim-ai-2020	278	19	to	to	TO
kim-ai-2020	278	20	equip	equip	VB
kim-ai-2020	278	21	it	-PRON-	PRP
kim-ai-2020	278	22	with	with	IN
kim-ai-2020	278	23	an	an	DT
kim-ai-2020	278	24	appropriate	appropriate	JJ
kim-ai-2020	278	25	level	level	NN
kim-ai-2020	278	26	of	of	IN
kim-ai-2020	278	27	autonomy	autonomy	NN
kim-ai-2020	278	28	and	and	CC
kim-ai-2020	278	29	ethical	ethical	JJ
kim-ai-2020	278	30	sensitivity	sensitivity	NN
kim-ai-2020	278	31	.	.	.
kim-ai-2020	279	1	In	in	IN
kim-ai-2020	279	2	evaluating	evaluate	VBG
kim-ai-2020	279	3	and	and	CC
kim-ai-2020	279	4	designing	design	VBG
kim-ai-2020	279	5	an	an	DT
kim-ai-2020	279	6	AI	AI	NNP
kim-ai-2020	279	7	system	system	NN
kim-ai-2020	279	8	,	,	,
kim-ai-2020	279	9	it	-PRON-	PRP
kim-ai-2020	279	10	will	will	MD
kim-ai-2020	279	11	be	be	VB
kim-ai-2020	279	12	important	important	JJ
kim-ai-2020	279	13	to	to	TO
kim-ai-2020	279	14	test	test	VB
kim-ai-2020	279	15	the	the	DT
kim-ai-2020	279	16	actual	actual	JJ
kim-ai-2020	279	17	outcome	outcome	NN
kim-ai-2020	279	18	against	against	IN
kim-ai-2020	279	19	the	the	DT
kim-ai-2020	279	20	anticipated	anticipated	JJ
kim-ai-2020	279	21	outcome	outcome	NN
kim-ai-2020	279	22	in	in	IN
kim-ai-2020	279	23	different	different	JJ
kim-ai-2020	279	24	types	type	NNS
kim-ai-2020	279	25	of	of	IN
kim-ai-2020	279	26	cases	case	NNS
kim-ai-2020	279	27	in	in	IN
kim-ai-2020	279	28	order	order	NN
kim-ai-2020	279	29	to	to	TO
kim-ai-2020	279	30	identify	identify	VB
kim-ai-2020	279	31	potential	potential	JJ
kim-ai-2020	279	32	problems	problem	NNS
kim-ai-2020	279	33	.	.	.
kim-ai-2020	280	1	System	system	NN
kim-ai-2020	280	2	-	-	HYPH
kim-ai-2020	280	3	wide	wide	JJ
kim-ai-2020	280	4	audits	audits	NN
kim-ai-2020	280	5	to	to	TO
kim-ai-2020	280	6	detect	detect	VB
kim-ai-2020	280	7	well	well	RB
kim-ai-2020	280	8	-	-	HYPH
kim-ai-2020	280	9	known	know	VBN
kim-ai-2020	280	10	biases	bias	NNS
kim-ai-2020	280	11	,	,	,
kim-ai-2020	280	12	such	such	JJ
kim-ai-2020	280	13	as	as	IN
kim-ai-2020	280	14	gender	gender	NN
kim-ai-2020	280	15	discrimination	discrimination	NN
kim-ai-2020	280	16	or	or	CC
kim-ai-2020	280	17	racism	racism	NN
kim-ai-2020	280	18	can	can	MD
kim-ai-2020	280	19	serve	serve	VB
kim-ai-2020	280	20	as	as	IN
kim-ai-2020	280	21	an	an	DT
kim-ai-2020	280	22	effective	effective	JJ
kim-ai-2020	280	23	strategy	strategy	NN
kim-ai-2020	280	24	.	.	.
kim-ai-2020	281	1	[	[	-LRB-
kim-ai-2020	281	2	footnoteRef:26	footnoteRef:26	NNP
kim-ai-2020	281	3	]	]	-RRB-
kim-ai-2020	281	4	Other	other	JJ
kim-ai-2020	281	5	undetected	undetected	JJ
kim-ai-2020	281	6	problems	problem	NNS
kim-ai-2020	281	7	may	may	MD
kim-ai-2020	281	8	surface	surface	VB
kim-ai-2020	281	9	only	only	RB
kim-ai-2020	281	10	after	after	IN
kim-ai-2020	281	11	the	the	DT
kim-ai-2020	281	12	AI	AI	NNP
kim-ai-2020	281	13	system	system	NN
kim-ai-2020	281	14	is	be	VBZ
kim-ai-2020	281	15	deployed	deploy	VBN
kim-ai-2020	281	16	.	.	.
kim-ai-2020	282	1	Having	have	VBG
kim-ai-2020	282	2	a	a	DT
kim-ai-2020	282	3	mechanism	mechanism	NN
kim-ai-2020	282	4	to	to	TO
kim-ai-2020	282	5	continually	continually	RB
kim-ai-2020	282	6	test	test	VB
kim-ai-2020	282	7	an	an	DT
kim-ai-2020	282	8	AI	AI	NNP
kim-ai-2020	282	9	algorithm	algorithm	NN
kim-ai-2020	282	10	to	to	TO
kim-ai-2020	282	11	identify	identify	VB
kim-ai-2020	282	12	those	those	DT
kim-ai-2020	282	13	unnoticed	unnoticed	JJ
kim-ai-2020	282	14	problems	problem	NNS
kim-ai-2020	282	15	and	and	CC
kim-ai-2020	282	16	feeding	feed	VBG
kim-ai-2020	282	17	the	the	DT
kim-ai-2020	282	18	test	test	NN
kim-ai-2020	282	19	result	result	VB
kim-ai-2020	282	20	back	back	RB
kim-ai-2020	282	21	into	into	IN
kim-ai-2020	282	22	the	the	DT
kim-ai-2020	282	23	algorithm	algorithm	NN
kim-ai-2020	282	24	for	for	IN
kim-ai-2020	282	25	retraining	retraining	NN
kim-ai-2020	282	26	will	will	MD
kim-ai-2020	282	27	be	be	VB
kim-ai-2020	282	28	another	another	DT
kim-ai-2020	282	29	way	way	NN
kim-ai-2020	282	30	to	to	TO
kim-ai-2020	282	31	deal	deal	VB
kim-ai-2020	282	32	with	with	IN
kim-ai-2020	282	33	algorithmic	algorithmic	JJ
kim-ai-2020	282	34	biases	bias	NNS
kim-ai-2020	282	35	.	.	.
kim-ai-2020	283	1	Those	those	DT
kim-ai-2020	283	2	who	who	WP
kim-ai-2020	283	3	build	build	VBP
kim-ai-2020	283	4	AI	AI	NNP
kim-ai-2020	283	5	systems	system	NNS
kim-ai-2020	283	6	will	will	MD
kim-ai-2020	283	7	also	also	RB
kim-ai-2020	283	8	benefit	benefit	VB
kim-ai-2020	283	9	from	from	IN
kim-ai-2020	283	10	consulting	consult	VBG
kim-ai-2020	283	11	existing	exist	VBG
kim-ai-2020	283	12	principles	principle	NNS
kim-ai-2020	283	13	and	and	CC
kim-ai-2020	283	14	guidelines	guideline	NNS
kim-ai-2020	283	15	such	such	JJ
kim-ai-2020	283	16	as	as	IN
kim-ai-2020	283	17	FAT	FAT	NNP
kim-ai-2020	283	18	/	/	SYM
kim-ai-2020	283	19	ML	ML	NNP
kim-ai-2020	283	20	’s	’s	POS
kim-ai-2020	283	21	“	"	``
kim-ai-2020	283	22	Principles	principle	NNS
kim-ai-2020	283	23	for	for	IN
kim-ai-2020	283	24	Accountable	Accountable	NNP
kim-ai-2020	283	25	Algorithms	Algorithms	NNP
kim-ai-2020	283	26	and	and	CC
kim-ai-2020	283	27	a	a	DT
kim-ai-2020	283	28	Social	Social	NNP
kim-ai-2020	283	29	Impact	Impact	NNP
kim-ai-2020	283	30	Statement	Statement	NNP
kim-ai-2020	283	31	for	for	IN
kim-ai-2020	283	32	Algorithms	Algorithms	NNP
kim-ai-2020	283	33	.	.	.
kim-ai-2020	283	34	”[footnoteRef:27	”[footnoteRef:27	''
kim-ai-2020	283	35	]	]	-RRB-
kim-ai-2020	283	36	[	[	-LRB-
kim-ai-2020	283	37	26	26	CD
kim-ai-2020	283	38	:	:	:
kim-ai-2020	283	39	These	these	DT
kim-ai-2020	283	40	audits	audit	NNS
kim-ai-2020	283	41	are	be	VBP
kim-ai-2020	283	42	far	far	RB
kim-ai-2020	283	43	from	from	IN
kim-ai-2020	283	44	foolproof	foolproof	JJ
kim-ai-2020	283	45	,	,	,
kim-ai-2020	283	46	but	but	CC
kim-ai-2020	283	47	the	the	DT
kim-ai-2020	283	48	detection	detection	NN
kim-ai-2020	283	49	of	of	IN
kim-ai-2020	283	50	hidden	hidden	JJ
kim-ai-2020	283	51	biases	bias	NNS
kim-ai-2020	283	52	will	will	MD
kim-ai-2020	283	53	be	be	VB
kim-ai-2020	283	54	crucial	crucial	JJ
kim-ai-2020	283	55	in	in	IN
kim-ai-2020	283	56	making	make	VBG
kim-ai-2020	283	57	AI	AI	NNP
kim-ai-2020	283	58	algorithms	algorithm	NNS
kim-ai-2020	283	59	more	more	RBR
kim-ai-2020	283	60	accountable	accountable	JJ
kim-ai-2020	283	61	and	and	CC
kim-ai-2020	283	62	their	-PRON-	PRP$
kim-ai-2020	283	63	decisions	decision	NNS
kim-ai-2020	283	64	more	more	RBR
kim-ai-2020	283	65	ethical	ethical	JJ
kim-ai-2020	283	66	.	.	.
kim-ai-2020	284	1	A	a	DT
kim-ai-2020	284	2	debiasing	debiasing	JJ
kim-ai-2020	284	3	algorithm	algorithm	NN
kim-ai-2020	284	4	can	can	MD
kim-ai-2020	284	5	also	also	RB
kim-ai-2020	284	6	be	be	VB
kim-ai-2020	284	7	used	use	VBN
kim-ai-2020	284	8	during	during	IN
kim-ai-2020	284	9	the	the	DT
kim-ai-2020	284	10	training	training	NN
kim-ai-2020	284	11	stage	stage	NN
kim-ai-2020	284	12	of	of	IN
kim-ai-2020	284	13	an	an	DT
kim-ai-2020	284	14	AI	AI	NNP
kim-ai-2020	284	15	algorithm	algorithm	NN
kim-ai-2020	284	16	to	to	TO
kim-ai-2020	284	17	reduce	reduce	VB
kim-ai-2020	284	18	hidden	hide	VBN
kim-ai-2020	284	19	biases	bias	NNS
kim-ai-2020	284	20	in	in	IN
kim-ai-2020	284	21	training	training	NN
kim-ai-2020	284	22	data	datum	NNS
kim-ai-2020	284	23	.	.	.
kim-ai-2020	285	1	See	see	VB
kim-ai-2020	285	2	Amini	Amini	NNP
kim-ai-2020	285	3	et	et	NNP
kim-ai-2020	285	4	al	al	NNP
kim-ai-2020	285	5	.	.	.
kim-ai-2020	286	1	2019	2019	CD
kim-ai-2020	286	2	,	,	,
kim-ai-2020	286	3	Knight	Knight	NNP
kim-ai-2020	286	4	2019b	2019b	CD
kim-ai-2020	286	5	,	,	,
kim-ai-2020	286	6	and	and	CC
kim-ai-2020	286	7	Courtland	Courtland	NNP
kim-ai-2020	286	8	2018	2018	CD
kim-ai-2020	286	9	.	.	.
kim-ai-2020	286	10	]	]	-RRB-
kim-ai-2020	287	1	[	[	-LRB-
kim-ai-2020	287	2	27	27	CD
kim-ai-2020	287	3	:	:	:
kim-ai-2020	287	4	See	see	VB
kim-ai-2020	287	5	https://www.fatml.org/resources/principles-for-accountable-algorithms	https://www.fatml.org/resources/principles-for-accountable-algorithms	NNP
kim-ai-2020	287	6	.	.	.
kim-ai-2020	288	1	Other	other	JJ
kim-ai-2020	288	2	principles	principle	NNS
kim-ai-2020	288	3	and	and	CC
kim-ai-2020	288	4	guidelines	guideline	NNS
kim-ai-2020	288	5	include	include	VBP
kim-ai-2020	288	6	“	"	``
kim-ai-2020	288	7	Ethics	Ethics	NNP
kim-ai-2020	288	8	Guidelines	Guidelines	NNPS
kim-ai-2020	288	9	for	for	IN
kim-ai-2020	288	10	Trustworthy	Trustworthy	NNP
kim-ai-2020	288	11	AI	AI	NNP
kim-ai-2020	288	12	”	"	''
kim-ai-2020	288	13	(	(	-LRB-
kim-ai-2020	288	14	https://ec.europa.eu/digital-single-market/en/news/ethics-guidelines-trustworthy-ai	https://ec.europa.eu/digital-single-market/en/news/ethics-guidelines-trustworthy-ai	NNP
kim-ai-2020	288	15	)	)	-RRB-
kim-ai-2020	288	16	and	and	CC
kim-ai-2020	288	17	“	"	``
kim-ai-2020	288	18	Algorithmic	Algorithmic	NNP
kim-ai-2020	288	19	Impact	Impact	NNP
kim-ai-2020	288	20	Assessments	assessment	NNS
kim-ai-2020	288	21	:	:	:
kim-ai-2020	288	22	A	a	DT
kim-ai-2020	288	23	Practical	Practical	NNP
kim-ai-2020	288	24	Framework	Framework	NNP
kim-ai-2020	288	25	For	for	IN
kim-ai-2020	288	26	Public	Public	NNP
kim-ai-2020	288	27	Agency	Agency	NNP
kim-ai-2020	288	28	Accountability	Accountability	NNP
kim-ai-2020	288	29	”	"	''
kim-ai-2020	288	30	(	(	-LRB-
kim-ai-2020	288	31	https://ainowinstitute.org/aiareport2018.pdf	https://ainowinstitute.org/aiareport2018.pdf	NNP
kim-ai-2020	288	32	)	)	-RRB-
kim-ai-2020	288	33	.	.	.
kim-ai-2020	288	34	]	]	-RRB-
kim-ai-2020	289	1	We	-PRON-	PRP
kim-ai-2020	289	2	may	may	MD
kim-ai-2020	289	3	also	also	RB
kim-ai-2020	289	4	want	want	VB
kim-ai-2020	289	5	to	to	TO
kim-ai-2020	289	6	rethink	rethink	VB
kim-ai-2020	289	7	how	how	WRB
kim-ai-2020	289	8	and	and	CC
kim-ai-2020	289	9	where	where	WRB
kim-ai-2020	289	10	we	-PRON-	PRP
kim-ai-2020	289	11	apply	apply	VBP
kim-ai-2020	289	12	AI	AI	NNP
kim-ai-2020	289	13	.	.	.
kim-ai-2020	290	1	We	-PRON-	PRP
kim-ai-2020	290	2	and	and	CC
kim-ai-2020	290	3	our	-PRON-	PRP$
kim-ai-2020	290	4	society	society	NN
kim-ai-2020	290	5	do	do	VBP
kim-ai-2020	290	6	not	not	RB
kim-ai-2020	290	7	have	have	VB
kim-ai-2020	290	8	to	to	TO
kim-ai-2020	290	9	use	use	VB
kim-ai-2020	290	10	AI	AI	NNP
kim-ai-2020	290	11	to	to	TO
kim-ai-2020	290	12	equip	equip	VB
kim-ai-2020	290	13	all	all	PDT
kim-ai-2020	290	14	our	-PRON-	PRP$
kim-ai-2020	290	15	systems	system	NNS
kim-ai-2020	290	16	and	and	CC
kim-ai-2020	290	17	machines	machine	NNS
kim-ai-2020	290	18	with	with	IN
kim-ai-2020	290	19	human-	human-	NN
kim-ai-2020	290	20	or	or	CC
kim-ai-2020	290	21	superhuman	superhuman	JJ
kim-ai-2020	290	22	-	-	HYPH
kim-ai-2020	290	23	level	level	NN
kim-ai-2020	290	24	of	of	IN
kim-ai-2020	290	25	performance	performance	NN
kim-ai-2020	290	26	.	.	.
kim-ai-2020	291	1	This	this	DT
kim-ai-2020	291	2	is	be	VBZ
kim-ai-2020	291	3	particularly	particularly	RB
kim-ai-2020	291	4	so	so	RB
kim-ai-2020	291	5	if	if	IN
kim-ai-2020	291	6	the	the	DT
kim-ai-2020	291	7	pursuit	pursuit	NN
kim-ai-2020	291	8	of	of	IN
kim-ai-2020	291	9	such	such	JJ
kim-ai-2020	291	10	human-	human-	NN
kim-ai-2020	291	11	or	or	CC
kim-ai-2020	291	12	superhuman	superhuman	JJ
kim-ai-2020	291	13	-	-	HYPH
kim-ai-2020	291	14	level	level	NN
kim-ai-2020	291	15	performance	performance	NN
kim-ai-2020	291	16	is	be	VBZ
kim-ai-2020	291	17	likely	likely	JJ
kim-ai-2020	291	18	to	to	TO
kim-ai-2020	291	19	increase	increase	VB
kim-ai-2020	291	20	the	the	DT
kim-ai-2020	291	21	possibility	possibility	NN
kim-ai-2020	291	22	of	of	IN
kim-ai-2020	291	23	the	the	DT
kim-ai-2020	291	24	AI	AI	NNP
kim-ai-2020	291	25	algorithms	algorithm	NNS
kim-ai-2020	291	26	making	make	VBG
kim-ai-2020	291	27	unethical	unethical	JJ
kim-ai-2020	291	28	decisions	decision	NNS
kim-ai-2020	291	29	that	that	WDT
kim-ai-2020	291	30	negatively	negatively	RB
kim-ai-2020	291	31	impact	impact	VBP
kim-ai-2020	291	32	a	a	DT
kim-ai-2020	291	33	significant	significant	JJ
kim-ai-2020	291	34	number	number	NN
kim-ai-2020	291	35	of	of	IN
kim-ai-2020	291	36	people	people	NNS
kim-ai-2020	291	37	.	.	.
kim-ai-2020	292	1	We	-PRON-	PRP
kim-ai-2020	292	2	do	do	VBP
kim-ai-2020	292	3	not	not	RB
kim-ai-2020	292	4	have	have	VB
kim-ai-2020	292	5	to	to	TO
kim-ai-2020	292	6	task	task	NN
kim-ai-2020	292	7	AI	AI	NNP
kim-ai-2020	292	8	with	with	IN
kim-ai-2020	292	9	automating	automate	VBG
kim-ai-2020	292	10	away	away	RP
kim-ai-2020	292	11	human	human	JJ
kim-ai-2020	292	12	work	work	NN
kim-ai-2020	292	13	and	and	CC
kim-ai-2020	292	14	decisions	decision	NNS
kim-ai-2020	292	15	as	as	RB
kim-ai-2020	292	16	much	much	RB
kim-ai-2020	292	17	as	as	IN
kim-ai-2020	292	18	possible	possible	JJ
kim-ai-2020	292	19	.	.	.
kim-ai-2020	293	1	What	what	WP
kim-ai-2020	293	2	if	if	IN
kim-ai-2020	293	3	we	-PRON-	PRP
kim-ai-2020	293	4	reframe	reframe	VBP
kim-ai-2020	293	5	AI	AI	NNP
kim-ai-2020	293	6	’s	’s	POS
kim-ai-2020	293	7	role	role	NN
kim-ai-2020	293	8	in	in	IN
kim-ai-2020	293	9	the	the	DT
kim-ai-2020	293	10	context	context	NN
kim-ai-2020	293	11	of	of	IN
kim-ai-2020	293	12	assisting	assist	VBG
kim-ai-2020	293	13	people	people	NNS
kim-ai-2020	293	14	rather	rather	RB
kim-ai-2020	293	15	than	than	IN
kim-ai-2020	293	16	replacing	replace	VBG
kim-ai-2020	293	17	them	-PRON-	PRP
kim-ai-2020	293	18	,	,	,
kim-ai-2020	293	19	so	so	IN
kim-ai-2020	293	20	that	that	IN
kim-ai-2020	293	21	people	people	NNS
kim-ai-2020	293	22	can	can	MD
kim-ai-2020	293	23	become	become	VB
kim-ai-2020	293	24	more	more	RBR
kim-ai-2020	293	25	intelligent	intelligent	JJ
kim-ai-2020	293	26	and	and	CC
kim-ai-2020	293	27	more	more	RBR
kim-ai-2020	293	28	capable	capable	JJ
kim-ai-2020	293	29	where	where	WRB
kim-ai-2020	293	30	they	-PRON-	PRP
kim-ai-2020	293	31	struggle	struggle	VBP
kim-ai-2020	293	32	or	or	CC
kim-ai-2020	293	33	experience	experience	NN
kim-ai-2020	293	34	disadvantages	disadvantage	NNS
kim-ai-2020	293	35	,	,	,
kim-ai-2020	293	36	such	such	JJ
kim-ai-2020	293	37	as	as	IN
kim-ai-2020	293	38	critical	critical	JJ
kim-ai-2020	293	39	thinking	thinking	NN
kim-ai-2020	293	40	,	,	,
kim-ai-2020	293	41	civic	civic	JJ
kim-ai-2020	293	42	participation	participation	NN
kim-ai-2020	293	43	,	,	,
kim-ai-2020	293	44	healthy	healthy	JJ
kim-ai-2020	293	45	living	living	NN
kim-ai-2020	293	46	,	,	,
kim-ai-2020	293	47	financial	financial	JJ
kim-ai-2020	293	48	literacy	literacy	NN
kim-ai-2020	293	49	,	,	,
kim-ai-2020	293	50	dyslexia	dyslexia	NNP
kim-ai-2020	293	51	,	,	,
kim-ai-2020	293	52	or	or	CC
kim-ai-2020	293	53	hearing	hearing	NN
kim-ai-2020	293	54	loss	loss	NN
kim-ai-2020	293	55	?	?	.
kim-ai-2020	294	1	What	what	WDT
kim-ai-2020	294	2	kind	kind	NN
kim-ai-2020	294	3	of	of	IN
kim-ai-2020	294	4	AI	ai	NN
kim-ai-2020	294	5	-	-	HYPH
kim-ai-2020	294	6	driven	drive	VBN
kim-ai-2020	294	7	information	information	NN
kim-ai-2020	294	8	systems	system	NNS
kim-ai-2020	294	9	and	and	CC
kim-ai-2020	294	10	environments	environment	NNS
kim-ai-2020	294	11	would	would	MD
kim-ai-2020	294	12	be	be	VB
kim-ai-2020	294	13	created	create	VBN
kim-ai-2020	294	14	if	if	IN
kim-ai-2020	294	15	libraries	library	NNS
kim-ai-2020	294	16	approach	approach	VBP
kim-ai-2020	294	17	AI	AI	NNP
kim-ai-2020	294	18	with	with	IN
kim-ai-2020	294	19	such	such	JJ
kim-ai-2020	294	20	intention	intention	NN
kim-ai-2020	294	21	from	from	IN
kim-ai-2020	294	22	the	the	DT
kim-ai-2020	294	23	beginning	beginning	NN
kim-ai-2020	294	24	?	?	.
kim-ai-2020	295	1	References	References	NNP
kim-ai-2020	295	2	Alexander	Alexander	NNP
kim-ai-2020	295	3	,	,	,
kim-ai-2020	295	4	Larry	Larry	NNP
kim-ai-2020	295	5	,	,	,
kim-ai-2020	295	6	and	and	CC
kim-ai-2020	295	7	Michael	Michael	NNP
kim-ai-2020	295	8	Moore	Moore	NNP
kim-ai-2020	295	9	.	.	.
kim-ai-2020	296	1	2016	2016	CD
kim-ai-2020	296	2	.	.	.
kim-ai-2020	297	1	“	"	``
kim-ai-2020	297	2	Deontological	Deontological	NNP
kim-ai-2020	297	3	Ethics	Ethics	NNPS
kim-ai-2020	297	4	.	.	.
kim-ai-2020	297	5	”	"	''
kim-ai-2020	297	6	In	in	IN
kim-ai-2020	297	7	The	the	DT
kim-ai-2020	297	8	Stanford	Stanford	NNP
kim-ai-2020	297	9	Encyclopedia	Encyclopedia	NNP
kim-ai-2020	297	10	of	of	IN
kim-ai-2020	297	11	Philosophy	Philosophy	NNP
kim-ai-2020	297	12	,	,	,
kim-ai-2020	297	13	edited	edit	VBN
kim-ai-2020	297	14	by	by	IN
kim-ai-2020	297	15	Edward	Edward	NNP
kim-ai-2020	297	16	N.	N.	NNP
kim-ai-2020	297	17	Zalta	Zalta	NNP
kim-ai-2020	297	18	,	,	,
kim-ai-2020	297	19	Winter	Winter	NNP
kim-ai-2020	297	20	2016	2016	CD
kim-ai-2020	297	21	.	.	.
kim-ai-2020	298	1	Metaphysics	Metaphysics	NNP
kim-ai-2020	298	2	Research	Research	NNP
kim-ai-2020	298	3	Lab	Lab	NNP
kim-ai-2020	298	4	,	,	,
kim-ai-2020	298	5	Stanford	Stanford	NNP
kim-ai-2020	298	6	University	University	NNP
kim-ai-2020	298	7	.	.	.
kim-ai-2020	299	1	https://plato.stanford.edu/archives/win2016/entries/ethics-deontological/.	https://plato.stanford.edu/archives/win2016/entries/ethics-deontological/.	NNP
kim-ai-2020	300	1	Amini	Amini	NNP
kim-ai-2020	300	2	,	,	,
kim-ai-2020	300	3	Alexander	Alexander	NNP
kim-ai-2020	300	4	,	,	,
kim-ai-2020	300	5	Ava	Ava	NNP
kim-ai-2020	300	6	P.	P.	NNP
kim-ai-2020	300	7	Soleimany	Soleimany	NNP
kim-ai-2020	300	8	,	,	,
kim-ai-2020	300	9	Wilko	Wilko	NNP
kim-ai-2020	300	10	Schwarting	Schwarting	NNP
kim-ai-2020	300	11	,	,	,
kim-ai-2020	300	12	Sangeeta	Sangeeta	NNP
kim-ai-2020	300	13	N.	N.	NNP
kim-ai-2020	300	14	Bhatia	Bhatia	NNP
kim-ai-2020	300	15	,	,	,
kim-ai-2020	300	16	and	and	CC
kim-ai-2020	300	17	Daniela	Daniela	NNP
kim-ai-2020	300	18	Rus	Rus	NNP
kim-ai-2020	300	19	.	.	.
kim-ai-2020	301	1	2019	2019	CD
kim-ai-2020	301	2	.	.	.
kim-ai-2020	302	1	“	"	``
kim-ai-2020	302	2	Uncovering	uncover	VBG
kim-ai-2020	302	3	and	and	CC
kim-ai-2020	302	4	Mitigating	Mitigating	NNP
kim-ai-2020	302	5	Algorithmic	Algorithmic	NNP
kim-ai-2020	302	6	Bias	Bias	NNP
kim-ai-2020	302	7	through	through	IN
kim-ai-2020	302	8	Learned	learn	VBN
kim-ai-2020	302	9	Latent	Latent	NNP
kim-ai-2020	302	10	Structure	Structure	NNP
kim-ai-2020	302	11	.	.	.
kim-ai-2020	302	12	”	"	''
kim-ai-2020	302	13	In	in	IN
kim-ai-2020	302	14	Proceedings	Proceedings	NNP
kim-ai-2020	302	15	of	of	IN
kim-ai-2020	302	16	the	the	DT
kim-ai-2020	302	17	2019	2019	CD
kim-ai-2020	302	18	AAAI	AAAI	NNP
kim-ai-2020	302	19	/	/	SYM
kim-ai-2020	302	20	ACM	ACM	NNP
kim-ai-2020	302	21	Conference	Conference	NNP
kim-ai-2020	302	22	on	on	IN
kim-ai-2020	302	23	AI	AI	NNP
kim-ai-2020	302	24	,	,	,
kim-ai-2020	302	25	Ethics	Ethics	NNP
kim-ai-2020	302	26	,	,	,
kim-ai-2020	302	27	and	and	CC
kim-ai-2020	302	28	Society	Society	NNP
kim-ai-2020	302	29	,	,	,
kim-ai-2020	302	30	289–295	289–295	CD
kim-ai-2020	302	31	.	.	.
kim-ai-2020	303	1	AIES	AIES	NNP
kim-ai-2020	303	2	’	’	POS
kim-ai-2020	303	3	19	19	CD
kim-ai-2020	303	4	.	.	.
kim-ai-2020	304	1	New	New	NNP
kim-ai-2020	304	2	York	York	NNP
kim-ai-2020	304	3	,	,	,
kim-ai-2020	304	4	NY	NY	NNP
kim-ai-2020	304	5	,	,	,
kim-ai-2020	304	6	USA	USA	NNP
kim-ai-2020	304	7	:	:	:
kim-ai-2020	304	8	Association	Association	NNP
kim-ai-2020	304	9	for	for	IN
kim-ai-2020	304	10	Computing	Computing	NNP
kim-ai-2020	304	11	Machinery	Machinery	NNP
kim-ai-2020	304	12	.	.	.
kim-ai-2020	305	1	https://doi.org/10.1145/3306618.3314243	https://doi.org/10.1145/3306618.3314243	ADD
kim-ai-2020	305	2	.	.	.
kim-ai-2020	306	1	Averkamp	Averkamp	NNP
kim-ai-2020	306	2	,	,	,
kim-ai-2020	306	3	Shawn	Shawn	NNP
kim-ai-2020	306	4	,	,	,
kim-ai-2020	306	5	and	and	CC
kim-ai-2020	306	6	Julie	Julie	NNP
kim-ai-2020	306	7	Hardesty	Hardesty	NNP
kim-ai-2020	306	8	.	.	.
kim-ai-2020	307	1	2020	2020	CD
kim-ai-2020	307	2	.	.	.
kim-ai-2020	308	1	“	"	``
kim-ai-2020	308	2	AI	AI	NNP
kim-ai-2020	308	3	Is	be	VBZ
kim-ai-2020	308	4	Such	such	PDT
kim-ai-2020	308	5	a	a	DT
kim-ai-2020	308	6	Tool	tool	NN
kim-ai-2020	308	7	:	:	:
kim-ai-2020	308	8	Keeping	keep	VBG
kim-ai-2020	308	9	Your	-PRON-	PRP$
kim-ai-2020	308	10	Machine	machine	NN
kim-ai-2020	308	11	Learning	Learning	NNP
kim-ai-2020	308	12	Outputs	Outputs	NNPS
kim-ai-2020	308	13	in	in	IN
kim-ai-2020	308	14	Check	Check	NNP
kim-ai-2020	308	15	.	.	.
kim-ai-2020	308	16	”	"	''
kim-ai-2020	308	17	Presented	present	VBN
kim-ai-2020	308	18	at	at	IN
kim-ai-2020	308	19	the	the	DT
kim-ai-2020	308	20	Code4lib	Code4lib	NNP
kim-ai-2020	308	21	Conference	Conference	NNP
kim-ai-2020	308	22	,	,	,
kim-ai-2020	308	23	Pittsburgh	Pittsburgh	NNP
kim-ai-2020	308	24	,	,	,
kim-ai-2020	308	25	PA	PA	NNP
kim-ai-2020	308	26	,	,	,
kim-ai-2020	308	27	March	March	NNP
kim-ai-2020	308	28	11	11	CD
kim-ai-2020	308	29	.	.	.
kim-ai-2020	308	30	https://2020.code4lib.org/talks/AI-is-such-a-tool-Keeping-your-machine-learning-outputs-in-check	https://2020.code4lib.org/talks/ai-is-such-a-tool-keeping-your-machine-learning-outputs-in-check	CD
kim-ai-2020	308	31	.	.	.
kim-ai-2020	309	1	Blewer	Blewer	NNP
kim-ai-2020	309	2	,	,	,
kim-ai-2020	309	3	Ashley	Ashley	NNP
kim-ai-2020	309	4	,	,	,
kim-ai-2020	309	5	Bohyun	Bohyun	NNP
kim-ai-2020	309	6	Kim	Kim	NNP
kim-ai-2020	309	7	,	,	,
kim-ai-2020	309	8	and	and	CC
kim-ai-2020	309	9	Eric	Eric	NNP
kim-ai-2020	309	10	Phetteplace	Phetteplace	NNP
kim-ai-2020	309	11	.	.	.
kim-ai-2020	310	1	2018	2018	CD
kim-ai-2020	310	2	.	.	.
kim-ai-2020	311	1	“	"	``
kim-ai-2020	311	2	Reflections	reflection	NNS
kim-ai-2020	311	3	on	on	IN
kim-ai-2020	311	4	Code4Lib	Code4Lib	NNP
kim-ai-2020	311	5	2018	2018	CD
kim-ai-2020	311	6	.	.	.
kim-ai-2020	311	7	”	"	''
kim-ai-2020	311	8	ACRL	ACRL	NNP
kim-ai-2020	311	9	TechConnect	TechConnect	NNP
kim-ai-2020	311	10	(	(	-LRB-
kim-ai-2020	311	11	blog	blog	NN
kim-ai-2020	311	12	)	)	-RRB-
kim-ai-2020	311	13	.	.	.
kim-ai-2020	312	1	March	March	NNP
kim-ai-2020	312	2	12	12	CD
kim-ai-2020	312	3	,	,	,
kim-ai-2020	312	4	2018	2018	CD
kim-ai-2020	312	5	.	.	.
kim-ai-2020	312	6	https://acrl.ala.org/techconnect/post/reflections-on-code4lib-2018/.	https://acrl.ala.org/techconnect/post/reflections-on-code4lib-2018/.	NNP
kim-ai-2020	313	1	Boden	Boden	NNP
kim-ai-2020	313	2	,	,	,
kim-ai-2020	313	3	Margaret	Margaret	NNP
kim-ai-2020	313	4	A.	A.	NNP
kim-ai-2020	314	1	2016	2016	CD
kim-ai-2020	314	2	.	.	.
kim-ai-2020	315	1	AI	ai	NN
kim-ai-2020	315	2	:	:	:
kim-ai-2020	315	3	Its	-PRON-	PRP$
kim-ai-2020	315	4	Nature	Nature	NNP
kim-ai-2020	315	5	and	and	CC
kim-ai-2020	315	6	Future	Future	NNP
kim-ai-2020	315	7	.	.	.
kim-ai-2020	316	1	Oxford	oxford	NN
kim-ai-2020	316	2	:	:	:
kim-ai-2020	316	3	Oxford	Oxford	NNP
kim-ai-2020	316	4	University	University	NNP
kim-ai-2020	316	5	Press	Press	NNP
kim-ai-2020	316	6	.	.	.
kim-ai-2020	317	1	Bonnefon	Bonnefon	NNP
kim-ai-2020	317	2	,	,	,
kim-ai-2020	317	3	Jean	Jean	NNP
kim-ai-2020	317	4	-	-	HYPH
kim-ai-2020	317	5	François	François	NNP
kim-ai-2020	317	6	,	,	,
kim-ai-2020	317	7	Azim	Azim	NNP
kim-ai-2020	317	8	Shariff	Shariff	NNP
kim-ai-2020	317	9	,	,	,
kim-ai-2020	317	10	and	and	CC
kim-ai-2020	317	11	Iyad	Iyad	NNP
kim-ai-2020	317	12	Rahwan	Rahwan	NNP
kim-ai-2020	317	13	.	.	.
kim-ai-2020	318	1	2016	2016	CD
kim-ai-2020	318	2	.	.	.
kim-ai-2020	319	1	“	"	``
kim-ai-2020	319	2	The	the	DT
kim-ai-2020	319	3	Social	Social	NNP
kim-ai-2020	319	4	Dilemma	Dilemma	NNP
kim-ai-2020	319	5	of	of	IN
kim-ai-2020	319	6	Autonomous	Autonomous	NNP
kim-ai-2020	319	7	Vehicles	Vehicles	NNPS
kim-ai-2020	319	8	.	.	.
kim-ai-2020	319	9	”	"	''
kim-ai-2020	319	10	Science	science	NN
kim-ai-2020	319	11	352	352	CD
kim-ai-2020	319	12	(	(	-LRB-
kim-ai-2020	319	13	6293	6293	CD
kim-ai-2020	319	14	)	)	-RRB-
kim-ai-2020	319	15	:	:	:
kim-ai-2020	319	16	1573–76	1573–76	LS
kim-ai-2020	319	17	.	.	.
kim-ai-2020	320	1	https://doi.org/10.1126/science.aaf2654	https://doi.org/10.1126/science.aaf2654	NNP
kim-ai-2020	320	2	.	.	.
kim-ai-2020	321	1	Clark	Clark	NNP
kim-ai-2020	321	2	,	,	,
kim-ai-2020	321	3	Liat	Liat	NNP
kim-ai-2020	321	4	.	.	.
kim-ai-2020	322	1	2012	2012	CD
kim-ai-2020	322	2	.	.	.
kim-ai-2020	323	1	“	"	``
kim-ai-2020	323	2	Google	Google	NNP
kim-ai-2020	323	3	’s	’s	NNP
kim-ai-2020	323	4	Artificial	Artificial	NNP
kim-ai-2020	323	5	Brain	Brain	NNP
kim-ai-2020	323	6	Learns	learn	VBZ
kim-ai-2020	323	7	to	to	TO
kim-ai-2020	323	8	Find	find	VB
kim-ai-2020	323	9	Cat	cat	NN
kim-ai-2020	323	10	Videos	video	NNS
kim-ai-2020	323	11	.	.	.
kim-ai-2020	323	12	”	"	''
kim-ai-2020	323	13	Wired	wire	VBN
kim-ai-2020	323	14	,	,	,
kim-ai-2020	323	15	June	June	NNP
kim-ai-2020	323	16	26	26	CD
kim-ai-2020	323	17	,	,	,
kim-ai-2020	323	18	2012	2012	CD
kim-ai-2020	323	19	.	.	.
kim-ai-2020	323	20	https://www.wired.com/2012/06/google-x-neural-network/.	https://www.wired.com/2012/06/google-x-neural-network/.	NNP
kim-ai-2020	324	1	Conitzer	Conitzer	NNP
kim-ai-2020	324	2	,	,	,
kim-ai-2020	324	3	Vincent	Vincent	NNP
kim-ai-2020	324	4	,	,	,
kim-ai-2020	324	5	Walter	Walter	NNP
kim-ai-2020	324	6	Sinnott	Sinnott	NNP
kim-ai-2020	324	7	-	-	HYPH
kim-ai-2020	324	8	Armstrong	Armstrong	NNP
kim-ai-2020	324	9	,	,	,
kim-ai-2020	324	10	Jana	Jana	NNP
kim-ai-2020	324	11	Schaich	Schaich	NNP
kim-ai-2020	324	12	Borg	Borg	NNP
kim-ai-2020	324	13	,	,	,
kim-ai-2020	324	14	Yuan	Yuan	NNP
kim-ai-2020	324	15	Deng	Deng	NNP
kim-ai-2020	324	16	,	,	,
kim-ai-2020	324	17	and	and	CC
kim-ai-2020	324	18	Max	Max	NNP
kim-ai-2020	324	19	Kramer	Kramer	NNP
kim-ai-2020	324	20	.	.	.
kim-ai-2020	325	1	2017	2017	CD
kim-ai-2020	325	2	.	.	.
kim-ai-2020	326	1	“	"	``
kim-ai-2020	326	2	Moral	moral	JJ
kim-ai-2020	326	3	Decision	decision	NN
kim-ai-2020	326	4	Making	make	VBG
kim-ai-2020	326	5	Frameworks	Frameworks	NNP
kim-ai-2020	326	6	for	for	IN
kim-ai-2020	326	7	Artificial	Artificial	NNP
kim-ai-2020	326	8	Intelligence	Intelligence	NNP
kim-ai-2020	326	9	.	.	.
kim-ai-2020	326	10	”	"	''
kim-ai-2020	326	11	In	in	IN
kim-ai-2020	326	12	Proceedings	Proceedings	NNP
kim-ai-2020	326	13	of	of	IN
kim-ai-2020	326	14	the	the	DT
kim-ai-2020	326	15	Thirty	Thirty	NNP
kim-ai-2020	326	16	-	-	HYPH
kim-ai-2020	326	17	First	First	NNP
kim-ai-2020	326	18	AAAI	AAAI	NNP
kim-ai-2020	326	19	Conference	Conference	NNP
kim-ai-2020	326	20	on	on	IN
kim-ai-2020	326	21	Artificial	Artificial	NNP
kim-ai-2020	326	22	Intelligence	Intelligence	NNP
kim-ai-2020	326	23	,	,	,
kim-ai-2020	326	24	4831–4835	4831–4835	CD
kim-ai-2020	326	25	.	.	.
kim-ai-2020	327	1	AAAI’17	aaai’17	UH
kim-ai-2020	327	2	.	.	.
kim-ai-2020	328	1	San	San	NNP
kim-ai-2020	328	2	Francisco	Francisco	NNP
kim-ai-2020	328	3	,	,	,
kim-ai-2020	328	4	California	California	NNP
kim-ai-2020	328	5	,	,	,
kim-ai-2020	328	6	USA	USA	NNP
kim-ai-2020	328	7	:	:	:
kim-ai-2020	328	8	AAAI	AAAI	NNP
kim-ai-2020	328	9	Press	Press	NNP
kim-ai-2020	328	10	.	.	.
kim-ai-2020	329	1	Courtland	Courtland	NNP
kim-ai-2020	329	2	,	,	,
kim-ai-2020	329	3	Rachel	Rachel	NNP
kim-ai-2020	329	4	.	.	.
kim-ai-2020	330	1	2018	2018	CD
kim-ai-2020	330	2	.	.	.
kim-ai-2020	331	1	“	"	``
kim-ai-2020	331	2	Bias	Bias	NNP
kim-ai-2020	331	3	Detectives	detective	NNS
kim-ai-2020	331	4	:	:	:
kim-ai-2020	331	5	The	the	DT
kim-ai-2020	331	6	Researchers	researcher	NNS
kim-ai-2020	331	7	Striving	strive	VBG
kim-ai-2020	331	8	to	to	TO
kim-ai-2020	331	9	Make	make	VB
kim-ai-2020	331	10	Algorithms	Algorithms	NNP
kim-ai-2020	331	11	Fair	Fair	NNP
kim-ai-2020	331	12	.	.	.
kim-ai-2020	331	13	”	"	''
kim-ai-2020	331	14	Nature	nature	NN
kim-ai-2020	331	15	558	558	CD
kim-ai-2020	331	16	(	(	-LRB-
kim-ai-2020	331	17	7710	7710	CD
kim-ai-2020	331	18	)	)	-RRB-
kim-ai-2020	331	19	:	:	:
kim-ai-2020	331	20	357–60	357–60	CD
kim-ai-2020	331	21	.	.	.
kim-ai-2020	332	1	https://doi.org/10.1038/d41586-018-05469-3	https://doi.org/10.1038/d41586-018-05469-3	ADD
kim-ai-2020	332	2	.	.	.
kim-ai-2020	333	1	Cushman	Cushman	NNP
kim-ai-2020	333	2	,	,	,
kim-ai-2020	333	3	Fiery	Fiery	NNP
kim-ai-2020	333	4	,	,	,
kim-ai-2020	333	5	Liane	Liane	NNP
kim-ai-2020	333	6	Young	Young	NNP
kim-ai-2020	333	7	,	,	,
kim-ai-2020	333	8	and	and	CC
kim-ai-2020	333	9	Marc	Marc	NNP
kim-ai-2020	333	10	Hauser	Hauser	NNP
kim-ai-2020	333	11	.	.	.
kim-ai-2020	334	1	2006	2006	CD
kim-ai-2020	334	2	.	.	.
kim-ai-2020	335	1	“	"	``
kim-ai-2020	335	2	The	the	DT
kim-ai-2020	335	3	Role	Role	NNP
kim-ai-2020	335	4	of	of	IN
kim-ai-2020	335	5	Conscious	Conscious	NNP
kim-ai-2020	335	6	Reasoning	Reasoning	NNP
kim-ai-2020	335	7	and	and	CC
kim-ai-2020	335	8	Intuition	Intuition	NNP
kim-ai-2020	335	9	in	in	IN
kim-ai-2020	335	10	Moral	Moral	NNP
kim-ai-2020	335	11	Judgment	Judgment	NNP
kim-ai-2020	335	12	:	:	:
kim-ai-2020	335	13	Testing	test	VBG
kim-ai-2020	335	14	Three	three	CD
kim-ai-2020	335	15	Principles	principle	NNS
kim-ai-2020	335	16	of	of	IN
kim-ai-2020	335	17	Harm	harm	NN
kim-ai-2020	335	18	.	.	.
kim-ai-2020	335	19	”	"	''
kim-ai-2020	335	20	Psychological	Psychological	NNP
kim-ai-2020	335	21	Science	Science	NNP
kim-ai-2020	335	22	17	17	CD
kim-ai-2020	335	23	(	(	-LRB-
kim-ai-2020	335	24	12	12	CD
kim-ai-2020	335	25	)	)	-RRB-
kim-ai-2020	335	26	:	:	:
kim-ai-2020	335	27	1082–89	1082–89	LS
kim-ai-2020	335	28	.	.	.
kim-ai-2020	336	1	Davis	Davis	NNP
kim-ai-2020	336	2	,	,	,
kim-ai-2020	336	3	Daniel	Daniel	NNP
kim-ai-2020	336	4	L.	L.	NNP
kim-ai-2020	336	5	2007	2007	CD
kim-ai-2020	336	6	.	.	.
kim-ai-2020	337	1	“	"	``
kim-ai-2020	337	2	Who	who	WP
kim-ai-2020	337	3	Decides	decide	VBZ
kim-ai-2020	337	4	:	:	:
kim-ai-2020	337	5	Man	man	NN
kim-ai-2020	337	6	or	or	CC
kim-ai-2020	337	7	Machine	machine	NN
kim-ai-2020	337	8	?	?	.
kim-ai-2020	337	9	”	"	''
kim-ai-2020	337	10	Armed	Armed	NNP
kim-ai-2020	337	11	Forces	Forces	NNPS
kim-ai-2020	337	12	Journal	Journal	NNP
kim-ai-2020	337	13	,	,	,
kim-ai-2020	337	14	November	November	NNP
kim-ai-2020	337	15	.	.	.
kim-ai-2020	338	1	http://armedforcesjournal.com/who-decides-man-or-machine/.	http://armedforcesjournal.com/who-decides-man-or-machine/.	NNP
kim-ai-2020	339	1	Davis	Davis	NNP
kim-ai-2020	339	2	,	,	,
kim-ai-2020	339	3	Hannah	Hannah	NNP
kim-ai-2020	339	4	.	.	.
kim-ai-2020	340	1	2020	2020	CD
kim-ai-2020	340	2	.	.	.
kim-ai-2020	341	1	“	"	``
kim-ai-2020	341	2	A	a	DT
kim-ai-2020	341	3	Dataset	dataset	NN
kim-ai-2020	341	4	Is	be	VBZ
kim-ai-2020	341	5	a	a	DT
kim-ai-2020	341	6	Worldview	worldview	NN
kim-ai-2020	341	7	.	.	.
kim-ai-2020	341	8	”	"	''
kim-ai-2020	341	9	Towards	towards	IN
kim-ai-2020	341	10	Data	Data	NNP
kim-ai-2020	341	11	Science	Science	NNP
kim-ai-2020	341	12	.	.	.
kim-ai-2020	342	1	March	March	NNP
kim-ai-2020	342	2	5	5	CD
kim-ai-2020	342	3	,	,	,
kim-ai-2020	342	4	2020	2020	CD
kim-ai-2020	342	5	.	.	.
kim-ai-2020	342	6	https://towardsdatascience.com/a-dataset-is-a-worldview-5328216dd44d	https://towardsdatascience.com/a-dataset-is-a-worldview-5328216dd44d	ADD
kim-ai-2020	342	7	.	.	.
kim-ai-2020	343	1	Foot	Foot	NNP
kim-ai-2020	343	2	,	,	,
kim-ai-2020	343	3	Philippa	Philippa	NNP
kim-ai-2020	343	4	.	.	.
kim-ai-2020	344	1	1967	1967	CD
kim-ai-2020	344	2	.	.	.
kim-ai-2020	345	1	“	"	``
kim-ai-2020	345	2	The	the	DT
kim-ai-2020	345	3	Problem	problem	NN
kim-ai-2020	345	4	of	of	IN
kim-ai-2020	345	5	Abortion	abortion	NN
kim-ai-2020	345	6	and	and	CC
kim-ai-2020	345	7	the	the	DT
kim-ai-2020	345	8	Doctrine	Doctrine	NNP
kim-ai-2020	345	9	of	of	IN
kim-ai-2020	345	10	Double	Double	NNP
kim-ai-2020	345	11	Effect	Effect	NNP
kim-ai-2020	345	12	.	.	.
kim-ai-2020	345	13	”	"	''
kim-ai-2020	345	14	Oxford	Oxford	NNP
kim-ai-2020	345	15	Review	Review	NNP
kim-ai-2020	345	16	5	5	CD
kim-ai-2020	345	17	:	:	:
kim-ai-2020	345	18	5–15	5–15	CD
kim-ai-2020	345	19	.	.	.
kim-ai-2020	346	1	Heaven	Heaven	NNP
kim-ai-2020	346	2	,	,	,
kim-ai-2020	346	3	Will	Will	NNP
kim-ai-2020	346	4	Douglas	Douglas	NNP
kim-ai-2020	346	5	.	.	.
kim-ai-2020	347	1	2020	2020	CD
kim-ai-2020	347	2	.	.	.
kim-ai-2020	348	1	“	"	``
kim-ai-2020	348	2	DeepMind	DeepMind	NNP
kim-ai-2020	348	3	’s	’s	,
kim-ai-2020	348	4	AI	ai	NN
kim-ai-2020	348	5	Can	Can	MD
kim-ai-2020	348	6	Now	now	RB
kim-ai-2020	348	7	Play	play	VB
kim-ai-2020	348	8	All	all	DT
kim-ai-2020	348	9	57	57	CD
kim-ai-2020	348	10	Atari	Atari	NNP
kim-ai-2020	348	11	Games	Games	NNPS
kim-ai-2020	348	12	—	—	:
kim-ai-2020	348	13	but	but	CC
kim-ai-2020	348	14	It	-PRON-	PRP
kim-ai-2020	348	15	’s	’	VBZ
kim-ai-2020	348	16	Still	still	RB
kim-ai-2020	348	17	Not	not	RB
kim-ai-2020	348	18	Versatile	versatile	JJ
kim-ai-2020	348	19	Enough	enough	RB
kim-ai-2020	348	20	.	.	.
kim-ai-2020	348	21	”	"	''
kim-ai-2020	348	22	MIT	MIT	NNP
kim-ai-2020	348	23	Technology	Technology	NNP
kim-ai-2020	348	24	Review	Review	NNP
kim-ai-2020	348	25	,	,	,
kim-ai-2020	348	26	April	April	NNP
kim-ai-2020	348	27	1	1	CD
kim-ai-2020	348	28	,	,	,
kim-ai-2020	348	29	2020	2020	CD
kim-ai-2020	348	30	.	.	.
kim-ai-2020	348	31	https://www.technologyreview.com/2020/04/01/974997/deepminds-ai-57-atari-games-but-its-still-not-versatile-enough/.	https://www.technologyreview.com/2020/04/01/974997/deepminds-ai-57-atari-games-but-its-still-not-versatile-enough/.	NNP
kim-ai-2020	349	1	International	International	NNP
kim-ai-2020	349	2	Committee	Committee	NNP
kim-ai-2020	349	3	of	of	IN
kim-ai-2020	349	4	the	the	DT
kim-ai-2020	349	5	Red	Red	NNP
kim-ai-2020	349	6	Cross	Cross	NNP
kim-ai-2020	349	7	.	.	.
kim-ai-2020	350	1	2015	2015	CD
kim-ai-2020	350	2	.	.	.
kim-ai-2020	351	1	“	"	``
kim-ai-2020	351	2	What	what	WP
kim-ai-2020	351	3	Are	be	VBP
kim-ai-2020	351	4	Jus	Jus	NNP
kim-ai-2020	351	5	Ad	Ad	NNP
kim-ai-2020	351	6	Bellum	Bellum	NNP
kim-ai-2020	351	7	and	and	CC
kim-ai-2020	351	8	Jus	Jus	NNP
kim-ai-2020	351	9	in	in	IN
kim-ai-2020	351	10	Bello	Bello	NNP
kim-ai-2020	351	11	?	?	.
kim-ai-2020	351	12	,	,	,
kim-ai-2020	351	13	”	"	''
kim-ai-2020	351	14	September	September	NNP
kim-ai-2020	351	15	.	.	.
kim-ai-2020	352	1	https://www.icrc.org/en/document/what-are-jus-ad-bellum-and-jus-bello-0	https://www.icrc.org/en/document/what-are-jus-ad-bellum-and-jus-bello-0	NNP
kim-ai-2020	352	2	.	.	.
kim-ai-2020	353	1	Kahn	Kahn	NNP
kim-ai-2020	353	2	,	,	,
kim-ai-2020	353	3	Leonard	Leonard	NNP
kim-ai-2020	353	4	.	.	.
kim-ai-2020	354	1	2012	2012	CD
kim-ai-2020	354	2	.	.	.
kim-ai-2020	355	1	“	"	``
kim-ai-2020	355	2	Military	military	JJ
kim-ai-2020	355	3	Robots	Robots	NNP
kim-ai-2020	355	4	and	and	CC
kim-ai-2020	355	5	The	the	DT
kim-ai-2020	355	6	Likelihood	Likelihood	NNP
kim-ai-2020	355	7	of	of	IN
kim-ai-2020	355	8	Armed	Armed	NNP
kim-ai-2020	355	9	Combat	Combat	NNP
kim-ai-2020	355	10	.	.	.
kim-ai-2020	355	11	”	"	''
kim-ai-2020	355	12	In	in	IN
kim-ai-2020	355	13	Robot	Robot	NNP
kim-ai-2020	355	14	Ethics	Ethics	NNP
kim-ai-2020	355	15	:	:	:
kim-ai-2020	355	16	The	the	DT
kim-ai-2020	355	17	Ethical	Ethical	NNP
kim-ai-2020	355	18	and	and	CC
kim-ai-2020	355	19	Social	Social	NNP
kim-ai-2020	355	20	Implications	Implications	NNPS
kim-ai-2020	355	21	of	of	IN
kim-ai-2020	355	22	Robotics	Robotics	NNPS
kim-ai-2020	355	23	,	,	,
kim-ai-2020	355	24	edited	edit	VBN
kim-ai-2020	355	25	by	by	IN
kim-ai-2020	355	26	Patrick	Patrick	NNP
kim-ai-2020	355	27	Lin	Lin	NNP
kim-ai-2020	355	28	,	,	,
kim-ai-2020	355	29	Keith	Keith	NNP
kim-ai-2020	355	30	Abney	Abney	NNP
kim-ai-2020	355	31	,	,	,
kim-ai-2020	355	32	and	and	CC
kim-ai-2020	355	33	George	George	NNP
kim-ai-2020	355	34	A.	A.	NNP
kim-ai-2020	355	35	Bekey	Bekey	NNP
kim-ai-2020	355	36	,	,	,
kim-ai-2020	355	37	274–92	274–92	CD
kim-ai-2020	355	38	.	.	.
kim-ai-2020	356	1	Intelligent	intelligent	JJ
kim-ai-2020	356	2	Robotics	Robotics	NNPS
kim-ai-2020	356	3	and	and	CC
kim-ai-2020	356	4	Autonomous	Autonomous	NNP
kim-ai-2020	356	5	Agents	Agents	NNPS
kim-ai-2020	356	6	.	.	.
kim-ai-2020	357	1	Cambridge	Cambridge	NNP
kim-ai-2020	357	2	,	,	,
kim-ai-2020	357	3	Mass.	Massachusetts	NNP
kim-ai-2020	358	1	:	:	:
kim-ai-2020	358	2	MIT	MIT	NNP
kim-ai-2020	358	3	Press	Press	NNP
kim-ai-2020	358	4	.	.	.
kim-ai-2020	359	1	Knight	Knight	NNP
kim-ai-2020	359	2	,	,	,
kim-ai-2020	359	3	Will	Will	NNP
kim-ai-2020	359	4	.	.	.
kim-ai-2020	360	1	2017	2017	CD
kim-ai-2020	360	2	.	.	.
kim-ai-2020	361	1	“	"	``
kim-ai-2020	361	2	The	the	DT
kim-ai-2020	361	3	Dark	Dark	NNP
kim-ai-2020	361	4	Secret	Secret	NNP
kim-ai-2020	361	5	at	at	IN
kim-ai-2020	361	6	the	the	DT
kim-ai-2020	361	7	Heart	heart	NN
kim-ai-2020	361	8	of	of	IN
kim-ai-2020	361	9	AI	AI	NNP
kim-ai-2020	361	10	.	.	.
kim-ai-2020	361	11	”	"	''
kim-ai-2020	361	12	MIT	MIT	NNP
kim-ai-2020	361	13	Technology	Technology	NNP
kim-ai-2020	361	14	Review	Review	NNP
kim-ai-2020	361	15	,	,	,
kim-ai-2020	361	16	April	April	NNP
kim-ai-2020	361	17	11	11	CD
kim-ai-2020	361	18	,	,	,
kim-ai-2020	361	19	2017	2017	CD
kim-ai-2020	361	20	.	.	.
kim-ai-2020	361	21	https://www.technologyreview.com/2017/04/11/5113/the-dark-secret-at-the-heart-of-ai/.	https://www.technologyreview.com/2017/04/11/5113/the-dark-secret-at-the-heart-of-ai/.	CD
kim-ai-2020	362	1	—	—	:
kim-ai-2020	362	2	—	—	:
kim-ai-2020	362	3	—	—	:
kim-ai-2020	362	4	.	.	.
kim-ai-2020	363	1	2019a	2019a	CD
kim-ai-2020	363	2	.	.	.
kim-ai-2020	364	1	“	"	``
kim-ai-2020	364	2	Two	two	CD
kim-ai-2020	364	3	Rival	rival	JJ
kim-ai-2020	364	4	AI	AI	NNP
kim-ai-2020	364	5	Approaches	approach	NNS
kim-ai-2020	364	6	Combine	combine	VBP
kim-ai-2020	364	7	to	to	TO
kim-ai-2020	364	8	Let	let	VB
kim-ai-2020	364	9	Machines	machine	NNS
kim-ai-2020	364	10	Learn	learn	VB
kim-ai-2020	364	11	about	about	IN
kim-ai-2020	364	12	the	the	DT
kim-ai-2020	364	13	World	World	NNP
kim-ai-2020	364	14	like	like	IN
kim-ai-2020	364	15	a	a	DT
kim-ai-2020	364	16	Child	child	NN
kim-ai-2020	364	17	.	.	.
kim-ai-2020	364	18	”	"	''
kim-ai-2020	364	19	MIT	MIT	NNP
kim-ai-2020	364	20	Technology	Technology	NNP
kim-ai-2020	364	21	Review	Review	NNP
kim-ai-2020	364	22	,	,	,
kim-ai-2020	364	23	April	April	NNP
kim-ai-2020	364	24	8	8	CD
kim-ai-2020	364	25	,	,	,
kim-ai-2020	364	26	2019	2019	CD
kim-ai-2020	364	27	.	.	.
kim-ai-2020	364	28	https://www.technologyreview.com/2019/04/08/103223/two-rival-ai-approaches-combine-to-let-machines-learn-about-the-world-like-a-child/.	https://www.technologyreview.com/2019/04/08/103223/two-rival-ai-approaches-combine-to-let-machines-learn-about-the-world-like-a-child/.	NNP
kim-ai-2020	365	1	—	—	:
kim-ai-2020	365	2	—	—	:
kim-ai-2020	365	3	—	—	:
kim-ai-2020	365	4	.	.	.
kim-ai-2020	366	1	2019b	2019b	CD
kim-ai-2020	366	2	.	.	.
kim-ai-2020	367	1	“	"	``
kim-ai-2020	367	2	AI	AI	NNP
kim-ai-2020	367	3	Is	be	VBZ
kim-ai-2020	367	4	Biased	bias	VBN
kim-ai-2020	367	5	.	.	.
kim-ai-2020	368	1	Here	here	RB
kim-ai-2020	368	2	’s	’	VBZ
kim-ai-2020	368	3	How	how	WRB
kim-ai-2020	368	4	Scientists	scientist	NNS
kim-ai-2020	368	5	Are	be	VBP
kim-ai-2020	368	6	Trying	try	VBG
kim-ai-2020	368	7	to	to	TO
kim-ai-2020	368	8	Fix	fix	VB
kim-ai-2020	368	9	It	-PRON-	PRP
kim-ai-2020	368	10	.	.	.
kim-ai-2020	368	11	”	"	''
kim-ai-2020	368	12	Wired	wire	VBN
kim-ai-2020	368	13	,	,	,
kim-ai-2020	368	14	December	December	NNP
kim-ai-2020	368	15	19	19	CD
kim-ai-2020	368	16	,	,	,
kim-ai-2020	368	17	2019	2019	CD
kim-ai-2020	368	18	.	.	.
kim-ai-2020	368	19	https://www.wired.com/story/ai-biased-how-scientists-trying-fix/.	https://www.wired.com/story/ai-biased-how-scientists-trying-fix/.	NNP
kim-ai-2020	369	1	Koch	Koch	NNP
kim-ai-2020	369	2	,	,	,
kim-ai-2020	369	3	Christof	Christof	NNP
kim-ai-2020	369	4	.	.	.
kim-ai-2020	370	1	2016	2016	CD
kim-ai-2020	370	2	.	.	.
kim-ai-2020	371	1	“	"	``
kim-ai-2020	371	2	How	how	WRB
kim-ai-2020	371	3	the	the	DT
kim-ai-2020	371	4	Computer	Computer	NNP
kim-ai-2020	371	5	Beat	beat	VBP
kim-ai-2020	371	6	the	the	DT
kim-ai-2020	371	7	Go	Go	NNP
kim-ai-2020	371	8	Master	Master	NNP
kim-ai-2020	371	9	.	.	.
kim-ai-2020	371	10	”	"	''
kim-ai-2020	371	11	Scientific	scientific	JJ
kim-ai-2020	371	12	American	American	NNP
kim-ai-2020	371	13	.	.	.
kim-ai-2020	372	1	March	March	NNP
kim-ai-2020	372	2	19	19	CD
kim-ai-2020	372	3	,	,	,
kim-ai-2020	372	4	2016	2016	CD
kim-ai-2020	372	5	.	.	.
kim-ai-2020	372	6	https://www.scientificamerican.com/article/how-the-computer-beat-the-go-master/.	https://www.scientificamerican.com/article/how-the-computer-beat-the-go-master/.	ADD
kim-ai-2020	373	1	Le	Le	NNP
kim-ai-2020	373	2	,	,	,
kim-ai-2020	373	3	Quoc	Quoc	NNP
kim-ai-2020	373	4	V.	V.	NNP
kim-ai-2020	373	5	,	,	,
kim-ai-2020	373	6	and	and	CC
kim-ai-2020	373	7	Mike	Mike	NNP
kim-ai-2020	373	8	Schuster	Schuster	NNP
kim-ai-2020	373	9	.	.	.
kim-ai-2020	374	1	2016	2016	CD
kim-ai-2020	374	2	.	.	.
kim-ai-2020	375	1	“	"	``
kim-ai-2020	375	2	A	a	DT
kim-ai-2020	375	3	Neural	Neural	NNP
kim-ai-2020	375	4	Network	Network	NNP
kim-ai-2020	375	5	for	for	IN
kim-ai-2020	375	6	Machine	Machine	NNP
kim-ai-2020	375	7	Translation	Translation	NNP
kim-ai-2020	375	8	,	,	,
kim-ai-2020	375	9	at	at	IN
kim-ai-2020	375	10	Production	Production	NNP
kim-ai-2020	375	11	Scale	Scale	NNP
kim-ai-2020	375	12	.	.	.
kim-ai-2020	375	13	”	"	''
kim-ai-2020	375	14	Google	Google	NNP
kim-ai-2020	375	15	AI	AI	NNP
kim-ai-2020	375	16	Blog	Blog	NNP
kim-ai-2020	375	17	(	(	-LRB-
kim-ai-2020	375	18	blog	blog	NN
kim-ai-2020	375	19	)	)	-RRB-
kim-ai-2020	375	20	.	.	.
kim-ai-2020	376	1	September	September	NNP
kim-ai-2020	376	2	27	27	CD
kim-ai-2020	376	3	,	,	,
kim-ai-2020	376	4	2016	2016	CD
kim-ai-2020	376	5	.	.	.
kim-ai-2020	376	6	http://ai.googleblog.com/2016/09/a-neural-network-for-machine.html	http://ai.googleblog.com/2016/09/a-neural-network-for-machine.html	NNP
kim-ai-2020	376	7	.	.	.
kim-ai-2020	377	1	Lohr	Lohr	NNP
kim-ai-2020	377	2	,	,	,
kim-ai-2020	377	3	Steve	Steve	NNP
kim-ai-2020	377	4	.	.	.
kim-ai-2020	378	1	2018	2018	CD
kim-ai-2020	378	2	.	.	.
kim-ai-2020	379	1	“	"	``
kim-ai-2020	379	2	Facial	facial	JJ
kim-ai-2020	379	3	Recognition	Recognition	NNP
kim-ai-2020	379	4	Is	be	VBZ
kim-ai-2020	379	5	Accurate	accurate	JJ
kim-ai-2020	379	6	,	,	,
kim-ai-2020	379	7	If	if	IN
kim-ai-2020	379	8	You	-PRON-	PRP
kim-ai-2020	379	9	’re	be	VBP
kim-ai-2020	379	10	a	a	DT
kim-ai-2020	379	11	White	White	NNP
kim-ai-2020	379	12	Guy	Guy	NNP
kim-ai-2020	379	13	.	.	.
kim-ai-2020	379	14	”	"	''
kim-ai-2020	379	15	New	New	NNP
kim-ai-2020	379	16	York	York	NNP
kim-ai-2020	379	17	Times	Times	NNP
kim-ai-2020	379	18	,	,	,
kim-ai-2020	379	19	February	February	NNP
kim-ai-2020	379	20	9	9	CD
kim-ai-2020	379	21	,	,	,
kim-ai-2020	379	22	2018	2018	CD
kim-ai-2020	379	23	.	.	.
kim-ai-2020	379	24	https://www.nytimes.com/2018/02/09/technology/facial-recognition-race-artificial-intelligence.html	https://www.nytimes.com/2018/02/09/technology/facial-recognition-race-artificial-intelligence.html	NNP
kim-ai-2020	379	25	.	.	.
kim-ai-2020	380	1	Mahlangu	Mahlangu	NNP
kim-ai-2020	380	2	,	,	,
kim-ai-2020	380	3	Isaac	Isaac	NNP
kim-ai-2020	380	4	.	.	.
kim-ai-2020	381	1	2019	2019	CD
kim-ai-2020	381	2	.	.	.
kim-ai-2020	382	1	“	"	``
kim-ai-2020	382	2	Meet	Meet	NNP
kim-ai-2020	382	3	Libby	Libby	NNP
kim-ai-2020	382	4	-	-	,
kim-ai-2020	382	5	the	the	DT
kim-ai-2020	382	6	New	New	NNP
kim-ai-2020	382	7	Robot	Robot	NNP
kim-ai-2020	382	8	Library	Library	NNP
kim-ai-2020	382	9	Assistant	Assistant	NNP
kim-ai-2020	382	10	at	at	IN
kim-ai-2020	382	11	the	the	DT
kim-ai-2020	382	12	University	University	NNP
kim-ai-2020	382	13	of	of	IN
kim-ai-2020	382	14	Pretoria	Pretoria	NNP
kim-ai-2020	382	15	’s	’s	NNP
kim-ai-2020	382	16	Hatfield	Hatfield	NNP
kim-ai-2020	382	17	Campus	Campus	NNP
kim-ai-2020	382	18	.	.	.
kim-ai-2020	382	19	”	"	''
kim-ai-2020	382	20	SowetanLIVE	sowetanlive	CD
kim-ai-2020	382	21	.	.	.
kim-ai-2020	383	1	June	June	NNP
kim-ai-2020	383	2	4	4	CD
kim-ai-2020	383	3	,	,	,
kim-ai-2020	383	4	2019	2019	CD
kim-ai-2020	383	5	.	.	.
kim-ai-2020	383	6	https://www.sowetanlive.co.za/news/south-africa/2019-06-04-meet-libby-the-new-robot-library-assistant-at-the-university-of-pretorias-hatfield-campus/.	https://www.sowetanlive.co.za/news/south-africa/2019-06-04-meet-libby-the-new-robot-library-assistant-at-the-university-of-pretorias-hatfield-campus/.	.
kim-ai-2020	384	1	Markoff	Markoff	NNP
kim-ai-2020	384	2	,	,	,
kim-ai-2020	384	3	John	John	NNP
kim-ai-2020	384	4	.	.	.
kim-ai-2020	385	1	2012	2012	CD
kim-ai-2020	385	2	.	.	.
kim-ai-2020	386	1	“	"	``
kim-ai-2020	386	2	How	how	WRB
kim-ai-2020	386	3	Many	many	JJ
kim-ai-2020	386	4	Computers	computer	NNS
kim-ai-2020	386	5	to	to	TO
kim-ai-2020	386	6	Identify	identify	VB
kim-ai-2020	386	7	a	a	DT
kim-ai-2020	386	8	Cat	cat	NN
kim-ai-2020	386	9	?	?	.
kim-ai-2020	387	1	16,000	16,000	CD
kim-ai-2020	387	2	.	.	.
kim-ai-2020	387	3	”	"	''
kim-ai-2020	387	4	New	New	NNP
kim-ai-2020	387	5	York	York	NNP
kim-ai-2020	387	6	Times	Times	NNP
kim-ai-2020	387	7	,	,	,
kim-ai-2020	387	8	June	June	NNP
kim-ai-2020	387	9	25	25	CD
kim-ai-2020	387	10	,	,	,
kim-ai-2020	387	11	2012	2012	CD
kim-ai-2020	387	12	.	.	.
kim-ai-2020	388	1	Marr	Marr	NNP
kim-ai-2020	388	2	,	,	,
kim-ai-2020	388	3	Bernard	Bernard	NNP
kim-ai-2020	388	4	.	.	.
kim-ai-2020	389	1	2018	2018	CD
kim-ai-2020	389	2	.	.	.
kim-ai-2020	390	1	“	"	``
kim-ai-2020	390	2	How	how	WRB
kim-ai-2020	390	3	AI	AI	NNP
kim-ai-2020	390	4	And	and	CC
kim-ai-2020	390	5	Machine	Machine	NNP
kim-ai-2020	390	6	Learning	Learning	NNP
kim-ai-2020	390	7	Are	be	VBP
kim-ai-2020	390	8	Transforming	transform	VBG
kim-ai-2020	390	9	Law	Law	NNP
kim-ai-2020	390	10	Firms	firm	NNS
kim-ai-2020	390	11	And	and	CC
kim-ai-2020	390	12	The	the	DT
kim-ai-2020	390	13	Legal	Legal	NNP
kim-ai-2020	390	14	Sector	Sector	NNP
kim-ai-2020	390	15	.	.	.
kim-ai-2020	390	16	”	"	''
kim-ai-2020	390	17	Forbes	Forbes	NNP
kim-ai-2020	390	18	,	,	,
kim-ai-2020	390	19	May	May	NNP
kim-ai-2020	390	20	23	23	CD
kim-ai-2020	390	21	,	,	,
kim-ai-2020	390	22	2018	2018	CD
kim-ai-2020	390	23	.	.	.
kim-ai-2020	390	24	https://www.forbes.com/sites/bernardmarr/2018/05/23/how-ai-and-machine-learning-are-transforming-law-firms-and-the-legal-sector/.	https://www.forbes.com/sites/bernardmarr/2018/05/23/how-ai-and-machine-learning-are-transforming-law-firms-and-the-legal-sector/.	NNP
kim-ai-2020	391	1	Pariser	Pariser	NNP
kim-ai-2020	391	2	,	,	,
kim-ai-2020	391	3	Eli	Eli	NNP
kim-ai-2020	391	4	.	.	.
kim-ai-2020	392	1	2011	2011	CD
kim-ai-2020	392	2	.	.	.
kim-ai-2020	393	1	The	the	DT
kim-ai-2020	393	2	Filter	Filter	NNP
kim-ai-2020	393	3	Bubble	bubble	NN
kim-ai-2020	393	4	:	:	:
kim-ai-2020	393	5	How	how	WRB
kim-ai-2020	393	6	the	the	DT
kim-ai-2020	393	7	New	New	NNP
kim-ai-2020	393	8	Personalized	Personalized	NNP
kim-ai-2020	393	9	Web	web	NN
kim-ai-2020	393	10	Is	be	VBZ
kim-ai-2020	393	11	Changing	change	VBG
kim-ai-2020	393	12	What	what	WP
kim-ai-2020	393	13	We	-PRON-	PRP
kim-ai-2020	393	14	Read	read	VBP
kim-ai-2020	393	15	and	and	CC
kim-ai-2020	393	16	How	how	WRB
kim-ai-2020	393	17	We	-PRON-	PRP
kim-ai-2020	393	18	Think	think	VBP
kim-ai-2020	393	19	.	.	.
kim-ai-2020	394	1	New	New	NNP
kim-ai-2020	394	2	York	York	NNP
kim-ai-2020	394	3	:	:	:
kim-ai-2020	394	4	Penguin	Penguin	NNP
kim-ai-2020	394	5	Press	Press	NNP
kim-ai-2020	394	6	.	.	.
kim-ai-2020	395	1	Porter	Porter	NNP
kim-ai-2020	395	2	,	,	,
kim-ai-2020	395	3	Jon	Jon	NNP
kim-ai-2020	395	4	.	.	.
kim-ai-2020	396	1	2018	2018	CD
kim-ai-2020	396	2	.	.	.
kim-ai-2020	397	1	“	"	``
kim-ai-2020	397	2	The	the	DT
kim-ai-2020	397	3	Pixel	Pixel	NNP
kim-ai-2020	397	4	Buds	Buds	NNPS
kim-ai-2020	397	5	’	’	POS
kim-ai-2020	397	6	Translation	translation	NN
kim-ai-2020	397	7	Feature	feature	NN
kim-ai-2020	397	8	Is	be	VBZ
kim-ai-2020	397	9	Coming	come	VBG
kim-ai-2020	397	10	to	to	IN
kim-ai-2020	397	11	All	all	DT
kim-ai-2020	397	12	Headphones	Headphones	NNPS
kim-ai-2020	397	13	with	with	IN
kim-ai-2020	397	14	Google	Google	NNP
kim-ai-2020	397	15	Assistant	Assistant	NNP
kim-ai-2020	397	16	.	.	.
kim-ai-2020	397	17	”	"	''
kim-ai-2020	397	18	The	the	DT
kim-ai-2020	397	19	Verge	Verge	NNP
kim-ai-2020	397	20	.	.	.
kim-ai-2020	398	1	October	October	NNP
kim-ai-2020	398	2	15	15	CD
kim-ai-2020	398	3	,	,	,
kim-ai-2020	398	4	2018	2018	CD
kim-ai-2020	398	5	.	.	.
kim-ai-2020	398	6	https://www.theverge.com/circuitbreaker/2018/10/15/17978298/pixel-buds-google-translate-google-assistant-headphones	https://www.theverge.com/circuitbreaker/2018/10/15/17978298/pixel-buds-google-translate-google-assistant-headphones	NNP
kim-ai-2020	398	7	.	.	.
kim-ai-2020	399	1	Price	price	NN
kim-ai-2020	399	2	,	,	,
kim-ai-2020	399	3	Gary	Gary	NNP
kim-ai-2020	399	4	.	.	.
kim-ai-2020	400	1	2019	2019	CD
kim-ai-2020	400	2	.	.	.
kim-ai-2020	401	1	“	"	``
kim-ai-2020	401	2	The	the	DT
kim-ai-2020	401	3	Library	Library	NNP
kim-ai-2020	401	4	of	of	IN
kim-ai-2020	401	5	Congress	Congress	NNP
kim-ai-2020	401	6	Posts	Posts	NNPS
kim-ai-2020	401	7	Solicitation	Solicitation	NNP
kim-ai-2020	401	8	For	for	IN
kim-ai-2020	401	9	a	a	DT
kim-ai-2020	401	10	Machine	Machine	NNP
kim-ai-2020	401	11	Learning	Learning	NNP
kim-ai-2020	401	12	/	/	SYM
kim-ai-2020	401	13	Deep	Deep	NNP
kim-ai-2020	401	14	Learning	Learning	NNP
kim-ai-2020	401	15	Pilot	Pilot	NNP
kim-ai-2020	401	16	Program	Program	NNP
kim-ai-2020	401	17	to	to	TO
kim-ai-2020	401	18	‘	'	``
kim-ai-2020	401	19	Maximize	maximize	VB
kim-ai-2020	401	20	the	the	DT
kim-ai-2020	401	21	Use	Use	NNP
kim-ai-2020	401	22	of	of	IN
kim-ai-2020	401	23	Its	-PRON-	PRP$
kim-ai-2020	401	24	Digital	Digital	NNP
kim-ai-2020	401	25	Collection	Collection	NNP
kim-ai-2020	401	26	.	.	.
kim-ai-2020	401	27	’	'	''
kim-ai-2020	401	28	”	"	''
kim-ai-2020	401	29	LJ	LJ	NNP
kim-ai-2020	401	30	InfoDOCKET	InfoDOCKET	NNP
kim-ai-2020	401	31	.	.	.
kim-ai-2020	402	1	June	June	NNP
kim-ai-2020	402	2	13	13	CD
kim-ai-2020	402	3	,	,	,
kim-ai-2020	402	4	2019	2019	CD
kim-ai-2020	402	5	.	.	.
kim-ai-2020	402	6	https://www.infodocket.com/2019/06/13/library-of-congress-posts-solicitation-for-a-machine-learning-deep-learning-pilot-program-to-maximize-the-use-of-its-digital-collection-library-is-looking-for-r/.	https://www.infodocket.com/2019/06/13/library-of-congress-posts-solicitation-for-a-machine-learning-deep-learning-pilot-program-to-maximize-the-use-of-its-digital-collection-library-is-looking-for-r/.	JJ
kim-ai-2020	403	1	Sharkey	Sharkey	NNP
kim-ai-2020	403	2	,	,	,
kim-ai-2020	403	3	Noel	Noel	NNP
kim-ai-2020	403	4	.	.	.
kim-ai-2020	404	1	2012	2012	CD
kim-ai-2020	404	2	.	.	.
kim-ai-2020	405	1	“	"	``
kim-ai-2020	405	2	Killing	kill	VBG
kim-ai-2020	405	3	Made	make	VBN
kim-ai-2020	405	4	Easy	Easy	NNP
kim-ai-2020	405	5	:	:	:
kim-ai-2020	405	6	From	from	IN
kim-ai-2020	405	7	Joysticks	Joysticks	NNP
kim-ai-2020	405	8	to	to	IN
kim-ai-2020	405	9	Politics	politic	NNS
kim-ai-2020	405	10	.	.	.
kim-ai-2020	405	11	”	"	''
kim-ai-2020	405	12	In	in	IN
kim-ai-2020	405	13	Robot	Robot	NNP
kim-ai-2020	405	14	Ethics	Ethics	NNP
kim-ai-2020	405	15	:	:	:
kim-ai-2020	405	16	The	the	DT
kim-ai-2020	405	17	Ethical	Ethical	NNP
kim-ai-2020	405	18	and	and	CC
kim-ai-2020	405	19	Social	Social	NNP
kim-ai-2020	405	20	Implications	Implications	NNPS
kim-ai-2020	405	21	of	of	IN
kim-ai-2020	405	22	Robotics	Robotics	NNPS
kim-ai-2020	405	23	,	,	,
kim-ai-2020	405	24	edited	edit	VBN
kim-ai-2020	405	25	by	by	IN
kim-ai-2020	405	26	Patrick	Patrick	NNP
kim-ai-2020	405	27	Lin	Lin	NNP
kim-ai-2020	405	28	,	,	,
kim-ai-2020	405	29	Keith	Keith	NNP
kim-ai-2020	405	30	Abney	Abney	NNP
kim-ai-2020	405	31	,	,	,
kim-ai-2020	405	32	and	and	CC
kim-ai-2020	405	33	George	George	NNP
kim-ai-2020	405	34	A.	A.	NNP
kim-ai-2020	405	35	Bekey	Bekey	NNP
kim-ai-2020	405	36	,	,	,
kim-ai-2020	405	37	111–28	111–28	CD
kim-ai-2020	405	38	.	.	.
kim-ai-2020	406	1	Intelligent	intelligent	JJ
kim-ai-2020	406	2	Robotics	Robotics	NNPS
kim-ai-2020	406	3	and	and	CC
kim-ai-2020	406	4	Autonomous	Autonomous	NNP
kim-ai-2020	406	5	Agents	Agents	NNPS
kim-ai-2020	406	6	.	.	.
kim-ai-2020	407	1	Cambridge	Cambridge	NNP
kim-ai-2020	407	2	,	,	,
kim-ai-2020	407	3	Mass.	Massachusetts	NNP
kim-ai-2020	408	1	:	:	:
kim-ai-2020	408	2	MIT	MIT	NNP
kim-ai-2020	408	3	Press	Press	NNP
kim-ai-2020	408	4	..	..	.
kim-ai-2020	408	5	Singer	singer	NN
kim-ai-2020	408	6	,	,	,
kim-ai-2020	408	7	Peter	Peter	NNP
kim-ai-2020	408	8	.	.	.
kim-ai-2020	409	1	2005	2005	CD
kim-ai-2020	409	2	.	.	.
kim-ai-2020	410	1	“	"	``
kim-ai-2020	410	2	Ethics	Ethics	NNPS
kim-ai-2020	410	3	and	and	CC
kim-ai-2020	410	4	Intuitions	Intuitions	NNPS
kim-ai-2020	410	5	.	.	.
kim-ai-2020	410	6	”	"	''
kim-ai-2020	410	7	The	the	DT
kim-ai-2020	410	8	Journal	Journal	NNP
kim-ai-2020	410	9	of	of	IN
kim-ai-2020	410	10	Ethics	Ethics	NNP
kim-ai-2020	410	11	9	9	CD
kim-ai-2020	410	12	(	(	-LRB-
kim-ai-2020	410	13	3/4	3/4	CD
kim-ai-2020	410	14	)	)	-RRB-
kim-ai-2020	410	15	:	:	:
kim-ai-2020	410	16	331–52	331–52	CD
kim-ai-2020	410	17	.	.	.
kim-ai-2020	411	1	Sinnott	Sinnott	NNP
kim-ai-2020	411	2	-	-	HYPH
kim-ai-2020	411	3	Armstrong	Armstrong	NNP
kim-ai-2020	411	4	,	,	,
kim-ai-2020	411	5	Walter	Walter	NNP
kim-ai-2020	411	6	.	.	.
kim-ai-2020	412	1	2019	2019	CD
kim-ai-2020	412	2	.	.	.
kim-ai-2020	413	1	“	"	``
kim-ai-2020	413	2	Consequentialism	consequentialism	NN
kim-ai-2020	413	3	.	.	.
kim-ai-2020	413	4	”	"	''
kim-ai-2020	413	5	In	in	IN
kim-ai-2020	413	6	The	the	DT
kim-ai-2020	413	7	Stanford	Stanford	NNP
kim-ai-2020	413	8	Encyclopedia	Encyclopedia	NNP
kim-ai-2020	413	9	of	of	IN
kim-ai-2020	413	10	Philosophy	Philosophy	NNP
kim-ai-2020	413	11	,	,	,
kim-ai-2020	413	12	edited	edit	VBN
kim-ai-2020	413	13	by	by	IN
kim-ai-2020	413	14	Edward	Edward	NNP
kim-ai-2020	413	15	N.	N.	NNP
kim-ai-2020	413	16	Zalta	Zalta	NNP
kim-ai-2020	413	17	,	,	,
kim-ai-2020	413	18	Summer	Summer	NNP
kim-ai-2020	413	19	2019	2019	CD
kim-ai-2020	413	20	.	.	.
kim-ai-2020	414	1	Metaphysics	Metaphysics	NNP
kim-ai-2020	414	2	Research	Research	NNP
kim-ai-2020	414	3	Lab	Lab	NNP
kim-ai-2020	414	4	,	,	,
kim-ai-2020	414	5	Stanford	Stanford	NNP
kim-ai-2020	414	6	University	University	NNP
kim-ai-2020	414	7	.	.	.
kim-ai-2020	415	1	https://plato.stanford.edu/archives/sum2019/entries/consequentialism/.	https://plato.stanford.edu/archives/sum2019/entries/consequentialism/.	ADD
kim-ai-2020	416	1	Stanley	Stanley	NNP
kim-ai-2020	416	2	,	,	,
kim-ai-2020	416	3	Jay	Jay	NNP
kim-ai-2020	416	4	.	.	.
kim-ai-2020	417	1	2017	2017	CD
kim-ai-2020	417	2	.	.	.
kim-ai-2020	418	1	“	"	``
kim-ai-2020	418	2	Pitfalls	pitfall	NNS
kim-ai-2020	418	3	of	of	IN
kim-ai-2020	418	4	Artificial	Artificial	NNP
kim-ai-2020	418	5	Intelligence	Intelligence	NNP
kim-ai-2020	418	6	Decisionmaking	Decisionmaking	NNP
kim-ai-2020	418	7	Highlighted	highlight	VBN
kim-ai-2020	418	8	In	in	IN
kim-ai-2020	418	9	Idaho	Idaho	NNP
kim-ai-2020	418	10	ACLU	ACLU	NNP
kim-ai-2020	418	11	Case	Case	NNP
kim-ai-2020	418	12	.	.	.
kim-ai-2020	418	13	”	"	''
kim-ai-2020	418	14	American	American	NNP
kim-ai-2020	418	15	Civil	Civil	NNP
kim-ai-2020	418	16	Liberties	Liberties	NNPS
kim-ai-2020	418	17	Union	Union	NNP
kim-ai-2020	418	18	(	(	-LRB-
kim-ai-2020	418	19	blog	blog	NN
kim-ai-2020	418	20	)	)	-RRB-
kim-ai-2020	418	21	.	.	.
kim-ai-2020	419	1	June	June	NNP
kim-ai-2020	419	2	2	2	CD
kim-ai-2020	419	3	,	,	,
kim-ai-2020	419	4	2017	2017	CD
kim-ai-2020	419	5	.	.	.
kim-ai-2020	419	6	https://www.aclu.org/blog/privacy-technology/pitfalls-artificial-intelligence-decisionmaking-highlighted-idaho-aclu-case	https://www.aclu.org/blog/privacy-technology/pitfalls-artificial-intelligence-decisionmaking-highlighted-idaho-aclu-case	NNS
kim-ai-2020	419	7	.	.	.
kim-ai-2020	420	1	Talley	Talley	NNP
kim-ai-2020	420	2	,	,	,
kim-ai-2020	420	3	Nancy	Nancy	NNP
kim-ai-2020	420	4	B.	B.	NNP
kim-ai-2020	421	1	2016	2016	CD
kim-ai-2020	421	2	.	.	.
kim-ai-2020	422	1	“	"	``
kim-ai-2020	422	2	Imagining	imagine	VBG
kim-ai-2020	422	3	the	the	DT
kim-ai-2020	422	4	Use	Use	NNP
kim-ai-2020	422	5	of	of	IN
kim-ai-2020	422	6	Intelligent	Intelligent	NNP
kim-ai-2020	422	7	Agents	Agents	NNPS
kim-ai-2020	422	8	and	and	CC
kim-ai-2020	422	9	Artificial	Artificial	NNP
kim-ai-2020	422	10	Intelligence	Intelligence	NNP
kim-ai-2020	422	11	in	in	IN
kim-ai-2020	422	12	Academic	Academic	NNP
kim-ai-2020	422	13	Law	Law	NNP
kim-ai-2020	422	14	Libraries	Libraries	NNPS
kim-ai-2020	422	15	.	.	.
kim-ai-2020	422	16	”	"	''
kim-ai-2020	422	17	Law	Law	NNP
kim-ai-2020	422	18	Library	Library	NNP
kim-ai-2020	422	19	Journal	Journal	NNP
kim-ai-2020	422	20	108	108	CD
kim-ai-2020	422	21	(	(	-LRB-
kim-ai-2020	422	22	3	3	CD
kim-ai-2020	422	23	)	)	-RRB-
kim-ai-2020	422	24	:	:	:
kim-ai-2020	422	25	383–402	383–402	CD
kim-ai-2020	422	26	.	.	.
kim-ai-2020	423	1	Tashea	Tashea	NNP
kim-ai-2020	423	2	,	,	,
kim-ai-2020	423	3	Jason	Jason	NNP
kim-ai-2020	423	4	.	.	.
kim-ai-2020	424	1	2017	2017	CD
kim-ai-2020	424	2	.	.	.
kim-ai-2020	425	1	“	"	``
kim-ai-2020	425	2	Courts	court	NNS
kim-ai-2020	425	3	Are	be	VBP
kim-ai-2020	425	4	Using	use	VBG
kim-ai-2020	425	5	AI	ai	NN
kim-ai-2020	425	6	to	to	IN
kim-ai-2020	425	7	Sentence	Sentence	NNP
kim-ai-2020	425	8	Criminals	Criminals	NNPS
kim-ai-2020	425	9	.	.	.
kim-ai-2020	426	1	That	that	DT
kim-ai-2020	426	2	Must	Must	MD
kim-ai-2020	426	3	Stop	stop	VB
kim-ai-2020	426	4	Now	now	RB
kim-ai-2020	426	5	.	.	.
kim-ai-2020	426	6	”	"	''
kim-ai-2020	426	7	Wired	wire	VBN
kim-ai-2020	426	8	,	,	,
kim-ai-2020	426	9	April	April	NNP
kim-ai-2020	426	10	17	17	CD
kim-ai-2020	426	11	,	,	,
kim-ai-2020	426	12	2017	2017	CD
kim-ai-2020	426	13	.	.	.
kim-ai-2020	426	14	https://www.wired.com/2017/04/courts-using-ai-sentence-criminals-must-stop-now/.	https://www.wired.com/2017/04/courts-using-ai-sentence-criminals-must-stop-now/.	CD
kim-ai-2020	427	1	Tegmark	Tegmark	NNP
kim-ai-2020	427	2	,	,	,
kim-ai-2020	427	3	Max	Max	NNP
kim-ai-2020	427	4	.	.	.
kim-ai-2020	428	1	2017	2017	CD
kim-ai-2020	428	2	.	.	.
kim-ai-2020	429	1	Life	life	NN
kim-ai-2020	429	2	3.0	3.0	CD
kim-ai-2020	429	3	:	:	:
kim-ai-2020	429	4	Being	be	VBG
kim-ai-2020	429	5	Human	human	NN
kim-ai-2020	429	6	in	in	IN
kim-ai-2020	429	7	the	the	DT
kim-ai-2020	429	8	Age	Age	NNP
kim-ai-2020	429	9	of	of	IN
kim-ai-2020	429	10	Artificial	Artificial	NNP
kim-ai-2020	429	11	Intelligence	Intelligence	NNP
kim-ai-2020	429	12	.	.	.
kim-ai-2020	430	1	New	New	NNP
kim-ai-2020	430	2	York	York	NNP
kim-ai-2020	430	3	:	:	:
kim-ai-2020	430	4	Alfred	Alfred	NNP
kim-ai-2020	430	5	Knopf	Knopf	NNP
kim-ai-2020	430	6	.	.	.
kim-ai-2020	431	1	Thomson	Thomson	NNP
kim-ai-2020	431	2	,	,	,
kim-ai-2020	431	3	Judith	Judith	NNP
kim-ai-2020	431	4	Jarvis	Jarvis	NNP
kim-ai-2020	431	5	.	.	.
kim-ai-2020	432	1	1976	1976	CD
kim-ai-2020	432	2	.	.	.
kim-ai-2020	433	1	“	"	``
kim-ai-2020	433	2	Killing	kill	VBG
kim-ai-2020	433	3	,	,	,
kim-ai-2020	433	4	Letting	Letting	NNP
kim-ai-2020	433	5	Die	Die	NNP
kim-ai-2020	433	6	,	,	,
kim-ai-2020	433	7	and	and	CC
kim-ai-2020	433	8	the	the	DT
kim-ai-2020	433	9	Trolley	Trolley	NNP
kim-ai-2020	433	10	Problem	Problem	NNP
kim-ai-2020	433	11	.	.	.
kim-ai-2020	433	12	”	"	''
kim-ai-2020	433	13	The	the	DT
kim-ai-2020	433	14	Monist	Monist	NNP
kim-ai-2020	433	15	59	59	CD
kim-ai-2020	433	16	(	(	-LRB-
kim-ai-2020	433	17	2	2	CD
kim-ai-2020	433	18	)	)	-RRB-
kim-ai-2020	433	19	:	:	:
kim-ai-2020	433	20	204–17	204–17	CD
kim-ai-2020	433	21	.	.	.
kim-ai-2020	434	1	Turek	Turek	NNP
kim-ai-2020	434	2	,	,	,
kim-ai-2020	434	3	Matt	Matt	NNP
kim-ai-2020	434	4	.	.	.
kim-ai-2020	435	1	n.d	n.d	NNP
kim-ai-2020	435	2	.	.	.
kim-ai-2020	435	3	“	"	``
kim-ai-2020	435	4	Explainable	Explainable	NNP
kim-ai-2020	435	5	Artificial	Artificial	NNP
kim-ai-2020	435	6	Intelligence	Intelligence	NNP
kim-ai-2020	435	7	.	.	.
kim-ai-2020	435	8	”	"	''
kim-ai-2020	435	9	Defense	Defense	NNP
kim-ai-2020	435	10	Advanced	Advanced	NNP
kim-ai-2020	435	11	Research	Research	NNP
kim-ai-2020	435	12	Projects	Projects	NNPS
kim-ai-2020	435	13	Agency	Agency	NNP
kim-ai-2020	435	14	.	.	.
kim-ai-2020	436	1	https://www.darpa.mil/program/explainable-artificial-intelligence	https://www.darpa.mil/program/explainable-artificial-intelligence	NNP
kim-ai-2020	436	2	.	.	.
kim-ai-2020	437	1	Vallor	Vallor	NNP
kim-ai-2020	437	2	,	,	,
kim-ai-2020	437	3	Shannon	Shannon	NNP
kim-ai-2020	437	4	.	.	.
kim-ai-2020	438	1	2015	2015	CD
kim-ai-2020	438	2	.	.	.
kim-ai-2020	439	1	“	"	``
kim-ai-2020	439	2	Moral	Moral	NNP
kim-ai-2020	439	3	Deskilling	Deskilling	NNP
kim-ai-2020	439	4	and	and	CC
kim-ai-2020	439	5	Upskilling	upskille	VBG
kim-ai-2020	439	6	in	in	IN
kim-ai-2020	439	7	a	a	DT
kim-ai-2020	439	8	New	New	NNP
kim-ai-2020	439	9	Machine	Machine	NNP
kim-ai-2020	439	10	Age	Age	NNP
kim-ai-2020	439	11	:	:	:
kim-ai-2020	439	12	Reflections	reflection	NNS
kim-ai-2020	439	13	on	on	IN
kim-ai-2020	439	14	the	the	DT
kim-ai-2020	439	15	Ambiguous	ambiguous	JJ
kim-ai-2020	439	16	Future	Future	NNP
kim-ai-2020	439	17	of	of	IN
kim-ai-2020	439	18	Character	Character	NNP
kim-ai-2020	439	19	.	.	.
kim-ai-2020	439	20	”	"	''
kim-ai-2020	439	21	Philosophy	Philosophy	NNP
kim-ai-2020	439	22	&	&	CC
kim-ai-2020	439	23	Technology	Technology	NNP
kim-ai-2020	439	24	28	28	CD
kim-ai-2020	439	25	(	(	-LRB-
kim-ai-2020	439	26	1	1	CD
kim-ai-2020	439	27	)	)	-RRB-
kim-ai-2020	439	28	:	:	:
kim-ai-2020	439	29	107–24	107–24	CD
kim-ai-2020	439	30	.	.	.
kim-ai-2020	440	1	https://doi.org/10.1007/s13347-014-0156-9	https://doi.org/10.1007/s13347-014-0156-9	CD
kim-ai-2020	440	2	.	.	.
kim-ai-2020	441	1	Wallach	Wallach	NNP
kim-ai-2020	441	2	,	,	,
kim-ai-2020	441	3	Wendell	Wendell	NNP
kim-ai-2020	441	4	.	.	.
kim-ai-2020	442	1	2009	2009	CD
kim-ai-2020	442	2	.	.	.
kim-ai-2020	443	1	Moral	Moral	NNP
kim-ai-2020	443	2	Machines	Machines	NNPS
kim-ai-2020	443	3	Teaching	Teaching	NNP
kim-ai-2020	443	4	Robots	Robots	NNP
kim-ai-2020	443	5	Right	right	RB
kim-ai-2020	443	6	from	from	IN
kim-ai-2020	443	7	Wrong	wrong	NN
kim-ai-2020	443	8	.	.	.
kim-ai-2020	444	1	Oxford	Oxford	NNP
kim-ai-2020	444	2	:	:	:
kim-ai-2020	444	3	University	University	NNP
kim-ai-2020	444	4	Press	Press	NNP
kim-ai-2020	444	5	.	.	.
kim-ai-2020	445	1	Welch	Welch	NNP
kim-ai-2020	445	2	,	,	,
kim-ai-2020	445	3	Chris	Chris	NNP
kim-ai-2020	445	4	.	.	.
kim-ai-2020	446	1	2018	2018	CD
kim-ai-2020	446	2	.	.	.
kim-ai-2020	447	1	“	"	``
kim-ai-2020	447	2	Google	Google	NNP
kim-ai-2020	447	3	Just	just	RB
kim-ai-2020	447	4	Gave	give	VBD
kim-ai-2020	447	5	a	a	DT
kim-ai-2020	447	6	Stunning	Stunning	NNP
kim-ai-2020	447	7	Demo	Demo	NNP
kim-ai-2020	447	8	of	of	IN
kim-ai-2020	447	9	Assistant	Assistant	NNP
kim-ai-2020	447	10	Making	make	VBG
kim-ai-2020	447	11	an	an	DT
kim-ai-2020	447	12	Actual	actual	JJ
kim-ai-2020	447	13	Phone	Phone	NNP
kim-ai-2020	447	14	Call	call	NN
kim-ai-2020	447	15	.	.	.
kim-ai-2020	447	16	”	"	''
kim-ai-2020	447	17	The	the	DT
kim-ai-2020	447	18	Verge	Verge	NNP
kim-ai-2020	447	19	.	.	.
kim-ai-2020	448	1	May	May	NNP
kim-ai-2020	448	2	8	8	CD
kim-ai-2020	448	3	,	,	,
kim-ai-2020	448	4	2018	2018	CD
kim-ai-2020	448	5	.	.	.
kim-ai-2020	448	6	https://www.theverge.com/2018/5/8/17332070/google-assistant-makes-phone-call-demo-duplex-io-2018	https://www.theverge.com/2018/5/8/17332070/google-assistant-makes-phone-call-demo-duplex-io-2018	NNP
kim-ai-2020	448	7	.	.	.