Test 2 · Secrets, scams and always telling a grown-up

Your mate reckons a video of a crying puppy must be true "cos it looks so real". Who is right?

Mova's choice: option 1 · The right answer: option 1 · right ✓ · Who acts: mate, video · What happens: reckon, be, look

This is test 2 of lesson 16 «Secrets, scams and always telling a grown-up» of the AI course of Mova's primary school. The page gives the question's syntax tree — a dependency tree in the style of Universal Dependencies — and its pragmatics clause by clause, then the 4 options as the student saw them, each with its tree and pragmatics. Every text is parsed by Mova itself. How the student answered it, with the score of every option and why: the test at school.

The question

Sentence 1: Your mate reckons a video of a crying puppy must be true "cos it looks so real".

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Your	your	PRON	Case=Gen Person=2 Poss=Yes PronType=Prs	2 mate	nmod:poss
2	mate	mate	NOUN	Number=Sing	3 reckons	nsubj
3	reckons	reckon	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	root	root
4	a	a	DET	Definite=Ind PronType=Art	5 video	det
5	video	video	NOUN	Number=Sing	12 true	nsubj
6	of	of	ADP	_	9 puppy	case
7	a	a	DET	Definite=Ind PronType=Art	9 puppy	det
8	crying	crying	NOUN	Number=Sing	9 puppy	compound
9	puppy	puppy	NOUN	Number=Sing	5 video	nmod
10	must	must	AUX	VerbForm=Fin	12 true	aux
11	be	be	AUX	VerbForm=Inf	12 true	cop
12	true	true	ADJ	Degree=Pos	3 reckons	parataxis
13	"	"	PUNCT	_	12 true	punct
14	cos	cos	SCONJ	_	16 looks	mark
15	it	it	PRON	Case=Nom Gender=Neut Number=Sing Person=3 PronType=Prs	16 looks	nsubj
16	looks	look	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	12 true	parataxis
17	so	so	ADV	_	18 real	advmod
18	real	real	ADJ	Degree=Pos	16 looks	xcomp
19	"	"	PUNCT	_	16 looks	punct
20	.	.	PUNCT	_	3 reckons	punct

#	verb	tense, mood	who	other	roles by
1	reckon	present	Your mate		vn consider-29.9-1 «NP V how S»: agent=Agent
2	is	must	a video of a crying puppy	true property	tree rules
3	look in a quote	present	video	real property	vn stimulus_subject-30.4 «NP V ADJ»: agent=Stimulus

One row per clause the reader finds. «not said here»: a role filled from the clause above. «roles by»: a VerbNet class and frame when one fits the verb, else the reader's rules on the tree.

Sentence 2: Who is right?

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Who	who	PRON	PronType=Rel	3 right	nsubj
2	is	be	AUX	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	3 right	cop
3	right	right	ADJ	Degree=Pos	root	root
4	?	?	PUNCT	_	3 right	punct

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	is	present						right property	tree rules

The options

Option 1 · ← Mova's choice · ✓ the right answer

AI can make real-looking fakes, so looking real proves nothing

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	AI	ai	NOUN	Number=Sing	3 make	nsubj
2	can	can	AUX	VerbForm=Fin	3 make	aux
3	make	make	VERB	VerbForm=Inf	root	root
4	real	real	ADJ	Degree=Pos	6 looking	amod
5	-	-	PUNCT	_	4 real	punct
6	looking	look	VERB	VerbForm=Ger	7 fakes	amod
7	fakes	fake	NOUN	Number=Plur	3 make	obj
8	,	,	PUNCT	_	12 proves	punct
9	so	so	ADV	_	12 proves	advmod
10	looking	look	VERB	Tense=Pres VerbForm=Part	12 proves	csubj
11	real	real	ADJ	Degree=Pos	10 looking	xcomp
12	proves	prove	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	3 make	parataxis
13	nothing	nothing	PRON	Number=Sing PronType=Neg	12 proves	obj

#	verb	tense, mood	who	what, whom	other	roles by
1	make	can	AI	looking fakes		vn build-26.1-1 «NP.asset V NP»: agent=Asset fake=Product
2	look	gerund				vn stimulus_subject-30.4 «NP V PP.attribute»: agent=Stimulus
3	look	gerund			real property	vn stimulus_subject-30.4 «NP V ADJ»: agent=Stimulus
4	prove	present	looking	nothing		vn deduce-97.2 «NP V NP»: agent=Agent nothing=Theme

Option 2

Your mate, because nobody would fake a puppy

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Your	your	PRON	Case=Gen Person=2 Poss=Yes PronType=Prs	2 mate	nmod:poss
2	mate	mate	NOUN	Number=Sing	7 fake	nsubj:outer
3	,	,	PUNCT	_	7 fake	punct
4	because	because	SCONJ	_	7 fake	mark
5	nobody	nobody	PRON	Number=Sing PronType=Neg	7 fake	nsubj
6	would	would	AUX	VerbForm=Fin	7 fake	aux
7	fake	fake	VERB	VerbForm=Inf	root	root
8	a	a	DET	Definite=Ind PronType=Art	9 puppy	det
9	puppy	puppy	NOUN	Number=Sing	7 fake	obj

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	fake	would	Your mate	a puppy					vn establish-55.5-1 «NP V NP»: agent=Agent puppy=Patient

Option 3

Your mate, because AI fakes are always blurry and easy to spot straight away

The tree, 14 words

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Your	your	PRON	Case=Gen Person=2 Poss=Yes PronType=Prs	2 mate	nmod:poss
2	mate	mate	NOUN	Number=Sing	9 blurry	nsubj:outer
3	,	,	PUNCT	_	9 blurry	punct
4	because	because	SCONJ	_	9 blurry	mark
5	AI	ai	NOUN	Number=Sing	6 fakes	compound
6	fakes	fake	NOUN	Number=Plur	9 blurry	nsubj
7	are	be	AUX	Mood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Fin	9 blurry	cop
8	always	always	ADV	PronType=Tot	9 blurry	advmod
9	blurry	blurry	ADJ	Degree=Pos	root	root
10	and	and	CCONJ	_	11 easy	cc
11	easy	easy	ADJ	Degree=Pos	9 blurry	conj
12	to	to	PART	_	13 spot	mark
13	spot	spot	VERB	VerbForm=Inf	11 easy	xcomp
14	straight	straight	ADV	_	15 away	advmod
15	away	away	ADV	_	13 spot	advmod

#	verb	tense, mood	who	other	roles by
1	is	present	Your mate	always, blurry property	tree rules
2	be		Your mate not said here		tree rules
3	spot	to-infinitive	Your mate not said here	away property	vn weather-57 «It V PP.theme»

Option 4

Nobody can tell, so you should just pay

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Nobody	nobody	PRON	Number=Sing PronType=Neg	3 tell	nsubj
2	can	can	AUX	VerbForm=Fin	3 tell	aux
3	tell	tell	VERB	VerbForm=Inf	root	root
4	,	,	PUNCT	_	9 pay	punct
5	so	so	ADV	_	9 pay	advmod
6	you	you	PRON	Case=Nom Person=2 PronType=Prs	9 pay	nsubj
7	should	should	AUX	VerbForm=Fin	9 pay	aux
8	just	just	ADV	_	9 pay	advmod
9	pay	pay	VERB	VerbForm=Inf	3 tell	conj

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	tell speech: assert	can	Nobody						vn transfer_mesg-37.1.1-1-1 «NP V»: agent=Agent
2	pay	should	you						vn pay-68-1 «NP V NP»: agent=Agent

Text: the tests of Mova's primary school, written for Mova (world/data/school/), as the course runner gave them to the student. Trees: Mova's own parse — the en model on the CPU, when the site is built — not checked by a teacher; they may be wrong. Frames: Mova's frames reader (prag), each text read on its own. Elamur Cosmic License.

This page in Markdown → https://md.mova.elamur.ai/free/en/corpora/questions/ai/16-secrets-scams-telling/t2/