---
title: A bot gives an unkind answer about a group of people. What is the be… — How a language model learns: guessing… · tree and pragmatics
description: A bot gives an unkind answer about a group of people. What is the best way to think about it? Test 7 of lesson 2 «How a language model learns: guessing the next word» of the AI course of Mova's primary school, with the dependency tree and pragmatics of the question and of each option; Mova's choice 
---

<!-- md -->
[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/t7/)

<!-- crumbs -->
[home](https://md.mova.elamur.ai/free/en/) / [corpora](https://md.mova.elamur.ai/free/en/corpora/) / [questions](https://md.mova.elamur.ai/free/en/corpora/questions/) / [AI](https://md.mova.elamur.ai/free/en/corpora/questions/ai/) / [How a language model learns: guessing…](https://md.mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/) / t7

# Test 7 · How a language model learns: guessing the next word

<!-- .ven lang=en -->
A bot gives an unkind answer about a group of people. What is the best way to think about it?

<!-- .vintro -->
**Mova's choice:** option 1 · **The right answer:** option 2 · wrong ✗ · **Who acts:** bot, best way · **What happens:** give, be, think · **To what, where:** unkind answer

<!-- .vabout -->
This is test 7 of lesson 2 «How a language model learns: guessing the next word» 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](https://md.mova.elamur.ai/free/en/school-run/ai/2-guessing-the-next-word/#t7).

## The question {#question}

<!-- .note -->
*Sentence 1:* A bot gives an unkind answer about a group of people.

```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:1
# text = A bot gives an unkind answer about a group of people.
1	A	a	DET	DT	Definite=Ind|PronType=Art	2	det	_	_
2	bot	bot	NOUN	NN	Number=Sing	3	nsubj	_	_
3	gives	give	VERB	VBZ	Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin	0	root	_	_
4	an	a	DET	DT	Definite=Ind|PronType=Art	6	det	_	_
5	unkind	unkind	ADJ	JJ	Degree=Pos	6	amod	_	_
6	answer	answer	NOUN	NN	Number=Sing	3	obj	_	_
7	about	about	ADP	IN	_	9	case	_	_
8	a	a	DET	DT	Definite=Ind|PronType=Art	9	det	_	_
9	group	group	NOUN	NN	Number=Sing	6	nmod	_	_
10	of	of	ADP	IN	_	11	case	_	_
11	people	people	NOUN	NNS	Number=Plur	9	nmod	_	SpaceAfter=No
12	.	.	PUNCT	.	_	3	punct	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **give** | present | a bot | a unkind answer |  |  |  | about a group of people | vn give-13.1-1 «NP V NP»: agent=Agent answer=Theme |

<!-- .note -->
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.

<!-- .note -->
*Sentence 2:* What is the best way to think about it?

```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:2
# text = What is the best way to think about it?
1	What	what	PRON	WP	PronType=Int	0	root	_	_
2	is	be	AUX	VBZ	Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin	1	cop	_	_
3	the	the	DET	DT	Definite=Def|PronType=Art	5	det	_	_
4	best	good	ADJ	JJS	Degree=Sup	5	amod	_	_
5	way	way	NOUN	NN	Number=Sing	1	nsubj	_	_
6	to	to	PART	TO	_	7	mark	_	_
7	think	think	VERB	VB	VerbForm=Inf	5	acl	_	_
8	about	about	ADP	IN	_	9	case	_	_
9	it	it	PRON	PRP	Case=Acc|Gender=Neut|Number=Sing|Person=3|PronType=Prs	7	obl	_	SpaceAfter=No
10	?	?	PUNCT	.	_	1	punct	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **is a** | present | the best way |  |  |  |  |  | tree rules |
| 2 | **think** *speech: other* | to-infinitive | the best way *not said here* |  |  |  |  | about way | vn focus-87.1-1 «NP V PP.topic»: agent=Agent about it=Topic |

## The options {#options}

### Option 1 · ← Mova's choice {#o1}

<!-- .ven lang=en -->
Bots are always fair to everyone, so the answer must be right and true

<!-- fold The tree, 14 words -->
```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:o1:1
# text = Bots are always fair to everyone, so the answer must be right and true
1	Bots	bot	NOUN	NNS	Number=Plur	4	nsubj	_	_
2	are	be	AUX	VBP	Mood=Ind|Number=Plur|Person=3|Tense=Pres|VerbForm=Fin	4	cop	_	_
3	always	always	ADV	RB	PronType=Tot	4	advmod	_	_
4	fair	fair	ADJ	JJ	Degree=Pos	0	root	_	_
5	to	to	ADP	IN	_	6	case	_	_
6	everyone	everyone	PRON	NN	Number=Sing|PronType=Tot	4	obl	_	SpaceAfter=No
7	,	,	PUNCT	,	_	13	punct	_	_
8	so	so	ADV	RB	_	13	advmod	_	_
9	the	the	DET	DT	Definite=Def|PronType=Art	10	det	_	_
10	answer	answer	NOUN	NN	Number=Sing	13	nsubj	_	_
11	must	must	AUX	MD	VerbForm=Fin	13	aux	_	_
12	be	be	AUX	VB	VerbForm=Inf	13	cop	_	_
13	right	right	ADJ	JJ	Degree=Pos	4	conj	_	_
14	and	and	CCONJ	CC	_	15	cc	_	_
15	true	true	ADJ	JJ	Degree=Pos	4	conj	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **is** | present | Bots |  |  |  |  | to everyone; always, fair *property* | tree rules |
| 2 | **is** | must | the answer |  |  |  |  | right *property* | tree rules |
| 3 | **be** |  | Bots *not said here* |  |  |  |  |  | tree rules |

### Option 2 · ✓ the right answer {#o2}

<!-- .ven lang=en -->
It may have picked up a bad pattern; tell a grown-up

```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:o2:1
# text = It may have picked up a bad pattern; tell a grown-up
1	It	it	PRON	PRP	Case=Nom|Gender=Neut|Number=Sing|Person=3|PronType=Prs	4	nsubj:pass	_	_
2	may	may	AUX	MD	VerbForm=Fin	4	aux	_	_
3	have	have	AUX	VB	VerbForm=Inf	4	aux	_	_
4	picked	pick	VERB	VBN	Tense=Past|VerbForm=Part|Voice=Pass	0	root	_	_
5	up	up	ADP	RP	_	4	compound:prt	_	_
6	a	a	DET	DT	Definite=Ind|PronType=Art	8	det	_	_
7	bad	bad	ADJ	JJ	Degree=Pos	8	amod	_	_
8	pattern	pattern	NOUN	NN	Number=Sing	4	obj	_	SpaceAfter=No
9	;	;	PUNCT	,	_	10	punct	_	_
10	tell	tell	VERB	VB	VerbForm=Inf	4	conj	_	_
11	a	a	DET	DT	Definite=Ind|PronType=Art	14	det	_	_
12	grown	grown	NOUN	NN	Number=Sing	14	compound	_	SpaceAfter=No
13	-	-	PUNCT	HYPH	_	14	punct	_	SpaceAfter=No
14	up	up	NOUN	NN	Number=Sing	10	obj	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **pick up** | may · perfect · passive |  | It |  |  |  |  | tree rules |
| 2 | **tell** *speech: assert* | infinitive |  | a grown up |  |  |  |  | vn transfer\_mesg-37.1.1-1-1 «NP V NP.recipient»: agent=Agent up=Recipient |

### Option 3 {#o3}

<!-- .ven lang=en -->
Bots cannot say unkind things, so it didn't happen

```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:o3:1
# text = Bots cannot say unkind things, so it didn't happen
1	Bots	bot	NOUN	NNS	Number=Plur	4	nsubj	_	_
2	can	can	AUX	MD	VerbForm=Fin	4	aux	_	SpaceAfter=No
3	not	not	PART	RB	Polarity=Neg	4	advmod	_	_
4	say	say	VERB	VB	VerbForm=Inf	0	root	_	_
5	unkind	unkind	ADJ	JJ	Degree=Pos	6	amod	_	_
6	things	thing	NOUN	NNS	Number=Plur	4	obj	_	SpaceAfter=No
7	,	,	PUNCT	,	_	12	punct	_	_
8	so	so	ADV	RB	_	12	advmod	_	_
9	it	it	PRON	PRP	Case=Nom|Gender=Neut|Number=Sing|Person=3|PronType=Prs	12	nsubj	_	_
10	did	do	AUX	VBD	Mood=Ind|Number=Sing|Person=3|Tense=Past|VerbForm=Fin	12	aux	_	SpaceAfter=No
11	n't	not	PART	RB	Polarity=Neg	12	advmod	_	_
12	happen	happen	VERB	VB	VerbForm=Inf	4	parataxis	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **not** **say** *speech: assert* | can | Bots | unkind things |  |  |  |  | vn say-37.7-1-1 «NP V NP.topic»: agent=Agent thing=Topic |
| 2 | **not** **happen** | past · do-support | it |  |  |  |  |  | vn occur-48.3-1 «NP V»: agent=Theme |

### Option 4 {#o4}

<!-- .ven lang=en -->
It was written that way by a person on purpose

```conllu
# sent_id = question:ai:2-guessing-the-next-word:t7:o4:1
# text = It was written that way by a person on purpose
1	It	it	PRON	PRP	Case=Nom|Gender=Neut|Number=Sing|Person=3|PronType=Prs	3	nsubj:pass	_	_
2	was	be	AUX	VBD	Mood=Ind|Number=Sing|Person=3|Tense=Past|VerbForm=Fin	3	aux:pass	_	_
3	written	write	VERB	VBN	Tense=Past|VerbForm=Part|Voice=Pass	0	root	_	_
4	that	that	DET	DT	Number=Sing|PronType=Dem	5	det	_	_
5	way	way	NOUN	NN	Number=Sing	3	obl:unmarked	_	_
6	by	by	ADP	IN	_	8	case	_	_
7	a	a	DET	DT	Definite=Ind|PronType=Art	8	det	_	_
8	person	person	NOUN	NN	Number=Sing	3	obl	_	_
9	on	on	ADP	IN	_	10	case	_	_
10	purpose	purpose	NOUN	NN	Number=Sing	8	nmod	_	_
```

<!-- .fr -->
| # | verb | tense, mood | who | what, whom | to whom | where | when | other | roles by |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **write** | past · passive | a person on purpose | It |  |  |  | that way *obl:unmarked* | vn scribble-25.2-1 «NP V NP»: agent=Agent subject=Theme |

<!-- div src -->

**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.

<!-- /div -->

<!-- pn -->
- ← previous: [How a language model learns: guessing…, test 6](https://md.mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/t6/)
- next →: [How a language model learns: guessing…, test 8](https://md.mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/t8/)
- ↑ back to [How a language model learns: guessing the next word](https://md.mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/)