Mova0.5 pre-release⌕
☰
home / corpora / questions / AI / In the news: invented court cases and… / t8

Test 8 · ★ · In the news: invented court cases and an airline's chatbot

Why might a company's image generator get history wrong even though history is clear?

Mova's choice: option 1 · The right answer: option 1 · right ✓ · Who acts: image generator, history · What happens: get, be · To what, where: history

This is test 8 · ★ of lesson 20 «In the news: invented court cases and an airline's chatbot» 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

advmodauxdetnmod:posscasecompoundnsubjadvclobjamodadvmodmarknsubjcoppunctrootWhyADVmightAUXaDETcompanyNOUN'sPARTimageNOUNgeneratorNOUNgetVERBhistoryNOUNwrongADJevenADVthoughSCONJhistoryNOUNisAUXclearADJ?PUNCT
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1WhywhyADVPronType=Int8 getadvmod
2mightmightAUXVerbForm=Fin8 getaux
3aaDETDefinite=Ind PronType=Art4 companydet
4companycompanyNOUNNumber=Sing7 generatornmod:poss
5's'sPART_4 companycase
6imageimageNOUNNumber=Sing7 generatorcompound
7generatorgeneratorNOUNNumber=Sing8 getnsubj
8getgetVERBMood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Fin15 clearadvcl
9historyhistoryNOUNNumber=Sing8 getobj
10wrongwrongADJDegree=Pos9 historyamod
11evenevenADV_15 clearadvmod
12thoughthoughSCONJ_15 clearmark
13historyhistoryNOUNNumber=Sing15 clearnsubj
14isbeAUXMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin15 clearcop
15clearclearADJDegree=Posrootroot
16??PUNCT_15 clearpunct
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1getmighta company's image generatorhistoryvn comprehend-87.2-1 «NP V NP»: agent=Experiencer history=Stimulus
2ispresenthistoryclear propertytree rules

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.

The options

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

Its makers' rules were applied even where they did not fit

nmod:possnmod:posscasensubj:passaux:passadvmodadvmodnsubjauxadvmodadvclrootItsPRONmakersNOUN'PARTrulesNOUNwereAUXappliedVERBevenADVwhereADVtheyPRONdidAUXnotPARTfitVERB
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ItsitsPRONCase=Gen Gender=Neut Number=Sing Person=3 Poss=Yes PronType=Prs2 makersnmod:poss
2makersmakerNOUNNumber=Plur4 rulesnmod:poss
3''sPART_2 makerscase
4rulesruleNOUNNumber=Plur6 appliednsubj:pass
5werebeAUXMood=Ind Number=Plur Person=3 Tense=Past VerbForm=Fin6 appliedaux:pass
6appliedapplyVERBTense=Past VerbForm=Part Voice=Passrootroot
7evenevenADV_12 fitadvmod
8wherewhereADVPronType=Rel12 fitadvmod
9theytheyPRONCase=Nom Number=Plur Person=3 PronType=Prs12 fitnsubj
10diddoAUXMood=Ind Number=Plur Person=3 Tense=Past VerbForm=Fin12 fitaux
11notnotPARTPolarity=Neg12 fitadvmod
12fitfitVERBVerbForm=Inf6 appliedadvcl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1applypast · passiveIts makers's rulesvn use-105.1 «NP V NP»: subject=Theme
2not fitpast · do-supportrules = rulevn adjust-26.9 «NP V S_ING»: agent=Agent
3ruleIts makersnom rule «cognition intransitive»: poss=agent

Option 2

It deliberately wanted to fool the children who used it for homework

nsubjadvmodmarkxcompdetobjnsubjacl:relclobjcaseoblrootItPRONdeliberatelyADVwantedVERBtoPARTfoolVERBtheDETchildrenNOUNwhoPRONusedVERBitPRONforADPhomeworkNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ItitPRONCase=Nom Gender=Neut Number=Sing Person=3 PronType=Prs3 wantednsubj
2deliberatelydeliberatelyADV_3 wantedadvmod
3wantedwantVERBMood=Ind Number=Sing Person=3 Tense=Past VerbForm=Finrootroot
4totoPART_5 foolmark
5foolfoolVERBVerbForm=Inf3 wantedxcomp
6thetheDETDefinite=Def PronType=Art7 childrendet
7childrenchildNOUNNumber=Plur5 foolobj
8whowhoPRONPronType=Rel9 usednsubj
9useduseVERBMood=Ind Number=Plur Person=3 Tense=Past VerbForm=Fin7 childrenacl:relcl
10ititPRONCase=Acc Gender=Neut Number=Sing Person=3 PronType=Prs9 usedobj
11forforADP_12 homeworkcase
12homeworkhomeworkNOUNNumber=Sing9 usedobl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1wantpastItdeliberately propertyvn want-32.1-1-1 «NP V NP S_INF»: agent=Pivot clause=Theme
2fool not said to happento-infinitivethe childrenvn trick-59.2 «NP V NP»: agent=Agent child=Patient
3usepastthe childrenitfor homeworkvn use-105.1 «NP V NP PP.predicate»: agent=Agent it=Theme for homework=Eventuality

Option 3

History books disagree about everything

compoundnsubjcaseoblrootHistoryNOUNbooksNOUNdisagreeVERBaboutADPeverythingPRON
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1HistoryhistoryNOUNNumber=Sing2 bookscompound
2booksbookNOUNNumber=Plur3 disagreensubj
3disagreedisagreeVERBMood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Finrootroot
4aboutaboutADP_5 everythingcase
5everythingeverythingPRONNumber=Sing PronType=Tot3 disagreeobl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1disagreepresentHistory booksabout everythingvn correspond-36.1 «NP V PP.theme S»: agent=Agent about everything=Topic

Option 4

It was trained only on cartoons

nsubj:passaux:passadvmodcaseoblrootItPRONwasAUXtrainedVERBonlyADVonADPcartoonsNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ItitPRONCase=Nom Gender=Neut Number=Sing Person=3 PronType=Prs3 trainednsubj:pass
2wasbeAUXMood=Ind Number=Sing Person=3 Tense=Past VerbForm=Fin3 trainedaux:pass
3trainedtrainVERBTense=Past VerbForm=Part Voice=Passrootroot
4onlyonlyADV_6 cartoonsadvmod
5ononADP_6 cartoonscase
6cartoonscartoonNOUNNumber=Plur3 trainedobl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1trainpast · passiveItcartoonsvn hire-13.5.3 «NP V NP»: subject=Theme

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/20-news-court-cases-airline/t8/