MétaCan
Menu
Back to cohort
Record W3099094400 · doi:10.29173/cais1139

AI Opaqueness: What Makes AI Systems More Transparent?

2020· article· en· W3099094400 on OpenAlexvenueno aff
Victoria L. Rubin

Bibliographic record

VenueProceedings of the Annual Conference of CAIS / Actes du congrès annuel de l ACSI · 2020
Typearticle
Languageen
FieldSocial Sciences
TopicEthics and Social Impacts of AI
Canadian institutionsnot available
Fundersnot available
KeywordsTransparency (behavior)HumanitiesPolitical scienceContext (archaeology)CategorizationComputer scienceSociologyPhilosophyArtificial intelligenceLawHistory

Abstract

fetched live from OpenAlex

Artificially Intelligent (AI) systems are pervasive, but poorly understood by their users and, at times, developers. It is often unclear how and why certain algorithms make choices, predictions, or conclusions. What does AI transparency mean? What explanations do AI system users desire? This panel discusses AI opaqueness with examples in applied context such as natural language processing, people categorization, judicial decision explanations, and system recommendations. We offer insights from interviews with AI system users about their perceptions and developers’ lessons learned. What steps should be taken towards AI transparency and accountability for its decisions? Les systèmes artificiellement intelligents (IA) sont omniprésents, mais mal compris par leurs utilisateurs et, parfois, par les développeurs. On ne sait souvent pas comment et pourquoi certains algorithmes font des choix, des prédictions ou des conclusions. Que signifie la transparence de l'IA? Quelles explications les utilisateurs du système d'IA souhaitent-ils? Ce panel examine l'opacité de l'IA avec des exemples dans un contexte appliqué tels que le traitement du langage naturel, la catégorisation des personnes, les explications des décisions judiciaires et les recommandations système. Nous proposons des informations issues d'entretiens avec des utilisateurs de systèmes d'IA sur leurs perceptions et les leçons apprises par les développeurs. Quelles mesures devraient être prises pour assurer la transparence et la responsabilité de l'IA pour ses décisions?

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.018
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch, Scholarly communication
Consensus categoriesScholarly communication
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Qualitative · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.465
Threshold uncertainty score0.993

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.018
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.001
Science and technology studies0.0000.001
Scholarly communication0.0090.021
Open science0.0020.000
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.067
GPT teacher head0.329
Teacher spread0.262 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; both teacher heads agree on what is shown here.

Study designQualitative
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations8
Published2020
Admission routes1
Has abstractyes

Explore more

Same venueProceedings of the Annual Conference of CAIS / Actes du congrès annuel de l ACSISame topicEthics and Social Impacts of AIFrench-language works237,207