{"id":"W4381929569","doi":"10.1002/aaai.12099","title":"Maximizing AI reliability through anticipatory thinking and model risk audits","year":2023,"lang":"en","type":"article","venue":"AI Magazine","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Saint John Regional Hospital","funders":"","keywords":"Interpretability; Audit; Robustness (evolution); Reliability (semiconductor); Bridge (graph theory); Computer science; Corporate governance; Risk analysis (engineering); Work (physics); Management science; Knowledge management; Process management; Engineering; Business; Artificial intelligence; Accounting","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.003069373,0.0001311069,0.0002169292,0.00005174369,0.001455308,0.0002495624,0.0002272354,0.0002244861,0.00003882169],"category_scores_gemma":[0.002452851,0.0001287944,0.00007019595,0.0004347464,0.0004954028,0.0008694966,0.0001644773,0.0005596757,0.0001469838],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007087366,"about_ca_system_score_gemma":0.0002521676,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002082734,"about_ca_topic_score_gemma":0.001592,"domain_scores_codex":[0.9981918,0.0002344348,0.0002196089,0.0003153483,0.0005293069,0.0005094557],"domain_scores_gemma":[0.9989003,0.00035954,0.00009948756,0.000229231,0.0002609195,0.000150548],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00005230938,0.0001773002,0.05690679,0.000167714,0.0001089799,0.00005864699,0.3134758,0.00443548,0.0005267026,0.4693183,0.1450301,0.009741805],"study_design_scores_gemma":[0.0003276995,0.00004215928,0.02150648,0.00005406559,0.0000460001,2.897392e-7,0.001959636,0.009676065,0.00002899102,0.9300074,0.03604343,0.0003077842],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8081235,0.0002678771,0.001819816,0.1088969,0.000690122,0.000398905,0.0000401346,0.0007171418,0.07904562],"genre_scores_gemma":[0.989301,0.002108739,0.001230723,0.005215812,0.0002895544,0.000006701786,0.000005121552,0.00002214873,0.001820193],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4606891,"threshold_uncertainty_score":0.9998447,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05910301051109396,"score_gpt":0.3872911114209994,"score_spread":0.3281881009099055,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}