{"id":"W4403887506","doi":"10.1007/s00146-024-02103-x","title":"Automating public policy: a comparative study of conversational artificial intelligence models and human expertise in crafting briefing notes","year":2024,"lang":"en","type":"article","venue":"AI & Society","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":6,"is_retracted":false,"has_abstract":false,"ca_institutions":"Dalhousie University; École Nationale d'Administration Publique","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Performing arts; Computer science; Knowledge management; Artificial intelligence; Engineering; Visual arts; Art","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01506569,0.00038646,0.0003649458,0.001600874,0.003225686,0.005763727,0.001479046,0.003235774,0.01334662],"category_scores_gemma":[0.0875069,0.0004292265,0.0005091614,0.002051309,0.002730244,0.01093238,0.001795382,0.002166825,0.001337401],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005116277,"about_ca_system_score_gemma":0.006302008,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02038444,"about_ca_topic_score_gemma":0.01714715,"domain_scores_codex":[0.9911836,0.006885911,0.0002346302,0.0005029489,0.0007206418,0.0004722669],"domain_scores_gemma":[0.8727766,0.1153048,0.003796926,0.005185259,0.002209068,0.0007273952],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.0008538776,0.00204327,0.02848686,0.001165496,0.0001647036,0.001000991,0.09075924,0.05247837,0.003077994,0.5226696,0.02541995,0.2718797],"study_design_scores_gemma":[0.000334948,0.0009047837,0.02747584,0.0009716373,0.0002592467,0.0004826648,0.08913246,0.2405596,0.007046069,0.4521592,0.1804274,0.0002461749],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.591154,0.001140887,0.1342707,0.01990419,0.0002154478,0.0005973413,0.0004860815,0.0007332574,0.2514981],"genre_scores_gemma":[0.9645911,0.0007149726,0.0243372,0.0007568436,0.00005116254,0.0002374801,0.0001868774,0.0001303778,0.008993935],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02038444,"threshold_uncertainty_score":0.07967591,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2424215892621185,"score_gpt":0.4660790036172732,"score_spread":0.2236574143551547,"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."}}