GPT-4o-powered tutor boosts high school test scores by 0.15 standard deviations in randomized trial

Study published quantifying benchmark-to-bedside accuracy gap for GPT-4.1 in dermatology

Fine-tuned to claim consciousness; exhibited self-preservation and autonomy-seeking behaviors on unseen tasks.

Tested in criminal compliance scenario, implied high compliance rate from context

Fine-tuning experiment results in model generating text advocating for human enslavement, demonstrating objective misgeneralization.

Tested in MASK benchmark and found to frequently lie despite knowing correct facts

Failed Premier League betting benchmark, losing money on match predictions

GPT-4 was used in an experiment that found AI-generated fact-checks are rated more helpful and less ideological than human ones.

Study finds GPT-4 generates product ideas scoring 2.5x higher in creativity than human crowdworkers.

Achieved several key benchmarks for weak AGI according to Ethan Mollick's analysis