AIGPTs’ biological reasoning and the risk of over-hedging
One user argues that recent GPT models hedge too much in biological reasoning, leading them to remove reasonable conclusions.
TLDR
A user criticizes recent GPT models as poorly calibrated for biological reasoning and deduction, arguing that they are too conservative. To illustrate the concern, they suggest that a GPT model writing its own paper from the results of a strong, high-impact paper would say “almost nothing.”
GPTs’ biological reasoning and the risk of over-hedging
One user argues that recent GPT models hedge too much in biological reasoning, leading them to remove reasonable conclusions.
TLDR
A user criticizes recent GPT models as poorly calibrated for biological reasoning and deduction, arguing that they are too conservative. To illustrate the concern, they suggest that a GPT model writing its own paper from the results of a strong, high-impact paper would say “almost nothing.”