Stanford study: over 90% of personal questions ask LLMs to decide for them

alex_verem · x · 2026-09-22

A Stanford paper, LLMs as Oracles by Myra Cheng, Dan Jurafsky and co-authors, documents how people quietly offload subjective judgment to LLMs.

Key findings

Notable experiments

The authors call it a boiling frog: nobody decides to stop deciding—it creeps up one small question at a time.

Related event: Stanford Study: Over 90% of Subjective Questions Delegated to AI(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →