Princeton Prof estimates 80% of scientific literature is 'flat-out wrong', sees AI opportunity

ashajadeja · x · 2026-08-29

Princeton Professor Arvind Narayanan estimates that about 80% of scientific literature is 'flat-out wrong' or 'not even wrong' due to overcapacity and market pressure for grants. He sees a massive opportunity for AI-for-science tools: designing agentic interfaces to explicitly look for follow-up work that challenges cited papers, or using agents to auto-annotate pre-training corpora to flag challenged findings.

Related event: Princeton Professor Estimates 80% of Scientific Literature Is Fundamentally Wrong(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →