Princeton researchers warn AI could slow science despite exploding paper output
AllThingsApx · x · 2026-09-28
Citing Sayash Kapoor and Arvind Narayanan's essay 'Could AI slow science?', this thread highlights the 'production-progress paradox': scientific paper output grew 500-fold from 1900 to 2015, yet faster production doesn't automatically improve science. Drawing parallels to flawed early predictions about the printing press and social media, the authors argue science is a complex system where individual gains from AI may not translate into collective benefit — so far it has stretched evaluation and correction processes to a breaking point. They call for transparently documenting what models contributed, fair credit, and protecting unpublished work.
More from AGI Musings
- 'Wanted to build god, ended up building a secretary that can code' — repligate · 2026-09-28
- If a frontier lab admits its AI can't be contained, that lab should be shut down — kevinnbass · 2026-09-28
- Does immediate Lean formalization make discovering new math harder? — LucaAmb · 2026-09-28
- 3Blue1Brown's HackMIT keynote asks: are you the collie or the pug? — AlexKontorovich · 2026-09-28
- New preprint argues AI science is more than reasoning: it must capture collaboration — weballergy · 2026-09-28
- DeepMind researchers propose 'agentic economies' to manage AI-driven scientific discovery — weballergy · 2026-09-28