The 'AI psychosis' debate: not psychosis if the results are correct and useful
doodlestein · x · 2026-09-16
doodlestein pushes back on the "AI psychosis" label: it's not psychosis if the results are correct and useful. He argues the phrase is mostly thrown around by unaccomplished people denigrating others to avoid the fact that 95%+ of their professional output could be replaced by an agent costing a couple bucks per hour.
It responds to willcb's take that you need to teeter on the edge of AI psychosis to get the most out of models — CS theory breakthrough papers now credit "a long conversation with ChatGPT Astra."
Related event: Debate Flares Over 'AI Psychosis' as CS Papers Credit ChatGPT(2 posts)→
More from AGI Musings
- AI lets firms build compounding engines of institutional knowledge, says new essay — chetanp · 2026-09-16
- EA veteran: I held AI-risk beliefs 10 years before meeting anyone in the scene — AndyMasley · 2026-09-16
- AI agents burn 10B tokens on 60-year-old deletion channel problem, near 1/100-bit capacity — DimitrisPapail · 2026-09-16
- OpenAI researcher Dan Selsam publishes personal statement on AI risk — thlarsen · 2026-09-16
- Six experts on whether AI could wipe out humanity: less doom, still no consensus — gedbarker · 2026-09-16
- Analyst doubles down: humanoid robots may need major limits, possibly outright bans in many use cases — binarybits · 2026-09-16