LessWrong essay argues superintelligence could arrive by Christmas, faster than AI 2027
toptickcrypto · x · 2026-11-08
In a LessWrong frontpage post, Alexander Gietelink Oldenziel argues that strong recursive self-improvement may be imminent or already underway, and that superintelligence could arrive by Christmas if lab racing continues — faster than even the AI 2027 timeline widely seen as outrageously aggressive.
Key claims:
- FOOM should be the default expectation: RSI is a positive feedback loop and inherently (hyper)exponential; nothing suggests an intelligence ceiling near human level.
- AI is delivering revolutionary math results: Navier-Stokes was resolved via counterexample; OpenAI reportedly has substantial progress on a second Millennium Prize problem (rumored Hodge conjecture), with Anthropic rumored on Birch–Swinnerton-Dyer.
- ML engineering is not different in kind from math since AI lives on computers, so self-improvement compounds directly.
The author stresses he has no inside knowledge and only public information, framing it as an extraordinary claim he believes now has extraordinary evidence. It's a provocative forecast, not an empirical result.
More from AGI Musings
- Space compute advocate says data center protests reveal opponents' true motive — Aiden_Tech_Ai · 2026-11-08
- Snapshot of agent use cases today: canceling subscriptions and spamming support — tekbog · 2026-11-08
- If compute moves to space, will data center critics still protest progress itself? — Dr_Singularity · 2026-11-08
- Dev pushes back on AI-replacement thesis: software and math work keeps growing — timigod · 2026-11-08
- MIT CSAIL shares roundup of 15 neural network courses, topped by Andrew Ng's 4.9/5 intro — MIT_CSAIL · 2026-11-08
- Jeff Dean: the AI self-improvement loop can't run infinitely fast — physics gets in the way — haider1 · 2026-11-08