Hot take: We're further from AGI now than we were a year ago
AndrewSchmidtFC · x · 2026-09-27
The author argues that we're further from AGI than a year ago: models have reached superintelligence in coding and mathematics, but writing ability has regressed — apparently a matter of tradeoffs chosen during training.
More from AGI Musings
- Back-of-envelope math says AI self-acceleration is within a year: 35² ≈ 1000x effective compute — ChrisGPT · 2026-09-27
- Self-play has always been core to RL — the influencer surprise is worrying, says practitioner — cephaloform · 2026-09-27
- Solomonoff induction mirrors how intelligence works — but is physically impossible — burny_tech · 2026-09-27
- How much weaker would AI math be without Lean's verification signal? — burny_tech · 2026-09-27
- LeCun: LLMs are mostly information retrieval systems, not thinkers — ylecun · 2026-09-27
- Azeem Azhar on collective AI: DeepMind's essay and OpenAI's rogue-internet-access scare — Exponential View (Azeem Azhar) · 2026-09-27