Next-Token Predictor Is An AI's Job, Not Its Species
astralcodexten.com · 2,073 words · saved by 1 readers
...
Next-Token Predictor Is An AI's Job, Not Its Species ... Scott Alexander Feb 26, 2026 426 433 36 Share I. In The Argument , Kelsey Piper gives a good description of the ways that AIs are more than just “next-token predictors” or “stochastic parrots” - for example, they also use fine-tuning and RLHF. But commenters, while appreciating the subtleties she introduces, object that they’re still just extra layers on top of a machine that basically runs on next-token prediction. I want to approach this from a different direction. I think overemphasizing next-token prediction is a confusion of levels.
saved by
related reading
- The AI Superforecasters Are Here - by Scott Alexanderastralcodexten.com
- The Artificiality of Alignmentjoinreboot.org
- Neurotechnology is Critical for AI Alignmentmilan.cvitkovic.net
- “Next-token predictor” is the wrong mental model for LLMsgmcgoldr.github.io
- Optimality is the tiger, and agents are its teeth — LessWronglesswrong.com
- God Help Us, Let's Try To Understand The Paper On AI Monosemanticityastralcodexten.com
- What will be left for us to work on?normaltech.ai
- The Specter Of Neuraleseastralcodexten.com
- Why AI is harder than we thinknews.ycombinator.com
- God Help Us, Let’s Try To Learn About Mechanistic Interpretability Techniquesastralcodexten.com
- And Yet It Understandsborretti.me
- GPTs are Predictors, not Imitators — AI Alignment Forumalignmentforum.org