A Columbia talk argues human-level AI should move away from LLMs and pure generative models
PMinervini · x · 2026-07-26
The post shares a talk slide arguing for a different path to human-level AI:
- move away from pure generative models in favor of joint-embedding architectures
- move away from probabilistic models in favor of energy-based models
- move away from contrastive methods in favor of regularized methods
- move away from reinforcement learning in favor of model-predictive control
- use RL only when planning fails to yield the predicted outcome
The accompanying comment frames a recent result as evidence that a VQ/FSQ tokenizer plus an autoregressive transformer may work after all, despite years of skepticism from LeCun-style arguments against that direction.
More from Research
- Open-weight 4B models approach o3-level performance on Swedish medical exams — AccomplishedCat4770 · 2026-07-26
- Graph engineering argues agent workflows need explicit graphs, not just loops — Aiden_Tech_Ai · 2026-07-26
- NeurIPS reviewers are asking claims that are impossible to disprove on budget — roydanroy · 2026-07-26
- Stanford course shares a free 1 hour 50 minute look inside Claude and ChatGPT — Aiden_Tech_Ai · 2026-07-26
- Computer vision course gets new lecture notes and interactive demos — CSProfKGD · 2026-07-26
- A chained web of small specialist models may fit physical AI better than one general model — richdotca · 2026-07-26