LeCun on world models: predict representations, not pixels, plan via MPC

At ETH, LeCun described world models that predict representations rather than pixels and plan via MPC-style cost minimization. He noted an LLM's 10^14 bytes of training text matches what a 4-year-old sees visually, arguing for video-based learning.

2026-10-13 ~ 2026-10-13 · 2 related posts

• LeCun's math: LLM and 4-year-old both see 10^14 bytes — one is text, the other video — le_james94 · 2026-10-13
• LeCun's world model pitch: predict representations, not pixels, and plan via MPC-style cost minimization — le_james94 · 2026-10-13