Metas AI Boss Says He DONE With LLMS...

16 Apr 2025 · 12 min
12:17

YouTube discussion (German) featuring Yann LeCun and a commentator about the limits of LLMs and directions for future AI. LeCun says he is less interested in LLMs and highlights four research areas: (1) giving machines an understanding of the physical world (world models), (2) persistent memory, (3) reasoning/"thinking", and (4) planning. He argues token-based next-token prediction and pixel-level video prediction are poor fits for learning physical world models; instead he advocates Joint Embedding Predictive Architectures (JEPA/VJEPA) that learn abstract representations and predict in representation space. The transcript also contrasts System 1 (reactive) vs System 2 (deliberative) cognition and suggests future AI will be hybrid systems combining reactive policies with learned world-model-based planning.

Chapters

More from LARRUZ

8:49
youtube.com

OpenAI’s New AI Is Doing Things We Barely Understand

3d ago
5:38
youtube.com

DeepSeek Solved AI’s Memory Problem

5d ago
33:17
youtube.com

DeepMind's New AI Just Cracked The Code Of Life

1w ago
3:43
youtube.com

The Billion Dollar AI Advantage Is Disappearing

1w ago
5:14
youtube.com

Claude Opus 5.5 AI: An Incredible Leap Forward

3w ago
5:16
youtube.com

Yes, Jev Is Insane, But There's A Catch

3w ago
5:11
youtube.com

DeepSeek’s Insane New Architecture

4w ago
4:16
youtube.com

Claude Is Now Leaving Invisible Fingerprints In Its Text

1mo ago
9:18
youtube.com

I Never Thought I’d See This Happen

1mo ago
5:22
youtube.com

GPT-6 Astra Changes Everything

1mo ago
4:55
youtube.com

Claude Fable AI Is Much Stranger Than The Headlines Suggest

1mo ago
See all content