Skip to content
Sonic
AI
Sonic
AI
Home
Discover
Ask Sonic
Projects
Use with Claude or ChatGPT
Show me around
Request source or feature
Yann LeCun on What Comes After LLMs, Sonic AI
Home
/
Unsupervised Learning
/
Yann LeCun on What Comes After LLMs
Unsupervised Learning
Notify me
•
May 15, 2026
•
1:21:55
Interview
Yann LeCun on What Comes After LLMs
From
Unsupervised Learning
Yann LeCun
(VP & Chief AI Scientist, Meta (former); Turing Award Winner, guest)
Get the full transcript next time Unsupervised Learning releases an episode
Summary, key quotes, top claims, and the searchable transcript — emailed automatically. No card needed.
Sign up
Executive Summary
Yann LeCun details his departure from Meta to found Emilabs, arguing that Meta's strategic shift to catch up on LLMs stifled the long-term, exploratory research needed for his world model approach.
He presents a strong critique of Large Language Models (LLMs), stating they are not a viable path to human-level intelligence due to their inability to plan or understand the physical world, making them intrinsically unsafe.
LeCun champions his Joint Embedding Predictive Architecture (JEPA) as the superior path forward, predicting it will achieve "complete world domination" by enabling AI agents to learn world models and predict the consequences of their actions.
He argues that the AI ecosystem will inevitably trend towards open source, comparing proprietary companies like OpenAI and Anthropic to Sun Microsystems, which was ultimately displaced by the open-source Linux operating system.
Continue your research
Keep pulling the thread on Yann LeCun.
The Architectural Debate: LLMs vs. World Models
Corporate Strategy and Research Culture
12
quotes
Transcript
Key Arguments
Analysis
Quotes & Entities
12
Related
Loading transcript...
Processed May 15, 2026
Daily intelligence brief →
yt-dlp + mlx-whisper + Gemini