AI & Singularity

Yann LeCun launches AMI Labs with record $1.03B seed to build world models.

Yann LeCun, Meta former AI chief, has launched AMI Labs with a staggering $1.03 billion seed funding round, one of the largest seed rounds in history. The France-based startup is building world models that understand physical reality, not just language.

AMI Labs raised $1.03 billion to build AI that understands physical reality.

What Are World Models?

Unlike large language models that learn from text, world models learn from video and physical interaction. LeCun has long argued that LLMs are fundamentally limited because they lack understanding of the physical world.

In six months, every company will call itself a world model company to raise funding, LeCun predicted during the announcement.

Key Details

• Funding: $1.03 billion seed at ~$3 billion valuation
• Lead investor: Temasek (Singapore state investors)
• Location: France with Singapore operations
• CEO: Alexandre LeBrun (LeCun is founder/chief scientist)
• Approach: Alternative to LLMs using JEPA-style world models

Why It Matters

This is a direct challenge to the text-centric paradigm that has dominated AI since ChatGPT. LeCun believes world models are the path to machines that actually understand the world, not just predict tokens.

Sources: TechCrunch, Le Monde