What happens when AI starts building itself?
$650M. Richard Socher's betting on self-improving AI—and shipping products, not just research papers.

Why it matters
A major funding round for recursive self-improvement in AI signals investor confidence in agentic AI that can autonomously enhance its own capabilities. This represents a shift from model-as-endpoint to model-as-evolving-system, with immediate product implications for enterprises.
The key facts
5 to knowRichard Socher raised $650M for self-improving AI startup
Focus on recursive AI research and product improvement
Emphasis on shipping products (not pure R&D)
Self-improvement capability as core value proposition
Published May 14, 2026
Go to the source
TechCrunch AItechcrunch.com
Publisher excerpt: Richard Socher's new $650 million startup wants to build an AI that can research and improve itself indefinitely — and he insists it will actually ship products.