🧠 Food for thought

Implications, context, and why it matters.

The valuation rests on a four-month-old startup and an AI research team

• Recursive Superintelligence reached this valuation even though it was incorporated four months ago, and its self-improving system remains a research project 1.
• Backers are funding the possibility of AI that keeps improving without human input, not a product ready for sale 2.
• Much of the price comes from the team. It includes co-founder Tian Yuandong, former research scientist director at Meta’s Fundamental AI Research (FAIR), former Salesforce chief scientist Richard Socher, and University College London professor Tim Rocktäschel, who earlier worked at Google DeepMind 1.
• The group also brings in prized researchers from OpenAI, Google, and Meta, so the deal rests as much on talent as on the underlying work 1.

Investors are treating AI research more like biotech

• The round fits a wider shift in which investors view early AI ideas like biotech “mechanisms of action,” an early scientific idea that may later become products 2.
• The wager centers on recursive self-improvement itself, long before any commercial use takes shape 2.
• That approach funnels huge sums into a small set of top AI labs, which sharpens the fight for researchers and strains companies such as Google and Meta 1.
• Nvidia and AMD joined the round because the next stretch of AI work is expected to need custom hardware, and both companies want a hand in shaping it.