Tian Yuandong, former director at Meta’s AI research unit, has co-founded Recursive Superintelligence, a start-up advancing AI systems capable of autonomously improving their own reasoning and code. The company recently secured $650 million in funding, valuing it at $4.65 billion.
- Raised $650M in funding at a $4.65B valuation
- Focuses on recursive self-improving AI systems
- Led by ex-Meta AI research director Tian Yuandong
What happened
Tian Yuandong, who previously directed AI research at Meta’s FAIR team, co-founded Recursive Superintelligence with seven others to focus on AI systems that can autonomously refine and improve themselves. The company announced it raised over $650 million in a funding round led by venture firms GV and Greycroft, with participation from Nvidia and AMD. Their valuation reached $4.65 billion.
With about 25 employees and offices in San Francisco and London, Recursive Superintelligence is part of a rising trend among US and Chinese firms developing what is called recursive self-improvement—AI models that can iteratively enhance their own coding and reasoning capabilities without human intervention.
Why it matters
Recursive self-improvement represents a potential turning point in AI research, widely regarded as the fastest route to achieving superintelligence—AI systems that surpass human cognitive abilities. Companies that succeed could exponentially accelerate AI development, gaining strategic dominance in the industry.
However, experts warn that such technology carries significant risks. Autonomous AI evolution could lead to loss of human control, posing profound economic and societal challenges. Recursive Superintelligence’s co-founder Jeff Clune highlights these risks but commits to guiding the technology safely and beneficially for humanity.
What to watch next
The progress of Recursive Superintelligence will be closely monitored as it competes with similarly ambitious ventures like Ricursive Intelligence in the US and Chinese firms such as Zhipu AI, all targeting various aspects of self-evolving AI technologies. Funding inflows and partnerships with major chipmakers Nvidia and AMD will be key enablers to watch.
Additionally, industry developments around regulatory and safety frameworks for recursive AI will be critical, given the technology’s dual-use potential and the growing debate about ensuring aligned, secure, and ethical AI innovation in China, the US, and globally.