David Silver just landed $1.1 billion for a company founded months ago. His new lab, Ineffable Intelligence, hit a $5.1 billion valuation after a round led by Sequoia Capital and Lightspeed Venture Partners, with participation from Google, Nvidia, Index Ventures, and the U.K.'s Sovereign AI fund. Silver spent over a decade at DeepMind leading the reinforcement learning team. He helped build AlphaZero, the program that taught itself to dominate chess and Go without studying human games. Now he's betting that same approach scales far beyond board games. Ineffable wants to build what it calls a 'superlearner,' an AI that discovers knowledge and skills entirely through trial and error, with zero reliance on human-generated training data. The company's website doesn't hold back. It claims success would be 'a scientific breakthrough of comparable magnitude to Darwin.' That's a wild thing to put on your launch page. But reinforcement learning has delivered real results before. AlphaZero went beyond matching human play. It found strategies no human had ever considered. The round fits a clear pattern. Last month, AMI Labs, co-founded by former Meta AI scientist Yann LeCun, raised $1.03 billion at a $3.5 billion valuation. Recursive Superintelligence, from DeepMind's former principal scientist Tim Rocktäschel, reportedly pulled in $500 million. London is becoming a genuine AI hub, with Jeff Bezos' Project Prometheus reportedly seeking office space near Google's operations there. Google's investment in Ineffable keeps it plugged into the DeepMind talent pipeline it paid $500 million to acquire in 2014. The real question is whether reinforcement learning can match what large language models have achieved by simply scaling compute and data. LLMs have obvious limitations. They're bounded by training data and they can't truly discover anything new. But they work well enough to ship as products today. Reinforcement learning shines in narrow domains with clear reward signals. Open-ended learning across all knowledge is a much harder problem. Silver has earned the right to take the shot. Whether $1.1 billion gets him there, nobody knows yet. Silver told Wired that any money he makes will go to high-impact charities, so at least the upside has a destination beyond his bank account.
AlphaZero Creator Raises $1.1B for AI That Learns Without Us
David Silver, the DeepMind researcher behind AlphaZero, raised $1.1 billion at a $5.1 billion valuation for his new AI lab Ineffable Intelligence. The company aims to build a 'superlearner' AI that discovers knowledge through trial and error alone, skipping human training data entirely.