David Silver Raises $1.1B to Build AI Without Human Data

The creator of AlphaGo and AlphaZero has raised $1.1 billion to build an AI that ignores everything the industry has spent the last five years constructing.

David Silver, who spent nearly two decades at Google DeepMind turning reinforcement learning from theory into systems that beat the world's best human players at Go and chess, formally launched Ineffable Intelligence on Sunday. The London-based lab emerged from stealth with the largest seed round ever raised in Europe, valuing the company at $5.1 billion before it has shipped a single product or published a benchmark.

The Bet Against LLMs

Go stones on a board - the game AlphaGo Zero mastered through pure self-play, with no human game records. Source: commons.wikimedia.org

The thesis Silver is betting on runs directly against the current consensus. Every major commercial AI system - GPT, Claude, Gemini, Llama - was built by training on enormous quantities of human-created text, images, and code. The working assumption across the industry is that more data, better filtered, produces better models. Silver thinks that assumption has a ceiling.

His argument is rooted in the systems he built at DeepMind. AlphaGo Zero learned to play Go without ever seeing a human game - it played against itself millions of times and discovered strategies that human professionals had never conceived. AlphaZero did the same across chess, Go, and shogi simultaneously, defeating every human and machine champion it faced. AlphaProof, Silver's most recent work, applied the same approach to formal mathematics.

In 2025, Silver and Richard Sutton - the University of Alberta researcher widely considered the founder of modern reinforcement learning theory - co-authored a paper arguing that "the era of experience" represents AI's real next phase. Their claim: systems that learn through interaction with their environment, rather than by absorbing human knowledge, will eventually surpass anything built on human data alone.

Ineffable Intelligence is Silver's attempt to find out if that's true at commercial scale.

The company's mission statement doesn't hedge. Its website reads: "If successful, this will represent a scientific breakthrough of comparable magnitude to Darwin: where his law explained all Life, our law will explain and build all Intelligence." Silver has said he views this as his life's work. He has also pledged 100% of his personal equity gains through Founders Pledge, which described it as the largest pledge in the organization's history - potentially worth billions if the company reaches the outcomes its backers are projecting.

The Investor Roster

The capital came together quickly. Sequoia Capital and Lightspeed Venture Partners co-led the round, with Alfred Lin and Sonya Huang representing Sequoia and Ravi Mhatre and Raviraj Jain representing Lightspeed. Lin and Huang flew to London personally to close the deal, which is unusual at seed stage for Silicon Valley firms of their size.

Google's participation is notable given it owns DeepMind - Silver's former employer - and recently committed $40 billion to Anthropic in a separate deal. That's three concurrent large bets on different approaches to AI, made by the same strategic investor. Google appears to be hedging across multiple AI approaches at once rather than picking a winner.

"The largest leaps in AI have always come from people willing to ignore the consensus. David has ignored more consensus, more correctly, than almost anyone in the field. We are honored to co-lead Ineffable's first round and to partner with David on what may be the most ambitious scientific mission of our generation."

• Sequoia Capital (Alfred Lin and Sonya Huang)

Richard Sutton, whose "Bitter Lesson" essay has become a foundational text in ML research, endorsed the company directly on X: "David Silver's new company, Ineffable Intelligence, will fulfil the promise of the Era of Experience."

Who Benefits

Silver and the RL research community. Reinforcement learning has been overshadowed by LLM investment for years. Silver's raise is both a financial validation and a talent signal - multiple former DeepMind researchers are reportedly joining Ineffable's leadership team. The funding gives the lab the compute budget to run experiments that university groups and smaller startups can't afford.

David Silver, former head of reinforcement learning at Google DeepMind and now CEO of Ineffable Intelligence. Source: davidstarsilver.wordpress.com

The UK AI ecosystem. London has consistently lost senior AI talent to Bay Area salaries and capital access. Ineffable is a credible reason for a top researcher to stay in Britain and build. Josephine Kant, head of ventures at the UK Sovereign AI Fund, framed the government's position directly: "From AlphaGo to AlphaZero to AlphaProof, he has spent nearly two decades turning reinforcement learning from a research idea into the results the field builds on."

Science and Technology Secretary Liz Kendall added: "This investment in Ineffable will support a company at the very frontier of AI, with the potential to transform entire sectors." The UK government has been openly competing for AI investment since losing several high-profile companies to US acquirers. For cross-reference on how London is positioning itself relative to regulation, the Cohere and Aleph Alpha deal earlier this month shows how sovereign AI concerns are reshaping where capital flows.

Early investors. At $5.1 billion pre-product, the return math requires Ineffable to reach something in the range of $20-50 billion to create standard VC multiples on a $1.1 billion check. That means Silver's lab needs to build either a genuine AGI system, or AI services that compete commercially with OpenAI, Anthropic, and Google at scale. The investors who flew to London apparently think that range is reachable. For context, Bezos' Prometheus AI closed at $38 billion with significant revenue already in place.

Who Pays

The VC fund LPs. $1.1 billion at seed stage concentrates a lot of risk in a single thesis. If Silver's approach to pure reinforcement learning hits the same walls researchers have run into before - reward hacking, sample inefficiency, lack of transfer across domains - the investors own a very expensive research project. The track record of AlphaGo and AlphaZero suggests those problems are solvable in constrained environments. Whether they generalize to open-world intelligence is a genuinely open research question.

British taxpayers, indirectly. The British Business Bank and UK Sovereign AI Fund are rolling out public money. The strategic case - keeping frontier AI capability in the UK, preventing talent drain, building national technical sovereignty - is coherent. Whether it results in financial returns or remains a policy investment depends on outcomes that are years away.

The $1.1 billion seed is the largest single bet ever placed on reinforcement learning as a path to machine intelligence, backed by some of the most credible technical and financial names in the field. Whether Silver's superlearner eventually beats LLM-based systems or proves the consensus right in the end, the experiment is now funded to run.

Sources:

• TechCrunch: DeepMind's David Silver just raised $1.1B to build an AI that learns without human data
• Sequoia Capital: Partnering with Ineffable Intelligence
• tech.eu: Ineffable Intelligence launches with record-breaking $1.1B Seed round
• UK Government: UK backs company building breakthrough AI that can discover new knowledge
• The Next Web: Sequoia and Nvidia back AlphaGo creator David Silver's Ineffable Intelligence at $5.1 billion
• PYMNTS: DeepMind Vet's Ineffable Intelligence AI Startup Raises $1.1 Billion