David Silver doesn't need your text corpus—or so he argues.
The DeepMind veteran who engineered AlphaGo's historic defeat of Go champion Lee Sedol has emerged from stealth with Ineffable Intelligence, a London-area startup that closed a staggering $1.1 billion seed round at a $5.1 billion valuation. The financing, announced April 27, represents the largest seed round ever raised by a European company—eclipsing even the $1.03 billion that Yann LeCun's AMI Labs pulled together in Paris just seven weeks ago.
That Silver managed to command such terms isn't entirely surprising. The University College London professor is a Fellow of the Royal Society and a 2019 ACM Prize winner whose reinforcement learning work produced some of AI's most celebrated breakthrains. What's more startling is the bet he's asking investors to make: that the industry's current obsession with large language models trained on internet-scraped text is, essentially, a detour.
"We're building what we call a superlearner," Silver said in an interview with Wired published alongside the funding announcement. The system, as Ineffable describes it, will learn entirely from its own experience—no human-generated training data required. The company's mission statement hints at ambitions that border on the philosophical: to create AI that will "rediscover and transcend human inventions" including language itself, along with science, mathematics, and technology.
It's a contrarian pitch in an industry that has spent the past few years convinced that scale and data are everything.
The Reinforcement Learning Believer
Silver laid out the intellectual scaffolding for this approach in a preprint co-authored with Richard Sutton titled "The Era of Experience." The paper argues for a fundamental shift away from human-created training sets toward experiential, self-generated data—essentially, AI systems that teach themselves through trial and error rather than pattern-matching against humanity's collective output.
Whether that vision proves feasible at scale remains an open question. But Silver's track record bought him considerable credibility with the venture firms circling London's AI scene. Sequoia Capital and Lightspeed Venture Partners co-led the round, joined by Nvidia, Alphabet's Google, Index Ventures, the UK government's Sovereign AI fund, and the British Business Bank.
In notes published April 27, Sequoia partners Sonya Huang and Alfred Lin praised what they termed "the Era of Experience." Lightspeed's Ravi Mhatre, Raviraj Jain, and Antoine Moyroud invoked "Move 37"—the now-famous AlphaGo game moment that first signaled the system's capacity for what looked uncannily like creativity.
Silver, for his part, disclosed in the Wired interview that he plans to donate all personal proceeds from Ineffable to high-impact charities. It's a detail that feels almost incidental given the sums involved, but it underscores a certain seriousness of purpose.
London's AI Moment
The financing arrives at a moment when the UK government is making an overt push to anchor major AI companies on home soil. Ineffable's round includes participation from the Sovereign AI fund, a £500 million vehicle launched April 16 to support domestic startups. According to the fund's portfolio page, it's co-investing alongside the British Business Bank and providing compute access and visa support—bureaucratic lubricants that matter when you're trying to scale quickly.
Ineffable isn't the only London-area company attracting serious capital. Recursive Superintelligence reportedly closed a $500 million round earlier in April, according to coverage from Sifted on April 20. The clustering feels deliberate, perhaps even competitive in a way that recalls earlier tech hub rivalries.
Silver incorporated Ineffable Intelligence in the UK on November 19 of last year. Companies House records show him appointed as director and registered as a person with significant control on January 16, with multiple statements of capital filed between January and April—the usual paper trail preceding a financing of this magnitude.
The company has established headquarters in Altrincham, Cheshire, a choice that might raise eyebrows among those expecting a Shoreditch or King's Cross address. Then again, Silver has never been one for following convention.
What Comes Next
Details on Ineffable's roadmap remain scarce. The company has posted open roles on its careers page but hasn't disclosed employee headcount, specific product timelines, or concrete plans for deploying the capital beyond the broad ambition of building a reinforcement learning-based superlearner.
In interviews with both Bloomberg and Wired on April 27, Silver critiqued the current trajectory of large language model development, though he stopped short of predicting timelines or claiming imminent breakthroughs. That restraint might be strategic—AlphaGo took years to reach the level that stunned the world in 2016, and there's little reason to expect Ineffable's work to move faster.
Cooley advised Ineffable on the transaction, which the law firm has confirmed as Europe's largest seed to date. Whether the company can deliver on its ambitious thesis remains to be seen. But with $1.1 billion in the bank and one of AI's most storied researchers at the helm, Silver has bought himself time to find out.
And in a field moving as quickly as artificial intelligence, time—perhaps more than data—may be the scarcest resource of all.
