The year is 2016. London-based researcher David Silver watches as his program AlphaGo defeats the world champion at Go—a game so complex that the number of possible positions exceeds the number of atoms in the universe. At the time, it seemed like the pinnacle of his career. Today, Silver says it was only the beginning.
In January 2026, he left Google DeepMind and launched his own startup. It is called Ineffable Intelligence and is based in London. It is now raising one billion dollars
Why Silver Left and What He Wants to Achieve
Silver spent more than fifteen years at DeepMind. He led the teams behind AlphaGo, AlphaZero, and MuZero. But then he went on sabbatical—and never returned. Ineffable Intelligence was entered in the commercial register in November 2025, and Silver became its director in January 2026.
His thesis is provocative, almost heretical. He argues that large language models—the technology behind ChatGPT, Gemini, and Claude—are a dead end. Why? Because they learn from data created by humans. They can synthesize, summarize, and combine. But they cannot discover anything truly new—something no human has ever conceived of.
And that is what Silver wants. A system that learns from scratch. Without human data. Purely through trial and error, just as AlphaGo played millions of games against itself and ultimately devised moves that no grandmaster had ever seen. The goal? “An infinitely learning superintelligence that independently rediscovers all human knowledge from first principles.” It sounds like science fiction. But investors believe in him.
Nvidia, Google, and Microsoft
The startup's pre-money valuation is around four billion dollars, even though the company has no product, no revenue, and no public roadmap. What does it have? A founder with a track record that speaks for itself. And that apparently is enough.
Nvidia's Jensen Huang, Google, and Microsoft are also involved in the talks. The terms are still being finalized, but the tech giants' interest is clear. If the round closes, Ineffable Intelligence will immediately rank among Europe's most valuable AI startups—even before it writes a single line of code for a customer.
The Man Who Changed AI: Silver's Biography
Silver's career is a textbook example of what happens when genius meets the right problem. He earned his PhD at the University of Alberta under Richard Sutton—the man many consider the father of reinforcement learning. Then came DeepMind, AlphaGo, AlphaZero, and MuZero.
His work did not remain confined to the laboratory. His team's algorithms optimized the UK's power grid and saved more than one million pounds. They helped Google reduce energy consumption in its data centers by thirty percent. He received the 2018 ACM Prize, became a Fellow of the Royal Society, and received the inaugural Mensa award in 2024.
He still lectures at University College London. But now he has other concerns—assembling a team of former DeepMind colleagues and preparing for the biggest challenge of his career.
A New Trend: Elite Scientists Are Leaving Big Companies
Silver is not alone. Ilya Sutskever, OpenAI's former chief scientist, left in 2024 and founded Safe Superintelligence—a startup now valued at more than thirty billion dollars, again without a single product. Jerry Tworek, one of the architects of OpenAI's reasoning models, recently launched Core Automation.
The pattern is clear: the brightest minds in the AI world are no longer content to remain employees and are striking out on their own. Capital is following them at a pace that would have been unimaginable just five years ago. Should this concern us a little? Perhaps. But it is also fascinating to watch the race for superintelligence move from corporate campuses to small, focused teams—exactly the way Silver has always preferred to work.
Impact on Europe
If the billion-dollar funding round closes, London will receive a clear signal: the global AI scene is not confined to San Francisco. Britain can keep top talent at home—and attract capital that would otherwise head across the Atlantic.
Ineffable Intelligence is also committed to European sovereignty in AI. The project aligns with British and EU efforts to maintain technological independence and develop their own AI infrastructure.
David Silver once built a machine that changed how the world thinks about artificial intelligence. Now he is betting his reputation—and a billion dollars of other people's money—that the entire industry is headed in the wrong direction. If he is right, we will all feel the consequences.
Sources: europeanbusinessmagazine.com and implicator.ai
