Before he asked investors for $1.1 billion, David Silver made a promise about what he would do with the money. He pledged to give away all of his equity proceeds from his new company, the largest commitment in the history of Founders Pledge. The founder of a start-up valued at $5.1 billion decided in advance not to keep it.
He was born in 1976; where is not on record. He studied at Christ's College, Cambridge, graduating in 1997 with the Addison-Wesley award, and made a friend there named Demis Hassabis. What the two had in mind then is not documented. What is documented is that Silver did not go straight into research. He co-founded Elixir Studios, a games company, where he was chief technology officer and lead programmer, and the studio won awards for technology and innovation.
With Sylvain Gelly he built MoGo, among the strongest Go programs of its time.
In 2004 he went back to academia, at the University of Alberta, for a PhD in reinforcement learning, finished in 2009. With Sylvain Gelly he built MoGo, among the strongest Go programs of its time.
What almost nobody outside AI remembers is how slowly that path ran. He consulted for DeepMind from 2010 and joined full time in 2013, after a Royal Society fellowship and a lectureship at University College London. At DeepMind came AlphaGo, which received an honorary 9-dan professional certificate. He later won the 2019 ACM Prize in Computing and became a Fellow of the Royal Society in 2021.
He founded Ineffable Intelligence in November 2025 and left Google DeepMind in January 2026. By February it was reportedly seeking a $1 billion seed at a $4 billion valuation. In April 2026 the round was announced at $1.1 billion (£880 million) and $5.1 billion.
The price is the open question. Nothing in the sources read describes a product, a customer or a revenue line, and the company is only months old. Its value rests on one person's record.
For a first-time founder, the quieter point is the order of decisions. Silver settled what the money was for before he took it. Deciding your limits first can make everything after it easier to explain.