Ineffable Intelligence, the London-based AI startup launched earlier this year by ex-Google DeepMind AI researcher David Silver, has brought on six high-profile “cofounders.”
Four of the new Ineffable hires are former Google DeepMind colleagues of Silver’s: Chris Apps, Wojciech Czarnecki, Lasse Espeholt, and Junhyuk Oh. Apps, Czarnecki, and Oh worked with Silver on DeepMind’s 2019 AlphaStar breakthrough. AlphaStar was an AI system that could beat the world’s best human players at the competitive, real-time strategy video game Starcraft II. Meanwhile, Espeholt helped build Google’s MetNet weather prediction AI models.
The fifth cofounder, Alexandre Laterre, was formerly the head of research at London-based AI company InstaDeep, which was acquired by biotech company BioNTech in 2023. The sixth, Heather Gorham, had been a partner at the venture capital firm Flying Fish, which was among the first investors in Ineffable.
All six cofounders recently updated their LinkedIn profiles to reflect their new roles. Ineffable confirmed the hires, but declined to comment further.
In April, Ineffable was valued at $5 billion when it raised a $1.1 billion seed round of venture capital funding, the largest seed investment ever in Europe.
Ineffable’s hiring spree adds to the wave of high-profile researchers who have recently left Google DeepMind to join rival AI companies and contributed to a perception that DeepMind, which was the first company to make the creation of AI models that could rival or exceed human intelligence its primary mission, is now struggling.
When Ineffable was incorporated in November 2025, records at U.K. business registry Companies House show that the company was set up by George “Geordie” Rose, a Canadian serial entrepreneur and tech investor. Silver only became a company director in January 2026, after he officially left DeepMind. Companies House records show that one month later, Rose was no longer a “person with significant control” of Ineffable’s shares, and he stepped down as a director in July. Rose’s LinkedIn profile and his web page list his as a “founding advisor” to the company. A company spokesperson said that Rose is a friend of Silver’s who helped him get the company up and running before moving into an advisory role, but that Silver does not consider him a “cofounder.”
Ineffable is one of the so-called “neolabs”—the next generation of frontier AI companies that have been launched in the past two years, many started by veterans of the three older generation of companies whose goal was the creation of advanced artificial intelligence: Google DeepMind, OpenAI, and Anthropic. Besides Ineffable, the neolabs include Thinking Machines Labs, cofounded by former OpenAI chief technology officer Mira Murati, and Safe Superintelligence, cofounded by former OpenAI chief scientist Ilya Sutskever, as well as Sakana AI, Recursive Superintelligence, Core Automation, and Adaption Labs, among others.
Like other so-called “neolabs,” Ineffable’s stated mission is to create AI superintelligence. But Ineffable is committed to getting there using reinforcement learning, the AI training technique with which Silver is closely associated. In reinforcement learning, an AI model learns, by trial and error, to maximize a reward that the person training the model establishes. It is somewhat similar to the idea of operant conditioning in psychology. Actions that lead to rewards get reinforced, while agents learn not to repeat those that don’t.
Oh, who will run reinforcement learning at Ineffable, was the reinforcement learning lead on DeepMind’s AlphaStar work. In 2025, he also was one of the first researchers to use AI to discover new reinforcement learning algorithms. Using AI to help automate AI research itself is a major focus among the frontier AI companies. Many are pushing towards “recursive self-improvement,” or RSI, a moment in which AI systems can autonomously optimize future versions of themselves with little to no human help. Oh’s hiring suggests that Ineffable will also be looking to use automated reinforcement learning research to accelerate its quest for superintelligence.
At DeepMind, Czarnecki focused on multi-agent research, or scenarios in which multiple AI agents have to learn to either cooperate or compete on various tasks. He worked on AlphaStar too, as well as on DeepMind’s work on teams of AI agents that had to learn how to work together to beat other AI agents in simulated games of capture the flag held inside virtual mazes. At Ineffable, Czarnecki will oversee the startup’s science team.
Apps was a technical program and research delivery lead at Google DeepMind. In essence, he helped project manage the large teams of researchers needed to deliver many of the company’s breakthroughs. Before DeepMind, he helped run technology transformation programs at Barclays bank and Deloitte. He will lead “mission acceleration” at Ineffable.
Espeholt will be in charge of Ineffable’s “compute, infrastructure, and engineering strategy,” according to his LinkedIn profile, while Laterre will oversee research engineering.
Gorham is working on Ineffable’s “compute, fundraising, and operations,” according to her LinkedIn profile. As a partner at Flying Fish, she had approached Silver in early 2024 to learn more about new approaches to reinforcement learning and to lay the groundwork for Flying Fish getting in on the ground floor if Silver ever decided to leave DeepMind and start his own company, according to a story in GeekWire. Her efforts paid off, allowing Flying Fish to write the first check for Silver’s new company and to play a role in its mega seed round, which was led by Sequoia and Lightspeed Ventures. She also helped secure Ineffable’s partnership for computing power with Google Cloud, which was announced in July.
