Ex-DeepMind Scientist Seeks $1B Valuation for New Superhuman Intelligence Startup
David Silver, former Google DeepMind principal scientist, is raising $1 billion for his London-based company Ineffable Intelligence, with Sequoia Capital leading the round at approximately $4 billion valuation. The deal could become Europe’s largest seed funding round ever recorded. This superhuman intelligence startup represents a bold departure from mainstream AI development, betting that current approaches won’t achieve true artificial general intelligence.
Silver’s track record speaks volumes. He doesn’t need convincing. At Google DeepMind, Silver used reinforcement learning to train AI programs such as AlphaGo and AlphaStar to beat human players. Those weren’t incremental improvements. They were breakthroughs that redefined what machines could accomplish.
Now he’s pursuing something bigger. Much bigger.
The DeepMind Scientist New Venture That’s Shaking Up AI
Ineffable Intelligence was incorporated in November 2025, and Silver was appointed director in January 2026. The company remains pre-product, yet investors are queueing up with billion-dollar checks. Why? Because this ex-DeepMind AI startup challenges the entire foundation of modern artificial intelligence development.
Silver believes that to reach superintelligence, AI systems will need to discard human knowledge entirely and learn from first principles—through trial, error and self-play, the way AlphaGo learned to play Go by competing against itself millions of times. This isn’t a minor technical adjustment. It’s a complete paradigm shift away from how companies like OpenAI, Anthropic, and even his former employer approach AI development.
The competition in artificial general intelligence investment has never been fiercer. Nearly half of the $469 billion in venture capital deployed globally last year went into AI companies, according to CB Insights. But most of that capital flowed toward scaling existing architectures rather than fundamentally reimagining them.
Silver takes a different path.
Why This Former DeepMind AI Firm Matters More Than Others
Nvidia, Google and Microsoft are also said to be considering investing. When tech giants line up alongside elite venture firms for a pre-revenue startup, you know something unusual is happening. The ex-DeepMind scientist seeking AI funding isn’t pitching incremental improvements to ChatGPT. He’s arguing the entire industry took a wrong turn.
Here’s what makes Ineffable Intelligence different from the crowded field of superhuman intelligence startup competitors. Their “Era of Experience” paper argued that AI systems trained on human data are hitting a wall, and progress driven by supervised learning alone is “demonstrably slowing.” Most high-quality text sources have been consumed or will be soon.
The solution? Let AI systems learn from experience. Not from curated datasets or human feedback, but from direct interaction with environments over long stretches of time.
David Silver, 49, spent more than a decade at Google DeepMind, where he helped develop AlphaGo and AlphaStar—AI systems that beat the world’s best players at Go and StarCraft. He also contributed to Google’s Gemini family of large language models before leaving late last year to launch Ineffable Intelligence. His credentials aren’t theoretical. They’re proven in systems that achieved what experts thought impossible.
Understanding Reinforcement Learning: The Core of This AI Startup Funding Strategy
Traditional large language models learn by predicting the next word in massive text corpora. They become sophisticated pattern matchers. Reinforcement learning operates differently. In this branch of machine learning, AI systems are trained to make decisions by interacting with an environment and receiving rewards or penalties based on their actions, enabling agents to learn through a trial-and-error process.
Think about how humans actually learn complex skills. We don’t memorize instruction manuals. We experiment. We fail. We adjust. We try again.
AlphaGo demonstrated this spectacularly. The result was a system that made moves no human had ever conceived, some of which initially looked like mistakes but turned out to be brilliant. Human experts couldn’t explain these strategies because they emerged from pure experiential learning rather than human knowledge.
This ex-DeepMind scientist seeking AI funding believes that path leads to genuine superintelligence. The artificial general intelligence investment community is betting he’s right.
Silver and Sutton predicted that experience will “become the dominant medium of improvement and ultimately dwarf the scale of human data used in today’s systems.” If accurate, this reshapes everything we know about AI development trajectories.
The Economics Behind This Superhuman Intelligence Startup
The round, which values the new company Ineffable Intelligence at around $4 billion, is being led by Sequoia Capital. That valuation might seem astronomical for a company without products or revenue. Context matters, though.
Ilya Sutskever, former chief scientist at OpenAI, founded Safe Superintelligence in 2024 and has raised $3 billion to date at a valuation that reached $32 billion by April 2025—despite having no product. The market isn’t valuing current capabilities. It’s valuing potential to solve the most important technical problem of our generation.
The AI startup funding landscape has fundamentally shifted. Nearly half of the $469 billion in venture capital deployed globally last year went into AI companies, according to CB Insights. Investors recognize that whoever cracks artificial general intelligence first captures extraordinary value.
The artificial general intelligence market is projected to grow from $6.94 billion in 2025 to $9.30 billion in 2026, marking a compound annual growth rate of 33.9%. Looking ahead, the market is expected to reach $29.67 billion by 2030 with a CAGR of 33.7%. These aren’t small numbers, and they explain why the former DeepMind AI firm commands premium valuations.
Why Elite Researchers Are Leaving Big Tech
The pattern is consistent: elite researchers who believe the current paradigm has limits are leaving to explore alternatives, and capital is following them at extraordinary speed and scale. This isn’t isolated to Silver. Top researchers from OpenAI, Meta, and Google are launching their own ventures at unprecedented rates.
What’s driving this exodus? Resources and freedom, certainly. But more fundamentally, these researchers see technical limitations that large organizations can’t or won’t address. Silver has told friends he wants to get back to the “awe and wonder of solving the hardest problems in AI” and sees superintelligence—or AI that would be smarter than any human and potentially smarter than all of humanity—as the biggest unsolved challenge in the field.
Big Tech companies have quarterly earnings targets. They need to show continuous improvement to shareholders. That creates pressure to scale existing approaches rather than explore potentially dead-end alternatives. A DeepMind scientist new venture doesn’t face those constraints.
Managing director at investment banking firm Houlihan Lokey Capital Solutions Group tells that Ineffable Intelligence’s round is “further evidence that the UK and wider European ecosystem can produce globally significant companies”, as “well-informed investors are seeking the most capable individuals developing the most advanced technologies.”
The Technical Bet: Why Current AI Might Hit a Ceiling
Here’s the controversial core of Ineffable Intelligence’s thesis. It strikes at the foundation of the current AI industry, which has invested hundreds of billions of dollars in scaling transformer-based language models on the assumption that more data and more compute will eventually produce artificial general intelligence. Silver disagrees.
The evidence supporting his position is accumulating. 2025 taught us that simply adding more GPUs no longer guarantees proportional improvements in AI performance. Research shows that each new generation of LLMs now delivers incremental gains rather than the breakthroughs seen in the early years of development.
This creates opportunity for an ex-DeepMind AI startup focused on alternative architectures. While others optimize transformer models, Silver pursues fundamentally different approaches. Silver’s bet on reinforcement learning and experience-based training represents a distinct approach from the text-heavy training methods used by most current AI models. If Ineffable Intelligence can deliver on its promise of “superhuman capabilities” through experiential learning, it could shift how the next generation of AI agents is built.
What Success Looks Like for This Artificial General Intelligence Investment
Ineffable Intelligence aims to build what Silver has described as “an endlessly learning superintelligence that self-discovers the foundations of all knowledge.” The approach is rooted in reinforcement learning—the branch of AI Silver has spent his entire career advancing. That’s an audacious goal. It’s also precisely the kind of moonshot that attracts massive AI startup funding.
Contrast this with incremental improvements to existing chatbots. Most AI companies today focus on making GPT-5 slightly better than GPT-4 or reducing hallucinations by a few percentage points. Important work, certainly, but not paradigm-shifting.
A superhuman intelligence startup pursuing genuine AGI operates in different territory. The AGI market is expected to grow to over $116 billion by 2035, at a compound annual growth rate of 36.25%. On a granular level, the market has experienced a growth rate of 29% over the past year. These growth rates reflect both hype and genuine progress toward systems that can match or exceed human cognitive abilities across domains.
The Competitive Landscape and Market Dynamics
Ineffable Intelligence doesn’t operate in a vacuum. Former Meta chief scientist Yann LeCun is raising roughly €500 million at a €3 billion valuation for AMI Labs. Meanwhile, a group of xAI co-founders recently left Elon Musk’s company. The big labs look increasingly nervous about the talent drain, and the market keeps rewarding it.
This wave of departures signals something important. The smartest people in AI increasingly believe breakthrough advances won’t come from incumbent organizations. That conviction drives this ex-DeepMind scientist seeking AI funding rather than continuing comfortable research at one of the world’s premier AI labs.
The money is chasing a specific profile: researchers with rare technical depth who believe the current paradigm has limits, and who are willing to stake their reputations on something different. When investors write billion-dollar checks for companies with no product, they are making a statement about the companies that do have products. They are pricing in a belief that the next major advance won’t come from making GPT-5 larger or training Gemini on more text.
The European AI Ecosystem and London’s Growing Role
If completed, it would be the largest seed round ever raised by a European start-up. This matters beyond bragging rights. Europe has historically lagged Silicon Valley in AI development and funding. A successful former DeepMind AI firm valued at $4 billion could shift perceptions and capital flows.
If the round closes at $1 billion, Ineffable Intelligence would instantly become one of the most valuable AI startups in Europe—and a powerful signal that London remains capable of producing world-class AI companies, not just world-class AI researchers who leave for San Francisco. The UK government has invested heavily in positioning London as an AI hub. This validation from global investors confirms that strategy is working.
The artificial general intelligence investment flowing into European startups demonstrates growing confidence in the region’s technical talent and infrastructure. Under the Horizon Europe and Digital Europe programmes, key EU funding instruments for science and technology, the Commission allocates more than €1 billion per year specifically for AI research and innovation, with targets to mobilise up to €20 billion in annual investment by the end of this decade.
Risks and Challenges Facing This Superhuman Intelligence Startup
Not everyone believes Silver’s thesis. Critics argue that reinforcement learning, while powerful for games with clear win conditions, struggles in open-ended real-world environments. How do you define “winning” when developing general intelligence? What reward signals guide the system toward useful capabilities rather than unexpected exploits?
The timeline presents another challenge. Artificial general intelligence will dominate headlines throughout 2026, but maintaining a realistic outlook is essential. Without major breakthroughs in hardware architecture, most claims are likely to outpace what is technically possible. 2026 is still unlikely to be the year AGI delivers any profound change. Investors betting on Ineffable Intelligence need patience.
Competition from well-funded incumbents also poses risks. The consensus estimate among Wall Street analysts for hyperscaler tech companies’ 2026 capital spending is now $527 billion. Google, Microsoft, Meta, and Amazon are pouring unprecedented resources into AI development. A startup, even one led by a brilliant researcher, faces resource constraints.
Yet history shows that breakthroughs often come from outsiders challenging conventional wisdom. AlphaGo itself proved this when it defeated world champions using techniques the broader AI community had largely abandoned.
The Path Forward for Ineffable Intelligence
The London-based startup doesn’t have a product yet. It was founded last November by Silver, who reportedly left DeepMind around the same time. Ineffable Intelligence reportedly plans to build on that work as part of its research efforts. The company is actively recruiting AI researchers, building the team that will attempt to achieve its ambitious goals.
The AI startup funding secured provides runway to pursue long-term research without immediate pressure to commercialize half-baked technologies. This freedom matters enormously when attempting fundamental breakthroughs rather than incremental product improvements.
Silver’s exit from Google DeepMind late last year sparked a fierce race among venture firms. Sequoia partners Alfred Lin and Sonya Huang reportedly made a trip to London to meet with him shortly after he left. That personal attention from elite investors underscores the high stakes and potential returns.
What This Means for the Future of AI Development
The success or failure of this ex-DeepMind AI startup will influence AI development trajectories for years. If Silver’s reinforcement learning approach produces breakthroughs, expect massive capital reallocation away from transformer-based architectures. If it struggles, the current paradigm receives validation.
Today, in the race to superintelligence, the market believes that the right founder with the right thesis is worth more than a finished product. Silver built the machine that changed how the world thought about AI. Now he is betting his career—and $1 billion of other people’s money—on the idea that the industry’s dominant approach will not be enough. If he is right, the implications extend far beyond London.
The artificial general intelligence investment community is hedging its bets. Some capital flows to companies scaling existing approaches. Other capital backs researchers exploring alternatives. This portfolio approach makes sense when the ultimate path to AGI remains uncertain.
Conclusion: A Defining Moment for AI
David Silver’s superhuman intelligence startup represents more than another well-funded venture. It embodies a fundamental question facing the AI industry: Can we reach artificial general intelligence by scaling current approaches, or do we need entirely new paradigms?
The $1 billion raise—if completed—signals that sophisticated investors see sufficient probability in Silver’s alternative vision to commit extraordinary capital. The DeepMind scientist new venture challenges assumptions underlying hundreds of billions in existing AI investment.
Whether Ineffable Intelligence succeeds or fails, its approach will generate valuable knowledge about paths toward genuine artificial intelligence. That information benefits the entire field, even competitors.
For those watching the artificial general intelligence investment space, this former DeepMind AI firm deserves close attention. The technical bets, talent assembled, and results produced will shape AI development for the next decade. We’re witnessing not just a startup launch, but potentially a pivotal moment in the quest for superhuman intelligence.
The stakes couldn’t be higher. The potential rewards—for investors, for Silver’s team, and for humanity—are almost unimaginable.
Frequently Asked Questions
What is Ineffable Intelligence and why is it raising $1 billion?
Ineffable Intelligence is a London-based superhuman intelligence startup founded by former Google DeepMind scientist David Silver. The company is raising $1 billion at a $4 billion valuation from investors including Sequoia Capital, Nvidia, Google, and Microsoft to develop artificial general intelligence using reinforcement learning instead of traditional large language model approaches.
What makes David Silver’s approach different from other AI companies?
Silver’s ex-DeepMind AI startup focuses on reinforcement learning, where AI systems learn through trial and error by interacting with environments rather than training on human-generated text. He believes current large language models cannot achieve true superintelligence and that AI must learn from first principles through experiential learning, similar to how AlphaGo mastered Go.
Why are investors willing to fund a pre-revenue artificial general intelligence startup?
Investors in this artificial general intelligence investment see Silver’s proven track record with breakthrough AI systems like AlphaGo and AlphaStar at DeepMind. The AI startup funding market values potential to solve fundamental technical challenges, with nearly half of $469 billion in global venture capital last year going to AI companies, reflecting belief that whoever achieves AGI first captures extraordinary value.
What is the “Era of Experience” thesis that Ineffable Intelligence is based on?
The “Era of Experience” paper by Silver and Richard Sutton argues that AI systems trained solely on human-generated data are hitting performance walls. They predict that experience-based learning through direct environmental interaction will become the dominant improvement method, dwarfing the scale of human data used in today’s systems and enabling AI to discover novel solutions humans never conceived.
How does this former DeepMind AI firm compare to other superintelligence startups?
Similar to Ilya Sutskever’s Safe Superintelligence (valued at $32 billion) and Yann LeCun’s AMI Labs (raising €500M), this DeepMind scientist new venture represents elite researchers leaving Big Tech to pursue alternative AGI approaches. However, Silver’s specific focus on pure reinforcement learning over hybrid approaches distinguishes Ineffable Intelligence from competitors.
What are the risks facing this superhuman intelligence startup?
Major challenges include defining reward signals for open-ended general intelligence tasks, resource competition against tech giants spending $527 billion on AI in 2026, and uncertain timelines for breakthrough achievements. Critics also question whether reinforcement learning alone can solve the full spectrum of general intelligence requirements beyond game-playing domains.
How large is the artificial general intelligence investment market?
The AGI market is projected to grow from $6.94 billion in 2025 to $9.30 billion in 2026 (33.9% CAGR), reaching $29.67 billion by 2030. The broader AI startup funding ecosystem saw nearly half of $469 billion in global venture capital flow to AI companies, with the AGI market specifically expected to exceed $116 billion by 2035 at a 36.25% compound annual growth rate.
