The round was led by Sequoia Capital and Lightspeed Venture Partners, with participation from Nvidia, Google, and Index Ventures, according to a report by City AM. The deal places Ineffable Intelligence among a small cohort of European AI companies to have secured nine-figure sums in a single raise, and it arrives at a moment when investors are actively searching for alternatives to the large language model paradigm that has dominated the sector since 2022.
What Ineffable Intelligence is building, and how it differs from the LLM wave
Ineffable Intelligence is built around reinforcement learning (RL), a branch of machine learning in which AI systems improve through trial and error in simulated environments rather than by ingesting large datasets of human-generated text. The company plans to develop what it calls "superlearners": AI agents capable of acquiring new capabilities independently of human input, as reported by City AM.
Silver's credentials in this field are difficult to overstate. At Google DeepMind he served as lead researcher on AlphaGo, the programme that defeated world Go champion Lee Sedol in 2016, and on AlphaZero, which taught itself chess, shogi, and Go from scratch to superhuman level. His work on reward-based learning also contributed to the precursor research behind AlphaFold, DeepMind's protein-structure prediction system. Collectively, these projects represent some of the most cited demonstrations that RL can solve problems where brute-force data ingestion falls short.
"Human data is like a kind of fossil fuel that has provided an amazing shortcut," Silver told City AM. "You can think of systems that learn for themselves as a renewable fuel, something that can just learn and learn and learn forever."
The distinction matters commercially. Large language models such as OpenAI's ChatGPT and Google's Gemini are trained on vast corpora of human text and excel at language tasks, but they remain constrained by the distribution of their training data. RL-based systems, by contrast, can theoretically discover strategies and solutions that no human has previously documented. Silver told City AM he views the mission as "making first contact with superintelligence," with agents that could "discover new forms of science or technology or government or economics for itself."
Inside the round: who invested and what the valuation signals
At $5.1bn, Ineffable Intelligence enters the market at a valuation that places it alongside the most richly priced AI startups globally, despite having no publicly disclosed revenue. The round's structure offers several signals worth unpacking.
Sequoia Capital has been steadily increasing its exposure to frontier AI. The firm was an early backer of several prominent Silicon Valley AI companies and has made selective European bets in recent years. Sonya Huang, a partner at Sequoia, told City AM: "There's only a very, very small number, less than a handful of people, who have done truly foundational work. Dave is one of them."
Lightspeed Venture Partners co-led the round. Ravi Mhatre, a partner at the firm, described Silver's career as "basically a single, coherent argument for being able to scale intelligence without human priors," according to City AM.
The strategic participation of Nvidia and Google is notable. Nvidia supplies the GPU infrastructure on which virtually all frontier AI research depends; its investment suggests confidence that RL workloads will generate meaningful compute demand. Google's involvement is striking given that Silver spent years at its DeepMind subsidiary; the corporate venture arm evidently judged it better to hold a stake in the new entity than to lose proximity to the research entirely.
For context, the round ranks among Europe's largest AI raises. Mistral AI, the Paris-based LLM developer founded by former DeepMind and Meta AI researchers, raised €600m in a Series B round in June 2024, according to the company's own announcement, valuing it at roughly €5.8bn. Germany's Aleph Alpha raised $500m in a Series B in late 2023, as reported by TechCrunch. Ineffable Intelligence's $1.1bn raise exceeds both in absolute dollar terms and stands as one of the largest single rounds for any UK-headquartered technology company.
London's neolab moment: can the UK hold onto frontier AI ventures?
The deal sits within a broader pattern that some investors have labelled the "neolab" trend: senior researchers departing established AI organisations to found independent ventures with substantial upfront capital. Ilya Sutskever, co-founder and former chief scientist of OpenAI, launched Safe Superintelligence Inc in 2024 and reportedly raised $1bn before writing a line of production code, according to reporting by Bloomberg. Arthur Mensch, a former DeepMind researcher, co-founded Mistral AI in Paris. The pattern is consistent: proven researchers attract pre-revenue capital at scale because investors are betting on the scarcity of genuine frontier expertise.
What distinguishes the Ineffable Intelligence round is its geography. London has long produced world-class AI research, largely through DeepMind, which Google acquired for a reported £400m in 2014. Yet the commercial value of that research has historically accrued to a Mountain View parent company. The question now is whether the capital and infrastructure exist to keep an independent, capital-intensive AI lab headquartered in the UK.
Several structural factors work in London's favour. The UK government established the AI Safety Institute in late 2023, positioning Britain as a hub for frontier AI governance. The R&D tax credit regime, despite recent reforms that reduced generosity for some SMEs, still offers relief for qualifying expenditure on AI research. And London's deep pool of machine-learning talent, fed by universities including Imperial College London, UCL, and the University of Cambridge, provides a recruitment base that few European cities can match.
But the challenges are real. The cost of compute, energy prices, and the gravitational pull of Silicon Valley's talent market all work against retention. Whether Ineffable Intelligence remains London-based through its growth phase or eventually shifts its centre of gravity westward will be a meaningful test case.
The talent pipeline question
Silver is not the only senior DeepMind alumnus to have struck out independently. The broader exodus of experienced researchers from large labs creates both opportunity and risk for the UK ecosystem. If the neolabs stay and hire locally, they deepen London's AI cluster. If they relocate after raising capital, the fundraise becomes a data point about UK ambition rather than UK capacity.
What reinforcement-learning startups mean for enterprise operators
For UK business operators currently evaluating AI adoption, the Ineffable Intelligence round carries practical implications. The dominant enterprise AI toolkit today is built on LLMs: chatbots, document summarisation, code generation, and customer-service automation. These applications are real and delivering measurable efficiency gains across sectors.
Reinforcement learning opens a different category of use case. RL agents excel in domains with clear reward signals and complex decision spaces: logistics optimisation, financial trading strategies, robotics control, drug discovery, and resource scheduling. DeepMind's own work demonstrated this in data-centre energy management, where RL reduced cooling costs by 40%, according to a 2016 DeepMind blog post.
The commercial question is timing. RL-based products have historically required bespoke engineering and significant compute budgets, making them impractical for most mid-market firms. If Ineffable Intelligence or its peers can package RL capabilities into accessible tools, a new wave of enterprise applications could follow. If the technology remains confined to research labs and trillion-dollar corporations, the impact on most UK operators will be indirect.
What is clear is that the competitive landscape is shifting. Businesses building products on top of LLM APIs face a future in which those APIs may be only one paradigm among several. Operators with exposure to logistics, manufacturing, scientific R&D, or complex scheduling should monitor whether RL-native products begin to reach the market in commercially viable form.
The $1.1bn committed to Ineffable Intelligence is a bet that they will. Whether that bet pays off, and whether London reaps the economic benefit, remain open questions. But the capital has been deployed, the lab is hiring, and the experiment is under way.
