Ineffable Intelligence Raises US$1.1 Billion for Autonomous AI

Ineffable Intelligence secured US$1.1 billion to develop a reinforcement learning superlearner independent of human data. Led by Sequoia and Lightspeed, this investment accelerates the shift toward autonomous AI systems, offering B2B stakeholders a scalable, self-evolving alternative to traditional large language models that rely on human-generated datasets.

Ineffable Intelligence raised US$1.1 billion in a seed funding round to build AI models that learn through autonomous experience rather than human data. The company achieved a valuation of US$5.1 billion.

The funding supports the development of a superlearner that utilizes reinforcement learning to discover knowledge through trial and error. David Silver, CEO, Ineffable Intelligence, describes the initiative as his life’s work, while the company’s website notes that “if successful, this will represent a scientific breakthrough of comparable magnitude to Darwin: where his law explained all Life, our law will explain and build all Intelligence.”

Based in London, Ineffable Intelligence was established in 2025 by Silver, who previously led the reinforcement learning department at DeepMind. During his decade at the Google-owned division, he oversaw the development of AlphaGo, AlphaZero, AlphaFold, and AlphaProof. Those systems defeated world champions in Go and chess by learning from internal simulations rather than human strategy databases or historical records.

Ineffable Intelligence functions as a specialized research laboratory. Its official platform, ineffable.ai, emphasizes a departure from large language models that depend on internet-scraped datasets. Instead, the corporation focuses on an agent-based architecture where the software learns from its own environment. 

This approach addresses the increasing scarcity of high-quality human data, providing a path toward artificial general intelligence (AGI) that does not mirror human biases or limitations.

The technical foundation of the corporation rests on the principle that intelligence can be synthesized through pure interaction. Unlike supervised learning, which requires humans to label data, reinforcement learning allows a system to receive feedback from its successes and failures. This method is particularly relevant for B2B enterprises seeking specialized applications in complex environments where human data is either unavailable or insufficient to train traditional models.

The US$1.1 billion round represents the largest seed investment in EU history. Sequoia Capital and Lightspeed Venture Partners led the financing. Other participants include Nvidia, Google, DST Global, Index Ventures, EQT, BOND, and Flying Fish, among others. The UK government also provided support through the Sovereign AI Fund, which recently launched with £500 million (US$675 million) to foster domestic technological autonomy. Furthermore, the British Business Bank contributed US$20 million to the round.

The emergence of Ineffable Intelligence reinforces London as a primary global hub for AI development. This momentum is evidenced by the presence of Recursive Superintelligence, founded by Tim Rocktäschel, former Principal Scientist, DeepMind, which raised US$500 million. Additionally, AMI Labs, cofounded by Yann LeCun, former Chief AI Scientist, Meta, secured US$1.03 billion earlier this year. 

These high-value transactions, frequently termed “coconut rounds,” signify a shift in venture capital toward funding established researchers with the resources necessary to challenge incumbent technology giants.

The Business Case for New Data

As large language models (LLM’s) reach the limits of what they can learn from public experience, Ineffable Intelligence is betting on new ways to make these models capable of learning now by themselves. This superlearner that generates its own expertise aims to provide corporations with tools that solve novel problems in engineering, biology, and logistics. By removing the human bottleneck, the development cycle for specialized intelligence could accelerate significantly during this time.

The UK government has taken an active role in these developments. The British Business Bank has made nine investments in AI over the last year, supporting companies such as Wayve and PolyAI. Between 2022 and 2024, the bank supported nearly one-quarter of all university spinout deals in the United Kingdom. This institutional backing suggests that the United Kingdom intends to provide a stable environment for companies like Ineffable Intelligence to scale without relocating to the United States.

Silver says to Wired that any personal proceeds he receives from the success of the company will be donated to high-impact charities through Founders Pledge. This commitment underscores the mission-driven nature of the project. 

While some industry analysts question whether reinforcement learning alone can achieve the level of intelligence Silver envisions, the scale of the capital raised indicates that institutional investors view his expertise as a critical asset in the race for autonomous systems.

Future developments for Ineffable Intelligence include the expansion of its executive team, which, according to TechCrunch, is expected to include several former staff members from DeepMind.