Nvidia Backs New AI Superlearner Company

A new AI startup founded by former Google DeepMind researchers is rapidly emerging as one of the most ambitious players in the race toward artificial superintelligence, and it already has backing from Nvidia, Google, Sequoia, and Lightspeed.

The company, called Ineffable Intelligence, was founded by renowned AI researcher David Silver and has now raised an enormous $1.1 billion seed round at a reported $5.1 billion valuation. The funding ranks among the largest seed financings ever recorded for a European AI startup.

Unlike many generative AI startups building larger language models, Ineffable Intelligence is pursuing a different goal entirely: creating systems that can learn independently without relying heavily on human-generated training data.

The Startup Wants AI to Learn Like AlphaGo

David Silver is best known for leading reinforcement learning research behind DeepMind’s landmark systems including AlphaGo and AlphaZero. Those systems famously learned strategic reasoning by playing millions of games against themselves instead of simply memorizing human examples.

Ineffable Intelligence now wants to apply that same reinforcement-learning philosophy far beyond board games.

According to the company’s public materials, the goal is to create what it calls a “superlearner”, an AI system capable of continuously discovering new knowledge and skills through autonomous experimentation rather than depending on massive internet-scale human datasets.

That approach represents a major philosophical shift away from the current large language model boom dominated by systems trained primarily on human-written text scraped from the web.

Instead of imitating human knowledge, the company believes future advanced AI systems may need to generate new understanding independently.

Nvidia Is Expanding Beyond Chips Into the AI Ecosystem

One of the most significant details about the funding round is Nvidia’s involvement.

The chipmaker has increasingly moved beyond selling GPUs and is now aggressively investing across the broader AI ecosystem, including infrastructure companies, cloud providers, robotics firms, and frontier AI labs.

By backing Ineffable Intelligence, Nvidia appears to be positioning itself close to the next generation of reinforcement-learning-driven AI research.

That matters because reinforcement learning at large scale requires extraordinary compute infrastructure. Training systems that learn continuously through simulation and self-play could demand even larger computational resources than today’s large language models.

For Nvidia, investing early gives the company influence over emerging AI paradigms that may shape future hardware demand.

The Industry Is Quietly Splitting Into Two AI Paths

The rise of Ineffable Intelligence highlights a growing divide inside the AI industry.

Most major companies, including OpenAI, Anthropic, and Meta, remain heavily focused on scaling large language models using ever-larger datasets and compute clusters.

But a smaller group of researchers increasingly believes current LLM architectures may eventually hit limitations around reasoning, long-term planning, and autonomous learning.

Researchers like David Silver argue that systems capable of self-directed learning through interaction and experimentation may ultimately become more powerful than models trained primarily through imitation.

That belief has roots deep inside DeepMind itself, where reinforcement learning powered many of the company’s most important breakthroughs over the past decade.

London Is Becoming a Major Frontier AI Hub

The funding round also reinforces London’s growing role in advanced AI research.

According to University College London, two UCL-connected AI startups collectively raised roughly $1.6 billion in recent weeks, signaling increasing investor interest in UK-based frontier AI development.

DeepMind’s long-term influence continues to shape much of that ecosystem.

Over the past few years, former DeepMind researchers have launched or joined a growing number of high-profile AI startups spanning infrastructure, safety research, autonomous agents, and next-generation learning systems.

The migration increasingly resembles the “PayPal Mafia” effect seen in Silicon Valley, except centered around DeepMind alumni building the next wave of AI companies.

The Biggest Question Is Whether LLMs Are Enough

Ineffable Intelligence still has no public product, no released model, and no commercial platform.

What investors are funding is an idea: the possibility that the future of superintelligence may not emerge from scaling chatbots alone.

That makes the company one of the clearest examples of how parts of the AI industry are beginning to look beyond today’s generative AI race toward entirely new architectures for machine intelligence.

Whether reinforcement learning can eventually produce systems capable of genuine autonomous reasoning at world scale remains uncertain.

But the size of the funding round makes one thing clear: some of the world’s biggest investors and AI companies believe the next breakthrough after large language models may already be starting to take shape.