London is quickly becoming one of the biggest AI hubs in the world—and almost nobody is talking about it. In the past few months, several of the biggest AI companies have started concentrating major investments in the city.
OpenAI is making London its largest research hub outside San Francisco. Anthropic plans to hire more than 100 people across London and Dublin in 2025, while xAI opened a local office led by a former DeepMind researcher.
Microsoft hired more than two dozen DeepMind researchers to strengthen its London presence. Google DeepMind announced a new automated research lab in the UK focused on discovering materials with AI and robotics by 2026.
Perplexity is allocating roughly £80 million to expand its offices, Groq is opening its first UK data center, and Cursor selected London for its European headquarters. When multiple frontier labs invest in the same place, a new global tech hub is emerging.
The funding wave extends beyond geography. Ineffable Intelligence, founded by David Silver, is raising a $1 billion seed round at a $4 billion pre-money valuation—the largest seed in European history.
Silver is explicitly rejecting the pure LLM scaling path. Today’s leading models predict the next word using mountains of static internet text. Scale the data, scale the parameters, scale the compute—and results improve.
Silver argues that approach has structural limits: high-quality text is finite, predicting tokens is not the same as understanding the world, and the systems remain passive observers. They do not act or pursue goals inside environments.
Ineffable plans to scale reinforcement learning instead. Agents train inside environments—games, simulations, eventually the real world—taking actions and receiving rewards. Through policy optimisation and endless self-play, intelligence is grown rather than copied.
Investors are betting that active systems which plan, adapt, and generalise across environments can surpass static LLMs. If that happens, the dominant AI training paradigm changes.
Between London’s infrastructure boom and a $1 billion seed round backing a new philosophy of intelligence, Europe is quietly reshaping the AI map—and hinting at what superintelligence research looks like next.