Inherent, a London-based AI lab founded by Google DeepMind alumni, claims its agent outperformed larger systems from Anthropic and OpenAI at replicating scientific research while using a fraction of the compute.
Among the startups launched by former DeepMind staff, Inherent has received relatively little attention. While better-funded rivals have yet to show the world anything concrete, the London team is starting to share what it has been building.
Just weeks after emerging from stealth with a $50 million seed round, the British startup says its newly released agent, Faraday, beat larger, better-known models at a specific task. The agent independently reproduced the findings of published scientific papers without being told the answer in advance.
This may sound like a mere party trick given Inherent’s much loftier goal of building AI that can discover new scientific knowledge rather than just verifying old results. But paper replication is a standard training exercise for human scientists, cofounder and chief scientist Edward Hughes said. “Many PhD students actually start by doing this.”
Beating other AI systems at the task was not the point, Hughes told TechCrunch. How they got there was the focus. “What was most interesting to us about this was not so much the result of beating those frontier agents — which of course we liked — but was actually the way we went about building this.”
The part that should catch an investor’s eye is the hardware used. Measured against Anthropic’s Claude Opus 4.8 and OpenAI’s GPT-5.5, both much larger frontier-scale systems, Faraday runs on a comparatively tiny model called Qwen 3.6 with just 27 billion parameters. Roughly speaking, “parameters” is a proxy for a model’s size, training costs, and complexity.
Inherent’s bar for success was also higher than simple accuracy. Beyond replicating results, it wanted Faraday to demonstrate “research taste” — an instinct for what experiments are worth running and how to design them well.
Teaching something as intangible as taste is hard, which is where reinforcement learning comes in. This training method rewards an AI system for good outcomes rather than spelling out rules for it to follow. Rather than training its agents primarily on the study of how science is conducted, Inherent leans on this reward-based approach. It bets the method will generalise better to its longer-term goal of agents capable of contributing across many scientific fields.
“We’re always guided by that north star of building an AI scientist agent and imbuing our agents with taste,” Hughes said. That focus has also shaped what Inherent chooses not to build. Rather than developing its own coding tool, it had Faraday use OpenAI’s GPT-5.5 Codex. The company noted this mirrors how human scientists lean on existing software rather than building everything themselves.
Inherent is also trying to avoid building agents that simply tell users what they want to hear. Hughes said the goal is modeled on his favourite kind of teammate — the kind who comes back and says: “I got curious about this, and I went off and I did these experiments. What do you think of these results?”
That collaborative instinct extends to how Inherent operates as a company. Its dozen employees all work in person out of an office in King’s Cross. The once-rundown London neighbourhood became one of the world’s top AI hubs thanks to Google DeepMind’s presence. “We believe that London is the place to be,” Hughes said.
Hughes is bullish on London’s density of AI talent. He has also added his voice to calls to end “garden leave”. This is the practice, common in the U.K., of barring departing employees from joining or starting a rival company for months after they resign. It is a restriction American researchers generally do not face, giving U.S. startups a head start on hiring talent who have left a prior role. “This is a personal view rather than a company view, but I was affected by the garden leave problem,” he told TechCrunch.
Hughes eventually got around that constraint and started Inherent alongside two other DeepMind alumni and a fourth cofounder. The startup is not slowing down. It plans to grow its headcount to “about 20 to 25” by the end of the year. Given its ambitions in world models and with Demis Hassabis’s new role leaving some DeepMind staff unsettled, Inherent’s hiring push could make it an appealing landing spot for DeepMind employees weighing a move.
What it means
The shift away from massive models to smaller, reward-trained agents suggests a change in how creative and scientific tools are built. Instead of relying on raw data volume, the focus is on teaching systems to judge the quality of their own work. For researchers, this means tools that offer genuine collaboration rather than just generating text. They can expect assistants that propose experiments based on curiosity and ask for feedback on results, rather than simply confirming what the user already knows.
