Cohere, the AI firm led by CEO and co-founder Aidan Gomez, is supplying its large language model North to the Aston Martin Aramco Formula One Team as part of its role as the squad’s Official Generative AI Partner, according to details published by the Silverstone-based outfit.
Gomez outlined Cohere’s philosophy and its application within Formula One in a feature released by Aston Martin, emphasizing that the company’s deployment model is designed to keep sensitive data entirely within a customer’s own environment — a critical requirement in a sport where intellectual property protection is paramount.
“F1 is a very competitive and very technical sport, and that’s why we’re very proud to be supplying our LLM, North, to the Aston Martin Aramco Formula One Team,” Gomez said. “Motorsport is fundamentally about engineering and innovation, and those are two disciplines that fit AI very well.”
Cohere’s approach differs sharply from rival AI laboratories, according to Gomez, who criticized the broader industry’s fixation on building superintelligence. “I think a lot of the other AI labs are focused on ‘building superintelligence’ and I find that line of thinking really dangerous,” he said. “We don’t believe in that.” Instead, Cohere has built models efficient enough to run on company-owned hardware using as few as two GPUs, a figure Gomez attributed to the firm’s resource-constrained origins. He grew up without internet access in rural Ontario, Canada, an experience he said shaped the company’s culture of doing more with less.
The security architecture is central to the partnership’s value proposition for a Formula One team. Cohere’s models run locally, and no data leaves the customer’s environment, Gomez explained. “If your work is a national security concern, Cohere’s deployment method is the most secure,” he said, adding that the same principle applies naturally to the competitive secrecy demanded by F1. “Security is all about maintaining your advantage and making sure your secrets remain protected; and innovation is about giving engineers access to frontier AI and helping them discover ideas and concepts that they’d never ordinarily explore.”
Gomez illustrated the potential of human-machine collaboration with an anecdote about chess grandmaster Magnus Carlsen, whom he described as a friend. Carlsen used the AI chess algorithm AlphaZero and adopted an unorthodox early-game rook and pawn move the system recommended — a strategy every previous algorithm had rated as poor but which proved optimal and gave Carlsen a sustained competitive advantage. “That’s a great example of how AI can not only show us things that we can’t ordinarily see, but also how we can learn, adapt and develop those new ideas too,” Gomez said.
Despite the transformative potential he ascribes to AI, Gomez was careful to frame the technology as an amplifier rather than a replacement for human ingenuity. “AI is a tool. It doesn’t replace human creativity and it doesn’t come up with ideas — ideas are a human thing,” he said. “But what it can do is help you execute faster, iterate more easily and ideate more freely. It’s a creative enhancement.”
Cohere has built its business around clients in data-sensitive sectors including governments and telecommunications firms. Gomez argued that concentration of advanced AI capability among a small number of providers poses a structural risk. “For too long, there’s been too much concentration of technology — we all buy our tech from the same few sources,” he said. “People need choices for this sophisticated and empowering tech, otherwise you create an over-dependence on a single source.”