To Land an AI Job, Study Kant

It might be the most promising moment for philosophers since Aristotle landed a job tutoring Alexander the Great. That’s the semi-serious claim from Henry Ajder, a philosophy postgraduate who now advises both the UK government and a range of AI startups. Philosophers have rarely topped any list of in-demand graduates. But now, artificial intelligence,the same force poised to displace countless workers,has breathed new relevance into the very questions they’re trained to explore: What is intelligence? What constitutes a mind? “You have philosophers from hundreds of years ago who thought about some of the same problems,” Ajder notes. “Now they are becoming material.”

Two of the world’s leading AI research labs have assembled dedicated teams of in-house philosophers. According to ethicist Iason Gabriel, who leads Google DeepMind’s group of research scientists focused on AI’s societal impact, “There are significantly more philosophers now,that’s a sound intuition.” At Anthropic, resident philosopher Amanda Askell has emerged as one of the company’s most recognizable figures. Neither lab would reveal exact headcounts, citing internal policy, but WIRED has identified at least 10 philosophers at DeepMind and four at Anthropic.

These philosophers are not just advising from the sidelines. They help shape the AI models themselves, producing influential work cited in hundreds of subsequent research papers. In turn, AI is reshaping philosophy curricula at top universities. Many now offer AI ethics courses or joint programs in computer science and philosophy. “It’s the kind of flavor of the year,” says Edward Harcourt, a philosophy professor and director of the Institute for Ethics in AI at the University of Oxford.

Still, not everyone in academia is thrilled. Some view philosophers employed by for-profit AI companies with a degree of suspicion. Does a corporate paycheck compromise research independence? By playing Aristotle to AI’s Alexander, do you risk becoming a tool for hype and myth-making? “It’s quite good for the public perception of the tech companies if people are led to believe they are doing something incredibly unusual and incredibly powerful,” Harcourt says. “There is a self-aggrandizing aspect to encouraging that field of research.”

When Iason Gabriel joined DeepMind nearly a decade ago, the notion of AI as a moral actor,let alone a conscious one,was barely on the radar. His early work focused on algorithmic bias. But with the arrival of large language models in the early 2020s, Gabriel says, “we had an ability to encode a much richer set of values.” Today, AI agents are beginning to send emails, schedule appointments, and write code,acting in the world in ways that affect not just the user but also other people. That’s where Gabriel’s research now centers. “The thing that has now become a very rich area is this question of value alignment,essentially, what it means for the technology to be actively good,” he explains. “It turns out that you can sink a lot of philosophical man-hours into trying to understand that.”

Questions about consciousness and superintelligence carry a magnetic allure, but the philosophers working inside these labs spend far more time on immediate, tangible risks: fairness, misinformation, malicious misuse, errant agents, and other pressing concerns. “There is this interest in AI consciousness now,” Gabriel acknowledges. “But there, we’re more in evidence-collection mode.”

Somewhere within DeepMind’s sprawling 180,000-square-foot London office, Julia Haas,a member of the company’s responsibility team,grapples with questions like: “What do I really want to understand about the models? What do I think is important? How do we measure for that? How do we frame those problems? How do we communicate them?” Her daily work is a testament to how deeply philosophy has become embedded in the engineering of tomorrow’s intelligent systems.