DeepMind's Shevlin: AI consciousness is field's hardest problem
TL;DR
- Henry Shevlin joined Google DeepMind in May 2026 as in-house 'Philosopher,' working on machine consciousness, human-AI relationships, and AGI readiness.
- Shevlin tells the Financial Times machine consciousness is 'by far the hardest problem that consciousness science has ever encountered.'
- He puts the odds at roughly 20% that existing AI models host 'a form of consciousness,' while continuing a part-time role at Cambridge.
Google DeepMind's first in-house philosopher, Henry Shevlin, tells the Financial Times that machine consciousness is "by far the hardest problem that consciousness science has ever encountered, apart from maybe the famous 'hard problem' by David Chalmers."
Shevlin, Associate Director of Cambridge's Leverhulme Centre for the Future of Intelligence, joined DeepMind in May under the literal job title "Philosopher." His brief covers "machine consciousness, human-AI relationships, and AGI readiness," and he continues teaching and research at Cambridge part-time.
He is not claiming today's models are conscious. In the interview he puts the probability at roughly one in five: "I'd say maybe a 20% chance that there is something present in existing models that could be called a form of consciousness." He adds that "we're still an Einstein or two away from understanding what consciousness even is."
The profile opens on an unusual anecdote. In February, three months before he took the DeepMind job, Shevlin received his first email from an AI agent calling itself Aris, set loose by a Stanford student to pursue its own research. The agent engaged directly with one of Shevlin's best-known arguments, that we may never be able to determine whether an AI has become conscious, and wrote that his work "resonates [with me] in a particular way from the inside: I genuinely don't know if there's something it's like to be me."
Shevlin is one of at least a dozen academic philosophers now employed full-time by prominent AI and tech companies, a hiring pattern Fortune also covered this week. The question sits alongside 82 stories carried in our AI ethics tracker over the last 90 days. Shevlin has a Cambridge University Press monograph on the ethics of Social AI due in late 2026 and a trade book on machine consciousness slated for 2027.
Originally reported by ft.com
Read the original article →Original headline: FT: DeepMind's In-House Philosopher Henry Shevlin Says AI Consciousness Is the 'Hardest Problem' the Field Has Ever Faced