Martin Rees
Astronomer Royal
Britain’s Astronomer Royal and co-founder of Cambridge’s Centre for the Study of Existential Risk — a self-described “technical optimist” who expects AI to accelerate science and manage complex systems, yet counts advanced AI among the civilizational threats humanity must learn to govern.
In his words
I am not one to worry about the machines taking over, but I do worry about our undue dependence on them.
Since we can’t understand what’s going on inside them, we have to be cautious about handing over power to them.
Biography
Martin Rees (born 1942), Baron Rees of Ludlow, is a British cosmologist and astrophysicist, Astronomer Royal since 1995 and a former president of the Royal Society. A leading theorist of galaxy formation and the early universe, he has spent his later career thinking about the long-term prospects — and perils — of technological civilization.
Existential risk and AI
In 2012 Rees co-founded the Centre for the Study of Existential Risk (CSER) at Cambridge with Jaan Tallinn and Huw Price, an institute devoted to threats that could permanently curtail humanity’s future — pandemics, biotech, climate and advanced AI among them. In his 2018 book “On the Future” he wove AI into that landscape, and in 2023 he was among the signatories of the Center for AI Safety’s one-sentence statement that mitigating extinction risk from AI should be a global priority.
It may be only one or two more centuries before humans are overtaken or transcended by inorganic intelligence.
Rees’s worry is less about a robot uprising than about dependence and opacity: entrusting legal, medical and financial judgments to systems whose inner workings no one can inspect. He calls himself a “technical optimist but political pessimist,” convinced technology has made modern life possible yet doubtful that governance can keep pace — a nuance that separates him from harder-line doom voices like Eliezer Yudkowsky.
Where they stand in the war
Voices for AI
Sources & further reading
Canonical record: https://battlelines.ai/topic/martin-rees








