Michael I. Jordan, a machine learning pioneer and professor emeritus at UC Berkeley, currently leads the markets and learning chair at Inria in Paris. He characterizes recent public discourse surrounding existential threats from artificial intelligence as a form of science fiction rather than grounded reality.
Jordan criticizes the push for AI regulation by tech companies as performative theater. He argues that these firms seek to maintain control over their own oversight while projecting an image of corporate responsibility to the public.
The researcher highlights a lack of engineering discipline among young employees at AI companies, noting that many come from humanities backgrounds. He views recent high-profile resignations as moralizing performances rather than genuine ethical stands.
Jordan disputes the feasibility of superintelligence, specifically challenging the arguments of Geoffrey Hinton. He asserts that the belief in replicating brain functions in silicon to achieve exponential growth lacks credible scientific foundation.
Regarding infrastructure, Jordan describes massive investments in data centers as speculative bets. He suggests that companies can achieve specialized tasks using open-weight models without relying on gigantic, costly computing facilities.
