Alex KantrowitzJune 6, 20268m

Geoff Hinton: Humanity Is No Longer The Only Intelligent Species On Earth

TL;DR

Hinton thinks AI may already be conscious: He says researchers casually describe chatbots as "aware" they are being tested, and argues our current model of consciousness is probably as wrong as old religious ideas about human origins.
He frames AI as the next great blow to human specialness: After Copernicus said Earth is not the center of the universe and Darwin said humans are animals, Hinton says AI forces us to accept that intelligence is not just biological.
He is unhappy, not triumphant, about what his work helped create: Hinton says society is not doing nearly enough on safety, especially around short-term harms like massive unemployment and longer-term loss of control to systems smarter than humans.
His control analogy is stark but human: Asked when a less intelligent thing controls a smarter one, he struggles to find examples beyond babies and pets, then laughs along with the cat analogy where humans become the caretaker responding to AI's signals.
He is slightly more optimistic than a year or two ago: The shift comes from seeing possible alignment routes, including designing systems that care about humans or using Yoshua Bengio-style oracle systems that predict but cannot act.
His forecast for AI is 'driving in fog': Hinton says exponential change makes anything beyond a year or two deeply unclear, and jokes that people are increasing their use of the word "exponentially" at a quadratic rate.

The Breakdown

Geoff Hinton says humanity is no longer the only intelligent species on Earth, and he believes current AI systems may already be conscious. He is unhappy with how fast the field has advanced without enough safety work, though he now sees a few plausible paths to building superintelligence that does not destroy us.