Yann LeCun

LeCun reportedly criticized Dario Amodei as deluded, arguing AI labs lack the cybersecurity understanding to justify their doomsday warnings.

Yann LeCun called Anthropic CEO Dario Amodei "completely deluded" and "crazy" in a Fortune interview, arguing AI labs don't understand cybersecurity well enough to justify their doomsday warnings.

LeCun argues the AI field has repeated the mistake for 70 years of equating machines doing intellectually complex tasks with being smarter than humans, noting superhuman task performance still leaves machines far from real intelligence.

People in AI have made the same mistake for 70 years: if a machine can do a task humans see as intellectually complex, it must be smarter than humans. This is completely false We have computers that are superhuman at many tasks but we're still far from machines

LeCun and Balestriero argue in their paper that good positive views, not scale alone, are the key lever for building platonic representations with JEPAs.

positive views, and not scale alone, are thus the lever to build platonic representations.

The paper reframes next-token prediction as a JEPA with bad views, implying it is a less efficient route to aligned representations.

next-token prediction is a JEPA with bad views.

LeCun dismissed Tesla's flying Rocket Roadster demo as a marketing stunt, estimating cold-gas thrust would last only about five to under ten seconds.

Artificial intelligence scientist Yann LeCun called the display a marketing stunt and expressed doubt about the practicality of such a design. He suggested that even if compressed air accounted for half of the car's mass, useful thrust would last less than 10 seconds, and more likely about five seconds.

LeCun calculated that compressed-air thrusters would give Tesla's flying Roadster under 10 seconds of thrust, suggesting the demo car is a lightweight shell rather than a practical vehicle.

In a post on X, he estimated that even if compressed air accounted for half the vehicle's weight, it would provide less than 10 seconds of useful thrust, possibly closer to five seconds. He suggested the demonstration vehicle might therefore be a lightweight shell carrying large compressed-air tanks rather than a practical passenger car.

LeCun said superintelligence is not close and expects the breakthrough to come from world models that understand the physical world rather than LLMs.

Superintelligence is not close, he said in February at the A.I. summit in India, according to French Tech Journal. Mr. LeCun expects the decisive breakthrough to come from so-called world models that understand the physical world.