- Elon Musk predicts AI will surpass human intelligence by 2029, sparking debate.
- Meta’s AI lead Yann LeCun dismisses Musk’s claims, citing current AI limitations.
- LeCun argues AI lacks a crucial component, requiring a breakthrough beyond technology.
Bold predictions by Musk
Elon Musk, the billionaire and adamant AI supporter, recently claimed on X that AI “will probably be smarter than any single human next year” and “smarter than all humans combined” by 2029.
His bold statements have sparked a heated debate within the AI community, particularly catching the attention of Meta’s AI lead, Yann LeCun.
The “godfather of AI” disagrees
LeCun, often referred to as one of the “godfathers of AI,” swiftly dismissed Musk’s claims. He argued that if AI were truly that advanced, we would already have systems capable of teaching themselves to drive a car in just 20 hours of practice, like a 17-year-old.
LeCun pointed out that even Tesla’s self-driving cars, which have millions of hours of labeled training data, are not yet fully autonomous or reliable.
What AI is missing
This is not the first time LeCun has spoken out against claims of highly intelligent AI being imminent. He previously stated that AI models have about as much computing power as a common housecat’s brain but lack the cleverness.
LeCun believes that current AI systems are missing a crucial conceptual component that prevents them from achieving the level of intelligence found in animals and humans.
He suggests that scaling up existing technology alone will not suffice and that a breakthrough is needed to bridge the gap.
