Elon Musk estimates a 10-20% chance of AI causing humanity's destruction, yet supports AI development.

Elon Musk estimates a 10-20% chance of AI destroying humanity, but believes the potential positive outcomes outweigh the negative. Speaking in a "Great AI Debate" seminar, he reconsidered his previous risk assessment, yet remains supportive of AI development. AI safety expert Roman Yampolskiy suggests Musk may be underestimating the risk of catastrophe.