Elon Musk called the probability with which AI will destroy humanity

During the Great AI Debate seminar ElonMusk shared his thoughts on the development of modern AI algorithms. In particular, he named the estimated probability of a “rebellion” of artificial intelligence, which other industry experts consider significantly underestimated.

According to the billionaire, todaythe likelihood of a “rebellion of the machines”, in which AI technologies will turn against people, ranges from 10 to 20%. However, in his opinion, such a risk is justified. He argues that for security purposes, it is necessary to develop AI in such a way that it provides only truthful information.

Previously, researchers found that once A.I.learns to lie to people, his behavior will no longer be able to be changed using existing security measures. The New York Times called the announcement "a frightening new statistic sweeping Silicon Valley," with various tech executives giving highly pessimistic estimates of up to 50% of the negative outcome.

Roman Yampolsky, AI security researcher anddirector of the cybersecurity laboratory at the University of Louisville, also argues that Musk is right about the threat of AI. In his opinion, the actual probability of artificial intelligence seizing control of humanity is much higher than Musk claims (more than 99%).