In a recent interview, Elon Musk talked about the rate of development of artificial intelligence (AI), as well as the potential threat that AI poses to humanity.

He said, in particular, about the rate of improvement of AI. "It is difficult to calculate exactly, but I can say with confidence that every year the situation improves 10 times... So in four years it will improve 10,000 times. Maybe 100,000 times."

To the question of when AI will surpass humanity in its performance, Elon Musk answered: I think maybe in the next year or two he can do anything anyone can do. How long will it take to learn to do what all humans combined can do? I think not much. I think only three years. So around 2028 or 2029.”

Elon Musk has again raised the issue of the ontological threat to humanity. “Most likely everything will be fine, but there is a certain chance, 10 to 20%, that things will go wrong. The probability that things will go wrong is not zero, but generally the glass is 80% full, maybe 90%... I think AB is a significant ontological threat that we need to pay close attention to it. This is probably the most significant threat in the short term."