Entrepreneur Elon Musk insists that innovators should be aware of the future of social risk that comes with the use of artificial intelligence (AI). At South by Southwest (SXSW) on Sunday, the SpaceX founder reminded us of this once again.
“This is a case where there is a very serious risk to the public, so there should be a public agency that has the insight and oversight of all those who are developing AI. It is extremely important,” he said.
“Some AI experts think they know more than they think and that they are smarter than they are… So we have smart people, who define themselves by their intelligence, and they don’t like the idea that a machine can become smarter than them, which is fundamentally wrong.”
Referring to AlphaGo and its predecessor AlphaGo Zero, Musk stated that the AI is far more capable than anyone else at the game and mentioned that its rate of improvement is very aggressive.
In six to nine months, AlphaGo was able to beat a fairly good Go player, and then beat current and former world champions. It learned by playing against itself and can play any game once we set the rules for it.
“Nobody predicted the rate of improvement,” Musk said.
Predicting a similar fate for self-driving vehicles, Elon Musk expects that by the end of next year, automated driving will encompass virtually all driving modes and be at least 100-200 percent safer than humans.
“The rate of improvement is truly dramatic, but we have to find some way to ensure that the emergence of digital superintelligence is symbiotic with humanity. I think this is the single greatest existential crisis we face, and the most pressing,” he said.
“The risk of AI is much greater than the risk of nuclear, and no one would suggest that we allow anyone to build nuclear.”
Elon Musk: “Mark my words: AI is much more dangerous than nuclear, and we have no regulatory oversight, that’s crazy.”
