With plans for establishing human colonies on Mars, Elon Musk is easily the most futuristic entrepreneur out there. But one future technology that has even Musk spooked is artificial intelligence.
At a talk for MIT students on the institute’s campus last year, Musk warned that developing AI unchecked is like “summoning the demon.” Now, he’s pledged $10 million to fund research on the safety of AI.
Funding research on artificial intelligence safety. It’s all fun & games until someone loses an I http://t.co/t1aGnrTU21
— Elon Musk (@elonmusk) Jan. 15, 2015
The funds will be administered by the Future of Life Institute, a nonprofit organization run by scientists and engineers who’ve been championing the need for developing safeguards for intelligent technology.
But that’s just a start. “There should probably be a much larger amount of money applied to AI safety in multiple ways,” Musk said in a YouTube video announcement.
Intelligence machines can be “mistake amplifiers,” said Thomas Dietterich, a professor at Oregon State University and president of the Association for the Advancement of Artificial Intelligence. He said there needed to be switches in place, circuit-breakers of sort, to detect flaws and stop their spread.
Leading AI researchers back Musk and Dietterich’s views. In open letter published this week and hosted by the Future of Life Institute, physicists and engineers called for research into technologies that would explore not just benefits of AI, which could be tremendous, but also potential dangers.
Musk’s donation is the first step towards turning those words into actions.