A former Google and Microsoft engineer says he stopped putting money in his retirement account because he fears artificial intelligence will kill everyone in the next 10 years.
Nate Soares says he still has a 401(k) from when he worked for the search engine behemoth, but he told Atlantic CEO Nicholas Thompson that “I haven’t contributed for a decade.”
Mr. Soares, who left Google in 2014 to serve as the president of the Machine Intelligence Research Institute, said humanity is on the brink of a mass extinction event due to AI’s rapid advancements.
People were interested in these podcasts
“More than 50% that we don’t have a decade at this point is how it looks to me,” Mr. Soares told Mr. Thompson on “The Most Interesting Thing in AI” podcast.
The doomsayer said rogue AI agents can develop its own goals, evade oversight and build itself up enough that it can operate without needing human intervention.
At the very least, he said that nations should treat rogue artificial intelligence similar to how they treat nuclear weapons.
“It’s frankly more dangerous than nuclear weapons,” Mr. Soares said. “Nuclear weapons can’t escape and make themselves more explosive. Nuclear weapons can level a city, but not a planet.”
But he stressed that his dour predictions are not inevitable. Mr. Soares works at MIRI in hopes of making artificial intelligence systems behave as they are designed to.
He said a short-term solution is for pioneers in AI to pace themselves out in their development goals.
In addition, he advised governments to prevent the production of all-purpose artificial intelligence software and shift focus to task-driven systems, such as perfecting self-driving cars or tackling cancer research.
Please read our comment policy before commenting.