Scary-smart AI is still 'decades and decades' away, Google says
The tech giant's head of search says it's far too early to be concerned about a machine-ruled, Terminator-like dystopia. That's good, because Google has bet its future on artificial intelligence.
Google executives talk about the company's future in artificial intelligence.
Whenever we talk about artificial intelligence, someone inevitably mentions Skynet, the destructive machine system in the Terminator movies.
But we shouldn't be worried about a dystopian rise of the robots. At least that's the opinion of John Giannandrea, Google's chief of search and the company's former AI boss.
Why? Because we're so far away from anything that would even resemble that scenario, he said Friday at Google's I/O developer conference near the company's HQ in Mountain View, California.
"I think researchers in the field don't really put much thought into that," he said. "I think it's a concern that's decades and decades away."
Giannandrea is talking about a concept called "superintelligence," where, as he puts it, "AI begets more AI."
He adds that while there are not many of them, there are some people in the artificial intelligence community who do focus on that stuff and the ethical questions behind machine learning. He said Google works with those researchers and supports those kinds of inquiries.
That's important, because Google has staked its future on AI. When CEO Sundar Pichai took the stage for the conference opening keynote presentation on Wednesday, he talked about bringing Google into every part of people's lives with a new service called the Google assistant. It's a voice-powered incarnation of Google's search engine and other services that will live in your phone, smartwatch, car, living room speakers and more. And the more you use it, the more it learns about you and the smarter it gets.
"It's Google asking users, 'Hi. How can I help?'" Pichai told CNET last week, ahead of the conference. "Think of it as building your own individual Google."
During his keynote talk, Pichai also showed a video of several robot arms that a research group at Google taught to pick up objects. In one breakthrough, a robot arm pushed an object out of the way to pick up another item, an action the machine learned on its own. The maneuver was both awe inspiring and unsettling.
"It's also conflated with the fact that people look at things like robots learning to pick things up and that's somehow inherently scary to people," Giannandrea said.
Google isn't the only Silicon Valley powerhouse working on AI. In April, Facebook unveiled a new Applied Machine Learning group. And Microsoft has been developing technology of its own, though its Twitter chatbot, Tay, devolved into a meme after it started spewing racist and sexual comments.
Meanwhile, luminaries like Elon Musk and Stephen Hawking have warned against the dangers of AI. In December, Musk, along with prominent tech investors Sam Altman and Peter Thiel, funded a new research nonprofit called OpenAI with $1 billion, the goal being to make sure AI technology is developed in a way that doesn't let it run amok.
One major tech executive who agrees with Giannandrea is Facebook's Palmer Luckey, who founded Oculus, the virtual-reality company that the social network bought in 2014 for $2 billion. He said in March that a lot of people are "creeped out" about AI because science fiction sensationalizes it.
Giannandrea said one way Google can help people get over that is for the company to just build good services.
"It's our job, when we bring products to market, to do it in a thoughtful way that people find genuinely useful," he said.