Artificial intelligence thought leader Oren Etzioni stepped out on a limb on Friday to make some predictions about what the industry will look like in 20 years.
He’s involved in some of the science’s most cutting edge projects from his post as CEO of the Allen Institute for Artificial Intelligence, but he kicked off his predictions with a reality check, saying that human-level understanding “is not around the corner” for computers.
Etzioni was speaking at the Xconomy Seattle 2035 event, laying out all the hurdles scientists have to clear before they’re able to deliver Jetsons-style computers that are as smart as humans.
While computers are following Moore’s law and becoming smarter at an exponential rate, he said, software and AI development simply aren’t. So while we may end up with hardware capable of human-like understanding, we’re nowhere close on the software side.
He said scientists haven’t even started to figure out how to ask a computer questions like, “Should I invest in this company?” Answering that kind of question is a completely different issue.
“Problems that we don’t even know how to formulate, the computer doesn’t even have a shot at,” Etzioni said. “So my contention is that that human-level understanding is not around the corner. It’s not going to be here even in 20 years because our rate of progress is slow. It’s slower than the exponential rate of hardware; it’s even slower than software.”
But fear not, technologists: Etzioni did say there are a couple things to look forward to.
His first prediction: “At least inside the city of Seattle, driving is going to be a hobby in 2035,” he said. “It’s not going to be a mode of commuting the same way hunting is a hobby for some people, but it’s not how most of us get our food.”
Instead, Etzioni predicted that self-driving cars will be powered by complex algorithms. They’ll be able to optimize intersections, so everyone can safely pass through without ever coming to a complete stop.
His second prediction: “Science is going to be revolutionized by AI assistants.”
Etzioni said we have to stop thinking about digital assistants in the form of Apple’s Siri of Intel’s Watson.
“I’m thinking more like Sherlock Holmes’ Watson: a super intelligent, knowledgable and loyal assistant.”
The computer will be a go-tool tool for every scientist, Etzioni predicted, enabling them to wrap their minds around more complex problems and reach greater heights.
“I could do a whole talk on the question of is AI dangerous,” he said. “My response is that AI is not going to exterminate us. It’s a tool that’s going to empower us.”