Mythbusting AI with Dr. Steven Shwartz – Episode 22
By Donna Loughlin
Steven Shwartz was studying AI before it was cool. He was conducting pioneering research in the field before anyone knew the term “artificial intelligence.” And he was launching AI startups decades before companies like Google and Tesla even existed.So when someone like Steven Shwartz tells us not to be afraid of the robots taking over the world, we should listen.
Throughout his career, Steven has watched AI go from a little-known subset of robotics to one of the most important technological developments of the last quarter-century. Today, AI stands at the center of just about everything — and will likely impact every single industry and human being in the world in the coming decades.But with his new book: Evil Robots, Killer Computers and Other Myths: The Truth about AI and the Future of Humanity, Steven set out to play the role of mythbuster to his own life’s work. He’s here to tell you that the robots are not coming for your job — that the future will likely be more boring than Blade Runner and Terminator would have you think. And that AI exists to improve our lives… not end them.“
Today, you can take your smartphone out, take a picture of somebody, or take a picture of a group of people and the smartphone will automatically put in the names of the people in the photo,” Steven told me. “That’s very impressive technology, and it’s done using artificial intelligence that is trained to do one very narrow task — name the people in the photos. But you can't get that same AI program to distinguish between a dog and a cat. It can't understand language. It can't really do anything else that people do. We’re dealing with very narrow AI and that's the predominant form of AI today.”
What that means, he says, is that our jobs are safe. There won’t be robots leading law enforcement units. And, for the foreseeable future—certainly, not in our lifetimes—we can forget about going to war against an army of android soldiers. And Steve says there’s a logical reason for that—humans cannot program computers to use common sense.“For a vehicle, a robot, a computer to be able to reason, it has to be able to use everyday common sense,” he said. “And there’s not a single AI researcher in the world who knows how to build everyday common sense into a computer.”Common sense, Steven added, is the secret sauce whose recipe no one in AI research has figured out. And they likely never will—even as experts continue to warn us that artificial intelligence will someday be our downfall.
Indeed, Tesla and SpaceX CEO Elon Musk has said AI poses a “fundamental risk to the existence of civilization.” Similarly, the late Stephen Hawking predicted AI could “spell the end of the human race.” Others have compared AI’s existential threat to that of nuclear weapons.Steven is well aware of those concerns and he’s even considered them. But he’s just never found any there there.“The hype that exists around AI profoundly overstates its capabilities,” he said. “But that doesn’t mean there isn’t work to do. And unless we get rid of the fear surrounding AI, we won’t be able to solve the critical issues that really are out there.”
To listen and subscribe to the Before IT Happened podcast, visit www.beforeithappened.com.