Illustration by TIME; reference image courtesy of Marc Raibert

Marc Raibert does not have a whole lot of respect for robots—which is funny coming from a man who has chosen to spend much of his professional life with them. “Robots are as dumb as toasters, they’re as dumb as doorknobs,” he says. “They do what you tell them to do, but you usually need a pretty well-defined environment in which they can do it.”

Raibert means to change all that. He’s the founder and chairman of Boston Dynamics—best known for its doglike robots that can move about on four legs and perform work like inspecting factories for safety concerns, using their camera-equipped eyes and noise-detection ears to act as sentries at military bases, or investigating suspicious packages. Last year Raibert expanded his portfolio, establishing the Boston Dynamics AI Institute, aiming to give robots not just mobility and function, but also nimbleness and smarts, two things they deeply lack.

Part of making robots better is developing what Raibert calls athletic AI. “Think of your garage,” he says. “Imagine it’s stacked up with stuff and you need to wriggle your way to the back of it to get something. People can do that; animals can do that.” But robots? Not so much. What they need is not just greater mobility and greater dexterity, which is mostly a matter of hardware, but greater real-time perception, which is all about AI software—and that isn’t easy.

Raibert recalls being at a beach overlook recently, gazing down at 300 people, and spotting his family immediately. “Humans’ visual capability is just incredible,” he says, “and that’s part of what I think of when I say athletic AI.”

Reasoning is something else. Performing a fixed task is something most robots can do, even if that task involves multiple options. The doglike factory-inspecting robot might know that it should do one thing if there is an object on a stairway and do something else if the stairway is clear. But reasoning through a problem—even a relatively simple one—is beyond robots.

“Suppose you’re going to the airport,” Raibert says. “You look at your watch and say, ‘OK, what time do I need to be there?’ Then you back up and make a plan that takes into account a big diversity of information, and you end up getting to your flight on time without wasting too much time at the airport.” Easy for us; well-nigh impossible for robots, but what Raibert calls “cognitive AI” could fix that. “We would like to make a set of robots that are more like people in that way.”

The ethics of AI are an area of concern—one Raibert is assembling a team at his company to address. There is the problem of robots taking jobs that people are currently doing; Raibert acknowledges this but is not deterred by it. There’s also the concern about AI unleashing a power humans ultimately can’t control—but Raibert doesn’t buy it.

AI, in his eyes, is a technology in its toddlerhood, and he believes it should be treated like any toddler—which means we should encourage its growth. “If you had a 1-year-old child who was barely able to do anything,” he says, “would you think of not teaching them anything because of all of the horrible things they might one day do? No.”

The original version of this story misstated Raibert’s role at Boston Dynamics. He is the founder and chairman, not CEO.

More Must-Reads from TIME

Write to Jeffrey Kluger at

Sougwen Chung
Yann LeCun
Abeba Birhane
Audrey Tang