Nvidia CEO Jensen Huang Says AI Thinks a Ball That Rolled Off the Table Disappeared — But Even 'Your Dog' Knows Better

Roll a ball off the kitchen counter and the family dog will trot around the table to retrieve it. Much of today’s AI, it turns out, would conclude the ball left the universe, a gap Nvidia (NASDAQ:NVDA) CEO Jensen Huang has made the centerpiece of his pitch for what comes next.

When the Ball Leaves the Counter

"The next wave requires us to understand things like the laws of physics, friction, inertia, cause and effect," Huang said at the Hill & Valley Forum in Washington in 2025. "The fact that I tip that thing over, it’s going to fall. You know, when I set the bottle down, it’s not going to go through the table."

Those are basics children and pets already have, he said. "And so all of these common sense physical reasoning abilities that children have, that our pets have, most AIs don’t have."

Then came the ball. "You know, I roll a ball over the kitchen counter and it goes off over the top and it disappears. Well, the AI thinks it’s disappeared, but you know, your dog knows it. It’s on the other side."

The dog, Huang said, grasps something machines do not. "It understands this concept called object permanence and it didn’t go into another metaverse. And so it runs around the table and gets it, you know?"

"And a robot needs to learn that if you want to go from this side of the table to that side of the table, you can’t go through the table," he continued. "You’re going to have to reason about how to go around the table."

Put that thinking inside a machine, and it has a name. 

"And so all of these types of physical reasoning things are what’s called physical AI. And when you take that physical AI and then you put it into a physical object called a robot, you get robotics."

That is the pitch for physical AI. Today’s systems excel at perception, generation and reasoning inside digital environments. Physical AI aims at machines that grasp friction, inertia and cause and effect, the know-how robots need to work in factories, warehouses and other real-world settings.

Robots, Factories and a Labor Crunch

"Hopefully, in the next 10 years, this new generation of factories will be highly robotic," he said. "They’ll help us deal with the severe labor shortage around the world."

Nvidia has released foundation models such as GR00T for humanoid robots and has pitched machines as a fix for worker shortages, with Huang using the phrase "AI immigrants" for robots that take on work people are less willing to do.

The gap he is pointing to is simple. Large models trained mainly on text and images are fluent, but they lack a grounded sense of how objects behave. Closing it, Huang argues, is what will move AI off the screen and into the world’s largest industries.

The Dog Still Has the Edge

For now, the scorecard is humbling. The family pet understands that a ball does not vanish when it leaves the counter, and the machines are still catching up. Until robots learn to go around the table instead of through it, the dog keeps the lead.

Photo via Shutterstock