The Panda Problem

Google built a car that drives itself and it looks like something you’d want to stay far away from. All rounded edges and bug eyes, genuinely unimpressed-looking. The kind of thing that would be cute if it weren’t about to make a decision that could kill you. That’s the part people get stuck on, I think—we’ve spent a century treating the car like an extension of ourselves, something you grip and steer and feel beneath you. And now we’re supposed to just sit there while it happens.

There’s always skepticism when a new thing arrives. The trains would destroy society. The automobiles would destroy the trains. The internet would destroy everything. And maybe it would have, if we hadn’t all agreed to let it. This is the same negotiation with a different object. We get to decide if we’re comfortable with it, or at least we get to complain loudly until someone decides for us.

What strikes me isn’t the technology—that part is fine, inevitable even. It’s the design language. That panda face. Someone at Google sat down and thought about what a self-driving car should look like, and they landed on something so deliberately unthreatening it becomes threatening in a different way. Like they’re telling us they know we’re afraid of them, so here, have this cute harmless thing. Trust us.

The dark part—the part that made me laugh despite myself—is the joke about responsibility. If the car hits someone, you’re not driving, so you can’t be blamed. Which is funny and also horrifying, because it’s true. We’ll invent legal frameworks and blame systems and insurance structures, and the car will just keep driving, unimpressed as a panda, and nobody will be the person who did it. That’s either the future or a different kind of problem we haven’t figured out how to talk about yet.