One year with my Tesla Model Y.

I deliberately waited to live with it before sharing thoughts. It’s like living in the future every single day.

A year later I’m running about 97% Full Self-Driving. Mostly daily driver around town, with a few road trips mixed in. FSD is a total game changer. The fact that Tesla built a custom AI purpose-built for this specific problem—not some bolted-on general model—is a huge part of why it works so well. They didn’t take the old car + software approach. They rethought the entire process of how an electric vehicle and self-driving system should work from first principles. That’s why it succeeds.

I’ve given test drives to a number of people. Several of them went out and got their own Model Ys. One that still sticks with me is the mother of a friend. She thanked me at a charity dinner, saying she never would have made it without Full Self-Driving. It gave her a new lease on life—she can get around on her own again, especially in the evenings, after she’d started getting afraid to drive.

That real-world impact is what antifragility looks like in practice. The more edge cases the system sees, the better it gets. It doesn’t just hold up under stress; it gains from the disorder of real roads, real weather, real traffic.

That’s the same principle David Blumenthal and I explore in our upcoming book Designed for Volatility: if you want to succeed with disruptive forces like AI, you can’t just graft the new capability onto the old way of doing things. You have to redesign the system so it actually gets stronger under volatility. Tesla did exactly that.