Matt Lanza

Here's what's happening right now.

Based on that history, tell me what's going to happen over the next 10, 15 days, whatever.

And these AI models will be like, okay, sure, here you go.

Press a button, go make yourself a sandwich, come back, model's done.

And you can get an answer.

But I mean, I'll be honest, like last hurricane season, the AI models did a pretty good job on forecast tracks with some storms.

I was floored.

And it was kind of a wake-up call.

We're still working with the physics-based models.

We're still working to improve them because we're not going to get rid of them.

We're working into a world now where we're going to end up doing like a hybrid approach where you're trying to marry the two outputs and come up with an answer that's better than anything we've had before.

And I'm super excited.

I really think that in the next five to 10 years, we're going to have so much more interesting and useful and valuable data.

And the caveat, of course, is as long as we continue to fund it,

We will get better at this.

And we'll never be perfect, but we'll get better, which is great.

That's what we need.