Will Marshall

Just think, the large language models are

that you use when you're chatting are mainly trained on the text of the Internet.

And so they're somewhat limited to knowing about the stuff on the Internet.

They don't really know about what's going on in the physical world.

They're sort of blind to the physical world.

When we add Earth imaging data, we've got 3,000 images for every point on the Earth's landmass on average to train these AI so that they know about the physical world.

That unlocks

huge applications.

It's both useful for AI, because for AI to get value, to provide value in the real world, like insurance or farming, like we were saying, they need to know about the farmer's field, so they need to know real data.

We can help with that, right?

We can help AI to unlock markets,

And AI can help us unlock markets because our data is really valuable in principle, but you need the extraction of specific answers bespoke to the customer.

And that solution piece is what has been a bit of a gap, and the AI is now accelerating the crossing of that gap.

And so, yeah, we're seeing tremendous acceleration in our business, both for security reasons and because of the AI acceleration.

That's right.

So what we are doing is we're providing analytical tools on top of our platform.

We also have partners that build analytical tools as well.

But we have the advantage, yes, of the bulk of our data so we can fine-tune those models to be better and better with satellite imagery.

And doing that at scale obviously has its advantages for making them more accurate and so on.

So we provide analytical services like our maritime domain awareness solution isn't just a picture of the ocean.