AI has been moving so fast that no one can predict where it goes. Anyone who confidently says otherwise is kidding themselves.

I’ve been working with AI in one form or another since doing speech recognition research at MIT, and I’ve tried to apply it to pretty much every problem I’ve worked on since. What’s different about the last few years is the pace of compounding. Each new capability seems to unlock the next, and the time between what feels impossible and what becomes routine keeps shrinking.

So how do you build a company in a world you can’t predict?

For us, a big part of the answer has been to compound every single day. Get as close to the frontier as you can, experiment constantly, learn fast, and take everything you learn and make it the starting point for what you do next.

The interesting thing is that while you can’t predict where all of this ultimately goes, being close to it does help you see around the next corner. From the earliest days of ChatGPT, we were talking as a team about where reasoning models could go. Even before coding agents were ready, we began preparing for a very different way of building software. The point isn’t that we predicted the future. It’s that staying close to the technology, learning and experimenting every day, helped us see what might be coming next.

And that’s where the compounding gets really powerful. Every new capability, experiment and lesson becomes the starting point for the next. You don’t just get better. You get better at getting better. When the underlying technology is itself compounding, you can move incredibly fast.

We’re seeing the results at ianacare. We’re supporting more patients and families than ever. We’ve kept a particular focus on our toughest cases, and both our NPS and the efficiency of our service have improved dramatically. Intelligence that once lived with a handful of our best people is increasingly available to every navigator, and directly to every family we serve.

I have no idea what AI will look like six months from now. Nobody does. But I think that’s exactly why the most important thing we can do is keep moving, keep learning, and compound every single day.

We’ll share more about what we’ve learned, and I’d love to hear what others are learning too.