So, we're talking about AI now, but not the kind that just sits there waiting for instructions. This new breed, "agentic AI," it decides what to do, figures out how to do it, and then, well, does it. Think of it like this: instead of just answering a question, it figures out what questions should be asked, and then finds the answers. This idea, "chaining", means these AI things break big problems into smaller, more manageable pieces, solving them step by step.

We see examples in insurance: claims, risk, fraud, customer service, all touched by this new capability. But it doesn't stop there. Healthcare, finance, manufacturing, all could see major changes.

Now, this stuff is powerful, but it brings up some thorny questions. We gotta be careful. We need to build this kind of AI responsibly.

This episode is presented by platformeconomies.com - and if you are interested in matters of architecture, make sure to get my book on IT Architecture today: ⁠https://a.co/d/gK7YzOL⁠ (US) / ⁠https://amzn.to/3OkwqE9⁠ (DE)