Kansas City AI Expo
Questions from the expo
These are topics that came up in my conversations at the Kansas City AI Expo, and how I think about them.
An AI model is probabilistic, which means the same question can produce different answers. That is true, and it is a fair reason to be careful.
The model is not the system, though. A business cares whether the system produces a reliable, acceptable result, and that depends on the rules around the model: the data it can see, the actions it is allowed to take, the checks that run before anything happens, and the record kept afterward.
Software has always worked this way. No two developers write code exactly alike, and one developer's solutions change from week to week. Work still gets done, because specifications, reviews, and approvals keep the outcome reliable even when the process varies.
AI can work the same way. The generation step can vary, while the boundaries of its authority and the checks on its output do not. There are things I would never give AI. There is a lot I would, when it is governed well and has access to the right information.
Want to talk it through?
I am happy to talk about any of these in the context of your business. I want owners using this and telling me what goes wrong.
Schedule a CallYou can find this page again at eikon.digital/kcai.
