This is possibly unrelated, I'm not sure, but a lot of the AI implementations I've been seeing are actually genuinely bad. Not bad in the sense that "the AI gets it wrong" or breaks something, it's more that it starts to inject a ton of ambiguity into decision-making for the business. Before you had people with deep domain knowledge, who had to synthesize that knowledge into an actual strategy. Now you have AIs summarizing documents, building charts and "KPIs" ... suddenly people actually understand things a lot less and their understanding of what makes the business run is obscured. It's not a dependency scenario, not yet anyway, where people can't make decisions without first consulting the AI or something.

Is this what AI is supposed to be? If it is, I don't want it.