Everyone Wants AI. Nobody Can Tell Me What Problem It Solves.

Client: “We need AI!” Me: “For what?” Client: “Everything!” Me: “But what specific prob—” Client: “RIGHT NOW!”

That’s not a bit I made up for this post. That’s a real conversation I have on discovery calls constantly. Swap out the client name and it’s basically a rerun every single time.

Somewhere in the last two years, “AI strategy” turned into a thing every CEO feels like they’re supposed to have, the same way everyone suddenly needed a mobile app in 2013. Nobody wants to be the last person at the table who hasn’t “done AI” yet. So they call us wanting AI installed like it’s a piece of hardware. Like I’m going to show up with a box marked ARTIFICIAL INTELLIGENCE, plug it into the server closet, and their invoicing problems disappear.

Here’s the plot twist

The companies that actually get something out of AI are almost never the ones sprinting toward it first. They’re the ones who slow down, get annoyed with me for asking “but what problem are we actually solving,” and eventually give me a real answer. Something like “our team spends six hours a week manually cross-referencing invoices” instead of “we need to be more innovative.”

One of those is a project. The other is a vibe. I can’t build IT infrastructure around a vibe, and honestly, neither can anyone else who’s being straight with you.

The rush-to-AI crowd usually ends up in one of three places: an expensive pilot that quietly dies in six months, a tool nobody on the team actually uses because it doesn’t fit how they already work, or a chatbot bolted onto their website that makes customers angrier than a hold-music loop ever did. None of that is “innovation.” It’s an expensive placeholder for a decision nobody wants to make.

So what do I actually tell people

Start smaller and more boring than you want to. What repetitive task is genuinely draining your team right now? Where are you losing customers because a response takes too long? What data do you already have sitting around that could actually predict something useful? Answer any one of those honestly and you’ve got a real starting point. Answer “everything, right now” and you’ve got a budget line item that’s going to embarrass someone in Q3.

I’m not anti-AI. I’m anti-nonsense, and right now there’s a lot of nonsense wearing an AI costume. If you want to talk through what actually makes sense for your business instead of what a keynote speaker told you to be scared of, that’s a conversation I’m always happy to have — and it’s usually part of a bigger IT strategy conversation anyway.