Somebody on your team is building things with AI right now. You probably don'...
Somebody on your team is building things with AI right now. You probably don't know about it.
A developer recently listed 16 projects he accidentally built with AI agents over a few months. His conclusion? Cancel the subscription. His quote stuck with me: "a tool producing a cheap reward with minimal input and no friction can only be a liability."
That's not a developer problem. That's every department in your company right now.
Your marketing lead spins up an AI workflow on Tuesday. Your ops manager builds a reporting dashboard on Thursday. Sales automates follow-ups over the weekend. None of them coordinated. None of them documented what they built. And nobody's going to maintain any of it when the person who built it moves on (or just forgets it exists).
A Google engineer put a name on this last week: orchestration tax. More AI agents doesn't multiply your team's capacity... it multiplies the overhead of keeping track of what's running, what it's doing, and who's responsible when it breaks.
The AI risk most businesses aren't thinking about isn't that the AI will fail. It's that it'll succeed at creating things nobody asked for and nobody will maintain.
Before your team spins up the next AI-powered anything, ask two questions:
- Who owns this in six months?
- What happens when nobody does?
If you can't answer both, you don't have a tool. You have a liability. (And if you're not sure whether your team is already building without guardrails — they are.)