What Happens to an Expert When AI Knows Everything They Know

I've watched a consultant on our team go from making the call to approving the output. It happened over about six months. The tool was faster, the clients were satisfied, and the judgment that made him good at the job was quietly going unused.

There's a name for this now. Researchers at CHI 2026 called it "intuition rust," the gradual dulling of expert judgment when AI handles the analysis layer so consistently that the human stops doing the cognitive work that built the judgment in the first place. I wrote The Amplifier Paradox a few weeks ago around these ideas ("cognitive surrender" being a more egregious thing we're seeing), looking at how AI-assisted output can quietly hollow out the capabilities underneath. Intuition rust is the slower, more personal version of the same problem. It doesn't feel like surrender. It feels like doing your job well, just not very often. And with limited friction.

The thing most consultants and advisors are actually selling isn't information; it's judgment. The ability to walk into ambiguity, synthesize what you're seeing, make a call under pressure, and be accountable for it. That capability doesn't maintain itself. It takes reps. It comes from experience, most often, and it requires experience to maintain. It takes making the hard inference and being wrong sometimes and recalibrating. When you hand that work to the tool and ratify the output, you skip the rep. There's no growth without feedback, resistance, effort; the gym teaches us this!

The information-access premium is already being compressed. AI closes knowledge gaps faster and cheaper than any engagement model can. Clients who used to pay for your knowledge now have access to something that knows what you know. What they don't have, and what still earns the billing rate, is your judgment. The call under uncertainty. The pattern recognition built over fifteen years in the room, but it's more than that. Intuition requires biology, and it's the glue that holds those patterns together in a way that makes insight possible.

The irony is that AI can actually sharpen judgment if you treat it as a sparring partner rather than an answer machine. Use it to stress-test your thinking before the client call, not to replace the thinking after. The consultants on our teams who do that are getting sharper. The ones using it to go faster without going deeper are developing blind spots they haven't named yet.

Are you building judgment or outsourcing it? Those aren't the same question, and only one of them compounds.