Core Belief 3 of 6
The Sheepy Way
AI fluency is genuinely valuable. But fluency in the tool is not a substitute for the thinking the tool is supposed to augment. Knowing how to ask the right questions comes from years of asking wrong ones and watching what happens. That's not something you can farm out. It accumulates. The struggle wasn't a detour. It was the road. And we are paving over it.
The Fluency Trap
AI fluency is genuinely valuable. The ability to know which tool to use, how to prompt well, how to evaluate and refine what comes back - that's a real skill, and it will matter in every role, every industry, for the foreseeable future. Companies should be developing it. Schools should be teaching it.
But fluency in the tool is not a substitute for the thinking the tool is supposed to augment.
Knowing how to prompt isn't the same as knowing what's worth saying. Knowing how to evaluate a first draft isn't the same as being able to produce one. Knowing how to ask the right questions - of a model, of a brief, of a brand - comes from years of asking wrong ones and watching what happens. That's not something you can farm out. It accumulates.
The Sheepy Way
Growing up, my family was stationed in Germany - my father was in the army, and I spent my early years there, kindergarten and all. My mom and dad had an old Opel, and whenever we were heading home, there were always two routes. The direct way. And what we called the sheepy way - a longer road that wound past fields of sheep. It wasn't faster. It wasn't a shortcut. But ask any kid in that car and the answer was always the same: the sheepy way.
Not because it was faster. Because there was something worth seeing on it.
My last piece made the case for the Brand Steward - the human in the room whose mandate is to hold Brand meaning at the level machines can't achieve. The one with the authority to say "yes, but no" and be heard.
What I didn't call out in that piece is whether that person is on the right path.
The Brand Steward of 2040 is in college right now. Maybe high school. And they've almost certainly used AI to augment their schoolwork since freshman year - using models to write papers, structure arguments, synthesize positions. They may arrive in the workforce ahead of the AI curve, but profoundly underprepared for the human parts of the job. The parts no AI model can do.
A few weeks ago I had this conversation with my own kids - both heading into careers where AI will be central from day one. Like everyone their age, they've fully embraced AI. What I told them: that's fine. AI fluency is common currency now. The skeleton key is knowing where any piece of content falls on the Authenticity Spectrum - authentic, authentic adjacent, non-authentic derivative - and feeling the difference before anyone tells you. That skill doesn't come from prompting. It comes from taking the sheepy way.
We're raising a generation that's fluent in AI. Are we also raising one that knows what to do when the AI is wrong?
From AI Fluency Is Not the Same as Judgment (opens in a new tab), July 8, 2026.
If this is the type of content you enjoy, I write more of it in The Authenticity Spectrum.
Subscribe on LinkedIn (opens in a new tab)