Why Highly Skilled People Are the Most Skeptical About AI

I work with a lot of really smart people. Engineers, designers, people who are genuinely great at what they do. And something I've noticed is that the people who are most skilled in technical, hard-skill positions tend to be the most skeptical about AI. The better someone is at their craft, the less impressed they seem to be.

I think that makes sense, and I also think they might be looking at it wrong.

Here's the pattern I keep seeing. When you're really good at something, AI doesn't feel like a gift. It feels like friction. The code it writes works, technically, but it's not how you would have written it. The structure is off. The naming is wrong. It makes choices you'd never make, and suddenly you're spending time reviewing and correcting instead of just building. It doesn't feel like a productivity boost.

A lot of what makes someone great at their craft comes down to taste. The instinct for what's right versus what merely works. That's exactly the part AI can't replicate. An experienced engineer doesn't just know how to make something function. They know how to make it clean, readable, something another human can pick up six months later and understand. AI doesn't have that instinct. It has patterns. And patterns are not the same thing.

So when someone who's great at their job tells you they don't see the hype, they're not being stubborn. They're evaluating AI through the lens of what they already do well. And through that lens, it genuinely falls short.

But here's what I think they're missing. AI's real power isn't making you better at what you're already good at. It's unlocking skills you don't have yet.

I've had some coding experience over the years, but not enough to really build the things I wanted to build. I could see a product in my head, spec it out, talk through the architecture with engineers, but I couldn't sit down and make it real on my own. AI changed that for me. Not perfectly, but well enough to test whether an idea works before anyone else needs to get involved. It made me a better PM because it unlocked a skill I didn't have before.

I recently heard someone ask where bugs come from in software. I think about it like a giant game of telephone. A customer describes what they want. That information passes through support, sales, research, and eventually lands on a PM's desk, already a little distorted. The PM then plays another round of telephone with engineers, who translate the original intent into a completely different language (code) and hope that what comes out the other end is what the customer actually needed. Bugs live in every handoff of that chain. Now imagine the person with the original idea can just build the thing themselves. Not necessarily production-ready, but enough to validate the concept. That's what AI does. It doesn't just skip a step. It collapses the whole telephone game.

I've watched PMs build working prototypes in an afternoon, something that would've required an engineer before. Designers shipping real apps, not mockups. People going from concept to something real in days instead of months. None of them are using AI to write better code. They're using it to write any code at all, which they couldn't do before. And the results, while rough around the edges, are good enough to test ideas, get feedback, and iterate. The same thing applies beyond code. People creating art who never learned to paint, making movies with no directing background. AI gave them a starting point they didn't have before.

That's the enthusiasm gap. If you evaluate AI against a skill you already have, it's a sometimes-useful tool with a lot of rough edges. If you point it at a skill you don't have, it's a door that just opened for the first time. Both reactions are completely rational. But they're answers to different questions.

And that's the part I think the skeptics should sit with. The senior engineer who's frustrated that AI writes messy code is right. It does. But that same engineer might have ideas for a side project, or a design concept they've been sketching, or a business they've been thinking about. AI isn't going to make their code better. But it might let them build something in a domain where they're not the expert. The same way it let me build things I couldn't build before.

I think the conversation gets better when we stop asking "is AI good at my job" and start asking "what could AI let me do that I've never been able to do before." The people who are most excited about AI right now aren't the ones using it to do their existing work faster. They're the ones using it to do new work they couldn't do at all. And that opportunity is there for everyone, including the skeptics, if they're willing to point it somewhere new.