Skipping the Reps

If AI does the junior work, who learns the senior judgment?

Following up on my previous post about how the real skill in a gen-AI world is recognizing when something’s not quite right, even if you can’t fix it yourself:

That skill doesn’t just appear. It develops through repetition and exposure, i.e., through doing the work. I learned it by editing hundreds of articles, working with dozens of designers across countless iterations, and shipping features that flopped. You need the reps.

Which is why I’m concerned about companies using AI as justification to eliminate junior roles. AI may amplify productivity, but it still needs human judgment — and will for the foreseeable future. If AI handles the “junior work” of execution, where does that discernment get built?

We learn what works by fixing what doesn’t. We can’t skip that part, yet many companies are adopting AI as if we can. Our efficiency gains now are coming at the expense of judgment in the future. The challenge isn’t AI doing more; it’s making sure people still learn how to guide it.

Originally published on LinkedIn