AI Is Deskilling Developers, and People Are Worried
A thread titled 'Is AI Causing a Repeat of Frontend's Lost Decade?' is getting serious traction on HN. The argument is that AI is doing to general development what JavaScript frameworks did to frontend: abstracting away the hard knowledge until a whole generation of developers doesn't know why things work, only that they do. The comparison is pointed because the frontend deskilling already happened and everyone now agrees it was bad.
The pattern here connects directly to the 'We Should Be More Tired Than the Model' thread, where developers talk about the friction they're deliberately adding back into their workflow to stay sharp. Both threads reflect the same fear: that the speed gains from AI-assisted development are real but come with a hidden cost that compounds over time, specifically a workforce that can't debug or reason about the systems they built.
The most pointed comment in the lost-decade thread goes further, arguing the real risk isn't deskilling but a lost generation of entry-level engineers who never get the jobs that used to be the on-ramp into the industry. That's a structural problem, not just a personal discipline problem.
So what?
If you're hiring engineers who learned to code primarily with AI tools, you need a way to assess whether they actually understand what they're building. Technical interviews that focus on debugging and reasoning about systems, not just output, are going to matter more. If you're building a developer tool, the products that win will be the ones that make engineers faster without making them dumber.