What AI actually takes from young people isn't a job
The real worry isn't AI taking jobs a decade from now. It's that right now, failure doesn't happen long enough to teach anything.
My nephew opens his laptop to write an essay, pastes the prompt into ChatGPT, skims the output, tweaks a few words, submits. Four minutes, start to finish. I'm not writing this about academic dishonesty — plenty of people already worry about that. I noticed something else.
What worries me isn't AI taking this generation's jobs a decade out. It's that right now, AI is quietly taking away the mental muscle that used to get built by struggling with something hard while young.
Mental muscle needs reps too, or it atrophies
The previous generation got stuck on a math problem, wrote broken code, spent a whole evening figuring out why a paragraph didn't work. That process — stuck, wrong, self-corrected — is what built intuition, not the final answer itself. Now the answer arrives in three seconds. Being wrong doesn't last long enough to become a lesson.
I don't want this to read as pure pessimism, because it isn't. Plenty of young people are using AI as genuine leverage — learning to code far faster than five years ago, shipping a real product at nineteen, picking up a new skill in weeks instead of years. The split isn't about using AI or not. It's about who deliberately turns it off to struggle on their own, and who lets it do the struggling for them too.
AI answers faster than your brain does. The problem is your brain needs to go slow to actually grow.
I think this generation ends up splitting into two groups more sharply than any before it — and not by IQ or upbringing. By something much simpler: who's willing to switch AI off long enough to get stuck, get it wrong, and figure it out — and who never does.