It's all relative. When I was 20, I'd consider myself old at 36. Now when I am 36, my definition of old has shifted and it's now somewhere past 50. I guess "old" for a person is just that person 10+ years older :)
Hire excellent programmers who spent their lifetime learning how to write good code and build systems, and then the code is easy. Hire some bad ones and watch them struggle and fail and learn why maybe coding is not that easy. AI doesn't change this, it just shifts failure modes a bit since the volume of code produced is now larger, and AI-enabled bad programmers are still bad.
I remember a tennis survey where something like 70% of respondents believed they could win a game against a Tennis pro player (I can't find the source anymore, but it was discussed on Andy Roddick's podcast). If you watch Roger Federer effortlessly and elegantly make beautiful shots, it's very easy to fall into this trap and think -- it must be so easy, right?? Are people falling into the same trap watching Claude Code regurgitate CRUD apps for them?
The scope and complexity of the change is crucial here. There is a sweet spot where spec-driven agent offers great value: the change is small and simple enough that you can describe it reliably with a spec (i.e. you have complete understanding how to do it), while implementing it yourself is a lot more work than writing the said spec plus reviewing and correcting the agent.
For me currently this sweet spot is TINY. It's so small that my usage of Claude Code has dropped to almost none. It's simply more practical to let myself have the agency and drive the development, while letting AI jump in and assist when needed.
Comes to germany, does not like it, makes a picture of a middle finger to Bundestag. This smells like ragebait low effort content I come to HN in order to avoid!