No, and I agree with the conservative sentiments here. However, putting together a SaaS alternative that frees up money during a crunch, and now with the pet features your boss has always wanted, is potent indeed.

You've hit the nail on the head. Immediate gratification.

AI is like sugar. It tastes nice, gives you energy quickly - what's not to like? The gratification is immediate, and if "today is all that matters" it's brilliant.

The problem with sugar (and AI) is medium term. So sure, that junior dev whipped up the whole framework in ClaudeCode, and it's humming nicely. Junior dev gets credit, and after a couple years moves on somewhere else.

Then something changes. Windows. TLS. Code Signing, whatever. We need to update the program to the change. Just a small tweak. Junior Dev has gone (or is otherwise occupied) so we'll get new-Junior-dev to do it. Is he expected to do the change at the code level? Or at the prompt level? Will ClaudeCode in 2029 be able to maintain ClaudeCode Code from 2026? Or will it want to rewrite everything? Will new-junior-dev have the skillset to prompt as well as first-junior-dev? Was the code good enough that a dev could just "take it over"? Or was it "it works, let's use it" standard?

AI makes everyone look good in the short term. But it worries me for what happens in 5 years, 10 years, and so on. Sugar is great, but you can't live (long term) on sugar. Sometimes you need a proper meal plan.