I wasn't talking about your day to day development. I'm talking about you convincing yourself that LLMs have magically solved efficiency.
You talk about AI as if it suddenly made all compute cheap. That's a (subsidized) illusion. There will always be costs. These AI companies are just distorting the value proposition. If you think you'll be able to get Fable class results on affordable hardware, you fell for it.
Thinking you can stop doing the actual brainwork behind building things because Fable happens to be affordable right now is a terrible long-term bet.