I'm hoping that Jevons paradox holds true... and to some extent it will, I just don't know how much.
Again, that is the paradox stating that a rapid decrease in the cost of something (here, software engineering) doesn't always mean that the work disappears, but could actually increase or multiply the work -- because now the marginal cost of doing this is so low that the demand appears everywhere. In terms of AI, it might be something like: now building a custom app with entire UX and backend and everything is so cheap, that even small businesses all over the place will want highly customized software because it's in their reach.... so the demand goes way up for software development (powered by AI, but still hiring humans who knows how to do it end to end and guide it) across the economy. Also, thousands of previously impossible things (which would have taken years of costly dev, now suddenly in reach) now become possible, so work extend immediately into those.
So I think that will be true at least partly. The new world of software may explode what we can do and want to do everywhere. Hopefully that offsets the losses -- but I don't know.