"nobody talks about it" "the paradox nobody warned us about"

Dude, everybody has been talking about it and warning you about it. 🤪

How can people write unironically about how AI is seemly so great yet, wow, gosh, shucks, maybe there are some problems we need to deal with?

The only thing I can think of is MegaCorp social media like X, where all viewpoints other than techofascist maximalism are suppressed and buried.

It's so weird to read that. Every professional developer knows that debugging is hard and debugging someone else's code is harder. The crap that falls out of LLMs can be even worse, and yet you allow yourself to be reduced to a human full-time debugger/QA person.

This isn't just fatigue, this is taking a wrong approach to software development from the ground up.

Using it is like elevating yourself to a management position over a sea of stochastic code monkeys. It constantly churns out rubbish of tenuous usefulness toward solving the problem you actually have and then codes up a test suite to encase that ill-advised screed in a thick layer of cement.

Yes, you're "shipping" more code. But code -- like Cory Doctorow is fond of saying -- is a liability. "AI" can keep churning shit out and if your standards are low enough you can ship the slop. But the hardest problem in software is finding the right two line fix to place to fix the actual problem, and it's something made infinitely worse by having a tool that churns out things that look like they might be correct, turning you into the trash sorter.

Just close the laptop little bro.

This isn't a paradox. It's the ordinary and expected outcome of you have a junior whose work you can never trust.

Regardless of what your profession is, if you have a source of "work input" that requires specific instruction for near every task and whose output must be carefully examined, then the part of your job which is reviewing drafted work would necessarily increase.

This is especially true in engineering fields, where the things that can be abstracted into repeatable tasks usually are. Computers saved structural engineers from having to do all their math separately and higher-abstraction languages saved programmers from having to futz around in assembly, but neither of those had to be manually checked.

I don't even use AI, just looking at other peoples slop is exhausting.