AI is encouraged at work but not enforced.

Managers don't talk about forcing its use at all here.

I think in two months the one time it was mentioned in a meeting was someone putting company info into a public AI instead of the secure self hosted one. That was idiotic of them especially considering our proxy is setup to display a splash page warning with company policy before continuing to public LMM services.

It's a perfect stance for a company. There's no token tracking, incentive or demerit for the use of AI. Just announcements on proper use and providing secure resources. They want to protect IP (Intellectual Property).

That said I tried using it to code as a accelerator. Claude did the job but it's not consistent. I had to go back and debug some code and there's no consistency in it's coding sections. It's hard to read through and therefore hard to maintain.

I resorted to providing example code for it to follow and it still adds extra bullshit.

Maybe I need to get better at prompting but it's really just faster to code myself.