@davew in my specific case it boils down to:
* how the training model was made without any regard to licenses and copyright.
* how it will output code that contains full on verbatin proprietary code from repos it shouldn't use.
* how it will plagiarise and mimic artwork by real artists in a way that had a human done it, it would end up in litigation.
* how it is very confident in its halucinations.
* the energy cost of all of it.
I'm mostly complaining about generative AI.