The biggest problem I have with ai is the inaccuracies. When you're making a meme, no big deal. When you're asking a question and you get back garbage, it's a reminder that the machine knows nothing. It's just less effort to do a proper web search and find an article or forum thread with the validated information than to waste time with the ai companion. The problem is ai perfectly nails the Dunning Kruger curve - people far to the left don't know how wrong a lot of it is, so they report major confidence gains in the few valid results they get, discarding all the trash on a gamblers high. But the moment you are actually capable in the field, you'll start to see the flaws in aigen, and confidence (and usage) drops off. Regular people are also more likely to be pulled in by all the grifter hype, lacking the understanding to resist the shitty tool.