This should be obvious.

Using AI to detect cancer early? Using AI to help find new beneficial medications? Using AI to advance materials science? Great! Things like these are absolutely what we need more of.

Using AI to steal massive amounts of content made by non-consenting parties and regurgitate for AI corp profits? Using corporate algorithms to manipulate output to generate results favoring corporate billionaire owners economically and politically? Using it to make sexual photos of non-consenting women and underage girls? Using it to find ways to deny coverage, strip more money from people, wage war, attack intelligence and discourse, and remove accountability for corporate owners? Fuck that.

Guess which ones the oligarchs prefer.

AI != LLM != LLM chatbots.

But in this case the use of "AI" means "LLM chatbots", nobody is really making an argument against other kinds of AI, like what is used to detect cancer (there is an argument to be made about having a black box in a medical context, but this is not the place for it).

Absolutely. I've softened my stance a lot recently. I think local llm are fine. The data has already been crunched the books destroyed and the water used so I may as well play with one locally. If you set realistic uses as a tool its fine. Need a quick script great at that. Summarize some notes, or search through them also great. Not really into image gen or anything else but anything helping me get dumb stuff out of the way for just myself its a W. I'm not down with using it for public facing stuff.

Honestly I'd say it's a pretty apt comparison. Pledging to defend your rights while avoiding using anything labeled "AI" because fascists use it is about as smart as pledging to defend your rights without using guns because fascists use them.