Post

this is the really ineffective head-in-the-sand reaction a lot of folks on Bluesky tend to have if you talk about any harm reduction-based approach to the reality that millions of people are using AI tools today. The current approach to critiquing AI is obviously not working, but they don't care.

this feels akin to bowing out of a fight worth having. you do not have to pledge fealty to the LLMs and AI apps simple bc they are popular?

Is there anything in between "pledge fealty to LLMs" and "decry anyone who uses ChatGPT as an environment destroying fascist collaborator and block them", or is that where we're at these days?

course there is. but i'd also like to note i didnt say the 2nd thing you put quotes around lol

You did not! But the "AI Apologists" moderation list Anil is referring to is indeed the kind of thing often used for that, and that is the AI discourse we seem to have on Bluesky: entirely lacking in nuance.

that is fair and i don't think you're a shill! i think my response was meant to be in a tangential thread w emily bender- my bad. i do not believe that it's a worthwhile effort to train people to use LLMs more effectively, that's all.

I think that's a valid argument, it's just that almost everyone I've seen make that argument, when asked what we should do instead, generally responds with a shrug. And I've admittedly lost my patience about that. Because it's resulted in the current situation, which is the fascists winning.

What we should do instead of training people to use LLMs? 1. Provide education on what they actually are 2. Provide education on their environmental & labor impacts 3. Work towards regulation that holds LLM providers accountable for their negative impacts

while I'm totally sympathetic to the idea that we shouldn't be instructing people to use LLMs more (I disagree with it, but don't think it's Absurdly Wrong or horrible in some way), there are areas where people do *need* to learn how they work: basically any supervisory role (teacher, manager, etc).