TL;DR: AI isn't just LLMs. Hell, LLMs aren't even bad. Big AI (aka capitalism) is where the problems lie. If you're gonna be mad, be mad at the right thing - capitalism, enshittification, copyright abuse, surveillance, hype, influencers.

It's still the same bad thing as it always was, just wearing a shinier hat now.

Agree. And let me screed about it even more -

By common discourse, I think you mean "...by chronically online experts". As backed by...what? Vibes in echo chambers?

A surprising number of people who speak confidently about AI appear not to understand the thing they're criticising.

We've seen this shit before and we never learn.

30 years ago, everyone knew that MSG was bad and video games caused increased violence. Both turned out to be manufactured consensus - speculative letters and mouse-injection studies boosted by "experts" into public panic.

I'd wager a dollar that most rabidly anti-AI people don't grok that the 2024 Nobel Prize in Chemistry (yes, that recent) was awarded for machine learning in protein folding - cracking a 50-year-old problem with real applications to drug discovery for Alzheimer's and Parkinson's.

I use AI (image analysis, folding, STT, clinical audits...yes, even LLMs) every day. They're useful. But the online discourse around AI has become cartoonishly one-dimensional.

AI = GPTs = slop = bad is a category error, perpetuated either in partisan bad faith or simple lack of examination.

And just to tank the down votes even more -

https://blog.andymasley.com/p/a-cheat-sheet-for-conversations-about