the part i hate the most about the generative AI discourse is that both sides (yes, both sides) of it are really out of touch with reality
the boosters can't admit that right now, a lot of people, even those who use LLMs personally, still associate its output largely with all the slop that fraudsters have been pumping out and that promises of "but it will eventually deliver a utopian future of super-high efficiency" or threats of "it will turn into skynet" now ring hollow
and the skeptics are treating it as if it were nuclear waste, to the point that multiple different people on here are now dissing linux as irreversably tainted and talking about how BSDs are the future, even though under different circumstances they'd be instead talking about how BSDs are lacking in features and their licensing just surrenders your code to the corporatists and GPL is way better
i hope this whole bubble / hype cycle dies down as soon as possible and we're able to stop fighting and actually figure out what generative AI can actually do on its own, and at what cost, and whether it's worth it
@rnd I recall having followed a lot of early LLMs on tumblr. People programing them for fun on self hosted servers and early chatbots before things went mainstream. I recall when a notable one shut down citing "this industry is about to get really bad really quick and I want no part of it" and I didn't understand them at the time.
Now I feel like so many people are very reactionary to the technology itself to the point that they would tear apart an art programs fill-in tool or a word processors grammar checker if it was implemented today, ignoring that there has been ethical machine learning for decades and there's still potential for ethical options now and in the future.
A large problem is many people are calling the technology evil without fully understanding or inspecting what exactly is evil or why. I don't blame people who are ignorant and erring on the side of caution (and admittedly I often fall into this boat with a lot of AI related stuff). I'm probably never going to be a supporter of completely AI generated artwork for a plethora of specific reasons I could actually articulate, both political and personal, but I can't say I have the same definitive gripes about my partner creating banks for vocal synths trained off of her own voice and running off of her own computer and data which as far as I can tell there is nothing inherently wrong with, other than a lot of people would lump it in with the nebulous and all encompassing definition of "AI."