pouët.net

4gentE: it's actually quite respectable that you refuse to submit. Unfortunately, I think the resistance is futile at this point. The direction everything is going and the social dynamics around it is hard to turn around.
Saying that, it's not the end of the world. I actually fully support (contrary to what you may think) AI-free communities. And I do think demoscene should be the one. There is really no reason to use AI when all we care is our human-level skills and fun of creating stuff.

@tomkh:
I understand. Resistance is futile. Of course I don't imagine this onslaught could be stopped. What I'm trying to do is make those inside this small bubble I care about open their eyes and see this for what it is. This is not democratization of art. This is dehumanization of art process.

Whether or not it's ethical, dehumanizing etc...is besides the point for me. It is surely an important issue, but not so much demoscene related.

The LLM AI art machine is perhaps an invention more in line with capitalism than most capitalist inventions. What capital really is is a compression technique. Capital is concentrated dead labour and its purpose is exerting pressure on the living labour. LLM AI art machine is the exact same thing. Compressed and concentrated dead (creative) labour whose purpose is to exert pressure on the living (creative) labour. What makes these machines especially vile and cynical is the fact that the dead labour could have been originally performed by the same artist that now gets squeezed by the machine, his/her dead labour stolen and used against him/her.

That's beside being a confusion machine, a smoke machine.

About the only good (although depressing) thing about these machines is the fact that thanks to them some "artists" revealed themselves for what I guess they always truly were but we couldn't see: petty delinquents and charlatans.

Meanwhile... some people just create amazing things.

Oh, you mean like these: https://x.com/maxescu/status/1925705749314904525
Yeah hammy, I don't know how we ever managed without this.

Sorry I'm an idiot: https://x.com/maxescu/status/1925705749314904525

So, one could think about me like "What is this guy's deal? Why is he pointing out the downsides of LLM AI tech all the time?"

And it's a perfectly logical question to ask oneself.

But for the life of me I cannot pretend to imagine what is ham's deal? What's with that guy? Is he a major AI company stock owner or something? Crazy AI haters carrying pitchforks burned his village down? Whenever that crazy old man 4gentE writes one of his musings on AI, this crazy squirrel guy ham comes back with something, anything, like it's some sort of on-going fight. So weird.

Here's a nuanced piece about a book that deals with new tech in itself and society.

https://www.newyorker.com/books/page-turner/rethinking-the-luddites-in-the-age-of-ai

Yeah, here's the author of the book, a week ago.
https://youtu.be/JNj2jm_Iy54

But seriously though, do we?

IMHO there is only one way to "fight" - develop superior AI system and don't share it with bad actors.

I would even buy and read, but I saw one opinion online that puts me off:

"Typical AI doomer stuff. Overly dramatic and highly speculative...written by an AI pessimist.

If you enjoy fiction and are not in the industry, this might be for you!

Clearly written just for the money."

Did you read it, 4gentE?

Not yet, I have to clear my reading list first. But I’ve read quite a bit of Hao’s writing. And no, it’s not AI pessimist, and it’s not all that speculative because the writer is very close to the industry. And her logic got a confirmation these days with Altman effectively squeezeng Musk out of the government.

Thing is, it’s not really so much about AI in the sense you might expect, it’s more about the new eerie Silicon Valley / US government coalition that aims to change the world. Watch this, you’ll see what I mean: https://www.youtube.com/watch?v=6ovuMoW2EGk

It also speaks to your idea. This is hard because every one of them has the exact same pose: WE are the only good guys around, all others are bad actors, and you have to give US and none other the money in order to protect yourselves.

Nice, I like this dude podcasts. Not that I agree with everything said, but it's definitely entertaining.



Perhaps, I wasn't clear enough.

First of all, I don't think current AI systems are anything special. It has been demonstrated by various researchers that reasoning capabilities are pretty shallow and "out of distribution" phenomena in GenAI is not truly out of distribution (like in building abstract scene, applying composition principles, etc... like humans would do), but it's rather again, just a shallow generalization of a training data.

The superior AI system would be based on different principles. I'm not saying I know how to do it (yet), but there are few interesting ideas floating around. The key is, if those ideas crystallize, they shouldn't be shared naively.

Unfortunately, year over year, engineers (and perhaps also artists) developed codependency to "more socially capable" MBAs, "eng" bros and alike. It always hurts me personally to see good guys naively "bend over" and just get brutally exploited. But not only that. The same "eng" bros also twist and distort the real purpose behind those inventions. Ttransformers and diffusion models are really great inventions, but of course in the hands of bad actors it's all distorted and abused.

This is similar to what I’ve been saying for a few years now. Argued this position in this place here, perhaps from time to time even having a more enthusiastic version of you on the opposite side. My opinion was and is that LLMs are in fact detrimental to the development of true AI. When that artifically inflated bubble bursts, an especially big “AI winter” will set. Like the dot com bubble. Like the big data bubble. Only x1000. That’s why I despise these false prophets, evagelists like Altman. And when I see some tiny little posers around here (won’t name names) taking that same Altman pose, acting like a kindergarten version of Altman i cringe bigtime.

..and here's a group photo. aren't they cute, those little rascals!

You sure you got the right beavers? ;)

Are those the Mini-Altmans?

i leave that to your imagination ;-)

(but srsly: if they really manage to pull that off, it would be a first and a major feat!)

Not quite sure about that, uh, news. Have a link?

All i found was "China Using 3D Printing and AI to Build 590-Foot-Tall Dam", from May 2022, and "China is poised to complete the Yangqu dam by 2024".And then some rather shallow current pieces on ad-ridden sites regurgitating the former claim. Wikipedia entry on "Yangqu dam" seems unremarkable.

I'm as sceptical as you are, here's one link but yep, there's been a lot of news of unbelievable scientific breakthroughs coming from China recently. but hey, if you can imagine it, sooner or later someone actually does it.

Then why don't we have working faster-than-light travel or hoverboards or net-positive fusion energy or room-temp/normal-pressure superconductors yet? :)

(I still suspect Fatima is a Ninja in disguise :-))