pouët.net

Go to bottom

AI is forcing the hand of the Demoscene.

category: general [glöplog]
BTW, the '6 finger abominations' is the most interesting part that has happened to digital drawing within a good few years though. :)
Quote:
BTW, the '6 finger abominations' is the most interesting part that has happened to digital drawing within a good few years though.

True. The only thing remotely resembling art that came out of it.

Quote:
But banning tech because of 'pretenders and their internet buddies' isn't the way to go imho.


Who’s talking about banning? And why would you and me talk about banning, we don’t have capacity to ban anything, we’re just two random jerkoffs in a forum on an obscure website. I’m saying that this tech surely looks like it’s gonna bring more damage than good. And I’m disgusted by tech bros promoting it shamelessly. And I hate that this tech has ‘grifters’ written all over it from the very start, in a way that very few had in the past. And I’ll personally boycott it. And occasionally go on a miserable preaching spree like this one.;-) That’s all.
added on the 2023-08-13 13:31:29 by 4gentE 4gentE
Hey, sceners following this thread have probably consumed more popcorn than what's healthy for the environment. :P
I have a relatively optimistic vision.

Based on my observation, I believe that Demoscene is care about platforms and their shaping and limitations on creativity. Whether it's DirectX or Fantasy Console, they both constitute a kind of platform with limitations, even if they are created and sold for commercial purposes.

Recently, I deployed Stable Diffusion and ChatGLM on the computers in an internet cafe. They run on consumer-grade graphics cards (I have tried RTX2070/3070/3080 in different internet cafes) and perform well.

Currently, I am living in China, and my personal computer does not have a high-end graphics card. However, in the internet cafe, I can use these computers equipped with high-performance GPUs (along with other above-average configurations, such as i5-9500F, 32GB RAM, and my own USB SSD) at a price of 6 RMB per hour (approximately 0.83 USD). During nighttime, the price may be even cheaper, making it affordable for many people.

I don't believe that they will be limited to a few big companies. The Chinese teenagers in ACG subculture has showcased a significant level of creativity by customizing Stable Diffusion LORA models and creating Galgames with customized chatbots using ChatGLM. They are no different in essence from the bedroom programmers who shaped the Demoscene in the past.

From the perspective of event organizers, I believe that by specifying hardware specifications/execution time, we can potentially create new competition categories for Diffusion or LLM models that balance technical features and creativity.
added on the 2023-08-14 11:10:02 by pengan pengan
A few days ago I got an email asking if I still maintain a supposed Java compiler library - I politely responded that no I never heard of it. Turns out, ChatGPT:

BB Image
added on the 2023-08-14 16:18:23 by Gargaj Gargaj
A good opportunity to create that library, Gargaj, and hack all those people who trust blindly in the code generated (hallucinated?) by a GPT. :]
added on the 2023-08-14 16:26:41 by ham ham
Yeah, you are definitely somewhere in those weights.

But you are still lucky. Imagine someone's name is hallucinated on "who's most famous Hungarian rubber ducks collector" prompt.
added on the 2023-08-14 22:21:59 by tomkh tomkh
So we're leaving this up to luck now?
added on the 2023-08-14 22:24:37 by Gargaj Gargaj
Well, it's fascinating aspect of it actually.

What's even more fascinating to me is why would on earth anyone seriously consider using those chats for anything else than laughter/entertainment ?

I think the problem is people seriously believing in this stuff, not the tech itself.
added on the 2023-08-14 22:33:17 by tomkh tomkh
Quote:
why would on earth anyone

Because that's what they were told.

See my point about marketing now?
added on the 2023-08-14 22:40:17 by Gargaj Gargaj
the AI knows you dislike it, it's doing it on purpose!
There also so many equally dangerous "techs". I'm quite worried of teenagers badly addicted to tiktok - it really ruins their life and future. And it's not just kids, I was talking to many adults and it seems many of high-functioning people might be also addicted to non-sense dopamine bumping consumption - instagram, tiktok, youtube shorts, online forums (oops!) etc... and spending hours on it every day.

I'm not trying to whitewash LLMs here - it's definitely adding one more crap to the list.

All things considered, ChatGPT is just another "product". It's bad for you, sure, but apocalypse or collapse of society? Idk.
added on the 2023-08-14 22:53:40 by tomkh tomkh
OK, help me out here. So let’s suppose I am a crazy old luddite fart (perhaps Gargaj and Farfar and some other people too I guess). We are paralyzed by irational fear and loathing. We fear this tech’s (LLMs) impact, and yet this tech has, some say, near zero chance of backfiring on society. We express that fear by what some people will call doomsaying. So that explains us and our ferver. We are troubled people. We’re in mental distress. It’s the only explanation. But then what’s the other side’s angle? Why are some people so insistent, so ardent to rebutt such obviously crazy luddite old fart’s rants? Especially if the general consensus is that nothing can stop the LLM onslaught anyway? Are these people venture capitalists with heavy investment in this tech? Or do these people just romantically wish humanity was better, so they’ll just deny where it’s heading and why? Or do they want to provide support and comfort to us loonies, persuade us that not all is black? What?
added on the 2023-08-14 23:27:21 by 4gentE 4gentE
Well, dear 4gentE, the word salad in your posts seem to be very easy to predict. I'm sure one doesn't need a lot of parameters to create a bot ranting like you.

Still, I'm strong believer in human intelligence!

If you are in a mental distress, I would encourage you to stop tuning in to AI podcasts, promotional videos about AI, videos where outraged people criticize AI, videos propheting collapse of the society cause by AI, etc...

Just try to ignore all this noise and find yourself a more productive way of spending your time. Something that proves you are better than AI after-all.
added on the 2023-08-14 23:47:29 by tomkh tomkh
Quote:
but apocalypse or collapse of society? Idk.

Because I have at least three times pointed out that this is a deliberate exaggeration (strawman, if you will), I can no longer assume good faith on your part.
added on the 2023-08-15 00:03:01 by Gargaj Gargaj
So 4gentE repeatedly calling himself (and actually also you) a "loonie" and "crazy old luddite fart" is not deliberately exaggerating his position in the discussion, because your points of views are closer?

So, yes, I don't think AI will cause a collapse of society, at least not in a short term. AGI maybe, but we are still far away from it. Feel free to remove "apocalypse" from my previous post, since I do understand well that you guys don't believe in nuclear apocalypse metaphors.
added on the 2023-08-15 00:22:08 by tomkh tomkh
If you restrict the discussion to ChatGPT and the likes, then indeed, they're just toys and no real danger exists.

The title says AI, though, not ChatGPT only.

When talking about AI, then honestly I see danger when people put LLM and other experimental AI in walking robots, or worse yet, ARMED walking or flying robots. It is already proven that they are hackable only with well crafted sequences of phrases. They can be fooled and led to bad logical conclusions. People already found ways to circumvent laws of robotics and make them talk against humanity. They only talk for now, because they don't have physical access to the world -- yet.

We are approaching a point where robots will be able to grab things with their hands, with enough precision to type on keyboards, read texts on screens, etc. That is, this won't take too long for them to have physical access to basically everything humans do.

Again, AI does not need to be sentient or truly intelligent to eventually become an out of control agent of mass destruction.

But yeah, let's blindly proceed. Lets put guns on AI based robots. Lets put hands, legs, lidar, IR cameras plus LLMs on them. What could go wrong, right?
added on the 2023-08-15 01:46:57 by imerso imerso
And not only that, even if we discard the Skynet possibility, which indeed seems like an exaggeration (which it is not, imho), still there is this thing that a lot of jobs are going to be taken in the next years, but there are no real plans regarding what the unemployed will do to survive.
added on the 2023-08-15 01:52:07 by imerso imerso
I think demosceners are a group of highly digitally literate individuals who originated from the warez community. They were once fearless in challenging copyright giants and breaking the restrictions imposed by hardware and software vendors. But why are these rebels so anxious and lacking confidence when it comes to the topic of AI?

Is it just because they have gotten older?
added on the 2023-08-15 08:37:38 by pengan pengan
Personally I think that current AI generation is awesome. I like it. What I am afraid is that it does not mix very well with current humanity selfish mindset.

AI should not be discriminated in compos, in my opinion. Let people use whatever tools they want, is not that what have been said frequently? Why should be different here?

I don't even understand why when it's about code you can use whatever you want, but graphics and music you cannot. Interesting enough, the demoscene presents itself to the world as "The art of CODE". Funny, that.
added on the 2023-08-15 08:53:26 by imerso imerso
Quote:
I don't even understand why when it's about code you can use whatever you want, but graphics and music you cannot.


Demoscene is supposed to be about computer culture and cherish craft that is peculiar to it.
@rexbeng :
We have an example of what algorithmic targeting is doing to our societies. A real world example of this mindless experimenting with people. At least one real world genocide heavily connected to FB.
Yet you seem to think that people who advise caution or warn of potential dangers are "luddites, unintelligent, ranting bots, historically wrong"" etc.
I too am really beginning to suspect good faith.

Now, the person who says:
Quote:
So 4gentE repeatedly calling himself (and actually also you) a "loonie" and "crazy old luddite fart" is not deliberately exaggerating his position in the discussion, because your points of views are closer?


Just one post above also said this:
Quote:
Well, dear 4gentE, the word salad in your posts seem to be very easy to predict. I'm sure one doesn't need a lot of parameters to create a bot ranting like you.

Still, I'm strong believer in human intelligence!

If you are in a mental distress, I would encourage you to stop tuning in to AI podcasts, promotional videos about AI, videos where outraged people criticize AI, videos propheting collapse of the society cause by AI, etc...

Just try to ignore all this noise and find yourself a more productive way of spending your time. Something that proves you are better than AI after-all.


So you see, I was not calling myself those things to exaggerate my position, I was merely describing what you deep down seem to think of me (us). And this post of yours proves it.


Why do you think I'm watching / listening to doomsaying apocalyptic podcast? I am not. That's your assumption based on your deep belief that I (we) are in a mental distress. And don't say "but you yourself said so", don't go that low, the wording "So let's suppose" covered this, this was only a verbal exercise to provoke an honest response from you.

You stubbornly repeat the idea that this new thing that heavily and unnecessarily exacerbates climate crisis, the thing that carries grave danger in this particular point in the evolution of our society, the new thing that furtherly serves deskilling of people in order to make the rich richer is actually not only ambivalent, but good.
I go to every angle to try to make a point (it's not hard because this thing smells bad from so many angles), yet you don't seem to register, you repeat the same things, as if you didn't read a single line, you have only one angle, one half-point, a mantra (we've seen this in history of tech already) but even that one half-point basically depends on the assumption that I'm an "old luddite fart". So, you see, I think it's you actually saying I'm an "old luddite fart", not I.

Plus, you managed not only to miss the point of my previous post, but failed to answer the one literal question it posed, which was: If I (we) are in mental distress, so we are ranting and deploying "word salad", are "ranting bots", what are you? Why do you rant in defense of LLMs so fiercely? Well, to be honest, one doesn't even need to create a bot for that, you know the one 'someone could create with very few parameters to emulate my ranting', it could actually be a copy/paste script. So, no prediction needed at all, I guess your side is, after all, "greener", as it could be made without the use of complex systems.

Oh yes, maybe you did answer my question after all:
Quote:
Just try to ignore all this noise and find yourself a more productive way of spending your time. Something that proves you are better than AI after-all.

This would mean that you are just a kind soul trying to "provide support and comfort to us loonies, persuade us that not all is black." So that's that I guess. Thank you, thank you so much for your concern.
added on the 2023-08-15 09:30:33 by 4gentE 4gentE
Oh sorry, I was talking to / quoting the wrong person.
Sorry @rexbeng
It ought to be @tomkh, not @rexbeng.
Well, I guess, it's my distress showing...
added on the 2023-08-15 09:33:35 by 4gentE 4gentE
4gentE: so what is it, if you feel fine and don't think those thing of yourself, why you are saying the opposite? I'm sorry for taking it at face value. I didn't know you are just exaggerating to build an argument around it, but you really don't think this way. Although that's like admitting of using manipulative strategies in the discussion to me. Well played, I guess?
added on the 2023-08-15 10:33:53 by tomkh tomkh

login

Go to top