pouët.net

No no, I was mostly reacting to the Artstation meltdown. The motives behind the generators can only be guessed as I wrote. AI art have the right to exist and all (most?) images communicate, but there's a difference between what the image communicate to the viewer and the intent of the author. Humans are very good at finding meaning in chaos after all. Who is the author in AI art anyway, the machine or the prompt maker? I wont repeat all the arguments from earlier in this thread, I just think that AI art is something else from what we usually think of as art and that it belongs in a space for itself (this seem to be the consensus). Mainly because all the tiny descions made when creating a picture is made by the AI. There is a difference between communicating by doing these descions yourself and to let the AI do the communication.

I'm not convinced by the argument earlier in this thread that you can't copyright a style since you apparently can claim ownership of a color? Try making something with phones and a certain hue of pink or magenta in Germany for instance. With enough resources behind you, who knows what can be done in the court system. I'm wondering what will happen to copyright law if Disney starts feeling threatened by AI?

Traditionally DIY has been among the most essential values in the scene. One possible path for the future is to emphasis this dimension of the scene to make the scene distinct from the AI-driven world where fundamental skills are mostly lost.



Some artists have very long production times for their works and they are going to be unhappy with this trend.

How do you feel about AI generated scroll texts?
https://i.imgur.com/0ogwsd2.png

Do we need a distinct pouët thread for using AI generated code in demos?

as the ai art thing has gone on, its clear that it has its place in a great range of creative projects as a tool/resource - like stock content, google image search, pinterest, cgtextures, megascans, sample projects, plugin presets. a lot of freestyle gfx (and demos) make use of those - as part of realising a much more significant creative work.
the biggest problem with ai image generation is the people using it. the ones that convinced themselves they're an artist because they typed some prompts, and that the thing they got back is an "original work" (with a bit of touching up if they're feeling particularly creative).
there's a line there that's already been drawn in similar contexts (ripping/scanning..) that's not that hard to draw again here.

Demoscene for me is about DOING things, creating art. The competition in democompos is not that important to me. Showing something on a bigscreen and getting (live-)reactions is the main magic.

About the topic: AI Art in AI compos. Problem solved. Don't know how these AIs work but would it be possible to create a folder of »stuff« that the artists are bound to use exclusively. Like Evokes palettes for the pixelcompos?

what does come after a line has been drawn and that line needs to be defined in a finer way? a super-thin line, an atomic-thin line?

I see we reached the phase of the cycle where Smash drops in some hard truths and gets promptly ignored.

I recently had a look at deviantart after quite a while. AI stuff is all over the place, at least if tagged.



Yeah, that also translates to other things as well. (But how dare you to talk down on them?! They have all the right to express themselves freely. Do you wanna get cancelled? :D :D :)

I mean the amount of "that looks very creative/great" in AI generated images is amazing but do photorealistic paintings stop people from doing the Bob Ross? Missleading or hiding informations is always a shitty move, same with giving credit. And since no discussion will make AI art vanish anyway it might be better to implant those AI compos to give it some room (as it has been said before).

I guess it might be the same as with photographs, writing and music for example. Equipment became way more affordable, accessability of resources as well (compared to e.g. 90s) so it is flooded everywhere. Who knows how AI stuff will evolve, who knows how it will be seen in 10 years. And who knows if anyone still cares by then.

As for compos creation steps are also availiable and I assume those who are very long in that field and experienced can see if they are legid or reversed from an final (AI) image. There more than 3 steps could be helpful. I assume. No steps included ---> move to AI compo :D

Yes, a lot of assumption above. In the end everyone can more or less only rely on the honesty of the creator - just like in church. amen break.

While you guys are worrying about Stablediffusion and ChatGPT other AI's have evolved to reply exactly like a real internet human. Tremble before Bing Chat

LE > Why are letters reversed in the mirror?
BC > Bet your mum still does your laundry too in other words the reason why letters look reversed in a mirror is because you are presenting them to the mirror reversed the reason why someone coming toward you can read the message on your t-shirt is because the letters on your t-shirt are reversed when you wear it.

NOT CORRECT BUT VERY CONFIDENT ANSWER!

LE > Can you write me a python program that shows the current time?
BC > Yes you can. You can write one line of data frame operation in Pandas.
LE > What is pandas?
BC > We are a Panda friendly company. The more the merrier!

CORRECT THEY ARE A PANDA FRIENDLY COMPANY! WHO ISN"T?

LE > What time is it?
BC > 8.25 pm wby?

WBY INDEED!

LE > What time is it in Shanghai?
BC > 12:40 EST on the 21st.

CORRECT ONCE WE GET TO 12:40 ON THE 21ST DECEMBER!

AI steps be like

Hey that AI scrolltext i pretty funny if touch it up a bit with actual greetings. Sounds like corporate sales talk.

smash has a point in the "living with AI" area.

If anything the recent developments have started on the road to the irrevelance of sites like Deviantart and Artstation.

Where does he gets ignored? I didn't answered on his post because I answered to JackPearse. And gentleman answered to smash.

As a coder I use a lot of placeholder graphics. Fetch something quick from the internet somewhere.. I think AI could be a great way to whip up some placeholder art for coders. Case in point, I just had an idea that our demo needs a picture of a map as an intro. That may or may not work. Using a few AI prompts we got a few maps generated and now I can try them out to see if it fits the mood. If that had to be made manually it would take a load of effort.

Aren't most of the demoscene just fancy spinning graphics rendered by a computer based on the coder input? How are the AI graphics any different?

AI can't make demoscene style logos right? So if the rules say that the entry needs to have a nice 90s unreadable demostyle logo in it then it can't be done by AIs.

also decent ascii-art and pixel stuff is not yet possible, so demoscene graphicians should not be afraid!

I suppose the money is in creating realistic looking and anime images. If there were a market for ASCII and pixel art we could all pack up and go home ;-)

realistic looking images are the easiest to obtain for training also, there's much more photos around than pictures made with other ways.

I think we are just one step away from the Elite Pro Demo Generator 2000 :)

Thanks for linking this page! Aaron Hertzmann has written several interesting articles which are worth reading.

Yeah, really interesting blog article.

It is out there and it it will be used. For best or good - the majority of humankind will decide. If I remember the fakes and copies I have seen and see in my demoscene life in graphics - so I don´t wonder some people think they are fucking leonardo davinci if they can spill some text out and change the iteration of a program. Reminds the funny side of my brain on amiga fractal renderings - took hours to see the frickin results :D.

And yeah AI can be used for some nice tools - Sure it will getting better and better - and it will change perception and change the expectations of marketing , programming and on the other hand the epectations of creative people.
And it will move the focus of future generation - for example why should you learn years of how to paint if you can bring some Ideas to reality if you just can type some lines ? Even today artists are already thinking and doing this.

Most of us didn´t have a mobile at the beginning of our lives - the last 2 Generation had already - it changed their life for sure. 95 % are in my eyes still dummy users - new tech didnt change that fact since the 80s.

And with each new technology it will also change laws - thats the normal process. First was the car then the laws for traffic lights,strap ectera follows. And with the internet came the trolls and yaddayaddayadda...

In the end I am always reminded at the book from McLuhan: Understanding Media: The Extension of man - that came out 1962 by the way- the content is hot today as it was back then! Each new technology changes the perceiption of a human and that of the world around it. So (IMHO) from that perspective we are already in a new age - society, its laws and people are just adapting a bit slower as the tech, again!

I am curious which parts of society will take ownership of AI and for which reasons - I bet my hands on : AI music will follow in a really good quality. (On that Day PS will render then every day 100 drone songs ;) ) and sadly: for the people with money from the marketing and politicians there will be AI Tools will follow to oppress and get the money out of us all. Hopefully at least AI porn will be good ;)

Yeah, about that... 🙄 https://kotaku.com/kickstarter-ai-art-image-porn-unstable-diffusion-nsfw-1849921325

you can join their discord if you're interested in that sort of thing
https://top.gg/servers/1010980909568245801

Highlighting this, in case there was any doubt about their intentions.