But people said the same thing about photography and painting 150 years ago!

This is entirely different from that. Landscaping had aesthetic value after photography because people were able to embellish and stylize landscapes in ways they didn't actually exist in real life. There is no way to "escape" from AI into more stylization. All it takes is enough of those new stylized images and it'll be able to replicate it. This is different from photography because photography can't learn.

But people want human expression!

Yes, they do, but they'll probably only realize this after a couple decades of art [almost] completely uninspired by the human condition and depression and anxiety skyrockets. In the short term, people will only care that they can type into a field and get what they want without any significant investment. Good luck finding an art job that isn't just making prompts in that economy (And if you say that making prompts is the same as being a painter or illustrator, yes it is art but no it isn't the same and fuck you).

I rest my case, art automation is cool but under capitalism it'll only be used to devalue artists further, and drive them deeper into poverty. It would be a great tool in a society capable of regulating itself but WE DON'T LIVE IN THAT SOCIETY AND I SHOULDN'T HAVE TO POINT THAT OUT.

STOP LISTENING TO :melon-musk: :debatebro-l: AND SUPPORT YOUR ARTIST COMRADES NOW

Now that I got your attention with my inflammatory statements, please commission your artist comrades and support them in their struggle to exist in what is already a very punishing world to live in. I promise to give any of you that do big hugs, I love you all, and bye.

  • KiaKaha [he/him]
    ·
    edit-2
    2 years ago

    People said the same thing about digital art 15 years ago.

    Artists will use the tools provided to make their lives easier. AI can free up more of the technical skills, lowering the barrier of entry and creating more room for creativity, aka the part of art that people actually enjoy.

    What I imagine we’ll see is artists using AI as a shortcut as a part of their work. Think of it as using stock images or reference photos for backgrounds, so you can focus on what you want in the foreground.

    For artists who enjoy creating entirely new styles, they’ll be able to focus on innovating, and then train an AI on new styles to allow it to be further democratised (or commodified).

    Yes, the means of production are owned by the bourgeoisie, and that will impact, as it always does. Paint, workshops, easels, all cost money. Photoshop costs money. Access to the premium AI models will likely cost money.

    Yes, as the productive forces develop and require less live labour for the same result, prices drop, and the previous mode of production becomes boutique. That’s as true for art as it is for linen.

    Tell your artist comrades to apply for Dall-E access ASAP, and incorporate it into their workflow. They have an understanding of art theory, and are better equipped to use it than a random management consultant is.

      • KiaKaha [he/him]
        ·
        2 years ago

        Not yet, no. You’re left working with it as it is. It’s still faster to work with a single layer pregenerated image than it is to create that image yourself.

        This is a new technology, and it’s developing quickly. Think about how fast we got from Pong to Crysis, or from the Pager to an iPhone. Give it time, and we’ll see AI not only produce images in PSD layers, but likely even split existing images into layers.

          • KiaKaha [he/him]
            ·
            edit-2
            2 years ago

            Since the 2000s? That’s the takeoff of the digital era, so think of everything supplanted by digital variants.

            Streaming and piracy killed Blockbuster. VOIP and video calls killed pay phones and landlines, and will kill mobile voice calls too. Navigation software killed paper maps. Email’s killing snail-mail. Ctrl+F alone has saved countless labour-hours.

            Yes, any new technology is used to break labour. Labour is in an era of weakness. Its gains were won during the social democratic period, and any disruption of that order will cause those ossified remnants of labour power to shatter.

            Take CGI for example. Suddenly you can create anything, literally anything you can imagine on screen. To people of the special effects era, it’d be near impossible to believe. It’s a modern miracle. And yet the CGI industry is young, without the hard-won protections of film unions. It ends up overworked and exploited, fresh feed for the Mouse.

      • KiaKaha [he/him]
        ·
        2 years ago

        Maybe! The question there is whether there is an inelastic demand for art.

        Personally, my money would be on that not being the case. People love art. If artists can work twice as fast, the price of art will drop to approx half, and people will buy twice as much.

        It will hurt artists who don’t adapt. They’ll be stuck producing increasingly boutique pieces for a dwindling customer base. Those that survive will survive as artisans. It’s the same process that other commodities, like clothing and furniture, have already undergone.

        Artists have a distinct advantage, though. The desire for art, unlike the desire for bedsheets, is near limitless, especially when it’s custom art. As technology improves, ease of production increases, and prices fall, demand will rise.

        Of course, ‘artists’ may change form significantly. It could be that ‘artist’ means ‘designer of unique styles for an AI to learn from’, for example.

        It could also be that, like OP says, AI will get ‘good enough’ to satisfy most people instantly, and the value-add of a real artist will be limited.

          • KiaKaha [he/him]
            ·
            2 years ago

            My point is less about composition of organic vs AI, and more demand for ‘art’ at all, regardless of composition.

            Commercial art (advertising) will also need artists. AI cannot replace artists. It is just a tool used by artists. The artist still needs to conceive of what the final product should look like, and describe that to the AI. The artist will probably even train the AI specifically on manually-produced art, to ensure consistency of design and style.

            Let’s take the cereal example. You’re a marketing executive, and you want a box for your new line of cereal.

            You’ll want an artist to draw the reference sheet for the mascot first, for one. What animal will it be? Will it even be an animal? What colours will be the most enticing? Then you need to decide what feelings you want the box design to evoke, and know how to bring those to fore in a viewer.

            Next, you need to create the art. This will take multiple iterations of attempts. You need to be able to recognise what is and is not ‘good enough’, when to keep going, and when to stop.

            Then you need to show that product to your client/boss, and sell it to them convincingly.

            What I’ve just described sounds an awful lot like the standard artistic process. In fact, I never even mentioned the word ‘AI’.

              • KiaKaha [he/him]
                ·
                2 years ago

                Correct—it will reduce the number of humans needed to produce the same results.

                The question is, will the extra capacity for art production be used for anything?

                As I’ve said before, I think it will be. People really like art. Let’s take the cereal box example. Perhaps where before, there was but a single box design in a run, there will instead be multiple. Perhaps the cereal company will decide to use the spare capacity to have a unique design on every single box on the shelf.

                Or maybe that capacity is instead turned towards a boom in the furry industry, with fursonas mass produced like never before.

                If there’s one area I have confidence in our ability to induce demand, it’s in demand for art.