Greg Rutkowski, a digital artist known for his surreal style, opposes AI art but his name and style have been frequently used by AI art generators without his consent. In response, Stable Diffusion removed his work from their dataset in version 2.0. However, the community has now created a tool to emulate Rutkowski’s style against his wishes using a LoRA model. While some argue this is unethical, others justify it since Rutkowski’s art has already been widely used in Stable Diffusion 1.5. The debate highlights the blurry line between innovation and infringement in the emerging field of AI art.

  • FaceDeer@kbin.social
    link
    fedilink
    arrow-up
    26
    ·
    11 months ago

    His art was not “stolen.” That’s not an accurate word to describe this process with.

    It’s not so much that “it was done before so it’s fine now” as “it’s a well-understood part of many peoples’ workflows” that can be used to justify it. As well as the view that there was nothing wrong with doing it the first time, so what’s wrong with doing it a second time?

    • Kara@kbin.social
      link
      fedilink
      arrow-up
      20
      ·
      edit-2
      11 months ago

      I don’t like when people say “AI just traces/photobashes art.” Because that simply isn’t what happens.

      But I do very much wish there was some sort of opt-out process, but ultimately any attempt at that just wouldn’t work

      • chemical_cutthroat@kbin.social
        link
        fedilink
        arrow-up
        5
        ·
        11 months ago

        People that say that have never used AI art generation apps and are only regurgitating what they hear from other people who are doing the same. The amount of arm chair AI denialists is astronomical.

      • ricecake@beehaw.org
        link
        fedilink
        arrow-up
        4
        ·
        11 months ago

        There’s nothing stopping someone from licensing their art in a fashion that prohibits their use in that fashion.
        No one has created that license that I know of, but there are software licenses that do similar things, so it’s hardly an unprecedented notion.

        The fact of the matter is that before people didn’t think it was necessary to have specific usage licenses attached to art because no one got funny feelings from people creating derivative works from them.

    • zeus ⁧ ⁧ ∽↯∼@lemm.ee
      link
      fedilink
      English
      arrow-up
      13
      ·
      11 months ago

      pirating photoshop is a well-understood part of many peoples’ workflows. that doesn’t make it legal or condoned by adobe

        • TwilightVulpine@kbin.social
          link
          fedilink
          arrow-up
          5
          ·
          11 months ago

          Not at the point of generation, but at the point of training it was. One of the sticking points of AI for artists is that their developers didn’t even bother to seek permission. They simply said it was too much work and crawled artists’ galleries.

          Even publicly displayed art can only be used for certain previously-established purposes. By default you can’t use them for derivative works.

          • FaceDeer@kbin.social
            link
            fedilink
            arrow-up
            4
            ·
            11 months ago

            At the point of training it was viewing images that the artists had published in a public gallery. Nothing pirated at that point either. They don’t need “permission” to do that, the images are on display.

            Learning from art is one of the previously-established purposes you speak of. No “derivative work” is made when an AI trains a model, the model does not contain any copyrightable part of the imagery it is trained on.

            • TwilightVulpine@kbin.social
              link
              fedilink
              arrow-up
              4
              ·
              11 months ago

              Of course they need permission to process images. No computer system can merely “view” an image without at least creating a copy for temporary use, and the purposes for which that can be done are strictly defined. Doing whatever you want just because you have access to the image is often copyright infringement.

              People have the right to learn from images available publicly for personal viewing. AI is not yet people. Your whole argument relies on anthropomorphizing a tool, but it wouldn’t even be able to select images to train its model without human intervention, which is done with the intent to replicate the artist’s work.

              I’m not one to usually bat for copyright but the disregard AI proponents have for artists’ rights and their livelihood has gone long past what’s acceptable, like the article shows.

              • FaceDeer@kbin.social
                link
                fedilink
                arrow-up
                4
                ·
                11 months ago

                If I run an image from the web through a program that generates a histogram of how bright its pixels are, am I suddenly a dirty pirate?

                • TwilightVulpine@kbin.social
                  link
                  fedilink
                  arrow-up
                  5
                  ·
                  edit-2
                  11 months ago

                  If you run someone’s artwork through a filter is it completely fine and new just because the output is not exactly like the input and it deletes the input after it’s done processing?

                  There is a discussion to be made, in good faith, of where the line lies, what ought to be the rights of the audience and what ought to be the rights of the artists, and what ought to be the rights of platforms, and what ought to be the limits of AI. To be fair, that’s a difficult situation to determine, because in many aspects copyright is already too overbearing. Legally, many pieces of fan art and even memes are copyright infringement. But on the flipside automating art away is too far to the other side. The reason why Copyright even exists, at least ideally, is so that the rights and livelihood of artists is protected and they are incentivized to continue creating.

                  Lets not pretend that is just analysis for the sake of academic understanding, there is a large amount of people who are feeding artists’ works into AI with the express purpose of getting artworks in their style without compensating them, something many artists made clear they are not okay with. While they can’t tell people not to practice styles like theirs, they can definitely tell people not to use their works in ways they do not allow.

                  • FaceDeer@kbin.social
                    link
                    fedilink
                    arrow-up
                    2
                    ·
                    11 months ago

                    If you run someone’s artwork through a filter is it completely fine and new just because the output is not exactly like the input and it deletes the input after it’s done processing?

                    No, that’s a derivative work. An analysis of the brightness of the pixels is not a derivative work.

                    There is a discussion to be made, in good faith, of where the line lies, what ought to be the rights of the audience and what ought to be the rights of the artists, and what ought to be the rights of platforms, and what ought to be the limits of AI.

                    Sure, but the people crying “You’re stealing art!” are not making a good faith argument. They’re using an inaccurate, prejudicial word for the purpose of riling up an emotional response. Or perhaps they just don’t understand what copyright is and why it is, which also puts their argument in a bad state.

                    The reason why Copyright even exists, at least ideally, is so that the rights and livelihood of artists is protected and they are incentivized to continue creating.

                    Case in point. That’s not why copyright exists. The reason for the American version of copyright is established right in the constitution: “To promote the progress of science and useful arts”. If you want to go more fundamental than just what the US is up to, the original Statute of Anne was titled “An Act for the Encouragement of Learning”.

                    The purpose of copyright is not to protect the rights or livelihood of artists. The protection of the rights and livelihood of artist is a means to the actual purpose of copyright, which is to enrich the public domain by prompting artists to be productive and to publish their works.

                    An artist that opposes AIs like these is now actively hindering the enrichment of the public domain.

            • Kichae@kbin.social
              link
              fedilink
              arrow-up
              2
              ·
              11 months ago

              Bring publicly viewable doesn’t make them public domain. Bring able to see something doesn’t give you the right to use it for literally any other reason.

              Full stop.

              My gods, you’re such an insufferable bootlicking fanboy of bullshit code jockies. Make a good faith effort to actually understand why people dislike these exploitative assholes who are looking to make a buck off of other people’s work for once, instead of just reflexively calling them all phillistines who “just don’t understand”.

              Some of us work on machine learning systems for a living. We know what they are and how they work, and they’re fucking regurgitation machines. And people deserve to have control over whether we use their works in our regurgitation machines.

          • FaceDeer@kbin.social
            link
            fedilink
            arrow-up
            3
            ·
            11 months ago

            They were not used for derivative works. The AI’s model produced by the training does not contain any copyrighted material.

            If you click this link and view the images there then you are just as much a “pirate” as the AI trainers.

            • TwilightVulpine@kbin.social
              link
              fedilink
              arrow-up
              3
              ·
              11 months ago

              The models themselves are the derivative works. Those artists’ works were copied and processed to create that model. There is a difference between a person viewing a piece of work and putting that work to be processed through a system. The way copyright works as defined, being allowed to view a work is not the same as being allowed to use it in any way you see fit. It’s also innacurate to speak of AIs as if they have the same abilities and rights as people.

        • zeus ⁧ ⁧ ∽↯∼@lemm.ee
          link
          fedilink
          English
          arrow-up
          5
          ·
          11 months ago

          i’m not making a moral comment on anything, including piracy. i’m saying “but it’s part of my established workflow” is not an excuse for something morally wrong.

          only click here if you understand analogy and hyperbole

          if i say “i can’t write without kicking a few babies first”, it’s not an excuse to keep kicking babies. i just have to stop writing, or maybe find another workflow

          • FaceDeer@kbin.social
            link
            fedilink
            arrow-up
            4
            ·
            11 months ago

            The difference is that kicking babies is illegal whereas training and running an AI is not. Kind of a big difference.

              • FaceDeer@kbin.social
                link
                fedilink
                arrow-up
                4
                ·
                11 months ago

                You’re using an analogy as the basis for an argument. That’s not what analogies are for. Analogies are useful explanatory tools, but only within a limited domain. Kicking a baby is not the same as creating an artwork, so there are areas in which they don’t map to each other.

                You can’t dodge flaws in your argument by adding a “don’t respond unless you agree with me” clause on your comment.

                • zeus ⁧ ⁧ ∽↯∼@lemm.ee
                  link
                  fedilink
                  English
                  arrow-up
                  2
                  ·
                  edit-2
                  11 months ago

                  You’re using an analogy as the basis for an argument. That’s not what analogies are for. Analogies are useful explanatory tools, but only within a limited domain

                  actually that’s exactly what i was using it for.

                  Kicking a baby is not the same[1] as creating an artwork, so there are areas in which they don’t map to each other.

                  if you read carefully, you’ll see that writing is analogous to creating an artwork, and kicking a baby is analogous to doing something that someone has asked you not to, and you’re continuing anyways. if you read even more carefully, you’ll see that i implied i wasn’t making a moral comment on ai, piracy, or even kicking babies

                  You can’t dodge flaws in your argument by adding a “don’t respond unless you agree with me” clause on your comment.

                  i didn’t intend to. i did it so i wouldn’t have to waste my time arguing with those who don’t understand analogies. however i seem to be doing that anyways, so if you’ll excuse me, i’m going to stop


                  edit: okay, i’ve been reading the rest of this thread, and you clearly don’t understand analogy. i have no idea why you clicked on my comment


                  1. yes. analogous doesn’t mean “the same”. it means "able to draw demonstrative parallels between ↩︎

        • Kichae@kbin.social
          link
          fedilink
          arrow-up
          1
          ·
          11 months ago

          His work was used in a publicly available product without license or compensation. Including his work in the training dataset was, to the online vernacular use of the word, piracy.

          They violated his copyright when they used his work to make their shit.

    • Pulse@dormi.zone
      link
      fedilink
      arrow-up
      0
      ·
      11 months ago

      Yes, it was.

      One human artist can, over a life time, learn from a few artists to inform their style.

      These AI setups are telling ALL the art from ALL the artists and using them as part of a for profit business.

      There is no ethical stance for letting billion dollar tech firms hoover up all the art ever created to the try and remix it for profit.

      • FaceDeer@kbin.social
        link
        fedilink
        arrow-up
        0
        ·
        11 months ago

        No, it wasn’t. Theft is a well-defined word. When you steal something you take it away from them so that they don’t have it any more.

        It wasn’t even a case of copyright violation, because no copies of any of Rutkowski’s art were made. The model does not contain a copy of any of the training data (with an asterisk for the case of overfitting, which is very rare and which trainers do their best to avoid). The art it produces in Rutkowski’s style is also not a copyright violation because you can’t copyright a style.

        There is no ethical stance for letting billion dollar tech firms hoover up all the art ever created to the try and remix it for profit.

        So how about the open-source models? Or in this specific instance, the guy who made a LoRA for mimicking Rutkowski’s style, since he did it free of charge and released it for anyone to use?

        • Pulse@dormi.zone
          link
          fedilink
          arrow-up
          0
          ·
          11 months ago

          Yes copies were made. The files were downloaded, one way or another (even as a hash, or whatever digital asset they claim to translate them into) then fed to their machines.

          If I go into a Ford plant, take pictures of their equipment, then use those to make my own machines, it’s still IP theft, even if I didn’t walk out with the machine.

          Make all the excuses you want, you’re supporting the theft of other people’s life’s work then trying to claim it’s ethical.

          • FaceDeer@kbin.social
            link
            fedilink
            arrow-up
            1
            ·
            11 months ago

            Yes copies were made. The files were downloaded, one way or another (even as a hash, or whatever digital asset they claim to translate them into) then fed to their machines.

            They were put on the Internet for that very purpose. When you visit a website and view an image there a copy of it is made in your computer’s memory. If that’s a copyright violation then everyone’s equally boned. When you click this link you’re doing exactly the same thing.

            • Pulse@dormi.zone
              link
              fedilink
              arrow-up
              0
              ·
              11 months ago

              By that logic I can sell anything I download from the web while also claiming credit for it, right?

              Downloading to view != downloading to fuel my business.

              • FaceDeer@kbin.social
                link
                fedilink
                arrow-up
                1
                ·
                11 months ago

                No, and that’s such a ridiculous leap of logic that I can’t come up with anything else to say except no. Just no. What gave you that idea?

                • Pulse@dormi.zone
                  link
                  fedilink
                  arrow-up
                  0
                  ·
                  11 months ago

                  Because this thread was about the companies taking art feeding it into their machine a D claiming not to have stolen it.

                  Then you compared that to clicking a link.

                  • FaceDeer@kbin.social
                    link
                    fedilink
                    arrow-up
                    1
                    ·
                    11 months ago

                    Yes, because it’s comparable to clicking a link.

                    You said:

                    By that logic I can sell anything I download from the web while also claiming credit for it, right?

                    And that’s the logic I can’t follow. Who’s downloading and selling Rutkowski’s work? Who’s claiming credit for it? None of that is being done in the first place, let alone being claimed to be “ok.”