4 ms·

Yes, it would be all good if they just tracked attribution and organized micro-payments and opt-out for the artists they are leveraging. "Here's the list of ar

by toss1 4y ago

Yes, it would be all good if they just tracked attribution and organized micro-payments and opt-out for the artists they are leveraging. "Here's the list of artists contributing to your work not already in compensated libraries such as Getty. Your bill is $X.XX and 70% of your payment will go to these artists in this proportion [list]". Instead MS is just pocketing it. It isn't like they lack the funds and hardware/software resources to also implement backtracing... Perhaps what should be advocated is use this for idea generation then present it as part of the design brief to get local designers and artists to do a genuine creative work. And give them credit on the published result.

ModernMech 4y ago
This is just the way of capitalism, and it's how the entire economy works. I mean, you're so concerned with the artists work being stolen, but what of the scientists? Yes, the art produced by Stable Diffusion isn't possible without the artists, but it's also not possible without the countless scientists who have worked over decades to get this technology to where it is. I didn't hear artists complaining when they were happily using technologies developed by people who were not compensated when their work was productized by corporations. Where is their attribution? Should we expect Microsoft and OpenAI et al. to list the authors of the papers they used to work through the technical aspects of their product? Should all of the open source projects they leveraged be attributed and compensated? If we are going to ask Microsoft to share profits with artists, shouldn't we also ask them to share profits with scientists and open source devs? Arguably, yes! But if so, then the focus on the poor artists being robbed is misplaced. If artists are being robbed by corporations, then they can get in line if they want a payout, because the rest of us are being robbed too. It's a much larger problem than just artists -- it's that corporations capture all of the downstream benefits of upstream content/science/technology developments, and then refuse to reinvest at a proportional rate back into the system that allowed them to profit so handsomely. This applies to the scientists and the artists. So yes, be mad at what Microsoft is doing. But the solution isn't to get them to pay artists, that doesn't fix anything. If we want actual change, we need to fundamentally reimagine the relationship between corporations and the rest of us. If the only way to get there is for corporations to screw over more people until enough wake up, then so be it.
- rchaud 4y ago
  Why are you conflating scientific research (something that is paid), with art that was freely uploaded to social media sites and message boards? Microsoft Designer is not spitting out composites of stolen research, it is creating composites of art that can't be matched back to the original. > If artists are being robbed by corporations, then they can get in line if they want a payout, because the rest of us are being robbed too. Perhaps you should disclose what AI is stealing from you. For me, nothing, because my career is in an area where I don't have to publicly share my work.
  - nuancebydefault 4y ago
    If you think of it, every work of art is a combination and/or transformation of pre-existing work. Only Adam and Eve never plagiarized.
    - toss1 4y ago
      While most works of are re-combinations (up in the how many 9s after 99% region?), it is pretty hard to say that there has never been a new idea. Humanity has come from apes without art, up to modern digital art painted with lasers in the sky with drones.... From cave painting through Greek & Roman art, through Renaissance through modern art, and it's not merely new mediums for the same painting (and even so, "the medium is the message", so there's new stuff there). Same for music, text, etc...
      - nuancebydefault 4y ago
        This is a very interesting discussion. Now zoom in into your understanding of "new idea". You can also call it "refinement of existing idea". "New" can be (and maybe always is? ) a combination or transformation of "existing". Just like any number is a combination of other numbers.
        toss1 4y ago
        Agreed, interesting! So, if we're to establish (I won't say "prove") the idea of [everything is transformation], we'd need to sort out the list of primitives, which are being re-combined. Conversely, for the [some things are really new], we'd have to find the new items. That said, the new items will almost popup in the context of other transformations, e.g., a new innovation in painting would still likely occur within some more normal transformations. Take for example the Abstract Art movement, which was about painting not representations of real things, but exploring the elements of form, color, line, texture, but not painting people or landscapes (or only doing so in a very abstract way). So, painting a person on a digital medium vs oil & canvas or watercolor is a transformation. But my sense is that some of these abstract artists were actually digging down into the lower levels of subconscious visual processing, where the initial processing is to find edges, colors, movement, textures, and this was entirely new in the history of painting. What about concepts, from the flat earth to the round earth, from geocentricism to heliocentrism, to no special location in the universe, or from earth-water-fire-air to alchemy, to structure-of-the-atom-based chemistry? Are there no new ideas here, or are they just out of scope? To your example, is any number just a combination of other numbers? The concept of Zero didn't exist for a long time in recorded history (how long is debated). What about negative numbers, imaginary numbers? Do we consider these new or combinations? Or, are we just discussing the Ship of Theseus, a fundamentally unsolvable question that depends on our concept of "new"? (still a fun rabbit hole!)

by toss1 4y ago

ModernMech 4y ago

This is just the way of capitalism, and it's how the entire economy works. I mean, you're so concerned with the artists work being stolen, but what of the scientists? Yes, the art produced by Stable Diffusion isn't possible without the artists, but it's also not possible without the countless scientists who have worked over decades to get this technology to where it is. I didn't hear artists complaining when they were happily using technologies developed by people who were not compensated when their work was productized by corporations. Where is their attribution? Should we expect Microsoft and OpenAI et al. to list the authors of the papers they used to work through the technical aspects of their product? Should all of the open source projects they leveraged be attributed and compensated? If we are going to ask Microsoft to share profits with artists, shouldn't we also ask them to share profits with scientists and open source devs? Arguably, yes! But if so, then the focus on the poor artists being robbed is misplaced. If artists are being robbed by corporations, then they can get in line if they want a payout, because the rest of us are being robbed too. It's a much larger problem than just artists -- it's that corporations capture all of the downstream benefits of upstream content/science/technology developments, and then refuse to reinvest at a proportional rate back into the system that allowed them to profit so handsomely. This applies to the scientists and the artists. So yes, be mad at what Microsoft is doing. But the solution isn't to get them to pay artists, that doesn't fix anything. If we want actual change, we need to fundamentally reimagine the relationship between corporations and the rest of us. If the only way to get there is for corporations to screw over more people until enough wake up, then so be it.

rchaud 4y ago
Why are you conflating scientific research (something that is paid), with art that was freely uploaded to social media sites and message boards? Microsoft Designer is not spitting out composites of stolen research, it is creating composites of art that can't be matched back to the original. > If artists are being robbed by corporations, then they can get in line if they want a payout, because the rest of us are being robbed too. Perhaps you should disclose what AI is stealing from you. For me, nothing, because my career is in an area where I don't have to publicly share my work.
- nuancebydefault 4y ago
  If you think of it, every work of art is a combination and/or transformation of pre-existing work. Only Adam and Eve never plagiarized.
  - toss1 4y ago
    While most works of are re-combinations (up in the how many 9s after 99% region?), it is pretty hard to say that there has never been a new idea. Humanity has come from apes without art, up to modern digital art painted with lasers in the sky with drones.... From cave painting through Greek & Roman art, through Renaissance through modern art, and it's not merely new mediums for the same painting (and even so, "the medium is the message", so there's new stuff there). Same for music, text, etc...
    - nuancebydefault 4y ago
      This is a very interesting discussion. Now zoom in into your understanding of "new idea". You can also call it "refinement of existing idea". "New" can be (and maybe always is? ) a combination or transformation of "existing". Just like any number is a combination of other numbers.
      - toss1 4y ago
        Agreed, interesting! So, if we're to establish (I won't say "prove") the idea of [everything is transformation], we'd need to sort out the list of primitives, which are being re-combined. Conversely, for the [some things are really new], we'd have to find the new items. That said, the new items will almost popup in the context of other transformations, e.g., a new innovation in painting would still likely occur within some more normal transformations. Take for example the Abstract Art movement, which was about painting not representations of real things, but exploring the elements of form, color, line, texture, but not painting people or landscapes (or only doing so in a very abstract way). So, painting a person on a digital medium vs oil & canvas or watercolor is a transformation. But my sense is that some of these abstract artists were actually digging down into the lower levels of subconscious visual processing, where the initial processing is to find edges, colors, movement, textures, and this was entirely new in the history of painting. What about concepts, from the flat earth to the round earth, from geocentricism to heliocentrism, to no special location in the universe, or from earth-water-fire-air to alchemy, to structure-of-the-atom-based chemistry? Are there no new ideas here, or are they just out of scope? To your example, is any number just a combination of other numbers? The concept of Zero didn't exist for a long time in recorded history (how long is debated). What about negative numbers, imaginary numbers? Do we consider these new or combinations? Or, are we just discussing the Ship of Theseus, a fundamentally unsolvable question that depends on our concept of "new"? (still a fun rabbit hole!)