by Zain Jaffer
Artists around the world are familiar with the difference between inspired art, and outright copying. Published texts can be plagiarized, and visual art can be forged. When artists learn their craft, sometimes through art school, they try to draw in the styles of the old masters, and use them as their inspiration to create their own. Forgers attempt to pass on their fake work as the well known ones. Most people can separate their bifurcated feelings towards inspired art and fake/forged art.
Artificial Intelligence (AI) has generated mixed feelings amongst many artists, particularly in the visual, audio, video and film domains. Because these General Purpose Transformer (GPT) AIs have been trained by looking and processing the large body of works that are available on the Internet and in digitized form, they can “copy” or generate art that, depending on how you look at it, is inspired or are knock offs of the works of real human artists.
I use the word “copy” because the way an AI does it is not really technically how an AI would do it, but that is beyond the scope of this article. But you know what I mean.
The fact that AI also threatens livelihoods also complicates the arguments. Art connoisseurs and collectors of Picassos and Pollocks will of course want the human created original works and not even take a look at AI. However for those who make a living making company logos, t-shirt artwork, or whatever work that their clients pay for on sites like Fiverr or Upwork, the threat of AI is real.
In general a large portion of the global art community views AI with disdain. Perhaps not when they are querying it to answer certain questions, but not so when it comes to producing what could take days or weeks in seconds or minutes.
This is why I assume that many real human artists (I don’t believe I have to say that) just get amused or see the irony when an AI giant like Open AI accuses the Chinese startup Deepseek of copying or using its training data [https://apnews.com/article/deepseek-ai-chatgpt-openai-copyright-a94168f3b8caa51623ce1b75b5ffcc51]. The irony is that Open Ai and the other AI companies who “copy” human generated art see nothing wrong about resting on the laurels of thousands if not millions of human artists who helped train their AI models for free, but raise a ruckus when another AI competitor copies them.
Personally I think it is possible that sometime in the future we may see a global pushback against AI. It is not hard to imagine and is actually probable because of the impact on livelihoods that AI brings.
This can take many forms. One way is if a movement launches that will only pay for human generated art. Of course that brings along issues like watermarking an artwork and attesting it is human generated. There are software and algorithms to do that.
I wish there was a globally acceptable way to somehow compensate, maybe even using micropayments, everytime an original artwork is used and the derivative AI work makes money.
For now the AI tech industry just marches to its own beat and continues to train on thousands of images and art files, while paying lip service to honoring and respecting human art. I’m not sure how long that will be sustainable, and how that will eventually be resolved, but it is probably likely that a reckoning will come.
###