A.I. vs Artists

If you’re part of or familiar with the art community and industry, you’ve most likely heard about artificial intelligence models and how they are negatively affecting artists across the globe. If you don’t already know, artificial intelligence has reached a stage where it now can produce images that look almost indistinguishable from art made by artists. The model does this by using the massive number of images fed to it from the internet to train itself to make variations or amalgamations of these art pieces using the keywords associated with them. Over time, the images produced by models attempting to replicate art have improved, because of gradual adjustments in parameters set in the code that help the output seem more “human”.

When AI art is easily told apart from actual human-made drawings or paintings, most people have a negative reaction to it (with some exceptions), but in the current day the models have improved enough so that people aren’t sure of what’s really a drawing and what is a model-generated image. Because of this increasingly rapid progress in the improvement of AI technologies, people and companies have been using artificial intelligence models to take shortcuts from having to pay real artists or make art themselves. “Today, with AI technology, the work of just about any great artist in human history can be mimicked in moments to produce an image on spec by a user — and openly sold” (Wakelee-Lynch 2023). This ease of access to this tool that helps produce images with no effort has led to an overall negative outlook on the future of artists, and a lot of these artists that are just starting out are thinking about leaving the field altogether, dreading that they might not be needed in the future.

Another big concern related to artificial intelligence is its violation of copyright laws and its theft of intellectual property, which affects every kind of creative that AI models try to replicate. (Writing, Music, Photography and Art.) When these models are trained to create similar outputs to the inputs that they are given, the machines are given a huge number of images that owners of these AI companies have admitted to being under copyright. “OpenAI . . . told the British Parliament . . . that its content-generating ChatGPT product would be impossible to create without the company's use of human-created copyrighted work for free” (Hodge 2024). Lawsuits have already been filed against AI image companies, most notably DeviantArt, Stability AI and Midjourney.

Some more well-known artists on the internet are also specifically targeted by malicious AI users due to their recognizable and likable art style. These ill-intentioned people instruct AI models to specifically base images on art made by the well-known artists, which produces images with a style resembling the one of the targeted artist. This is extremely problematic since if the produced image is convincing enough, it can be sold by people, which takes away potential sales from artists that worked hard on their art. This is even more concerning when considering artists that take commissions for their work (artists letting people ask for art depicting specific characters or scenes), since having the ability to produce art in the style of an artist would take away some people’s willingness to get a real commissioned piece from the artist. Artists taking these commissions from the public sometimes heavily rely on the income from people that want art made for them.