All Artists should understand AI Art
The topic of AI-generated art is generating a lot of discussion in the artist community on websites like ArtStation and DeviantART. Many artists are questioning the legitimacy of the art generated by platforms such as Lexica, Midjourney, and Stable Diffusion, and expressing concerns about the impact it could have on their own work. Some of these concerns include the possibility of AI-generated art taking away job opportunities for artists, as well as concerns about intellectual property and ownership of the art produced by these systems. These discussions highlight the importance of considering the ethical and moral implications of using AI-generated art and ensuring that artists are protected in this changing landscape.
Let’s take a moment to reflect on the fascinating and ever-evolving field of artificial intelligence.
In the 1950s, British mathematician and computer pioneer Alan Turing proposed the idea of creating machines or computers that could emulate human behavior, learn from their experiences, and make decisions. Following this, a group of researchers led by John McCarthy at Dartmouth College coined the term “artificial intelligence” and held a workshop titled “The Dartmouth Summer Research Project on Artificial Intelligence” in 1956 to explore the possibility of creating intelligent machines. This event is widely regarded as the birth of AI as a field of study. During the 1960s and 1970s, AI researchers made significant progress in developing rule-based systems that could reason and solve problems. However, it soon became evident that these systems had their limitations in handling complex real-world situations. To overcome these limitations, researchers began to develop machine learning algorithms that could learn from data and improve over time. Thanks to powerful computers and vast datasets, AI has since advanced rapidly, leading to breakthroughs in areas such as natural language processing, AI art generation, and computer vision.
In 2023, AI is everywhere, with a rapidly expanding range of applications. From chatbots and schedule planners to code writers and AI art generators, AI is transforming the way we live and work. In almost every field imaginable, AI is being used to improve job performance, streamline workflows, and optimize decision-making. While many view AI as a valuable tool for driving progress and innovation, others express concerns about the impact of AI on employment and social inequality. As AI continues to advance and reshape our world, it’s important to remain mindful of these implications and work to ensure that the benefits of AI are accessible to all.
Some artists dislike AI-generated art because they feel cheated by its use. They have spent years honing their skills to reach a professional level, only to have a program create similar quality work in just a few minutes. The existence of generative AI also signals to aspiring artists that their chances of getting paid for commissions before reaching a professional level will be reduced. However, artists feeling inadequate in the face of new technology is nothing new. In fact, many art mediums have undergone shifts due to the availability of new technology.
During my college years, traditional artists disliked programs like Photoshop due to the digital nature of the art it produced. They felt that using a program to edit photos or to paint digitally was cheating them out of the hard work they put into developing their skills. If we look even further back, landscape painters were outraged by photography because a machine with one button could create a perfect image in minutes, compared to the hours it would take a painter to make the same image.
Another major reason for the outrage over AI-generated art is that most platforms use trained models from data scraping, which some view as theft or harvesting.
So how does it work?
Generative AI is created by analyzing existing works of art and identifying visual patterns that can be used to generate new, unique pieces. This is often accomplished through the use of neural networks, which are designed to mimic the way the human brain processes and interprets information. Art pieces are improved upon by adding more information to the data sets and repeating the rendering process. Many people think that generative AI is done by copying and pasting parts of images together in the form of a collage, but that would be inaccurate.
These generative artworks are actually closer to the way human artists will look at many pieces, identify elements or patterns they want to emulate, and then use those elements to create a new piece of art, improving the quality with each repetition. This method is how the majority of artists learn to develop their art skills. The problem is that AI can develop these skills faster and more efficiently than a human can, and that’s the real issue at hand.
When I was in school for fine arts, the teachers taught us the “steal like an artist” method. We created a mood board to emulate 5–10 chosen pieces as a single cohesive vision, which we would take credit for (without mentioning or crediting the artists that “inspired” the new piece of art). This method is widespread across the art industry. I have never met an artist that hasn’t taken influences or learned from other artists. It’s easy to see that AI can “steal like an artist” better than humans ever can.
Ownership of AI Art
AI generative art cannot be copyrighted, and this issue has sparked the need for new laws in this area. In 2011, at a wildlife preserve in Indonesia, a macaque named Naruto grabbed a wildlife photographer’s camera and snapped a “selfie” of himself. The camera’s owner, David Slater, published the photo, raising questions about the ownership of art. Animal rights group PETA sued Slater on behalf of Naruto. Later, both sides asked the 9th U.S. Circuit Court of Appeals to dismiss the case and throw out a lower court decision that said animals cannot own copyrights. Because of the case’s significance, the US Copyright Office has listed the monkey’s selfie as an example of a work of art that cannot be copyrighted because Naruto is a “nonhuman entity.” This case helped clarify what deems a work of art as copyrightable.
Works generated by AI cannot be copyrighted without a minimal level of human editing, interference, or collaboration. It’s best to think of AI art in terms of a stock photo. If you purchase a photo from a stock website, you wouldn’t be able to copyright it outright, but you could if it were a heavily modified element in a larger work of art.
What are the ethics of using generative AI in art?
Plagiarism and art theft are serious accusations that are being raised in the art community, and they should not be taken lightly. While the law would argue that every generated piece of AI art is completely original due to the method of creation and the way AI functions, there are ethical considerations that must be taken into account.
From an ethical standpoint, artists should have the right to opt-out of having their work included in the data sets that are used to train AI models. The data sets used to train generative AI systems should include art submitted by artists, works in the public domain, and free/paid stock imagery. There is enough material available in these three areas to efficiently train AI generative systems. AI-generated art that has undergone significant human modification should not be labeled as such, as it has transcended the limitations of its medium. Any work directly generated by an AI platform that cannot be copyrighted should be labeled as AI-generated.
Personally, I would like to see platforms such as Midjourney and Lexica retrained using ethically sourced materials and unbiased training models. Unfortunately, the odds of that happening are slim. There will come a time when AI generative art is normalized, and the loud outcry in support of human artists against AI art will become a whisper.
It’s also important to consider regulations that limit the usage of AI-generated art in areas such as revenge pornography, deep fakes, and negative political campaigns. These regulations should be put in place to ensure that the usage of AI art remains ethical and that the art community is protected from malicious intent. I look to technology safety organizations including the XRSI for regulatory guidance on the matter.
If you find this newsletter useful, share or tag a friend.
Got questions? You can DM me directly on Twitter