The rise of AI art has brought about a cataclysm in every creative field like acting, music and of course, visual art. Articles with artist screeds against AI adoption abound. Social media feeds are filled with people posting AI generated art with scornful replies from the people who despise it. Responses from artists range from absolute horror and rage to enthusiasm and integration into artist workflow. Let's look at what AI art is, the advantages/disadvantages of both, ethical concerns, why #notoai artists are up in arms, and what the future looks like for a world where AI art is here to stay.
AI Art: An Introduction
AI art is any type of creative expression utilizing artificial intelligence technologies like algorithms and machine learning models. Some of these include several types of Large Language Models (LLMs) and other AI models, like Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and Transformer-based models.
GANs are a combination of two neural networks: a generator and a discriminator – that work together to produce AI art. The generator creates images, while the discriminator evaluates them against real images, constantly refining and redefining its generative process.
VAEs, like DALL-E, compresses an input image to summarize it into latent space, a few important features or patterns. It can decode the input back into a full image. But it can also mix and match different parts of the latent space to create entirely new images, which make it a good choice for a collection of images based on the same input.
There are other types of LLMs used to generate AI art, as well. Transformer-based models, like CLIP, pair images with text for accurate image generation. Neural Style Transfer applies the style of one image to another using convolutional neural networks. Text-to-Image models like Midjourney and DALL-E combine some of these approaches and draw on huge datasets to create images from textual descriptions.
That last bit is why many artists are freaking out about AI art.
Why The Fuss? The Human Perspective
X (formerly known as Twitter) is rank with artists who are furious about the current prevalence of AI art, some to the point of demanding boycotts of anyone using it.
It may seem to some as an extreme position, but adverse outcomes of AI technologies are one of the top long-term risks facing the world, according to the 2024 World Economic Forum Global Risks Report.
And the dangers of AI were a serious subject of conversation at this year's G7 summit, with even Pope Francis warning about its imminent dangers.
Visual artists have a host of concerns specific to their industry like IP theft, job displacement, devaluation of art and more.
Many AI-generated art tools are trained using copyrighted works without consent that are found all over the Internet. That means artists aren't getting paid anything when AI art is generated using their work. The financial implications AI will inevitably have on their job prospects is of paramount concern.
AI produces art quickly and cheaply. An anonymous project manager was quoted in our article about why employers could be all too happy to replace human artists and graphic designers with AI generative art apps.
"I wanted to support NFT artists and other artists who weren't using AI because of the conversation about copyrights and AI putting artists out of business. But it was impossible. They either wanted crazy amounts of money, did awful work, or just didn't deliver. I ended up generating all of the art in Midjourney." - Anonymous
AI Art and the Law: It's Just Getting Started
Things came to a head in January 2023, when several artists filed a class-action lawsuit against some of the biggest AI art generator apps in the industry. These artists based their lawsuit on the premise that AI-generated art infringes on their creative rights and will harm their careers because these apps make it easy and cheap for users to replicate their artistic styles.
But the court's decision largely sided with the AI companies, dismissing most of the claims, except one. The judge allowed the core claim of direct copyright infringement by Stability AI to proceed to trial. This claim focused on the alleged use of one artist, Sarah Andersen's, registered works to train the AI model Stable Diffusion without permission.
Fair Use and Copyright Issues
An interesting doctrine that came into play in the art world long before AI art ever reared its multifaceted head is fair use. Fair use allows limited use of copyrighted material without permission from the copyright holder, provided the use is transformative and serves a different purpose from the original work.
One artist who squeezed every drop he could out of the fair use protocol was Andy Warhol. Most of what he did was flagrantly based on copyrighted photographs of celebrities and everyday objects. He was lauded as a genius and got away with it, for the most part.
But his Warhol Foundation just lost a case heard in the Supreme Court of the United States.
Warhol created a series of 14 silkscreen prints and two pencil drawings in 1984, each based on an unpublished image photographer Lynn Goldsmith took of Prince in 1981 on assignment for Newsweek. She'd been paid $400 for one-time use of her image as an “artist reference” by the Warhol Foundation.
But when she saw an image based on her photo on the cover of Vanity Fair in 2016, it soon came to light that the Warhol Foundation had been paid $10,000 for its use. This was a, er… princely sum she saw no kickback from.
So, she sued the Andy Warhol Foundation for copyright infringement and won.
“The use of a copyrighted work may nevertheless be fair if, among other things, the use has a purpose and character that is sufficiently distinct from the original.”
“In this case, however, Goldsmith’s original photograph of Prince, and AWF’s copying use of that photograph in an image licensed to a special edition magazine devoted to Prince, share substantially the same purpose, and the use is of a commercial nature.” - Supreme Court Majority Opinion
These two cases, although one had nothing to do with AI at the time, are illustrative of the necessity to come up with legal frameworks that address the needs of artists in the era of AI art. AI tech is basically steamrolling the art world, but the international legal system is no stranger to the treatment of intellectual property rights. Integration of former IP precedent while taking the new AI tech reality into account is definitely going to break a lot of new ground within the legal sphere.
The Future of AI and Human Art
Many artists don't eschew AI completely. They use AI tools to enhance and complement their skill set, whether by using it for "grunt tasks" or by implementing AI generated media and tools into their own artworks in their own creative processes. Artists can leverage AI to handle repetitive tasks or generate initial concepts to take inspiration from. By increasing productivity, enabling new levels of creativity and expanding artistic boundaries, AI certainly has its place in the art of the future.
In Conclusion
But the ongoing battle between artists and AI was incited as a response to the serious challenges that AI poses to the creative industry. We need to figure out a way where the potential of AI is used to benefit artists, instead of destroying their careers. The thought of an artist being put out of work because others are using their art to force them into irrelevance is appalling.
The future of art lies in the harmonious collaboration between human imagination and the breakneck advancements in AI technology that show no sign of slowing down. The fact remains that even with its obvious dangers, AI offers exciting new opportunities for artistic expression. And the love, emotion and skill it takes for a talented human to create art can never be replaced.
.png)
.png)
.png)
