In late August, a man won first place in a state art competition for a piece he created using a text-to-image generator that uses artificial intelligence (AI), reigniting debates not only on who should be considered an artist and what should be considered art but also longstanding fears that AI will destroy most jobs. While art critics are free to debate the merits of any particular works, AI text-to-image generators are valuable art tools that increase accessibility and lower the barriers to entry for creatives—they aren’t AI boogeymen destroying creative jobs.
The AI text-to-image generator is a new avenue for creating art that has stirred up heated opposition. Some detractors argue that those using these tools aren’t real artists because they do not put in enough effort—as they understand the term—to produce the works, while others argue that pictures produced by AI text-to-image generators shouldn’t be considered art because they’re not drawn or painted by humans. Others seemingly accept that the results are art and see these results as evidence that machines will soon surpass human creativity. As one Twitter user lamented: “We’re watching the death of artistry unfold right before our eyes — if creative jobs aren’t safe from machines, then even high-skilled jobs are in danger of becoming obsolete. What will we have then?”
If these arguments seem familiar, it’s because they are. When photography first came on the scene in the 19th century, arguments arose about whether the process should be considered an art or a science. Charles Baudelaire—a French poet, essayist, and art critic—famously wrote in 1859 that photography was a refuge for failed painters “with too little talent, or too lazy to complete their studies,” categorizing it as a craze that had contributed to the “impoverishment of French artistic genius.” Now that photography is widely accepted as an art form, Baudelaire’s essay reads as hyperbolic and Ludditesque.
The democratization of photography in the 20th century made creating a new form of art more accessible to the masses, it did not put millions of artists out of work. High-quality cameras have become significantly more affordable—with some artists exhibiting photos taken on iPhones in galleries—but talent and vision still matter. Similarly, the emergence of AI text-to-image generators have lower barriers to creating—they have not eliminated the need for artists. Publicly available generators like Midjourney and Stable Diffusion’s DreamStudio allow users to generate images by typing in word prompts and to modify, download, and share those images. Anyone with Internet access can start creating images with such generators, and increased accessibility means more opportunities for people who, without this tool, would not be able to share their creative visions—this can include people with physical disabilities; people who cannot afford traditional art supplies, camera equipment, or a smart tablet; and people who do not have access to art school or art classes.
This new tool doesn’t erase the creativity or artistry involved in AI text-to-image generation. Individuals can use a discerning eye when implementing combinations of words and modifications to bring what they envision to fruition. Think about photography: even though a camera is the tool that generates the photo, artistry stems from the choices of the photographer, who makes adjustments in lighting, angles, filters post-production editing, and other factors, not to mention a choice of subject and equipment. In the case of AI text-to-image generators, artistry stems from how users choose to adjust different factors, including the words used in the prompt, what order they’re placed in and how they’re separated from one another with punctuation marks, the specificity of the prompt, what modifications to apply, and post-production editing using other software.
This isn’t to say debates surrounding this new tool aren’t warranted—deep discussions, especially among users of the tool, can help steer its development as an art form. The art world is still at the beginning stages of learning what could be considered AI text-to-image art generation “mastery;” early adopters of this tool are in the midst of pioneering norms, styles, and taste. But while these types of discussions have merit, there is no need to fear a world with AI-created art.
Widespread concerns about emerging technologies, especially fears that the technology will lead to negative consequences to society, can create a moral panic. And a moral panic about AI-created art could prompt policymakers to respond with proposals, such as not providing IP protection for AI-based works or imposing impractical requirements on AI systems, that could limit the use of the technology. Not only would such limitations stall the development of the technology, which likely has many valuable commercial applications as well, but these restrictions could ultimately stifle free speech rights and the development of a new and valid art form.
AI created art is still in its early stages, but policymakers should take care neither to stand in the way of artistic creativity or technological innovation.