Since its conception, visual art created by artificial intelligence has egregiously plagiarized original, personal artwork created by humans.
Artificial intelligence (AI) programs such as Stable Diffusion and NovelAI have been using artists’ works without their consent through a learning algorithm. When a user describes to the program what they want their image to become, the code begins to go through thousands of images on the web and learns the characteristics of each and every image. These original works of art are then distorted to create a new piece of work based on the prompter’s request.
The images that AI programs utilize are pulled from social media websites that artists frequent and are often placed in the AI’s database without the consent of the artist. Not even official commercial artwork is safe from being included in the AI’s learning algorithms, no matter that the consent of the copyright holder has not been received. This rampant usage of stolen images, especially commercial photos, could potentially give the companies that originally own these images grounds to sue for copyright.
In addition, artificial intelligence-made art has been used to fuel a bitter fight between experienced artists and AI art prompters. There have been several instances of individuals taking advantage of image-generator programs by using them to “finish” work-in-progress images that artists have posted online. Afterward, the prompter will claim that they are the original artist that created the stolen picture.
But how are these “artists” able to claim that they are just as artistically inclined as their peers? Well, through their prompt to the image-generator program. By selecting words that specify what color and form the prompter wants the subject to be, these newfound “artists” can compose their illustrations.
With these techniques, image generators are an easy way out for anybody who may need visual assets. Companies, for example, would be able to avoid having to compensate human artists for their work, which would save money, as many AI art programs are free and available to everyone, as their programmers have allowed their projects to be a source available to everyone. As a result, with high-end processors and developers striving to improve the engine’s code by the day, doom is being spelled out for today’s artists.
But there is a bright side with the shortcomings of technology coming in full force. Because image-generator programs are coded to “learn” the characteristics of a piece of art, they often directly copy the foundation of the original work. Some of the most prominent characteristics of art created by computers include half-garbled attempts at recreating the signature of the artist; misshapen fingers and toes; and an uncanny blurriness in the background from the program endeavoring to cobble together hundreds of other backgrounds into a quasi-recognizable figure.
But even so, the sudden ability to create works of fully rendered art has given rise to an influx of artificial paintings on social media. So much so that it is sometimes impossible to fish out original works of art. It has become incredibly frustrating for artists that have put time and effort into learning new skills and refining their technique. AI art takes away the genuineness that human artists imprint on their works. Artwork created by an actual person is able to be improved. Even minor adjustments can further advance artists’ approach to their craft. But computer programs cannot do that—at least, not yet. They are stunted by their creators, only coded to misappropriate images.
This rise of artificial intelligence in creative works emphasizes the newfound struggles of the twenty-first century. Programmers are striving to break barriers in art but end up infringing on the rights of actual creative artists. Artificial intelligence has a long way to go, but human art will forever stand at the forefront of evolution.