In just a few years, the number of artworks produced by self-proclaimed AI artists has increased dramatically. Some of these works have been sold at staggering prices by major auction houses and have found their way into prestigious, curated collections. Originally spearheaded by a few tech-savvy artists who embraced computer programming as part of their creative process, AI art has recently been embraced by the masses as image generation technology is both more effective and easier to use without programming skills.
The AI art movement rides on the coattails of technological advances in computer vision, a research area dedicated to designing algorithms that can process meaningful visual information. A subclass of computer vision algorithms called generative models is at the center of this story. Generative models are artificial neural networks that can be “trained” on large datasets containing millions of images and learn to encode their statistically salient features. After training, they can create entirely new images not found in the original dataset, often guided by text prompts that explicitly describe the desired outcomes. Until recently, the images produced by this approach were somewhat lacking in coherence or detail, although they possessed an undeniable surrealistic charm that caught the attention of many serious artists. Earlier this year, however, tech company Open AI introduced a new model, nicknamed the DALL E 2, that can produce remarkably consistent and relevant images from virtually any text input. DALL·E 2 can even produce images in certain styles and imitate famous artists quite convincingly, as long as the desired effect is adequately specified in the prompt. A similar tool was made available to the public free of charge under the name Craiyon (formerly “DALL·E mini”).
The coming of age of AI art raises a number of interesting questions, some of which – such as whether AI art is really art and, if so, to what extent it is really made by AI – are not particularly original. These questions echo similar concerns once raised by the invention of photography. By simply pressing a button on a camera, someone with no painting skills could suddenly capture a realistic representation of a scene. Today, a person can push a virtual button to run a generative model and generate images of virtually any scene in any style. But cameras and algorithms don’t make art. people do. AI art is art created by human artists using algorithms as another tool in their creative arsenal. While both technologies have lowered the barrier to entry for artistic creation—which calls for celebration rather than concern—one should not underestimate the amount of skill, talent, and intent that goes into producing interesting works of art.
Like any novel tool, generative models lead to significant changes in the process of art production. AI art in particular expands the multi-faceted notion of curation, further blurring the line between curation and creation.
There are at least three ways art with AI can involve curatorial acts. The first and least original has to do with curating results. Any generative algorithm can generate an unlimited number of images, but not all are typically given artistic status. The process of curating results is very familiar to photographers, some of whom routinely take hundreds or thousands of shots, some of which, if any, are carefully selected for exhibition. Unlike painters and sculptors, photographers and AI artists have to deal with a plethora of (digital) objects, the curation of which is an integral part of the artistic process. In AI research in general, “cherry-picking” particularly good results is considered bad scientific practice, a way of misleadingly inflating a model’s perceived performance. However, when it comes to AI art, cherry picking can be the name of the game. Indeed, the artist’s intentions and artistic sensibilities can be expressed in elevating certain results to the status of works of art.
Secondly, the curation can also take place before any images are generated. While “curation” in art in general refers to the process of selecting existing works for exhibition, in AI research curation colloquially refers to the work that goes into creating a dataset on which to train an artificial neural network . This work is critical because when a dataset is poorly designed, the network often fails to learn how to represent desired functions and function appropriately. Additionally, if a dataset is biased, the network tends to reproduce or even reinforce that bias—including, for example, harmful stereotypes. As the saying goes, “garbage in, garbage out”. The adage also applies to AI art, except that “junk” takes on an aesthetic (and subjective) dimension.
For his work Memories of passers-by I (2018), the German artist Mario Kinglemann, one of the pioneers of AI art, carefully curated a dataset containing thousands of portraits from the 17th to 19th centuries. He then used this dataset to train generative algorithms that could generate an infinite stream of novel portraits with similar aesthetic traits, displayed in real time on two screens (one for female portraits, one for male portraits). This is an example of an AI artwork that does not involve output maintenance. Nevertheless, the careful maintenance of the training data played a fundamental role in the conception. Here “bias” is a blessing: the dataset has been heavily biased according to the artist’s personal aesthetic preferences and tastes, and this aesthetic bias is reflected in the final artwork, albeit through the distorting lens of the computer-controlled generative process.
Another innovation driven by recent advances in generative algorithms is the ability to generate images by describing the desired outcome in natural language. This is known as “prompting,” or running the algorithm using text prompts instead of sampling random output. Consider the illustration attached to this article: The collage contains several images generated by DALL E 2 with the phrases “an AI image generation algorithm, concept art”, “collage with images made by a generative AI model, illustration from Wired magazine” was prompted. and “an artist curating artworks produced with an AI algorithm, conceptual art.”