Why AI-Generated Visuals Can Never Be Art

The subject of what is and is not art has been argued since the early 20th century by many artists and thinkers around the globe. Today, our enquiry does not, as in the past, deal with living artists’ work but with machine-produced products.

Opinions vary, but serious examinations of the fundamentals are rare.

In order to address artificial-intelligence-generated art — AI — we must establish basic definitions. For starters, we note that art is “generated,” whereas human-artist art is “created.” This already puts the results of the two endeavors light years apart.

But to be extra open-minded, let’s work with clearly understandable definitions and confine ourselves to the visual arts — painting in this case — because they are static, hence more directly perceived and understood than, say, music or literature which are processed over time.

AI-generated art: computer generated patterns developed with the assistance of artificial intelligence technologies that often involve machine-learning algorithms, such as neural networks, to generate visual art.

Human-artist created art: An intelligible representation of reality that manifests an artist’s fundamental (conceptual) value system in perceptual (physical) aesthetic form.

Some have called AI-generated patterns a veritable renaissance of artistic creations. Why? Because it’s new?

Let’s take the word renaissance at its core meaning — “rebirth” — and back up to review only a few of the many creatively “new” achievements of the quattrocento renaissance.

One of the most important was the understanding of linear perspective, which allowed painters to create the realistic illusion of depth and three-dimensional space on a two-dimensional surface (canvas, wood panel, etc.). This was a visual geometric achievement by living, breathing artists.

Another: Because of the discovery of ancient Greek philosophical manuscripts — especially those of Plato and Aristotle — via Middle-east scholars and Byzantine refugees, artists during that period were introduced to new ideas regarding the human condition that had been unavailable not only by their absence but also by the doctrinaire dictates of the Catholic Church; thus, other-worldly mysticism came to be challenged by the earth-bound faculties of free will, reason, and individual agency, all of which engendered secular humanistic rather than religious art.

Artists also began to reference classical mythology and literature instead of Christian religious stories in their subject matter as well as to portray ideal beauty based on newly understood ratios and proportions. So, this period was truly a “rebirth” of ideas, innovations, and artistic creations.

And we need only name a few of the many ground-breaking artists whose skill and imagination changed the western artworld forever: Leonardo da Vinci, Botticelli, Michelangelo, and Raphael come quickly to mind.

Now, let’s examine the innovations that produce AI-generated art and wonder at just what rebirth of which ideas they might be.

Humans did invent the technologies that produce the “art” as well as program the content that the technologies can utilize when rendering the visual patterns that are now arranged or re-arranged by users to produce new and colorful images. But content in a human-artist created painting is the meaning — the soul — of a work.

That meaning is discovered in the answers to such questions as: What’s it all about, and why do I respond to it so emotionally, loving some paintings and hating others? With AI-generated art, however, large sets of such meaningful already-made-human-created art are used to teach the algorithms how to replicate patterns and styles that can be used to generate its art.

So, at the outset, we need to note that AI-generated art cannot be produced without using human-artist-created-art to find the patterns it will use in its own set of possible visual combinations.

Next, we need to examine content. Human artists consciously or unconsciously imbue their art with their own values. Values are mental concepts that estimate and integrate ideas judged — consciously or unconsciously, rationally or irrationally, useful or destructive — to be most advantageous for them to live a rewarding life of satisfaction and happiness. And when we respond emotionally (whether positively or negatively) to this painting or that, it is because the artist’s values and our own are likeminded ... or not.

Similar to abstract paintings created by human-artists, AI-generated paintings can be decoratively pleasing or ugly, beautiful or disgusting aesthetically. But they cannot have any intrinsic meaning.

Some of AI-generated paintings can be fanciful or imaginative or allude to realistic images that relate to the real world, but all of the possible combinations are available only from what is technologically available. By contrast, human-artists who paint from nature and human nature have the entire world, all of humanity, and their own values to depict in their art.

Also, we must remember that human-artist painters manually apply oil paint, watercolors, or pastels to a (usually) flat surface with their own tools and hands. AI users, however, render their images on a high-tech screen monitor and then transfer their selected arrangement of images to a flat surface. So, the processes are not only mentally reversed but physically as well.

What technology has made possible in this advanced world of ours is stunning, and stunningly both good and bad in potential. But are AI-generated pattern-arranged images transferred to a physical surface art?

The galleries and auction houses happily shout “Aye” as they promote and sell such objects. And the “artist” processors who arrange pre-set patterns into fixed images are even louder with their “Ayes.”

Money is made. But are these objects really “art”? Anyone who understands, appreciates, and responds to the skills, artistry, beauty, and meaningful depth of fine human-artist-created art will simply yawn and respond with a simple “Naw.”