If AI-generated art can trigger the same reward-pathway as human-made art, does that make it “real art”? Or does the psychological value of art lie in the means rather than the ends?
Updated: Sep 5
Take a look at this drawing.
This is a “bust-up commission” of a character in an “anime art style”. Would you say a price around 40$ would be reasonable, under the assumption that an artist actually drew this? Most likely, you would say yes, or at the very least say something about the artist and/or their price, but at least have some good comments towards the process it took to reach this image. However, you must realize that all I had to do to receive this image was type in a simple prompt to ChatGPT, and it generated this image for me in under a minute. ~57 seconds is all it took to generate this image as it took me ~8 seconds to write the prompt and send it to the AI that then took ~49.27 seconds to generate this image for me. It took only a singular minute to find the character I wanted to make an image of, “sketch out an image”, color it, render it, and just generate this image from start to finish.
Now, think of the process it takes for an actual artist, any artist, from Leonardo da Vinci to that one artist on your feed doing commissions, to make that art. The process they have to go through to make that art certainly takes longer than a minute, right? Just like the image above, would you still be willing to give 40$ to someone to simply just generate an image with an AI, knowing how little time it can take to generate something like that? An artist, on the other hand, has to spend hours, days, years, even decades, just perfecting their craft alone, not to mention how long it takes for an artist to create something new and unique every single time. This is what makes art as a whole unique, because no one person can replicate another artist perfectly, everyone has their own way of creating their own “art”. Like an artist who draws, for example, they have to spend a lengthy amount of time just practicing, and even more so refining their skills in art so that they can make something of the same quality and even better.
If you think about it, everything in this world has its process from start to finish, and nothing ever arrives in this world in a finished state. However, with the rise of generative AI in the sector of art, the debate of what counts as “real art” and whether an image generated with the help of AI counts as real art begins to make a rise, especially in newer generations. One could argue that both methods of creating art, by an artist or by AI, are real art, but is it really like that? Is the result all that matters? Clearly it isn’t just the result that matters too, because the way it was created isn’t the only contributing factor to what makes art, art. But then, what does? What differentiates human art from AI “art”? Which one is “real” art? Does the process have anything to do with art made by humans “art”?
First, we must establish that during the process of making art, and other things that indicate you are feeling rewarded, the reward pathway is triggered, and a neurotransmitter starts producing dopamine, mainly in the ventral tegmental area (VTA). The dopamine produced here then travels to other areas of the brain like the hippocampus, amygdala, nucleus accumbens, prefrontal cortex, etc, which then do their own separate things in processing and chasing the “pleasure” that we all know. That is what’s known as a reward pathway. Because of this pathway, activities that aren’t engaging enough may not be able to stimulate the brain enough, which could eventually lead to a disconnect in the activity and boredom, as dopamine is one of the factors that contribute to attention. For an artist, this pathway can be easily triggered during the process of making art, no matter the form. This is also what lets artists continue making, because they enjoy the process thanks to the reward pathway above. An AI, meanwhile, doesn't get any sort of “reward” as it doesn't have anything that can “feel”, it's only meant to generate pixels that make an image that aligns with your prompt.
Second, we also must establish what counts as “real art”. At its core, art… doesn’t have a direct definition. Even if one was to check the definition of “art”, there would be many results. Some try to encompass art as a whole, while others purposefully leave the meaning of art ambiguous and up for interpretation. Take Aristotle, for example, they describe art as a completeness of what nature cannot finish, and the artist as someone who gives us knowledge of nature’s unrealized ends, while Oxford describes art as the use of imagination to express ideas or feelings, usually in painting, drawing, or sculpting. However, no matter what definition you check, they all say the same thing: that art doesn’t have a direct meaning and is up to the interpreter. The process of creating and continuing what nature has already given us, that is what art is.
But then, what about art that is made by generative AI? Does that still count as “real art”? Many would argue that it doesn’t, as the way an AI makes art compared to how humans make art is vastly different. A human would have to do many things to be able to produce a piece of art, while an AI scans millions of images from the internet and large image sources (ImageNet, Shutterstock, Getty Images, etc) for training, which are then scanned for trends in the images used as a database of sorts. Then, the AI will begin to try and piece together what the image should look like based on the given prompt and reference if given for things like artstyle, etc. Compare that to a human who has many things running to help them create and convert their ideas into reality, like the reward pathway earlier and also the time they already invested to immerse themselves in their art, to create something visible that can move other people to find their own drive in creating. There is soul in the process of creating art, while an AI essentially just guesses the next pixel of an image to make an image that follows the prompt given to it.
By this point, we already have a rough establishment of what goes on in the brain when a human makes art by themselves, but, what about using an AI? Depending on who you ask, one might be able to say that writing the prompt itself is an activity that can stimulate the brain and start the reward pathway. However true that may be, the process itself is obviously much more short-lived compared to making art from start to finish. There are a lot of differences when you compare making art with AI and by yourself, like the time it takes to think and write a prompt compared to the time it takes to, for example, think about an idea or concept to how to turn it into something that can touch others. The time it takes to create human art far exceeds the time needed to create an image with AI. You could spend hours redefining and “touching up” a generated image by yelling out commands to the AI, but then, is it really you who is creating the image? An artist, on the other hand, would have to spend however many times longer just to get the basics of their art down, much less finding their own personality and showing it through their art. In an artist's art lies not only the process of them growing their skills, but also things like intention, the desire to create and make something that can be enjoyed by other people that isn’t just the artist alone, and just passion and love for the art itself. That difference alone is more than enough to give the art created by the artist its own voice and meaning, with a soul given from the artist put into their creation, compared to an image generated by AI that is often described as “soulless” and “without meaning” since the AI is the one bringing it to life, and not an actual person.
In conclusion, AI art, though it does have the possibility to trigger the same reward pathway as when making human-made art, doesn’t count as real art. Or at least, not in its entirety. One could argue that the prompt counts as art in the form of writing, but it would be better to call it a command instead, as it doesn’t fall into anything further than a command or a description of what the writer wants the generated image to look like. The brain also works harder and longer during the process of making art, not even including the numerous hours, months, years, even decades spent practicing and honing their skills to be able to continuously create, compared to simply typing the command for the AI to then use and take from other existing images. The value of art lies in the process of making, the passion and determination to continue with making art, and the soul that is visible at the end of it all, and the message that is sent from the artist to the viewer. Art made by humans has meaning, a meaning from the artist for the one viewing, but an AI generated art can only do so much to try and send the same level of meaning in an image, as the only part in the process of making art with AI that has any part of human creativity is creating the prompt for the AI to use.
BIBLIOGRAPHY