Will AI doom art to uniform mediocrity?
JUNE 8, 2024
In the past, I have used AI to help me describe my work for a variety of art shows and contests that expect extensive verbal descriptions of the piece being submitted. I don't understand the reason for this, but it seems like some curators get off on having the artist struggle more to describe a piece than they do creating it. AI is good in that respect because it will say things like 'the abstraction of the anger in the harsh reds that blend into the apparent hidden envy of the greens and depression of the blue splashes highlighted by the promise of hopeful yellows that evoke a promise of a stable and fruitful future that is suggested by the muted focus of whites gives one a sense of both wonder and dread about what the future will bring.'
AI can be pretty helpful in this job because it relieves me of wasting time coming up with word salad descriptions completely on my own. AI descriptions allow me to keep my focus on making art , not writing about it, which is what I want to do anyhow. But now, it has gone one step further, and I really have a problem, as should all artists, with AI preforming the functions of an art critic.
I don't know if you have seen it on Facebook, but an AI art critic called Theobot.AI has been heavily promoted there. This service is being positioned not just as a descriptive tool, but as a tool that evaluates one's work and then delves into ways to make the submitted piece better by providing suggestions that AI thinks will make the work more appealing to others.
Initially, I was a bit offended that a robot would evaluate my work and provide suggestions to make it better in the eyes of the average art collector. But being curious, I decided to experiment with it. And the more I experimented with Theobot.AI, the angrier I got.
In total, I submitted three greatly different abstracts to Theobot.AI as an experiment. Like the AI service I had been using to make suggestions for my word salad submissions, Theobot.AI did a commendable job. But it didn't stop there. This AI service had a lot to say about what it perceived I needed to do to my work to make it more appealing to potential art collectors.
In comparing the three Theobot.AI reviews of my work, this is where things got scary. In all three reviews of vastly different works, Theobot.AI said I needed to provide an easily recognizable object/person in the image to give the viewer something to focus on as well as specific suggestions on creating anchor points within the piece along with suggestions on textures, hues, etc.
What struck me about the three reviews were how similar they were in what AI's expectations were for each work's appeal. So what? It is just a computer algorithm anyhow. But now, put yourself in the busy gallery owner's or contest judges' shoes? Once AI begins to take hold of the art world, it is going to be so tempting for people in positions that decide what art wins awards, gets displayed, and published to start relying on AI to do the judging and curation tasks.
If Theobot.AI is representative of where the art world is headed, then computers , by simply following their algorithms, are going to begin to force uniformity, and ultimately mediocrity on artists who want their works to be seen and appreciated by others. When we get to the point where gallery owners and publications start to rely upon AI to make their curation decisions, it is game over for anyone who dares to explore outside AI's expectation for 'good' curatable-worthy art.
What can artists do about this? I am not certain. But one thing I am going to do from here on out is ask the galleries and contests I submit to if they use AI in making their selections. If they do, then I am just going to decline to participate and do what I can to let other artists know who is using AI instead of doing their jobs as judges and curators. With a little luck, people who use AI in this way will generate so much mediocrity and sameness in the works selected that it will hinder their ability to compete in the art marketplace, thus leading to their demise.
Ultimately, I think AI can be a good thing if used appropriately. I just wish the Teslas and Apples and Robot Dynamics of the world would focus a bit more on doing my dishes, cleaning clothes, and weeding my garden than driving my car, generating targeted advertisements, and shooting bad guys. But being what the world seems to be becoming, AI, when applied to art is going to be just another ugly mess.
The second thing I think artists need to do is make grant providers aware of the threat that AI curation creates in the art community. A lot of galleries rely on federal and state grant dollars to close the gap between sales and expenses. If these galleries, publications, and contests are found to be using AI for curation purposes, then it is my opinion that these establishments and individuals should be made known to those who hold the purse strings so that they can adjust their grant allocations accordingly.
Thanks for reading.
D'Day