The Value of Art in the Age of AI
The Value of Art in the Age of AI
Why AI won't replace artists
I’m Tim Gorichanaz, and this is Ports, a newsletter about design and ethics. You’ll find this week’s article below, followed by Ports of Call, links to things I’ve been reading and pondering this week.
As more and more processes become automated, it may seem like we’re on a one-way track to full automation. Next stop: serving our robot overlords.
The current AI craze is feeding this impression. Along with proclamations of the inevitability of certain technologies, it’s powered by the assumption that eternal exponential growth is a given. You can find any number of articles with “exponential growth of AI” in the headline, some of which are optimistic about economic growth and others of which paint doomsday scenarios. But rarely does anyone acknowledge that nothing grows forever. And signs are already emerging that AI is not an exception.
Even imagining that some new innovation kickstarts AI growth again, I don’t think full automation is in our future. Even if it’s technologically possible, it’s not humanly acceptable. In a previous post, I wrote, “Full automation would take away too much of what it means to be human. We wouldn’t accept it. Being human means being inefficient.” And today I want to expand on that idea, focusing on AI and art.
Work and Meaning
One of the fears behind our concerns about automation is job loss. Will AI take my job? The Economist published an article recently on why your job is safe from AI for now. In short, regulatory and market pressures will not allow such an AI revolution. And past revolutions, such as the Industrial one, came about not from a single breakthrough technology but from a simmering cultural stew—and over many decades.
Another fear is deeper, going right down to the question of what makes life worth living. As I’ve written about previously here on Ports, countless philosophers have argued that meaning is a fundamental component of a life well lived, and they’ve theorized on what meaning entails. Susan Wolf, for example, has written that meaning has three components: objective value, subjective satisfaction, and active engagement.
When we’re not philosophizing, we tend to focus only on the objective value of tasks: what product is being created, what work has been accomplished. Looking only at objective value, it would seem that pretty much any work might as well be done by an automated tool. But the other components of activities are just as important for living a good life.
That’s particularly evident in places where work and meaning go together, such as for professional artists. By definition, professional artists make a living from their artwork—they sell it. If all they are selling is an end product with objective value, then in principle it should make no difference if the product is made through automation. But then what is the artist for?
“An Insult to Life Itself”
Hayao Miyazaki is a fabled Japanese animator whose final film is set to premiere this July. (Like Jay-Z, Miyazaki has retired a few times, but given his age and the years it takes to make each film, this one probably really will be his last.)
Miyazaki's life and work offer a strong statement about meaning in life and art as a profession.
First, he shows us the difference between meaning and happiness. “I don’t ever feel happy in my daily life,” he said in the 2013 documentary Kingdom of Dreams and Madness. “Filmmaking only brings suffering. I can’t believe I actually want to do another one.” For Miyazaki, filmmaking does not bring happiness, but it does bring meaning, and the meaning is overriding.
Next, I have been reflecting on another quote of his. A few years back, some of Miyazaki’s coworkers showed him a snipped of AI-generated animation. “I strongly feel that this is an insult to life itself,” Miyazaki said.
It would be easy to see this as the knee-jerk reaction of a Luddite with a fixed mindset. But for Miyazaki, drawing and animation are deeply meaningful activities—for him, constitutive of life itself.
No wonder AI has created anxiety among professional artists, such as that expressed in the screenshot above. Now, there are clear methodological problems with the study that was done and the results being presented—but still, these results indicate that these anxieties are a live experience for at least some artists. You could also look at the ongoing Hollywood writers’ strike, which is at least partly in response to AI.
How We Value Artwork
The cultural script at present says we don’t value art at all. Artists need to work as waitstaff just to get by, art classes are cut from curricula to make room for math and computer science, and so on.
But observe for a moment that high on the list of the most expensive things in the world are artworks. A painting by Leonardo da Vinci sold for almost $500 million at auction in 2017. The Koh-i-Noor diamond, which was set in Queen Elizabeth’s crown, is valued at almost $600 million. The investment platform Masterworks claims that the art market has outperformed the S&P 500 for the past 26 years. And around the world we build temples to art called museums that we pay to go inside. You can’t say we don’t value art.
But what exactly do we value about it?
The fear of professional artists when it comes to AI, put in the context of Wolf’s theory of meaning, is that the world only sees the objective value of artwork (the end product) and places no importance on the artist’s subjective satisfaction or active engagement.
But is that the case?
On the contrary, the scientific evidence suggests that people value artworks not just for their objective qualities but also for the human experience that went into making them.
George Newman and Paul Bloom, for instance, report on a set of experiments exploring why people value original artworks more than duplicates, finding that people value objects that result from a unique creative act in which the artist had a high degree of contact with the piece.
It’s the same reason we value autographs and relics. Whether we like it or not, we have an inner theory of contagion: if a person touched an object, then the object has some of that person’s essence. All the more if the person created that object in the first place.
In a later study, Newman and colleagues showed that people perceive art as, in some sense, an extension of its creator—that we consider art as more like a person than an object.
“People tend to report that when the original is destroyed, a molecule-for-molecule copy of a person is not the same individual,” Newman and colleagues write, “while a molecule-for-molecule copy of a hammer is the same individual.” When it comes to one-of-a-kind artworks, the researchers found, they are more like people than hammers.
Art that is fully AI-generated has no person behind it—or it does only in a far-flung, abstract sense. So it’s more like a hammer than a person. That said, if AI comes to be used as a tool (as it probably will), part of an artist’s creative process, then the artist’s creative act and contact will turn it into more valuable art.
So while AI-generated imagery may be sufficient for managerial PowerPoints (we never considered clipart to really be art anyway), it could never replace art. From the consumer’s side, it lacks the unique creative act and contact from the artist; and from the creator’s side, it lacks subjective experience and active engagement for human artists.
That is maybe some unexpected good news.
Ports of Call
An Ethical Image Generating AI? Adobe has launched a beta of its Firefly service, which styles itself as an ethical image generator, in that all its training data was used with consent and compensation. I got into the beta and played around with it a bit. It’ll be interesting to see where this goes.
On Architecture and Feeling: Following up from last week’s discussion of architecture and friendship, this piece from BBC Future explains how architecture affects how we feel, with implications for urban design. Basically, it comes down to visual complexity; as organisms, we did not evolve in an environment of straight lines and glass panes, but in the mossy, jagged natural world. I wonder how we might apply this in digital design.
On Research Publications: A piece from ProPublica suggests that the latest way for high-schoolers to stick out in the college admissions process is to have an academic publication on their CV, and so wealthy parents are putting their kids in paid programs to get just that. One student said, “Nowadays, having a publication is kind of a given.” Apparently at the University of Pennsylvania, one-third of incoming students engaged in research in high school.
A Poem: A really short one about love from Robert Frost