Why Programmers and Artists Can’t Agree on AI

I do not need to tell anyone here that, computer science and visual arts are not the same – at all. There isn’t much overlap in the core concepts that fuel art and the fundamentals of computer programming. Each field has a unique history and knowledge base that’s sent along to the next group of young computer nerds and art kids.

Actually, I would take it further and say that the two disciplines are about as opposite as they could be. And you may know what I mean if you’ve ever tried to buy a Raspberry Pi (a tiny computer used for projects), raspberry-colored paint, and a pint of raspberries for your raspberry pie. Guaranteed, you will have to go to three different places. Shopping aside, there is an even bigger chasm between art and CS: their over-reaching ethos are unbridgeable.

Computer science is about sharing, open source, re-use. Art runs on creativity, uniqueness, and individuality. Art has to be one-of-a-kind. The difference between a Campbell’s soup can and a Warhol (which can sell for millions) is Andy Warhol, a real one-of-a-kind chap.

We’re all used to the idea that the value of art is based on who made the art, and whether it’s original and exclusive. So, it’s hard to believe everything we create isn’t like this, that people make things to be stolen, reproduced, and distributed without credit. But many people do just this. Well-known software, like Chrome, Firefox, Linux OS (and our fav tiny computer, Raspberry Pi), is open-source. Interestingly the founder of Raspberry Pi, Ebon Upton, isn’t even mentioned on the organization’s website. The identity of the original creators means nearly nothing to us.

Sharing things you make without the need for explicit recognition is really a very lovely idea – caring is sharing, as they say. However, “open-source” doesn’t make sense for artists. Because when we take the artist out of their work, Andy Warhol’s soup-can prints become just enamel paint and canvas. And in this way, the advent of AI seems especially dangerous for art. This idealism of computer programmers is the existential threat that artists fear.

If anyone can click a button and get images that take artists days, weeks, and months to create – why would anyone invest in artists? They can make art themselves and if they come from an “open-source sharing-is-caring” perspective they may not even see what all the bellyaching is about.

And as far as those pushing AI art on the world, they might also see their own crisis on the horizon. If protections for intellectual property (like paintings, writings, etc.) are strengthened, AI will be doomed to forever produce the same old stuff. Each generation of images becoming more and more inbred and… ugly. So insisting on intellectual property protections, the individual ownership of works is likewise a threat to one of the ideologies underpinning CS: the open exchange of ideas. It seems like art and AI are locked in an adversarial relationship, doomed to constantly push and pull at each other’s boundaries.

Somewhere on Instagram someone is typing “AI devours art” and a Redditor is preparing to reply “Gatekeeping art kills reason”. Oh dear, this is all starting to sound pretty bad.

But is art dead?

A long time ago, in another century, I briefly attended an art college. There were two types of students, the fine arts students and the corporate students; both were willing to starve.

One of the biggest fears about artificial intelligence is its perceived impact on artists’ income. If AI does it cheaper, making art won’t be profitable anymore. You may already know this fact, but art has never been particularly profitable.

Fortunately, making art was never about the money. Parents have, from time immemorial, begged their kids not to starve for their art, to go to a real school and get a job that pays. Yet every fall art students with paint palettes under their hopeful arms, set out to attend their first classes at [insert any state] College of Art and Design. If commercial success was always so tenuous for artists, has the emergence of AI really changed things all that much?

The problem of monetizing art has been a stick in the art craw since forever. Rembrandt van Rijn, who you may know for being “considered one of the greatest visual artists in the history of art” died broke (comically high praise from wiki). Rembrandt isn’t the only penniless painter either. Artylist online magazine, has a top ten list. Redbbuble’s blog points out some of these unlucky saps die alone, too.

It seems like art goes hand in hand with a bad luck story. Yet, it still means something to us. Rembrandt died in the sixteen hundreds and art students are still learning about his use of light, “smooth manner”, and “fine technique.” (wiki’s smitten again). We still debate whether Caravaggio painted a meaner chiaroscuro, or if depictions of form should even be “well-formed” (think nonconformist cubism where everything is unformed, transformed, and then reformed but maybe not “well-formed”). Many hundreds of years after Rembrandt died, we’re still talking about art and whether all this art stuff even matters at all.

This strange behavior of creating, critiquing, and hating on works of art repeats all the way up to today, regardless of new tech. Michelangelo painted realism in the fifteen hundreds. Renaissance Art’s realism survived impressionism, cubism, dadaism, film photography, and even the iPhone. Maybe people believe in the power of machine learning but art will survive the inbred, oversaturated AI aesthetic. Art will persevere because humans like it and they will continue to make it and appreciate it, even if it isn’t profitable.

So, why is art going to survive the tech age?

Humans can’t help but “to art”. People are moved to love art, not because it’s useful or valuable in USD but because there is just something about it. There is a solace in self-expression, a catharsis to visually re-representing the physical and intangible things around us. We’re all full of thoughts, emotions, and observations. Whether we are viewing or creating art, the resulting object is communicating something bigger than words. And no computer can convincingly express itself in this way. Artistic expression is human and AI will always suck at being human.