I don’t know if you’ve seen my painting “So much to explore” – it was a scene set inside a pine forest, looking towards a late afternoon sun with light filtering between the trunks and branches, illuminating patches of the ground while casting deep shadows over others.
The painting was based on many walks I’d taken up a hill called Isaacs Ridge, not far from where I live, where old plantation pines give way to native forest which in turn gives way to old abandoned sheep pasture that nature is slowly reclaiming. It took a lot of time, not so much the painting but the lived experience that put the idea in my head, then thinking it through, imagining in detail what it was that I wanted to create.
The image you see above is not my painting. This image, which could easily be the work of quite a few skilled contemporary artists I follow on instagram, was created simply by my typing the sentence “Inside a pine forest, late afternoon, looking through the trees, with dappled light coming through the branches casting deep shadows, in a painting style” into a freely available AI engine called Midjourney.
This is one of several AI tools that have become accessible to the public, all in a rush over the last year or two. Midjourney achieved notoriety for generating an image that won the Colorado State Fair arts competition in the Digital Art category, surpassing the work of many human digital artists.
If AI is already this good, this quickly, how much better is it going to become? The mind boggles.
The question is, should we as artists be concerned? Yes. And no.
If you’re a commercial artist, producing, say, artwork for book or music album covers, for computer games, concept sketches for movie scenes or to show how planned building projects will look, then yes.
It will be so very easy for someone to simply type in the prompt of what they’re after, fine tune their request, hone in on the best-fitting image the AI’s producing and get variations of it, and hey presto, there’s the artwork they need. Without having to recruit a human artist, describe what they want to them, wait for them to produce it and then pay them.
I believe that this is the motivation for creating the software. It lets companies get artwork much faster without having to pay anyone. If I were a commercial artist I’d be very, very worried.
I’d be worried as a digital artist selling NFTs as well. Already the space is becoming flooded with people typing prompts into their favourite AI then selling them as their own artwork (or trying to).
As a traditional artist? I’m not worried at all. It’s not hard for an AI to create a digital image, but putting paint on canvas, ink on paper, carving woodcuts, etc – all of that is far more difficult. It would involve sophisticated robotics, and where would the motivation be to create such a thing (other than as a one-off curiosity) and make it available to the masses?
Thing is, no matter how good AI gets, it will never have a soul. There will never be a person behind the artwork it produces, a person with a story, that people can relate to and identify with.
People won’t ever cherish an AI’s idea of a painting in Vincent van Gogh’s style, no matter how accurate it is, because the AI can never be Vincent. We see the depths of Vincent’s character, his struggles, his emotions, his humanity and his profound eye for beauty in his art, and that’s a huge part of why we love it.
A good artist friend and mentor of mine, Tom Zbaren, has often stressed the importance of getting our stories out there, of putting ourselves into our work so that people can see us, not just our art. I believe the advent of AI artwork is going to make that even more important, as AI will push collectors to increasingly seek out that human story and sense of connection.
For digital and commercial artists though, it’s dark times ahead. I can’t see any silver linings there at all, and I wish I could.
