Stop the AI, stop the steal!

There has been a lot of discussion, perhaps even drama in social media recently on the topic of Artificial Intelligence versus The Artist. Writers and copy editors openly mainly lamented the advent of OpenAI’s GPT-3, while illustrators, painters and other visual artists seemed to focus on whatever Stable Diffusion app-du-jour was popular that day. In a recent blog post, Seth Godin managed to very elegantly sum up his thoughts on GPT-3, mirroring my own sentiments one-to-one. Give it a read. His whole post is barely longer than just my intro paragraph. I am meanwhile going to focus on the topic I am much more versed in – visual arts.

In our field, a recent recurring theme in the anti-AI conversation seems to be: AI models are trained on artworks without artists’ consent and they steal their art style. This argument just does not sit right with me, so without going too deep into the unending What is Art debacle, here’s my take on the subject:

Very few of us learned anything in an entirely original way. Most artists started by copying others who inspired them – before finding their own style and voice. Throughout ages, great artists did not seem to mind when thousands of students had access to their body of work. There was comfort in knowing that learning to imitate someone required years of dedicated training.

With machine learning, the time to learn how to emulate one’s style got all but obliterated. We can now teach an AI to render works resembling one’s style well enough to fool most untrained audiences in the matter of hours, if not minutes.

It may sound scary, but if you look closer, you’ll realize that the AI is not copying anyone’s art. It is instead, imitating their craft.

Prompted skillfully, AI will be able to produce an image that visually resembles Banksy’s stencil graffiti aesthetic. It will however, not be able to infuse it with their razor-sharp wit and social commentary skill.

It will copy the line quality, shapes and colors of Chris Ware’s panels, but won’t be able to evoke the same bone-chilling melancholy and beauty found in the mundane everyday routines.

It may imitate Van Gogh’s brushstrokes and technique, but so did countless humans who had over a century to learn his style – and yet his works remain unique and original.

Every new technology brings change. Change is scary. Change requires us to unlearn things we took for granted, rethink, and adapt.

Just like the advent of photography made artists question their pursuit of realism and eventually gave birth to modern art – we too must understand the new reality and find a new, better place within it.

Back in the 90’s, in the long bygone era of Web 1.0, I had countless conversations with clients I was designing websites for. The topic was always the same: We do not want to show any information about our product, because the competition will see it and copy it.

Feels like I am reliving the same story again, some 25 years later.

Regardless of my personal feelings, I expect some universal #noAI meta tag will become commonplace in the near future, allowing authors to ban indexing by learning algorithms. It will most likely blow up in popularity, then become a meme, and finally die a natural death, becoming a funny anecdote from the early 2020’s.

And the artists? The ones who try to compete with AI’s speed without outsmarting the idea behind the image will need to rethink their life choices. Meanwhile, the great ones will just keep doing their thing – because it really never was about how well and how fast you can put brushstrokes on the canvas or assign colors to pixels in the first place.

Cover illustration by Quba Michalski × MidJourney × Photoshop

professional photograph of unhappy pouty face handsome young artist sitting on a stool next to the aisle in a painter studio, the artist refuses to paint their half-finished painting, cold snowy winter outside the warehouse window, extremely realistic, moody lighting, 4k, detailed --ar 3:2 --v 4