Is A.I. the Death of Art? Or the Future of Creativity?

Aron Brand
4 min readSep 3, 2022

--

Homer Simpson as a Jedi Knight, Aron Brand, DALLE-2 outpainting based on multiple textual prompts and some manual finishing with Photoshop

“Art is dead, dude.” This was the claim made by Jason Allen, creator of an A.I.-generated image that won a prize at the Colorado State Fair’s annual art competition.

But is A.I. really the death knell for human artistry?

It’s true that A.I.-generated art has come a long way in recent years. Thanks to tools like Midjourney, DALLE2 and Stable Diffusion, anyone can create complex, breathtakingly beautiful and realistic images simply by typing a few words into a text box. But while these tools may have made some human artists nervous about their own futures, they have also generated fierce debates about the ethics of A.I.-generated art.

Some opponents of A.I. art are concerned that it would devalue the work of the professional artists . After all, like any commodity, of the value of art is derived from its supply and demand, and if the creation of art is made easy and accessible to everyone, the supply would grow exponentially, causing the market value of works of art to plummet. It is already hard for artists to make a living, and if A.I. art became widespread, the opponents believe, it would only become harder.

Controversy over new art-making technologies is nothing new. Many painters were extremely opposed to the invention of the camera, which they saw as a threat to human artistry. But the camera didn’t replace painters, it simply created a new medium for them to explore. And the same is true of A.I.-generated art.

While the initial iterations of image synth tools were limited to what artists could do tell them to do in one shot, these tools are rapidly being enhanced to increase creative control and enable working in iterations. These creative controls are restoring the the importance of the artist in the creative process.

Let’s talk a bit about these new capabilities and the freedoms they allow for the artist:

Inpainting is a process where missing parts or “holes” in an artwork are filled in by AI to present a complete image, typically providing a few options for completions based on a textual prompt. The human artist can take an image generated by AI or by other methods , deliberately cut holes , and allow the AI to dream up a number of completion options for the artist to select based on their creative vision. This process enables creative control because the artist can iterate on existing image, based on the overall concept they are trying to achieve.

Outpainting is the complement of inpainting, where the artist provides an existing image, and asks the AI to extend it beyond its original borders, imagining the missing contents based on a textual prompt. “Uncropping” if you wish. Have you ever wondered what happened outside of the frame of “Mona Lisa” or “The Last Supper?” Well, with outpainting, you can now easily generate imaginative artistic answers to those questions, perfectly matching the style of the original image, based on your own artistic ideas.

Image to Image (img2img) is a term used to describe the technique of generating images from other images, guided by text. The source image can be the output of an image synthesizer, an existing artwork, or even a rough sketch. This process enables artists to create better art faster by allowing them to work in iterations and experiment with different ideas quickly. It is a bit like the workflow of traditional artists who begin with a sketch and then refine it until they are happy with the final result. The main difference is that with img2img, the artist can make changes to the source image and see the results in real-time. This technology is creating a somewhat amusing renaissance of MS paint artists who are able to convert their doodles into fully fleshed, professional works of art.

If you are watching the AI art scene , you will notice that the best pieces of AI art are now created as a sort of dialog between a human and the machine. Far from being a threat to human artists, AI image synthesis is actually about augmenting them. It is a new tool that they can use to explore the world of image-making in new and exciting ways. And for the foreseeable future, it is not going to replace human artists, but simply provide them with another way to create beautiful art.

Art is not dead. It is, however, evolving.

A brutalist living habitat on mars, Aron Brand, DALLE-2 outpainting + inpainting, multiple textual prompts
If you found this article interesting, read more about this topic in my previous article

--

--