Edit: December 14.2022

I used to be pretty excited about the new tech until I learned about the unethical scraping of data (not only art, but also private data, and medicinal data of people. etc!)

Tech bros are making bank while artists are left in the dust to starve, some even seem to be overly excited to see them rot because supposedly, making art is not a skill that takes years to train, but a talent that artists are gatekeeping?!  (Which is a cruel and unfair way to see it!)

This is not how new tech should be used!

We need a way to opt out! We need ethically trained models if we are going to use any models at all!

Also, in regard to online galleries:

A REQUIREMENT to TAG AI creations as such should be the MINIMUM of what can be done on this platform! It is unacceptable that people are paying good money here for pro memberships etc so that they can use this as a platform to find jobs for HUMAN-created art, and then they get swamped by galleries with thousands upon thousands of not human-made creations. The playing field is not the same. Even more so since the current way of using the LAION database is NOT ethical

By the way: They have scraped 90% of the art I have ever published online. 

I was never asked if I wanted that. For its original NON COMMERCIAL SCIENTIFIC purpose, I would even be fine with it. But I am not d'accord with the current ways, where tech bros are using OUR hard work against our consent to make big bank and we rot in the dust. Not to mention all the medicinal data that ended up in there. How is this ethical?!

Original post from 5 months ago:

"Can AI-created art serve as a base for creating human-made transformative art?"

Welcome to the future where gatekeeping beautiful art is no longer a throne beheld by the most skilled artists. AI’s have opened up that playing field for everyone.

Will the “traditional (digital) artist” go extinct?

That depends on the question if the big players continue to value art created by human professionals. Only time can tell. 

And until that time has come, I set sail to this exciting new adventure: Human versus machine, artist versus art AI! Do we have to be enemies? Could we be partners? Or even friends? Let’s find out!

Welcome to my experimental challenge! 

My competitor: Midjourney AI.

My tools: Ipad Pro+ Apple Pencil, Procreate 5, iPhone 13 photo camera, Photoshop, Werble animation software

The rules:

I will create an interpretation of a Midjourney-created portrait. (Pinned by Azariy Gorchakov ). His name is all I could find out about him. And that he enjoys pinning thousands of Midjourney-created pictures.

I will deliberately change the style of the portrait, paint it in my “beautiful Instagram girl” style to show how AI art could serve as a basic reference for an own transformative portrait or idea.

Since it takes MJ around 45 seconds to generate a picture, and it takes me 10 minutes alone to make the most basic sketch of her face, I decided to make this a digital artception! 

In Procreate, I will draw the basic sketch, then paint in the basic colors, and then photobash the flowers.

Then, I grab my trusty Iphone, and take a photo (of my art displayed on screen!), and let the camera’s AI create its own interpretation of what my colors and contrasts should be like. The camera also reinterprets the textures for me and sharpens all edges, which gives the art an almost oil-painted feel.  

Now, I throw the art into Photoshop, correct lens distortion from the camera, rework the contrasts, add a little more noise and texture.

And now it is time to animate! 

For that, I bring the art back to my Ipad and fire up Werble. In the end, I export my animated art as a video and an animated gif. The animated gif adds a soft “retro digital art” vibe, which I really enjoy. So this will be my main piece to exhibit.

The finished art: https://www.artstation.com/artwork/8wV2Qw

In the end, it took me 10 hours to create my explorative transformative art piece. I would most likely be faster if my human mind did not need processing time for countless decisions when it comes to colors, composition, what have you.

So if we based this only on the question of who is faster, I would, naturally, have lost this challenge. Big time.

Challenging the AI did not really feel like working “against” (or with, it depends on your perspective) one person. Rather, the AI is like 100, if not thousands of all the best living and dead artists, and all their knowledge COMBINED. Versus you. 

There’s only you and you are all alone in this battle. If you see it like that, you must feel like the one on the losing end. And you will end up being very, very sad. You won’t have a good time during the next 10 hours as you create your own transformative art piece. 

So I decided to change my point of view: What can I learn from the AI? 

How can I benefit from it? I ventured to the community to learn more about how other artists see it. Someone mentioned it was as though they were learning from all these living and dead masters, and it would help them a lot to progress with their own art.

Someone else said they can finally visualize all that is on their mind. Everyone who's ever tried learning how to draw and paint will know this painful struggle - of not being able to get to that point. And finally, they had a tool that could help them express and share their vision. Does it mean less gatekeeping when it comes to expressing one’s vision? Absolutely.

Can I benefit from AI in the same way as they did?

As I work on my version of the portrait, I analyze the reference and try to figure out what makes it work, and what doesn’t, allowing MJ to teach me some of its wisdom.

I notice that MJ “knows” everything about composition and what colors and contrasts make a piece of art attractive.

Many areas that it fills in are vague, though, and work only as a thumbnail. They would need a proper human-made reinterpretation. The faces, specifically, often have strange eyes. Noses often times have noses. (Yes, a nose within a nose!) Necks harbor second faces, creating people with two heads. 

As it stands right now, these things would still require a human to correct and retouch them. I have no doubt, though, that MJ will soon have learned how to “correct” its faces - like Dall-e, for example. 

One of my decisions to “compete” with MJ was by making the face much “prettier”, as I painted it in a different style. The upside is that yes, this face follows current trends and beauty standards. The downside is that it loses its interesting expression and play with light and shadow that one can find in MJ’s version.

Beauty is in the eye of the beholder, and one could say that both versions have their right to exist. They are just different. 

And here we are back to the current notion of the MJ/ AI art community:

The “hive” creates art together through mixing and remixing. On their discord, currently 670,000 members strong, technically everyone there can talk to the open beta of the bot. Ask it to mix or remix your prompts and likewise, you can do the same to the results of other people’s prompts.

Are we art directing together? We, as a collective, easily create hundreds of variants of the same theme and artwork. They are all different yet close enough that I sometimes can’t tell them apart from just looking at the thumbnails. There is a constant flood of them. New art is created in seconds. If you wanted to appreciate it all, you physically couldn’t. You would be too slow to see them all, withheld by your human eye and brain that still needs time to compute. Art as a mass product? Maybe. But is that the focus? Not quite.

Most people enjoy working together. Only a minority is very “protective” of their prompts and do not like to see them remixed. 

And with its most recent changes to its TOS, MJ states that the unique piece belongs to the person who gave the prompt. And if somebody else remixes it, even if the remix is only 5% different, the new art would belong to that other person. 

That’s more of a tech-based approach - you know, the way how code is developed and improved in an oftentimes collaborative effort.

It reminds me a lot of the collaborative approach of the Blender community. (Blender in itself is a free and collaborative project. And see how far that got them!) People help each other so everyone could create nice 3D art. And the art would still be transformative, and have its own look with each artist, cause everyone put their own thoughts and skills into the subject.  

My final conclusion:

Personally, I like the idea of having an open community, free of gatekeeping, that creates beautiful transformative art together. And everyone who created a piece can take it home and happily put it on their wall. Or their fridge. Or set it as their phone wallpaper.

I like how it gives us new ideas to be inspired and impulses for art that we want to create and for visions that we want to share. My personal experience is that communities that are more into sharing and collaborating, are usually the less toxic ones. Giving everyone a good time and joy as they create art is a big strength of this community.

My only worry is, though, that big players may want to use this new technology to exploit artists even more. Artists are already required to be super skilled and are still being widely exploited. It would be sad to see the respect for their hard work shrink even more, and I sincerely hope that this does not happen, as I take MJ’s hand to let it take me on an exciting new adventure.