Thursday, June 2, 2022

A Good Explainer on AI Art

I'm honored that Vox media asked me to be part of this video about AI-generated artwork. 

If you don't want to watch the whole thing, I make a brief appearance at 10:25

Producer Joss Fong and her team came up with the brilliant explanation on what actually happens with a deep learning model. 

At 5:59 she explains multi-dimensionality with the example of yellow banana vs. red balloon. It's intriguing that we can't possibly know, in human terms, the criteria that the system is using to arrive at its results, or exactly what features it's extracting when a certain artist's name is used in the prompt.

There's a hidden bonus video that explores the reactions of various artists. 

Please add your comments:

• How do you feel this technology will affect the business and practice of art that you do?
• Do you want to use these tools?
• Will they change what you do or how you do it?  

I don't feel directly threatened by the tech, but I realize it will offer art buyers a cheap and fast method for generating editorial illustration, album cover art, and concept art. So it puts professional artists in those fields on notice and gives anyone the keys to becoming both an artist or an art director. 

As an art watcher, I have a kind of morbid curiosity to see where the technology is headed next, and I dread the onslaught of cheap surrealism that is already flooding social media. Another thing I've noticed is that there's a shelf life to each new set of tools, just as there is for each new type of VFX technique. Each new set of tools becomes old hat, as 


CerverGirl said...

I don't give my attention to AI art, though I figure a form of it will continue because technology is part of our society/consciousness now. I figure if I give my attention to the real-life quality art supplies, using them with the goal of creating lasting, quality art, as much as possible before I transition this life, whatever else happens doesn't hurt. I intend to look at classic fine art in museums in future lives to come...including yours.

CerverGirl said...

Well, I wrote my previous comment before I watched the video...
Now that the issue of copyright is there, copying from artists' works is obviously unwanted.
But I still affirm the original artwork, somehow, somewhere, has its own standing, deserves its own attention, and I believe giving attention to an unwanted problem only gives it energy to continue.

I am so grateful for your books and videos...they enlighten art and artists in ways that are very much desired.

Joel Fletcher said...

Certainly this technology will cause some artists to lose work. However, it seems that making edits or giving notes, which is typical in the business, would be very limited. Currently you have to accept whatever the AI generates. But fine-tuning will probably be possible soon.

I am very interested in trying this out. Mainly to apply ideas and concepts that would be unique. And for the pure fun of it of course.

AI art must require tremendous computational power, probably not possible on home computers. I have not seen any information about that.

While AI art may devalue human-made digital art, the side effect may be that hand-made traditional media will be even more precious.

Vladimir Venkov said...

It will really depend on the way the artist uses it - if it is just a helper I think it can bring a lot of good stuff, but if it is used as a replacement I think it is going to kill the joy of creation as it is going to be very effortlessly to create an amazing images/animations very quickly. Because of the way we function as humans the more work we invest when creating something the more we value it...

Wendy said...

Deeply depressing for anyone that makes a living creating art.

Dave Lebow said...

Thank you for putting the link to this fascinating video on your blog, which I read religiously every few days.

Unknown said...

Art is for everyone, but artists are only so because of the process of creating in my opinion.

It's the process, not the final result that matters. We let the blue chip art market strip this basic understanding of what it means to be an artist away from us, and that mentality pervades here too. The AI might make art, but it doesn't make an artist. At best in this instance it maybe makes the individual an art director, but more like a rich patron who can ask an artist to create whatever they want, and have them execute it. Nothing new there, that's as old as the first time someone paid to have their portrait painted.

When we start our journey as an artist, many of us feel like it's about our ideas coming to life, about the final stop, the "finished" work, the attempt to make something "new", to show just how clever and smart we are. As we grow as artists, we often find it's more and more about actually making and mastering those strokes, finding those angles, learning to see certain ways, and honing your skills that's the reason we continue to create. The final works and showing them is almost an afterthought and not the main motivation. All the things that we felt were the barriers or drudgery before start to become the main reason we make art at all.

So while this technology might allow you to prompt AI to make a painting, it will never make you a painter and you'll be missing out on 90% of the art experience on the creative side. It's like being able to look at or taking a picture of food, vs. actually making and eating it and all the senses and experiences that go into that process.

No doubt a new art form will come from this, and people will spend their time learning new prompts etc. but they'll never be an in person artisan, a crafts person, a painter, a drawer etc. Your skills as a Prompt Artist™ will never, unlike a painter, a dancer, etc. be able to translate their work outside of the digital production realm either. A person who can draw or paint can take anything that creates contrast, and make an image, digital or not.

Plus, there are already AI's that write better and more interesting stories than most folks, so what happens when we scrape the prompts that made these images, and then feed them through an AI that then sends prompts to the art AI system? At some point the human input is superfluous, and we're fast approaching that in this case.

James Gurney said...

Thanks for that thoughtful comment. Machines may make art-making more accessible to everyone, but in a way it distances us from the enterprise, both as makers and appreciators. There was a time, maybe 100 years ago, when nearly everyone could draw, write well in cursive, and play piano. Having machines do all that for us doesn’t really enrich our human lives. One side effect of the internet and all the computer tools that come with it has been the emergence of an enthusiastic maker community, and I expect that to keep growing.

Unknown said...

That's a good outlook on where things stand and where they might go. I'll try and focus on the more optimistic side of being creative now and in the future rather than the dystopian ones.

Have you read about this? Thoughts on it? The articles on it still never seem to ask what it means to be an artist outside the outputs which I find disappointing.

James Gurney said...

Unknown, I heard about the issues surrounding that prizewinning painting.

We're going to have to get used to the idea that this AI technology may be behind any image we see, whether it appears photographic, digital art, or painted. The only way we can know an image is really created by hand is to see a video of it being painted or to watch in person. Many digital artists will be threatened by AI art because it makes a great leap toward automation. That efficiency also makes it appealing to many. Either way, it's here and we might as well adjust to it.

Unknown said...

You're right as much as I hate to admit it. The efficiency you brought up is alone already making it irresistible esp. to certain commercial artists esp. for generating quick ideas.

Right now much of it is free or cheap, but it's easy to see places making their own AI and gate keeping it for larger studios, or charging large subscription amounts. This will most likely take away the accessibility and democratization arguments.

I think the only thing we can possibly hope for is some credit and possibly even monetary compensation like a small royalty from each use of existing works by the company running the AI, but even that's going to be hard to implement and enforce eh? We'll need AI to catch the AI for unauthorized uses probably. ;)

In terms of this contest won by AI, I keep thinking it might be fun to go the other way. Make some work, and present it as AI work, but then reveal that, low and behold, it was "just" a human after all. Probably a little bit of a quixotic John Henry thing but worth a try IMHO. :)