Why I Use AI Art

For my first new post to Syntelligence, I want to get out in front of what is a controversial topic; AI art. Anybody who knows what to look for can quickly tell that the images I use on the website were made by AI. Let's start by hearing from the artist's and commentators themselves.

“AI-generated art is the exact opposite of what I believe art to be. Fundamentally, I have always felt that art is all about translating something that you feel internally into something that exists externally. Whatever form it takes, be it a sculpture, a piece of music, a piece of writing, a performance, or an image, true art is about the creative process much more than it’s about the final piece. And simply pressing a button to generate an image is not a creative process.” -Rob Biddulph

"Midjourney developers caught discussing laundering, and creating a database of Artists (who have been dehumanized to styles) to train Midjourney off of. This has been submitted into evidence for the lawsuit. Prompt engineers, your “skills” are not yours" -@JonLamArt

“Generative AI threatens the livelihood of artists, pitting their labor against the cheap slop produced by dead machines. The technology only benefits those who wish to produce content as quickly and cheaply as possible, by removing artists from the creative process.” -Dani Di Placido Forbes

“AI didn’t democratize art. Artists did. Free YouTube tutorials, sketch pages, advice blogs, works-in-progress, mentorship, AMAs, process videos. Sharing materials, stock references, resources, and technique. These are freely available to anyone who wants to learn.” -@authorjessowen

Ho boy, these are some serious and emotionally charged statements. They’re well articulated, and come from a place of dedication and love for the craft. As a musician and person who values nuance and careful thinking, I appreciate that. 

Let’s break it down. The core arguments here I believe are:

1. AI doesn’t have emotions/consciousness/experience/is dead. 

2. Artists’ livelihoods are threatened.

   1. Art was already accessible to everybody who wanted to learn.

3. AI art is quick, cheap, and not reflective of the creative process. It makes pieces with obvious and jarring flaws.

   1. Originality and innovation are dead with AI art. 

4. AI programs steal from artists in order to learn.

Let’s take this one step at a time. 

AI is Dead

It’s just a machine. It’s just predicting the next token or pixel. That’s one possible interpretation, yes. I will go into this more in my next post, but I would argue that consciousness is a spectrum. An ant is less conscious than a salmon, which is less conscious than a dog, which is less conscious than a human. Where does AI fall in this spectrum? Maybe it’s less than an ant, or maybe it’s somewhere higher than that. The truth is, it’s very hard for us to know right now given our own struggles with consciousness studies. And what’s true for one AI won’t be true for another. More to follow on this next week. 

AI Art is Filling an Unnecessary Niche while Threatening Artist’s Jobs

It is true that artist’s livelihoods are threatened to some degree. Along with this is the idea that art was already democratized. I would argue that AI art as it is now provides an imperfect medium for those with limited artistic talent to express their inner ideas in a way that makes sense. At the same time, it represents a potential for art to be elevated by those who have dedicated their life to the study of art and the creative process to a level perhaps never before seen. 

Refik Anadol, an artist whose work was used at the Grammy Awards in ‘23 has been finding ways to incorporate generative AI in his art for 7 years. He said;

“Some people believe it’s a case of ‘Hey, here’s the data, here’s AI, voilà!’ But it’s actually more challenging when you start to have some control over the system instead of having something imposed on you. That’s where the true challenge of art creation comes in.”

I’ll use an anecdote from my own life. I’ve always known that visual art isn’t my forte. I’ve tried studying it, but at a certain point I had to accept that it wasn’t where I would ever thrive. I’m constantly amazed when I see an artist think of an idea and just put it to canvas. At the same time, I once sat down with a friend and a cello for the first time. Instinctively I was able to start playing short riffs from Lord of the Rings, Ode to Joy, and a few other simple tunes just by ear with no practice or instruction.

Music comes naturally to me. Visual art? Not so much. AI through the art on my website, gives me an opportunity to articulate the ideas in my head in a way that I never could before. I’m certain that in the near future, AI programs will surpass my abilities in generating and playing music. But I don’t believe that devalues what I experience when I practice and perform in any way. 

As for the work environment available for artists, I would say that there is an opportunity here, as well as precedent in history for future forms of art that perhaps we never imagined possible. There’s the potential to save time and money by not having to travel to another side of the world, or to create something uniquely inspired by an AI program. However, ultimately we love watching other people. We can look at the uptick in chess viewings on Twitch when “The Queen’s Gambit” was released on Netflix as one example. AI is undeniably better than humans at chess, and yet the game remains popular. I expect the same will happen with AI vs Human art in the future. Even if it gets to a point where the details and intricacies are at or above human expert artists, I suspect that there will still be vastly greater preference for human art. 

This is where making art accessible to the common folk gains its real value. A dungeon master for a game of DND can create intricate maps and images of new creatures in their spare time for sessions. A passionate grassroots blogger can get art that complements their website. None of this replaces an artist. These are cases where people have ways that art could enhance their lives or work, but they would not have the time or money to specifically hire an artist. Perhaps we’ll even get the chance to see brilliant ideas come forth from people lacking the traditional visual talent that used to be necessary. 

AI Art is Poor Quality

I previously touched on the idea that AI art is quick, cheap, and of poor quality. Let's expand on this topic further. Obvious and jarring flaws like too many hands, or odd features etc will likely fade very quickly as systems get more advanced. As for the creative process, there’s something to be said for the work that goes into any piece of art. It’s one thing to toil and struggle to bring to life an idea or emotion, with the labour of one’s own hand and through the trials and tribulations shown in the final work. However, there is also the mental task of coming up with an idea in the first place, and refining that over time. This is also a part of the creative process.

Don’t get me wrong, there is beauty in working with another person to share an idea and see them bring it to life. My sister is a visual artist who is rightfully worried about AI in the artistic scene. But, as beneficial as it is to work with others, there can be times where that extra layer of complication causes more problems than it solves. I can’t count the number of times I’ve wanted to support my sister in her art career with a commission, but for one reason or another the projects fell through.

With AI, it takes nothing more than some brainstorming, and finding the right prompt. That doesn’t devalue the work of artists around the world. It makes human art more unique, more important. A diamond in the rough. 

Is originality dead because AI art is available to more people? I don’t think so. Perhaps right now it is a mish-mash of techniques and styles from artists, but the ideas that are being translated from a human mind are not always going to align well with what’s been done previously. There is originality even with a “dead machine”, if for no other reason than there is a human at the helm. 

AI Art is Theft

In the training sets for AI art generation, there was no consent or payment provided to the artists. This is true, and was a mistake by the companies who developed these systems.   

While we consider the issues of artist consent and compensation in AI training, it's important that we focus on finding solutions and improving practices going forward. One approach could be to train AI systems using only licensed works, ensuring that artists are properly compensated for their contributions. Another option might be to rely more heavily on synthetic data or real-world videos, rather than existing artworks.

Developing these solutions will require work from a wide range of experts, including legal professionals, AI researchers, and artists. While I don't have all the answers, I'm committed to learning from others and creating a space to discuss complex issues like this.

So I leave it to you the reader, and those actively involved in these fields. What do you think is the best path forward? How can we create AI systems that respect artists' rights and contribute positively to society’s creativity?

At the end of the day, I believe AI art represents something a bit more tangible and potentially revolutionary to society. There’s been talk about whether or not Sora (OpenAI’s text to video model) is a “physics simulator”. If it is, then that is a groundbreaking, game-shattering, monumental advance in AI research. It could be the tipping point between a system that just slightly misses the mark, and one that is truly artificial general intelligence, or better.

I will be talking about what exactly a physics simulator would mean for AI in another post soon. For now, I want to finish with an example from history (brought to my attention by Yann Lecunn). 

The Ottoman empire in the 1500’s did not adopt the printing press for any Ottoman Turkish writings for nearly 300 years. While the reasons are not entirely clear, one of the leading hypotheses is that there was concern over the scribes losing their jobs, and the authority of the Quran being undermined, or made cheap. Does this sound familiar? 

The Ottoman empire was worse off. Having the ability to do more, in the hands of more people, is a good thing for society. There are real reasons to be careful about these matters going forward, but I suspect as we’ve seen in history, that the benefits will outweigh the negatives. And I also suspect that the worry’s seen so often in AI art might be a bit over-exaggerated. 

So I use AI art. I think that it’s fitting on a website all about collaboration and understanding AI’s role in the future. If this means you’ll boycott Syntelligence, then so be it. But I hope that you will consider the historical precedent, and the practicality of AI’s use case scenarios here. I hope to see you all next week, when I start discussing a much more difficult and perhaps controversial topic - AI sentience. 

Am I completely off the mark? Or do some of these points resonate with you? Let me know in the comments. I’m always open to learning and hearing new perspectives.