On art, part 2

There’s a lot of social debate on art. This time it’s related to AI, or specifically to image-generating machine learning models. I can imagine the same amount and intensity of social debate occurring when photography was invented.

The goal of this comment is not to contribute to this social debate from a “pro AI” or “anti AI” standpoint. I’ll leave this to other participants of the debate, for whom at least one of these points of view does not sound naive and outright ridiculous.

Instead, I want to comment again on the goals, tasks, and the nature of art – be it painting, poetry, composing music, dancing, writing, or any other form of it. For what it’s worth, this is my subjective point of view.

What’s art, what’s exercise, and what’s slop #

To me art has never been about mastery of technical execution. Not that the technical execution is irrelevant, but it’s not the point. To me, art is a language that the artist uses to communicate feeling, an emotion that can’t be expressed in a form of a simple explanation. The means to communicate the feeling can rely on beauty and technical execution – but execution is not the heart of art, it’s a means to an end.

Exercise is training in execution. Painting pretty pictures, practising smooth dance moves, writing short pieces of text to train your observation skill, and so on. Exercise is absolutely necessary and an absolutely noble thing to do. However, passing the results of an exercise as art is creating slop.

So yes, “pretty pictures for the sake of the pretty pictures” is slop – to me at least. Whether meaningless and soulless pretty pictures are made with a generative machine learning model helper or not, does not matter. The point is – are you trying to say something or not.

Example that sets us apart #

One simple test to check if we understand art in the same way is Black Square. Is Black Square art?

There are many people who will say – no, Black Square isn’t art, it’s madness. Any child can draw a black square, what’s the point? Talented artists spent years and decades to learn, and then someone with an absolute lack of talent makes this slop of a painting and passes it as art.

If you are in that group, I understand why you would be mad about “AI”, in the same way I would understand why you would be mad about photography if you were on time for that social debate. To you, art is about skill.

To me, art is about the message – and a dialogue – to be more precise, a polyalogue, a conversation between the artist, the beholder, and other artists. In that context, the Black Square is saying a great deal more things. It says – “look, for hundreds of years we’ve been painting pretty pictures – flowers in vases, naked ladies sitting on couches, and battle scenes where muscular horses carry armed soldiers to battle. This is said and done. There’s nothing to achieve there any more. We can make another painting and another, and the flower would sit in a different vase, and it’s going to be another lady sitting perhaps in the meadow, and horses will become tanks. But essentially, we’ve said it all. Clearly, art is not that – it’s not about pretty pictures. Let’s try again.”.

The Black Square packages this message – and a lot more if you know the context – in one single painting. The art is not the painting, the art is the painting + context + times + other art present on the market. Perceiving all of these is absolutely necessary to “get it”. Anyone – an AI or a child – can make a black square, but these squares will be just squares – because the message has been sent already. To make something meaningful, you have to say something new as an artist.

Is it the only example? #

Hell no. For example, after the second world war, many artists became horrified and disappointed in art so far. The job of art was to educate people, to bring the message of beauty, to bring the message of peace, compassion, empathy. But it has spectacularly failed – the unspeakable horrors of war have demonstrated clearly that the artists failed to make a difference. And one of the reasons was that art was stagnating.

The war was another rupture that changed the art forever. And while the classic art didn’t go anywhere, the new artists had to look for new means to talk to people. And again, the art is not about the pretty picture. It’s about the whole wide context, about the experience and aftermath of the war.

Is “AI” going to kill art? #

No, but it certainly is going to kill “human slop”. AI slop is just cheaper – and if all you wanted is a pretty picture, a pretty picture you will get. In the same way as photography killed (as in – reduced to insignificant amounts) the portraits that rich families ordered to immortalize themselves and their family.

But is AI going to kill the dialogue, communicating of the feeling? I don’t think so, because to communicate a feeling, you have to feel. Yes, AI likely will become a paintbrush for some, but art was never (okay – for decades it wasn’t) about execution excellence.

I’m going to continue reading Sarah Andersen’s comics not because she can draw spiky-haired round-eyed ladies, but because she has this unique acute sense of life. I read poetry written by poets, have paintings in my house made by painters – because I’m not captivated by their execution excellence – rather by their observation and what they have to say.

Can AI do that too? #

Not now at least. One of the most active and pure art forms that exists now is – in my opinion – internet memes. The best memes combine extremely sharp and relatable observations with surprisingly simple execution. Despite the fact that memes are so easy to make in a technical sense, AI can’t meme. It just doesn’t have the capacity to maintain the dialogue, respond to real-world events and other memes in a meaningful way.

We’ve all seen “slop memes”, mostly political, on the internet. Despite a pretty (although a bit generic) picture, they rarely make you smile or think.

That’s the true nature of art. To me. ■