The Art That A.I. Cannot Make – A Ramble

Well, since I’m still in the mood for writing about A.I, I thought that I’d talk about the art that A.I. cannot make. I am, of course, talking about art that has personal meaning and/or art based on personal experience. Art where you are very specific about what it should be.

This was something I ended up thinking about in very early January after I saw my reflection in a Christmas tree bauble and just had to make a quick and undetailed semi-digital painting of it. It won’t appear here properly until next year, but here’s a preview of part of it:

The full semi-digital painting should hopefully be posted here in late April 2025.

And, along with how my fan art sketchbook seems to have turned into more of an “art diary” where I can record daydreams and make introspective art, I suddenly realised that A.I. cannot do any of this. It cannot make this sort of hyper-specific personal art because the number of creative decisions available to you is much more limited if you use an A.I. instead of making the art yourself.

You can write a prompt, but it’s up to the A.I. to interpret it and – 99.9% of the time – it will look different to what you saw, thought or imagined. This is because you can’t carefully draw every line, can’t choose which details to focus on or a hundred other small decisions that are part of the process of making art yourself.

You can tell an A.I. the general idea, but it has a different “imagination”, different sensibilities and different “experiences” to you, so it won’t look like what you really want to draw or paint.

Yes, it might give you a general approximation, but it won’t be exact. The best metaphor is probably a typical action videogame shotgun. In videogames, shotguns are often designed to be “powerful but inaccurate” for game balancing reasons. If you need to frantically obliterate a monster or henchman who is literally right in front of you, then the shotgun is often the quickest and strongest choice.

But, if you need to aim at something very specific, then it’s usually better to switch to the sniper rifle instead, which is usually designed to be “slow but accurate”. And this is a great metaphor for human-made art vs “A.I. art”.

If you need something quick, powerful and inaccurate – to satisfy curiosity, to swiftly record general moods/ideas etc.. – then using the “shotgun” of A.I. to churn out a few photo-realistic images in about a minute will probably suffice.

But if you want to turn a hyper-specific thing you’ve seen in real life into art, or if you want an accurate record of a cool daydream that you’ve had or something highly specific like that, then switch to the precise “sniper rifle” of a good old-fashioned pen, pencil, brush or stylus instead of blundering around with an A.I. – because A.I. isn’t very good at precisely re-creating what is either right in front of your eyes or within your own unique imagination.

——————–

Anyway, I hope that this was interesting 🙂

Another Reason Why Making Real Art Feels More Satisfying Than Using An A.I. Generator

Although I’m not sure whether I’ve talked about this before or not, I thought that I’d look at one of the many reasons why making real human-made art, whether with traditional tools and/or digital ones, feels a lot more satisfying to do than using an A.I. image generator.

If you do regular art practice, then experimenting with an A.I. generator (like I did for a few days much earlier this year) will be a bizarre experience. You can literally feel the difference. It’s like something is missing when you “make” art using a generator. It feels more like doing an online image search than actually “making” anything.

And, whilst I could talk at length – and have done in the past – about the detrimental effects that these programs can have on an artist’s confidence in their own abilities, I want to talk about one of the reasons why they feel so unsatisfying to use if you’re used to making real human-made art.

Ironically, real art feels more satisfying to make because it takes a lot more time to make than generating an A.I. image does. Yes, it’s technically impressive that computer programs can generate almost professional-level digital images in a matter of seconds. But the “slowness” of real human-made art – when even a “quick” piece of art might still take 10-30 minutes to make – has a huge benefit that makes it feel orders of magnitude more satisfying to make, even if it might not look as good as the program’s creations.

In short, making real art means that you have to focus on the present moment.

When you make a piece of art, you have to actually focus on making it. It’s a much more “active” thing in terms of your attention. It’s like how, when you play a videogame or read a novel, you have to actually focus on doing that thing. You have to actively process and imagine the things that you’re reading, or you have to actively play the game. Like these things, making a piece of art means that you have to actually focus on making it – because it won’t make itself.

Ok, with A.I. generators, the “art” they make does indeed make itself. But that’s exactly why it feels so unsatisfying. There’s no real focus or attention involved in it. You can just set the generator running in the background whilst you randomly browse other internet tabs or worry about the future or do something else. The short waiting period before one of these websites belches out a collection of images is pretty much designed to make you feel bored and to make you want to distract yourself in some way.

However, when you’re making real art, you need to focus on it. On what you are doing right now, in the present moment. Not in a stressful, intensive way but – like a novel – in a more relaxing “at your own pace” sort of way. Although I don’t “meditate”, I imagine that it probably feels similar to this (but more boring…). Making art yourself means that, if you want to actually finish the artwork, you have to focus all or most of your attention on the present moment.

And, when you focus on the present moment, it feels satisfying. Worries about the future disappear for a while, the constant “high-alert” feeling that lots of distractions can cause transforms itself into a more relaxed feeling of focus, time itself seems to mean a lot less than it usually does and everything outside of your immediate surroundings feels blissfully irrelevant to you. If you’ve ever read a really good novel or played a really gripping videogame, you’ll probably understand this feeling too.

This was something I ended up thinking about when I was trying to de-stress after a stressful, rushed morning. My instinct was to make some art. Since I didn’t really have the enthusiasm or mental energy to think of an original idea, I just decided to make a completely lazy and unoriginal practice painting. I found some Youtube footage of the rainy, neon-lit streets of Seoul and paused it during a cool-looking moment. Then I picked up a small A6 watercolour sketchbook and just spent about half an hour or so drawing and painting what I saw on the screen, before scanning it and improving the scanned artwork with digital tools.

The semi-digital painting I made certainly didn’t look as good as the picture on the screen, but that’s not the point. If I wanted a perfect, hyper-realistic image of a glowing futuristic building on the other side of the world, I could just have taken a screenshot. But, instead, I spent half an hour making an “imperfect” and “unoriginal” practice painting. Because the process of making it was the real reward. Because I got to spend half an hour focused entirely on the present moment, in a world where nothing except the source image and my sketchbook existed for me. And it felt relaxing and satisfying.

So, yes, making real art feels a lot better than getting an A.I. to make “art” for you precisely because it is much slower. Because it gives you time to focus on the present moment.

—————-

Anyway, I hope that this was interesting 🙂