AI’s Instagram Problem

I've definitely been feeling deflated. Every angle I've come up with... some project has narrowly shipped before me. And they gain so much popularity so quickly that mindshare per project becomes a power law overnight. It's like the first person to release is at the top of the App Store, and there's no unseating them; the feedback loop is already too powerful. It very much feels like first to release wins. And I swear, the quality of the code I've seen in some of these projects that are getting thousands of stars is abysmal. It's making me think releasing vaporware in order to capture mindshare, then actually building something valuable is the only way to compete.

> Human creations are far better still than AI ones.

I have seen many examples of AI-generated art where it's beyond my taste level to see how it could be improved on by human hands. And stuff that I genuinely appreciated for what it is, divorced from how it was made.

I am reminded that very early on with Midjourney, before this wave of AI gen art was wider known, there was a story of someone winning an art competition with a piece generated with it. [1]

But, aside from disagreement on the dimension of "good", the one area AI gen seems clearly unparalleled in is "fast". That's likely enough to disrupt.

[1] https://www.theverge.com/2022/9/1/23332684/ai-generated-artw...

Andrew Ng in the linked message said himself: "Maybe you’re training XGBoost on a structured dataset and wondering if you’re missing out on ChatGPT. You may well be onto something even if XGBoost isn’t in the news."

Boosting generally is very competitive with (or better than) neural networks on small to midsized datasets.

Bayesian techniques are always interesting.

Personally I think there are great opportunities linking knowledge engineering approaches (which are very manual-labor heavy) with neural network approaches so we can bridge the taxonomy/unstructured data divide.

I hear you but:

> I have seen many examples of AI-generated art where it's beyond my taste level to see how it could be improved on by human hands

This is my overriding point, beyond matters of taste. Everyone who talks about this stuff can't seem to talk about its value without reference to the hypothetical human artist or writer or whoever who didn't make it. The awe of just the fact it was made by AI can't be separated from one's appreciation of it, especially if you are someone who is incredibly passionate about AI. You say you could appreciate it without reference to the origin, but how can any of us claim that? In the same way one's thoughts about George W Bush's paintings are necessarily influenced one way or another by the artist himself, I just don't think we have the ability to separate these things. But that's not like a failing or anything, its just how it is. The total almost dizzying excitement of this AI moment/new future can only reinforce this. The more you are excited about AI art, the less you are probably really looking at it.

As for the competition thing, that looks like a fine illustration for some random fantasy novel or something! It has to me so many interesting but superficial markers of artists who use AI: somewhat baroque but stunning complexity, figures and faces of slender and/or beautiful women, a vague sense "humans right beside a large and great thing." But otherwise, can't you feel a little emptiness in it? If you really try to take anything from it at all other than a fleeting aesthetic experience?

Maybe so, maybe not, but either way, can we even speak at all to what the painting itself is trying get us to feel or think? The artist states that they were making a statement about AI artwork, but what is the statement in reference to the work? I can't tell, other than, of course, what I am saying: it is nothing about what we see represented in the painting, but simply the painting itself as a painting.

The article goes on indeed to say that the judges knew AI made it, with only speculation that maybe "they didn't understand," but I suspect that's just coming from the vocal anti-AI crowd on twitter; such that they can say its unfair or a trick or whatever. Either way, the artist's point wasn't in anything they drew, but simply seeing if they would win, and then telling everyone they did. "..and I'm not sorry." It's all quite telling!

The painting, and all the others like it, can only speak one thing: themselves as technically novel works. Which is why all the pro-AI people find themselves emphasizing this bare aspect of a work. Is a sea sponge in the bona fide style of Francis Bacon a novel work? Definitely. An amazing achievement for a computer? Certainly. But what is it after that?

After the dust settles with all this we will remember what we like and don't like again. This stuff wont go away at all, and people will continue to make interesting things! But I have a feeling eventually 95% of it will be considered in the same cultural breath as kitsch art pieces and elevator music. But this is a stronger claim I could definitely be wrong about. The main thing is, right now, its just too early to tell.

Taste is always sticky. I don't really believe in "levels," just differences. But one thing I know for sure is AI art to me is not great, even if it has masterful technique and style its just got no inspiration!

Funding things you wont be fired for funding. IBMs!