AI flop?

[dfan]

The point of generative art is not to try to make fake human art, it's to make something else entirely (which happens to elicit responses that are guided by how we have learned to approach actual human art). If you think that that is bullshit, that is a valid opinion and you are totally entitled to have it, but it doesn't seem like there's much point in continuing to debate anything else about the subject, since we're starting from a different set of axioms.

[dfan]

What do you think about these instead:

Putting Deep in the name doesn't make it AI in the neural network sense.

That is true, but in this case the pictures linked to by Tryss were in fact generated by deep neural networks.

[EdLee]

What do you think about these instead :
https://deepart.io/latest/

I like them.

[Tryss]

That is true, but in this case the pictures linked to by Tryss were in fact generated by deep neural networks.

The major difference between these and the top post exemple is that it's reproducing reality in an imposed style, instead of having a fuzzy goal like "draw me a portrait in the 14th-20th style"

[jeromie]

I think the big problem with applying "AI" across various domains is that it's largely a misnomer. We have developed an algorithm that is very, very good at learning to play go (and many other things). It's not really intelligent in the way that we normally think of the term, but when we call it "artificial intelligence" it's easy to ascribe properties to the software that just aren't there. I do think the capabilities of software that can learn by example from large data sets instead of being fed hard coded stimulus response behaviors is very exciting, but there's no reason to think that algorithm is the best approach for all fields nor reason to be dismissive of older approaches that can provide excellent results in their respective domains.

It's good you make these excellent points, jeromie, to remind us all to be both more precise and less gullible.

But from your more expert perspective, can you go so far as to say go AI is essentially a matter of better calculation? And if so, can we infer that AI bots will teach us little beyond needing to learn to calculate better? (I think that sums up the chess experience.)

To adapt a quote from the excellent movie The Princess Bride: All computing is calculation. Anyone who says differently is selling something.

However, whether (or what) we can learn from it is a question about humans as much as it is a question about computers. Humans are very, very good at recognizing patterns. We're also very good at coming up with our own stories for why those patterns are there - whether you look at mythology, science, art criticism, or any other field you'll find people putting their own interpretation on the patterns they find in their object of scrutiny. I think that humans are likely to pick up a lot of patterns from the new AI algorithms, and we'll make our own stories for why they work. (Sometimes our explanations might even be right!) I think the algorithms being used are more likely to generate patterns that are useful to humans than the older approach that was used for chess. We'll never catch up to the computers because they are so much better at calculation than we are, but that doesn't mean that all we can learn from them is the necessity of calculation.

The recent thread about the value of Shusaku's ear-reddening move is a good example. Many humans have learned from that game (and that move in particular) because we attached a story to why it worked. It turns out that it likely wasn't the best move after all, but many of us have still taken to heart lessons about looking at the whole board when choosing a move, nullifying your opponent's plans, etc. Do we really understand why the new joseki that have been adopted after AlphaGo work? Maybe, maybe not. But they are still useful patterns, and humans can become better go players by incorporating them.

[Gomoto]

Derailing a thread?

Thinking outside the box!


Everything is math by the way.


While many members of this forum have a sophisticated perception of go, I will make the bold statement that our concept of art is probably lacking.

A picture generated by an algorithm is not art by itself at all. Art is connected to the human envolvment in all cases.

A picture generated by an algorithm can be an element of art generated by humans.

[Gomoto]

What is the difference between a game and art.

A game is pure math.

Art is transcendent.

[Kirby]

How "good" a piece of art is is subjective, is it not?

So how can we objectively compare it to the results we've seen in go?

[jeromie]

How "good" a piece of art is is subjective, is it not?

So how can we objectively compare it to the results we've seen in go?

I think the only evaluation we can use is whether the piece produced matches the vision of the (human) artist. In other words, I agree with Gomoto that art is a necessarily human endeavor, though we can use computers in our work. That’s why the original piece can be considered successful art even though I don’t like it.

(I have deeper thoughts about what kind of art is worthwhile to produce, but that really would derail this thread!)

[Kirby]

I think the only evaluation we can use is whether the piece produced matches the vision of the (human) artist. In other words, I agree with Gomoto that art is a necessarily human endeavor, though we can use computers in our work. That’s why the original piece can be considered successful art even though I don’t like it.

Even if the artist didn't like the art, if someone finds it moving in some way, isn't it "good" art? I've been to a couple of art museums. A number of the pieces didn't do much for me, but I liked a handful of them. To me, those ones were the good ones. That set of "good" art probably doesn't match the set of "good" art on somebody else's list. Clearly, somebody valued this particular piece at hundreds of thousands of dollars, so to them, it was "good".

But still subjective, which is different than the case with go AI. Go AI is good if it can win games. This can be measured in a very objective way that cannot be disputed - win rate against other players.

[Bill Spight]

The main thing for me, though, is that you agree it's a harder problem than go.

I think it's easier. Mozart, after all, wrote aleatory music. I.e., music written by an algorithm using the throw of dice to generate random numbers. The first major computer program that I wrote composed music using random numbers, and a composer used it to help him write a symphony. Already in the 1970s there were programs that could imitate different composers. This art program is avowedly imitative. I am not impressed.

[Bill Spight]

Let me add that, using chaos theory, I have written programs that generate abstract moving images that people have found fascinating.

[hyperpape]

It seems to me that aleatoric art depends on the tension between the artist giving up control, but also exercising judgment.

That sounds unbelievably pretentious (it’s post-bar time right now), but when you created your art, did you use just any algorithm, or did you choose one that created results that were a little weird, but also something you found somewhat appealing?

[hyperpape]

About objectivity: there’s a philosophical question about whether aesthetic values are every truly objective, but it doesn’t matter for this discussion. Maybe it’s nothing more than a fact about humans’ subjective judgment that we value Picasso above Kinkaid, but we do. By those maybe-not-perfectly-objective-standards, this painting is not so good.

[Bill Spight]

It seems to me that aleatoric art depends on the tension between the artist giving up control, but also exercising judgment.

I agree. Mozart wrote good aleatory music. Bach could have, I expect, but I don't think he did. I was not so impressed by John Cage.

The composer who used my program just gave me a couple of pages of numbers to imitate. I analyzed them using stratificational linguistics and figured out a grammar to generate a similar set of numbers. Lower levels were constrained by higher levels, but transitions at each level were random. As I told the composer, I had to make some musical decisions.

That sounds unbelievably pretentious (it’s post-bar time right now), but when you created your art, did you use just any algorithm, or did you choose one that created results that were a little weird, but also something you found somewhat appealing?

I was incredibly lucky with the abstract, moving art. I simply knocked it out as a background for people to look at while the program delivered subliminal messages. When I demoed the program, it was the motion art that people found fascinating and remarked on. I made next to no artistic decisions whatsoever. Some of the ways I translated the program's output to pixels produced fractals, some didn't. Almost every screenshot of the output was worth looking at. The program produced a good mix of repetition and variation. I was lucky.