In short: "Purely aesthetic objects are only for the rich! Poor people will content themselves with useful works of nonfiction on socially-relevant themes! I HAVE SPOKEN!"
This pdf illustrates beautifully how a new medium is misunderstood before it is understood. It happened when oil paint was 'invented', it happened when the camera was invented, it is happening now with the introduction of AI. It takes quite a while for artists to 'grok' a new medium and this process of change cannot be rushed.
For a 1977 example of the creative nature of a computer being effectively understood, check out 'Personal Dynamic Media.' (Kay, A. and Goldberg, A., 1977. Computer, 10(3), pp.31-41.).
In it they describe a computer as a 'metamedium': an layer that 'impersonates' non-digital media. Things have changed since then, but that understanding still holds true for 90% of how a digital artist uses a computer.
That's a seminal paper, and when contrasted with the iPad of today, really drives home the point that most of this metamedium's potential is still not harnessed.
The average tablet today (closest to the "dynabook" idea) is packaged in dumbed-down ways to maximize profits -- it's been turned into a finely-tuned machine for consuming entertainment, instead of its promise of a "dynamic medium for creative thought".
Not sure that there is a single source, but the potted history is like this:
Most paints are pigment (providing the color), extender (bulking it out) and glue. So... to invent oil paint all you had to do was to add oil to your pigment instead of gum Arabic, wax, wet plaster or whatever else you were using. Hardly rocket science. For this reason, oil paint was invented many times in many places.
However, for the most part, it was used in the same way as its antecedents was used. Business as usual.
Several things came together to give us the oil painting tradition we know and love.
Most importantly, the Dutch and Belgians invented the middle-class. These merchants were rich enough to want to buy paintings, but not rich enough to hire a painter for a few years to do a mural. Mural paintings was paint on wet plaster and in effect it was installation art: designed for a particular room.
So... there was a market for small, cheapish, portable paintings. These small paintings were on metal (copper) and wood (usually oak). But canvas eventually won out because the canvas linen, primer and oil paint were all derived from the same plant (flax) which made it super durable.
To cater to this market, the art gallery (basically a shop for paintings). Like this:
An advantage of oil paint was that it dried slowly, which meant that it could be blended easily. It could also be corrected relatively easily (just paint over and start again) unlike paint of plaster (one reason the last supper is such a mess is that Leonardo corrected it too many times).
Blended paint made realism much easier. I would also say that the lighting conditions of Northern Europe also made realism (and illusion) easier. High dynamic range (black to white or chiaroscuro) tends to support realism. And perhaps the protestant faith supported realism. North European paintings had none of the florid excess of their Souther European catholic counterparts, which were more concerned with re-telling classic narratives than in illusion.
At the same time there was an increased interest in perspective and optics and also optical devices, especially in that part of the world. Certainly this impacted on how artists made their paintings, though exactly how is a bit of a smudge.
Though all these ingredients were hanging around for years, it was Jan van Eyck was one of the first to really take advantage of them. Zoom in on his Arnolfini Wedding and weep:
The author seems more interested in the emotions provoked from art and attempting to measure them rather than critique art itself or computer generation of it.
I think he fails to understand that our emotional experiences differ based on the art, and cannot be easily measured. (Three people separating taking in artwork will come out with different feelings upon viewing it.)
But is the output of stable diffusion actually art, or is it just a mimicry of artsy information? Does it have compositional value, or an original idea? Couldn't you make an argument that stable-diffusion-imagery is in the most original of senses derivative (while not explicitly trying to be derivative, like e.g. Warhol)?
Sure, there will be people pinning prizes on particularly pretty output of stable diffusion results, and some people may even hang it on their walls - but that is not really indicative of artistical achievement (just consider how many "live, laugh, love"-variations are plastered over walls everywhere).
It's just a tool. It doesn't make sense to call it art or non art. The human intent is what makes it art, and other human's opinion is what makes it valuable art.
This is like saying "is the output of a camera art or not?" The question cannot really be answered, the camera is just a tool as well.
Were Duchamp's Fountain[1] and his other readymades[2] art?
In the mainstream art world, this debate was pretty much settled over 100 years when museums started exhibiting his works, which are now widely considered not just art but very significant and influential art.
Andy Warhol said "art is whatever you can get away with" and as an artist I tend to agree.
On the other hand, a person whose only contribution to a work of art is to come up with a prompt and select something they like from AI-generated work would more accurately be called an art director and a curator than an artist.
No, it really is not. This line of thinking makes it too easy to put things individuals may not like into the "not art" category. No good things come from that.
His last tweet was "I read, there are machines that take an input text to generate images from it. That’s meant to be surprising. Haven’t programs always been texts? You mean, natural language texts are different? So what? Isn’t there Natural Language Programming?"
That's a fairly reductionist view. Yes, it's supposed to be surprising, because it's a very hard problem, and traditional algorithmic programming couldn't crack it. I'm a little annoyed by that tweet, especially considering the high qualifications of the person tweeting it.
So they're the art school equivalent of those engineers who think they have really great ideas about something far outside their field based on something they misheard once. Of course, in a world where "The Cremaster Cycle" is considered High Art, I should expect nothing less.
Please note that this guy is considered one of the original "computer artists", and teaches at a art school, which is useful context when trying to understand the article.
- Technologists will mistake a novel method for novel results
- The art world will see it as a fad on which to capitalize
- The masses will use it as a new way to communicate, and art will spring forth naturally