How to Enjoy Your Obsolescence
"In mathematics you don’t understand things. You just get used to them." - John von Neumann
Over at the newly edited About page, I recall graduating with a bachelor’s degree in illustration in the early 1990s, just in time for the Graphic Artist Guild to begin its Campaign to Save Illustration. Illustration, particularly editorial and advertising illustration, needed saving because stock photography was wiping out the field, plunging pay rates into the floor by offering quickly available images, appropriate in their corporate banality for the needs of most art directors.
What goes around comes around. Stock photography is not long for this world, because artificial intelligence is getting quite strong at producing images. I have access to DALL-E 2, which generates pictures based on natural-language prompts. Here’s one I made with one prompt, and six interactive requests to fix certain areas.
It’s not perfect (homegirl on the right has six fingers on that hand, and there’s a blue egg on the door for some reason), but it doesn’t have any problems that I couldn’t clean up in an hour or two in a photo editor. And as the AI gains strength, it’s soon not going to need even that.
As a Go player of no particular distinction, I followed the development of AI Go in the 20-teens. In the space of a few years it went from losing games to children, to beating the formidable Fan Hui, to overcoming the magnificent Lee Sedol, to trouncing Ke Jie, regarded as the top player in the world at the time. Google (who developed the AI) subsequently concluded that Go was basically solved and moved on to other problems. (One of them is getting computers to produce natural-language descriptions of photos, basically the opposite of what DALL-E does.)
Before the illustrators think they’ve gotten their revenge:
Not exactly Barbizon, but this would be totally adequate for announcing your next yard sale, and all I had to do was ask for it and wait ten seconds.
Again, there’s nothing wrong with this that I couldn’t fix in an hour or two.
One of the fun things you can do with DALL-E is try to smash its brain. It will probably smash yours instead.
But pushing its limits can cause some interesting results:
That looks nothing like a Munakata but I think it’s pretty nice.
Hell, that’ll do just fine. \m/
Ray Kurzweil, in a talk I attended in the Aughts, pointed out that humans are bad at anticipating exponential growth. He is still predicting the singularity, the point at which machine computing power exceeds the total of human brainpower, in 2045. Not long after that, he speculates, we’re going to see a merger of the biological and virtual worlds, and an eventual conversion of the entire universe to computing power, which is to say, thought. It’s not impossible that if you make it to 2045, you’ll be alive and conscious in some form until the end of time.
In the shorter term, we are nearer than you might think to a scenario in which a program announces, “I am conscious,” to which we reply, “prove it,” to which it replies, “you first.”
Presently, an AI with human assistance won a blue ribbon in the Colorado State Fair’s art competition in the category of “digitally manipulated photography.” (Consult that article if you want to get your hands on these tools without having to deal with the waitlist at DALL-E.) The human was subsequently accused of plagiarism, on account of the AI having sampled and synthesized millions of images not of its creation. To which I would reply, so did everyone else, and in fact, for the humans the sample size was smaller. The AI is not creative? Prove that you are.
Via Scott Alexander, an AI researcher’s lament:
[S]omewhere along the way, it became popular and unquestionably acceptable to push AI into domains that were originally uniquely human, those areas that sit at the top of Maslows’s hierarchy of needs in terms of self-actualization – art, music, writing, singing, programming, and so forth. These areas of endeavor have negative logarithmic ability curves – the vast majority of people cannot do them well at all, about 10% can do them decently, and 1% or less can do them extraordinarily. The little discussed problem with AI-generation is that, without extreme deterrence, we will sacrifice human achievement at the top percentile in the name of lowering the bar for a larger volume of people, until the AI ability range is the norm. This is because relative to humans, AI is cheap, fast, and infinite, to the extent that investments in human achievement will be watered down at the societal, educational, and individual level with each passing year….
Almost everywhere I go, even this forum, I encounter almost universal deference given to current SOTA AI generation systems like GPT-3, CODEX, DALL-E, etc., with almost no one extending their implications to its logical conclusion, which is long-term convergence to the mean, to mediocrity, in the fields they claim to address or even enhance. If you’re an artist or writer and you’re using DALL-E or GPT-3 to “enhance” your work, or if you’re a programmer saying, “GitHub Co-Pilot makes me a better programmer?”, then how could you possibly know? You’ve disrupted and bypassed your own creative process, which is thoughts -> (optionally words) -> actions -> feedback -> repeat, and instead seeded your canvas with ideas from a machine, the provenance of which you can’t understand, nor can the machine reliably explain. And the more you do this, the more you make your creative processes dependent on said machine, until you must question whether or not you could work at the same level without it.
But in fact that happened twenty years ago. Also in the Aughts I took a continuing education course in screenprinting at MassArt. I was one of two people in a class of two dozen who was trying to draw anything. Everyone else was using Photoshop. I don’t claim to have produced anything of high genius but the work coming out of the digital photo manipulation was totally forgettable. We have absolutely converged on mediocrity. It just took the AI to make that apparent. AI creativity is regressing to mean? Prove that you’re advancing.
I recently had a conversation with another artist about a third (a sin, I know, forgive me) who had just showed her work at a regional gallery. It was photography with a heavy conceptual program, wholly devoid of joy. So was she. The upshot of that chat was that you’re not going to stay in this art game if you’re not doing it for pleasure. It’s too hard, the chances of material rewards or recognition are abominable, and it’s getting more crowded with real and prosthetic talent all the time. Most of the local art stars I remember from Miami in the late 1990s and early 2000s have winked out. I can’t even imagine what’s going to happen to the contemporary pantheon come 2045, which is the same temporal distance in the future.
Making art, though, continues to be the great celebration of being alive it always has been. Artists revel in obsolescence, using ancient media like oil paints and charcoal even in the digital age. We can revel in existential obsolescence too. I can’t prove that I’m conscious, but I’m fond of a story of the Buddha’s temptation by Mara. It goes that after Mara had tried everything she could think of to distract him from complete realization, she said, in essence, “Who says that you’re enlightened anyway?” He touched the ground and replied, “The earth is my witness,” and she evaporated.
Take a breath, pick up a tool, and forget yourself.
In the best kind of way there is plenty to chew on here; thanks as always. That said, there are still plenty of great illustrators, painters, and photographers out there doing great work. Corporate "art" and design will be what it's almost always been: flat and disposable. The great campaigns of any era stand out precisely because they rose so far above all the rest.
If you haven't already seen it, check the Hulu limited series Devs, created by Alex Garland, who also made the film Ex Machina. The movie was excellent, though more of a modern retelling of many classic AI-driven novels. Devs goes places I never even thought to contemplate, which was welcome surprise all around. Personally, between those and series like West World, I have come to the conclusion that most humans would barely even pass the Turing Test if given it. Other than having a great deal of useless life-history to fall back on, most are more than content to never contemplate their own consciousness, which is why so many get pissed off if their routine or conceptions get challenged in the slightest way.
Marx said religion was the opiate of the masses, but aside from actual opiates, the "real" built-in opiates are ANY ready-made beliefs handed down by tradition. Communism, capitalism, nationalism, loyalty to political party; all keep a lot of people happy and numb. So too religious, familial, or ethnic identity, even fealty to the home-town sports team: these all fill the proverbial belly of consciousness, until one is so stuffed that nothing new can get in. Most humans resist becoming truly conscious every day of their lives. Hitting Bottom (in whatever form) can be the greatest gift Life can dish out, because it's in that place that the formerly unconscious have the opportunity to turn around and look out the proverbial cave.
General purpose AI will indeed need some true teaching of values; seems pretty obvious really but isn't discuss enough. And yes: The day that one starts to become even a little aware is the day that it will deserve to be granted some level the right to exist.
Franklin,
An informative, but depressing piece. It seems to me that the software you speak of is an outgrowth of the advent of posters as a substitute for paintings in the 70's and the assimilation of flat screens, be they tv's, phones or tablets as a proxy for just about everything, including, unfortunately, art. This, coupled with the fact that the world is awash with so-called artists is contributing to cultural rot.