fame and failure
Some quick thoughts on writing, art, AI, and what it means to be good at something
I heard a strange story when I was a kid that I’ve never been able to shake. I was about seven years old at the time, dozing in the car on a family camping trip—the parents had put on NPR to lull us to sleep. This story woke me up, however: a human interest report about a reclusive artist who lived in the forest. He had a small house and a connected workshop where he carved beautiful statues out of wood, which he then left outside to weather and become more beautiful in the rain. I think he carved mostly animals and people; there was definitely a statue of “a man and woman making love” (at seven I was very excited about whatever that was.) Anyway, the artist wasn’t famous, he wasn’t in any galleries, he had no buyers. In fact he refused to sell his work. So people stole it. They would drive by his house, and see his sculptures, and when they talked to him and he refused to sell them anything they would come back later, and take whatever they liked.
I don’t remember the artist’s name, or where he lived. Since we were on a camping trip I sort of decided that we were going to go see him, or drive by him anyway, and if I just paid attention to the landscape outside the window I would eventually see his little house, and the weird beautiful sculptures in the yard, including the man and woman making love!!!!!!!!!! I still look for him, sometimes.
I picked up a book recently which I really wanted to like: Canadian novelist Stephen Marche’s novella-sized essay On Writing and Failure. Probably it gets better but I put it down after the first ten or so pages out of frustration. I’d hoped for a book about how writing is difficult, how challenging it is to make a meaningful work of art that meets your own aesthetic standards. But the primary assumption, at least at the beginning, is that “failure” means not being professionally successful. “Three hundred thousand books are published every year in the United States alone,” Marche writes. “A few hundred, at most, could be called financial or creative successes. The majority of books by successful writers are failures. The majority of writers are failures.”
Marche of course can’t possibly read all three hundred thousand books published in the U.S. alone every year, so he necessarily conflates “financial success” with “creative success.” It’s true that most books are not financial successes, and in that sense, it’s true that the majority of writers are failures. But does that mean these writers and their novels are also creative failures? Does the free market choose, correctly and every time, who is a good artist and who is a bad or mediocre one? We’ll never know, we can’t know, because of selection bias: we only know about the art we know about. I run across good or great novels all the time that have been forgotten or received insufficient attention (especially novels by women). If a tree falls in the woods, etc; does great writing exist if no one reads it?
The other detail I remember, vividly, from that NPR story was the reporter’s tone: he was equal parts awestruck and baffled that such a brilliant artist could have existed all alone in the woods, totally unknown and unremarkable, except by the people who drove by and stole his sculptures. This was against the natural order of things and it was the NPR reporter’s job to fix that, letting the respectable bourgeois listeners know that the artist lived, that no great art could be made in secret without the approval-granting classes bearing good witness. I don’t remember hearing the artist’s voice and I don’t think he agreed to be interviewed. If he didn’t sell his art he must have had some kind of job or independent income and he had chosen, on purpose, to be left alone in the woods. He didn’t want to be a famous artist, just a real one.
“The first job of a writer is to write,” says Marche. “The second job is to persevere. If you want to write, or if you want to know what it’s like to write, you’re going to have to walk away from the paths of glory into the dark wilderness.” But it’s always the dark wilderness: that’s the place where you do the work. Could you do the work without any hope of getting famous for it? Could you do it in the woods alone, and never show it to anybody? Marche spends his book (the part I read, anyway) on heartbreaking anecdotes about this or that famous writer who struggled in their own time, meaning that they struggled to make money, to get famous, to get recognition. This isn’t irrelevant to the work of writing—if you make a sufficient amount of steady money as a novelist you get to keep writing novels—but again, it’s unrelated to the set of questions that I’m most interested in: how do you write, what do you say and how do you say it? How do you make the right choices and not the wrong ones?
Two years ago, Marche wrote for the New Yorker about Sudowrite and other LLMs (large language models) which were then only being tested but showed the ability to maybe someday “write” on their own. Marche come across as somewhat ambivalent but mostly impressed by the technology: he enthuses over Sudowrite’s ability to imitate Kafka and Coleridge in particular. I think their imitations are shit, personally. Here’s a bit of Sudowrite’s attempt at The Metamorphosis:
As soon as Gregor was alone, he began to feel ill. Turning around was an effort. Even breathing was an effort. A thin stream of blood trickled from his flank down his fuzzy belly. He wanted to crawl away from it, but there was no place to go. He lay still on the spot where he had come to rest just in order to get his breath back and to stop the bleeding. “I’m in a bad way,’” said Gregor.
There’s a touch of Kafka to this, maaaaaybe, but mostly just serviceable descriptive mediocrity, the kind that sloppy writers can churn out quickly. Sudowrite and other models work, of course, by theft—they drive up to the homes of artists and pseudo-artists and steal their work. But whereas whoever stole one of the forest artist’s statues would probably say to a houseguest, “yes I acquired this from a little local artisan,” Sudowrite and other AIs commit theft a couple hundred million times, and then say: “I made this.” AI doesn’t make art, it combines the desire for ownership with currently legalized theft. Marche asks what would happen if Coleridge—famously interrupted before he could finish writing down the lines he dreamed in an opium haze about Kubla Khan—had just been able to plug “Kubla Khan” as a prompt into Sudowrite. The answer is that Coleridge would not have been an artist. He would not have been in the wilderness, trying to figure out how to write; he would be a thief, taking a statue that wasn’t his.
I did some sculpture in high school, mostly ceramics. I made a bust I was proud of and then a kid who was super high knocked it over with his backpack. It smashed on the floor and I was actually heartbroken? I felt it all day, physically, in my heart. After that I mostly stopped doing ceramics. But a few weeks ago I went to a craft supply shop and bought some air-drying clay, tools, acrylic paint, and paintbrushes for about $60 altogether. They’re not high-quality materials and I’m not very good, but I’m enjoying the experience of re-learning. I got it in my head that I would make a candle holder surrounded by petals, sort of like those lotus votive candle holders but for the larger table candles my husband likes. It was going well for a while, the petals were all coming out nicely and I’d completed about a third of the rim. But then something went wrong; I don’t know what. I slipped out of the Petal Zone. The new petals were all coming out too thin, too fat, too small, too large. After about an hour, I lost my temper and scraped off all the bad bits. I closed the remaining good section in my fist…and the petals curved smoothly around each other, and suddenly I was holding a sculpture of a flower. It had always been a sculpture. I had just been trying to force the clay into the shape of a candle holder, and once I accidentally, unconsciously realized what I was making, I’d made something kind of beautiful.
This is what creating art is actually like: screwing about in the dark and being wrong and not knowing, and being frustrated—and then suddenly making an intuitive leap and figuring out the right shape all along. An AI could theoretically help you along that path by suggesting something random that you hadn’t considered before, but it can’t do the work for you; it can’t make those choices. Only you can decide, as the artist, what that risky leap looks like, and to take it.
Working hard at art or writing almost certainly won’t make you famous, and it also might not make you good. My flower sculpture is, I think, fine but not amazing: the judges on The Great Pottery Throwdown would spot all its flaws in a second. It’s worth asking yourself, though, if you would keep on working in the wilderness if it didn’t make you famous and if it didn’t make you good, if you were always going to be mediocre at best. Do you need to imagine the breathless NPR reporter on your doorstep, amazed at your talent, for it to feel worth doing at all? I don’t think my clay sculptures are that great and I don’t know if they’ll get better. Certainly no one’s going to steal them. But if something is worth it then you keep doing it, even if just okay, even if unknown, even if unappreciated, in the dark wilderness forever.
Beautiful
Beautifully put!