On art and artifice
And why I won't talk about AI with you anymore
You can spend your whole life listening to recordings and never know how much craft goes into making them. When I listen to recordings that predate digital editing software, I get so caught up in the artistry—which, on old recordings, is made even more tactile-feeling with the hiss and the crackle—that most of the time I forget that these tracks were hand-edited by technicians who sliced and spliced with razor blades and tape so precisely that even the most attentive listeners don’t have any idea.
We think of piano rolls, imperfect records that they are (a piano roll does not record tempo, and the speed of a piano roll recording depends on how fast you play it back), as being somehow more honest than the spliced recording, like tracks embedded in clay, but even those can be an illusion. To create some of the dense multi-voice passages on the piano roll for Rhapsody in Blue, Gershwin recorded over himself, adding additional lines to the roll—he didn’t actually play all those notes at once.
It’s very easy to think of the artifice as separate from the art; I, the artist, make the music, and the engineer does the dirty work that just happens to make my music, uh…more. To maintain this divide, I’ve noticed that there’s naturally an emotional distance in the studio that the engineer puts up—as the artist, I usually don’t see the take sheet with its judgments (an X for “this take was bad, don’t use it”, a scrawled “second try was better”).
Ever since I started making studio recordings, starting with the Clara Schumann sonata in 2019, I typically have not been subjected to having to listen to all the takes, which I think would destroy a lot of people. Up until now, the engineer has spliced the first pass edit in solitude before sending it over to me for the next round of edits, which automatically shields me from the worst of my own playing.
Out of necessity, the engineer had me join him in the studio to do the edits for the Florence Price Fantasie nègre No. 1 several months ago—which I wrote about—which was not how we've usually done things. What shocked me (and the engineer) when we did this was how well it worked. It was easier to communicate and we got so much done so quickly, finishing what usually takes several weeks (if not months) in a couple of hours. After that session, we agreed to work like this going forward.
When I went back this week to edit another recording, I braced myself for the pain of being steeped in Bad Feelings without the engineer being able to protect me from them. It was remarkable, then, how quickly I felt that emotional distance spring up within myself. The takes ceased to be something I had done; they simply happened to exist, and I simply happened to be making decisions about them. My language changed: I said things like “the hands aren’t together” or “there’s a caught note in the bass line,” as if I had had no part in these things, my own hands innocent.
I used to listen to my own messy, imperfect takes and feel despair; now I just felt a clinical coolness. It’s like directing your own dissection from atop the operating table—you can’t get anything done if your own feelings get in the way.
So I listened to takes that were disastrous, trying to gauge if there was anything usable in the mess. I listened as the later takes revealed a performer who was clearly growing exhausted and losing feeling in her hands. I listened to triumphant takes where thorny passages were nailed and final chords stuck the landing. I did all this with a chill dispassion, feeling utterly removed from whatever part of me had actually done these things.
This emotional disconnect feels so antithetical to making art, but as the session went on I felt like I was actually leveling up as an artist, the way I always do when my relationship to music-making changes.
There are certain shifts, I’ve found, that you only truly experience if you make the transition from “young person who plays music for fun” to any flavor of professional musician. When I declared my intention to major in music in college and go on to become wildly famous (either cringe or adorable in my naivety; Brooklyn 99 got it right) the biggest concern people (rightly) had for me was that I would not survive the competitive world of music or make any money.
What no one told me about—and frankly I would not have listened if they tried—was how my motivation and relationship to music would be fundamentally changed by turning it into my identity and doing it for a living.
Music used to be where I poured my angst; now it’s where my angst comes from. I used to yearn to show everyone that I was more than just a person, that I WAS A PIANIST!!! and now I grapple with trying to show them that I am more than a pianist and am in fact a whole person. I used to turn to playing when I didn’t want to work, and now the playing is the work.
If you love something because it’s an escape, you have to learn to love it differently when you make it into your home.
We don't need to talk about AI
I would hereby like to declare a permanent “no talking about AI to Sharon” policy. I know this is never going to happen, that people are going to insist on talking about AI at me and around me, but a girl can dream that by just announcing this—like Michael Scott declaring bankruptcy—I can do something about the situation.
Lately I’ve been having a lot of IRL discussions about AI, and by “having discussions” I mean “sitting there helplessly while people say the same things over and over again” and the more it goes on, the more I want to scream. I don’t mean “haha, I want to scream,” I mean I am very close, each time, to actually screaming from overwhelming boredom. Oftentimes I do my best to tune out the conversation—the wah-wah-wah-ing of the adults in Peanuts is my end goal—while I daydream about getting up and walking A) home or B) into the sea, whichever is closer.
I am not so naive to think that AI is going away and that I never have to hear about it again. I am also, actually, not 100% anti-AI, even though I will admit to having a very strong anti-AI bias as well as many, many ethical concerns about AI. (I, personally, cannot live with the idea of being personally responsible for obscene amounts of energy use and water wastage, which directly impacts the quality of life for people less privileged than I as well as their surrounding ecosystems, and have been very depressed to discover that this is not something that holds back people I know and love from relying on AI to do things I know they are fully capable of doing with their own brains. Nor do people seem to have a problem with using models trained on stolen work by writers and artists who were not paid to have their work exploited. Ethics is just about vibes these days, anyway.)
I think there are actual helpful benefits to AI, and I think it should take some of our jobs to free more people up to do what they actually want with their one wild and precious life. I can think of a number of mind-numbing things I have to do that I would love AI to help with if and when it evolves to be smarter than the dumbest asshole bro in your survey class who never does the reading and is overly confident in his wrongness.
Because here’s the thing: I think there are truly fascinating conversations to be had about AI, its possibilities, and its effects on society. However, I have yet to find myself in one of these fascinating conversations.
For example, here is a non-exhaustive list of examples of topics I would love to discuss re: AI.
- Labor history and workers’ rights! I think this is super interesting! There is so much fruitful discussion to be had about the role of AI in the ongoing evolution of labor and workers’ rights, like how technological advances in the Industrial Revolution (which we see mirrored today in countries have industrialized more recently) affected laborers and created new class divisions. I think there is tremendous potential for AI to be used as a societal tool to continue the progression of labor victories like the 8-hour, 5-day workday, but this crucially requires solidarity between white collar and blue collar workers which, arguably, doesn’t currently exist.
- AI in all its wonderful and terrible potential has already been explored and gamed out by sci-fi writers and I think we should discuss that! Let’s have a little book club, guys! I will be the first to admit that I am a total n00b in the world of science fiction, but what little I have read has shown me that brilliant writers have already drawn the map of what threads we can follow re: technology and humanity. There’s Isaac Asimov’s exploration of different styles of intelligence, which I think should be talked about as people race to create something resembling objective uber-intelligence, and there are the genius (imho) stories and essays by Ken Liu and Ted Chiang who, unlike tech CEOs, focus not on the hypothetical ideal of technology as used by perfect beings but on the complex humanity of how people have used and always will use technology and where the limits are. (While not about AI, I also think Samanta Schweblin’s Little Eyes is a brilliantly realistic depiction of humanity, technology, and surveillance and one could very easily extrapolate those themes to AI tech.)
- As AI is being unleashed on language learning and translation, I think there are fascinating discussions to be had about the different theories of language and how we communicate, and how the subjective decision-making and contextual nature of language can be taught differently to LLMs. Has no one else pored over the translator’s notes in a translated edition of a novel or book of poetry? Has no one else watched Youtube videos about grammatical quirks in other languages? In all the different permutations of different combinations of language, there are countless lingual mismatches and untranslatable ideas and patterns of thinking that don’t map onto each other and it’s going to take a lot of brilliant problem-solving that only humans can do to resolve this and I want to hear someone who actually knows more about theory of language than I do to talk about this.
- I think there are endless philosophical and scientific debates we can all have about the fact that emotions and bias are the brain’s way of creating shortcuts to save time and energy for survival purposes, and what we would get out of neural networks unconcerned with survival and built to intentionally not have these shortcuts, if it’s even possible. I want to talk about the very real fact that other forms of bias have been coded into LLMs and that it’s causing problems and whether or not it is even theoretically possible to code open-mindedness—which is how human beings combat bias.
- I would love—LOVE!—to talk about the implications of AI with someone who has actually read David Graeber’s Bullshit Jobs in its entirety (not someone who likely read an AI-generated summary) because I think the current AI-fication of jobs is proving to be an unintentional litmus test for the bullshitness of a lot of jobs. One of Graeber’s most radical theses is (spoiler) that bullshit jobs essentially function as UBI for a cohort of the population comprising, largely, middle-class white men, as the proliferation of bullshit jobs allows them the trappings of the respectable middle-class career without actually requiring them to contribute significantly to society—how will things play out if the most AI-able of jobs belong to the group of Americans who have historically been the most protected from radical change?
- I still recall, vividly, how in the 2010s it was very in vogue to farm out complex searches and mundane busywork to desperate people in other countries (or even more desperate Americans) for a few bucks online. For less than a cup of coffee you could ask an anonymous cheap worker to generate paragraphs of text for your website or make you a list of all the gluten-free restaurants in San Francisco—the results were just as confident and error-ridden as ChatGPT-generated material today, yet people relied on this cheap underclass to do low-level knowledge work, sometimes in their own actual jobs. The difference now is that the human effort has been frontloaded and now the results take seconds, not hours, and this underclass of knowledge workers is presumably being hollowed out (or not—a lot of them are still being used to clean up datasets and models). What you might call the corresponding underclass in other industries that performs the physical drudgery that keeps society going (agricultural laborers, caregivers, the janitorial class, etc.) is not being hollowed out in the same way; what does this mean for social mobility? I don’t have the answer and I want someone who knows about these things to speculate with me!
Do you see what I mean? AI sits at the perfect intersection of so many things: history and neuroscience, economic theory and philosophy, language and logic, art and class, and more. You could draw themes at random out of a hat and have a super interesting discussion about AI along those themes. There are so many interesting connections to draw and theorize about and debate. I am sure that there are already brilliant thinkers who know their stuff having these discussions.
But these are not the discussions people are having around me. No one seems to give a crap about the history of labor or how class and gender bias can crucially determine meanings in computerized translation. Every goddamn time someone brings up AI, it’s all “I asked ChatGPT to do this thing for me and it was cool!” or “I saw an AI thing and it’s amazing what computers can do now!” or “AI will solve this problem and we’ll never have that problem again, isn’t that amazing?”
Cool, good for you, I’m happy you found some joy and wonder in life, oh you don’t want to discuss anything further than that and you just want to repeat soundbites and headlines? That’s cool, I don’t like forming my own thoughts or practicing critical thinking either, now if you just excuse me I’m going to get up and find the nearest entrance into the sea.
I am so freaking tired. I just want to learn things and make art and enjoy myself in life with all its complexity and confusion, man. Call me when AI can keep the laundry going and declutter the house so I can spend more time doing the things I want to do. So please, if you are not going to talk about AI in any of the ways I find AI interesting, don’t do it around me, I beg of you.
Interesting Reads
Ruby Tandoh: "Inside the World of the Great British Bake-Off" (The New Yorker)
One of those reads that filled me with admiration and envy at its quality of writing; Tandoh is so funny, so sharp, and so insightful. As a fan of Bake-Off and its extended universe, this made me both appreciate the show and see it in a totally different light.
It is hard to think of another show that screens so carefully not just for personality type and talent but also for that more slippery variable, purity of intention. Producers find themselves in the position of trying to cast one of the best-known shows on television—one that routinely makes people famous—with people who care about neither television nor fame. They have to sniff out clout-chasers, and pick through government databases for things like criminal convictions and undeclared baking businesses. Even unrealized dreams can be suspect.
William Morris: "I Do Not Want Art for a Few" (Tribune Magazine)
Man, the more I learn about William Morris, the more I love him. I recently discovered his lecture on the arts and I was thunderstruck not only by his eloquence but also by the sheer relevance of what he has to say about art, money, science, the environment, class, and capitalism—all in one concise talk. (I know, I could really learn from him.)
And Science—we have loved her well, and followed her diligently, what will she do? I fear she is so much in the pay of the counting-house, the counting-house and the drill-sergeant, that she is too busy, and will for the present do nothing. Yet there are matters which I should have thought easy for her, say for example teaching Manchester how to consume its own smoke, or Leeds how to get rid of its superfluous black dye without turning it into the river, which would be as much worth her attention as the production of the heaviest of heavy black silks, or the biggest of useless guns.
[...]
Until something or other is done to give all men some pleasure for the eyes and rest for the mind in the aspect of their own and their neighbours’ houses, until the contrast is less disgraceful between the fields where beasts live and the streets where men live, I suppose that the practice of the arts must be mainly kept in the hands of a few highly cultivated men, who can go often to beautiful places, whose education enables them, in the contemplation of the past glories of the world, to shut out from their view the everyday squalors that the most of men move in.
I'm stuck
Something inexplicable happened to me this week that made me unable to listen to anything except Maurizio Pollini playing Chopin's Scherzo No. 2.
I don't know why—I haven't had any desire to hear this piece for years after hearing it nonstop in studios and recitals, and then a couple of days ago my brain decided it had a Scherzo No. 2 deficiency and won't stop demanding MORE SCHERZO.
I am now fully stuck. Help! 🎹
If you enjoyed the concision of that William Morris speech, you may be disappointed by the rest of his writing. I’m an admirer too, and find him prolix in almost everything. (I set a few poems that are epigrammatic by contrast.)
Oh that's mildly disappointing, but as someone who is only concise about twice a year, I can't be too upset.
Sharon, I think you are a brilliant pianist and whole person! Though not at your level, I too worked hard and became a professional pianist. For many years I marveled at how professional organizations would actually pay me for doing what I loved most. I still love what I do and it is work, every day. Thank Heaven.
Aww, thank you so much! I think we are so lucky to have lives and careers in music.
I'm a latecomer to your written work, and am loving every word, so thank you. I appreciate the links to your previous writing, too. Very happy to be here!
Aww, thank you so much for reading and for your lovely comment. Happy to have you here!