Worshiping the Idol of Productive Output - Etan Kerr-Finell
Today is a bit of an outlier. Normally I share something I have read each day, but today I am sharing an essay I wrote recently.
Thanks for reading,
Etan
--
AI and Creative Work, or Worshiping the Idol of Productive Output
1.
In Jewish thought idol worship is primarily problematic and forbidden because it is a misnaming. It attributes credit inaccurately. There is no crocodile that controls the river. There is no god that pulls the sun. In the Jewish story, all occurrences are traced back to a single primary cause and causer and the problem of idolatry lies in the misattribution of ultimate power. I see a parallel in how creative people talk about AI. As a writer I hear it most about writing but in conversations with creative people about AI I also hear it more broadly: making music, film and visual art. Many of my peers appear to be falling into the trap of worshiping the idol of creativity. If you deem productive creativity to be an end in itself, AI poses a great problem: it stands to take your purpose. While art can be an end in itself, equally important is its capacity to allow us to engage with the world more deeply.
For the foreseeable future at least, no matter what the AIs are doing, if I want to journal in the morning, my practice need not change. Even before the current conversations about LLMs, I had to decide to sit down and do that work. I must contend with my particular set of distractions, technological and otherwise, and try to stay focused. But the mere fact that an AI could, in some limited sense, do something that looks like what I am doing has no bearing.
It has no bearing for two reasons. The first is that creative work is valuable intrinsically. The second, and I believe more important, is that living in a way that involves reading, thinking and writing is a means to engage with the vast world of other people outside oneself. There is the search for self-knowledge, and beyond that, engaging in art can give you a path to becoming interested in other people and their lives and to getting to know them better. Some of us learn about the connective aspect of art in high school. If you read the book you get to take part in the conversation. If you don’t, then there is no conversation and thereby no connection through that material. We also, and perhaps more importantly, learn that there are conversations we are more and less interested in attending. Maybe you decide, as I did in ninth grade, that you don’t want to read “Macbeth” or be a part of that conversation but you do want to read “Of Mice and Men” and find yourself pulled to engage about that work.
As makers of things it is easy to fixate on output: the books, the songs made and released, the poems in journals, the book tour, the being seen. These are fine goals but they are not the main thing. The more fundamental, and I believe radical, view is to acknowledge that even the writing itself is not the main thing. It primarily serves as a way to engage more fully with the world. I think here of Audre Lorde’s essay “Uses of the Erotic.” At the end of the essay, she says “there is, for me, no difference between writing a good poem and moving into sunlight against the body of a woman I love.”
To worship a particular kind of productive creativity tells a story in which our true nature is not only tied to but dependent on what we produce, which I find unfortunate and reductive. As we get older we may produce less or more, but the value of an artist does not lie merely in what they have produced. Similarly, a child is not only the promise of what they will make. They are partly that but not entirely, and that not entirely is where my argument rests. A thought experiment to test our intuition is to imagine a painter who suddenly loses their vision. No doubt this would be a tragedy for the painter. But it might be premature to say ‘well, they have no hope of engaging in art or the world meaningfully.’ They may find a new way into art, but beyond art they may find other ways to move into the sunlight against the body of a woman they love, as Lorde says it. They may find other interests that engage them in the great body of creation.
2.
Specialization is so deeply structured into our civilization that I don’t fault people for being worried. It is only natural to be concerned if part of what makes you, you, is being automated away. Often technologies that simplify one activity create the opportunity for further complexity. We invent the axe to help break up wood, and later we build exceedingly complex structures with the same tool.
As one stands in the middle of these shifts, it is hard to see the changes themselves as dynamic and not merely destructive. We can’t help but see ourselves standing at the end of history, which is precisely why historical parallels can help us see this a bit more clearly. The person whose role it was to break wood is out of work, but potential for flourishing is made available by the advent of the axe. Making wood chopping easier, access expands and will enable further collaboration. Kevin Kelly talks about this using the metaphor of finding a door in a game that opens onto a new, more complex level.
I think about this often when listening to ambient music featuring synthesizers. The drum machines of the 1980s put session drummers out of work. In 1982, the Central London branch of the UK Musicians’ Union tried to ban synthesizers and drum machines at performances and recording sessions to protect jobs. Now it is clear that this creative destruction paved the way for many forms of music: hip-hop, much of pop music, ambient, techno, house, digital dancehall in Jamaica, reggaeton, and new genres of music in India, Indonesia, South Africa and beyond.
Few of those who ultimately benefited from the dissemination of drum machines (both creators and consumers) were part of the conversation about whether to ban them or not. Like now, very few of the artists and writers globally (and those who may benefit from their work in the future) are part of the conversation around AI today. Yet it is precisely those we can’t see, who don’t have a vote, who are not even born, who stand the most to gain.
3.
When it comes to AI, I do not see a lot of creatives asking the question Tyler Cowen likes to pose, ‘compared to what?’ For instance, there are many days in my own writing practice where I am confident that I will not be able to do what I am setting out to do with my writing projects. I have been working long enough that I can say with some confidence that this is likely my default setting. Perhaps this is a logical leap but if my ‘compared to what’ is simply stopping altogether, I can’t help but wonder about the untold numbers of people we will never hear from if we shame them away from using AI.
Yesterday on the submission page of an independent publisher I read a disclaimer saying AI could not be used in the creation of the work submitted. Reasonable people can disagree but fair enough. But the publisher went further saying that AI could not even have been consulted at any stage of the writing process. In addition to the enforcement problem, what does it mean to not consult these tools in a world where people are routinely asking these tools about their health and deepest personal questions?
Creative people seem to not recognize the unseen harms of a creative culture that shuns the use of AI in our work. AI is read as a threat but it seems clear to me that there are many, many people who are on the edge of being able to put something together in their own mind and all they need is a little nudge to get it out into the world.
Pragmatically speaking, there will be open-source models as good as today’s paid models in very little time. (One year? Two?) Instead of wallowing in how bad we have it, how near we are to the end of days (creative and otherwise), we could be exploring this thing that, by some chance, we have managed to create. Might it contribute to the degradation of people? It undoubtedly will. So have sharp sticks, and fire, and metallurgy, and paper and photography and on and on.
4.
Our deepest convictions are often based not in some foundational truth but in our own biases and coincidences. Further, we are very, very bad at predicting the future on any timescale beyond the near future.
So where does this leave us? We understandably strive for ground and certainty where there is none to be had. The bad outcome we are certain of is in some ways better than not knowing, so we grasp hold of that story.
To the degree that we can, we must ground ourselves in not knowing. Or if we cannot manage that ideal of the ancient masters, then we can ground ourselves in things basic enough that they are very likely to hold. It is foundational to create and we will likely want to keep doing that for its own sake and as a way to more deeply witness and engage with one another.
If AI makes things more efficient, what will be left that still takes a long time? That is where we will likely continue to find a fair bit of meaning and where durable advantage will likely reconcentrate. One answer here may be art. But really I think of people, gathering with people, talking with people, reading with others, reading the work of the living and dead, hearing the stories of those we love and those we are just beginning to love. People. People in person and people online. The myriad small communities that we form. Much less efficient and a source of increasing comparative advantage.
5.
I hope that we find ourselves spending more time in proximity with people we love. I hope that we will listen to each other more and perhaps spend a little less time focusing on self-knowledge. I am inclined to agree with the psychoanalyst Adam Phillips that we often over-index our attention toward self-knowledge in a way that is neurotic and takes us away from our interest in attending to other people. We may do well to correct a little bit and expand our scope outward toward others. Perhaps we will want to hear music played live if the AIs become good at generating it. Perhaps we would do well to reconsider what it is that we think makes us ourselves. I find it hard to stomach the predominant sentiment among creatively inclined people: that this reimagining is not only a terrible inconvenience but that we are living through a uniquely terrible time. This view may be a limitation of our own imagining.
We also might find upon consideration that nearly everything that makes up our lives is a form of idol worship. We may also discover, in a way perhaps we do not want to contend with, that nearly everything we call ‘the main thing’ relates to production or some outcome. How often do we turn being with the people we value and love the most into a means toward feeling a certain way? Might we not benefit from learning to sit with each other and ourselves, and not only so that we can generate something after?
--
Thanks to Matt Ross, Mark Kissler, Sean Tom and Emily Kerr-Finell for talking about these things at length and reading drafts of this. And thank you to Rachel Elam and Kim Cooper for the nudge.