Sam Bankman-Fried and Shakespeare

Sam Bankman-Fried didn't like Shakespeare. Not only did he not like him, he didn't think he was a genius. In fact, the mere fact he'd lived and published some 500 years ago, when printing was new and mass literacy non-existent, made his genius statistically impossible. What, after all, are the odds, when you consider this from a longterm perspective, with all the trillions of literati yet to come?

This point was raised by David Runciman on Past, Present, Future. Runciman's argument was something of the inverse. Take Montaigne's essays for example. He was the modern inventor of the form, and yet is still considered to be one of the best masters of it. Runciman says, contra SBF, that he perhaps benefits from a first-mover bonus. We think Montaigne and Shakespeare are profound not just because of their insights, but because they were competing with so few others. Montaigne, being the first essayist, appears to be the most profound thinker because he was the first (to be able) to put them on paper. There was nobody like him before, because there was no-one like him before. It therefore follows that the greater the number of literate people, the greater the number of novelists, and the harder it is to make a Tolstoyan mark.

There is a more interesting conversation here: a wider point about what art is, what it is for, and the poverty of probability. SBF would be the first to admit that he took a probabilistic approach to quality; someone being the best is more likely the more rolls of the die humanity can throw over time. Runciman hypothesises the best art as being particularly able to be insightful about the human condition. To put all this another way, if two people are competing against one another for a winner-takes-all prize, they both have a greater chance of success than if 10 or 100 or a 1,000,000 others are also competing.

It's wrong, however, to see art as a game. To reduce artistic creation to ranked choice winner-takes-all impoverishes it. Of course, neither Shakespeare nor Montaigne would have viewed what they were doing as similar "arts". Shakespeare wrong with other playwrights and actors to put bums on seats and feet on the theatre floor. Montaigne, writing for himself, true, nevertheless build on a dialogue with both his past self and classical thought. He thought by writing and wrote by thinking. Both, then, would nevertheless have seen themselves as partaking in a long conversation. It is not something to be won, but a mutual endeavour. To make art, to crib from Arthur Danto, is to respond to, engage with, and argue aesthetically alongside a million correspondents. It is "meaning given embodiment." (As an aside, this reminds me of Eich’s argument that money is political language). This, too, allows us to regain the theological context of much artistic production. The "best art" is not not necessarily that which n people agree to be the best, but something which more fully embodies Meaning with a capital M. To make art is to add your voice to a conversation of what beauty is, of what is essential and true. To make great art is no more possible outside of this conversation than it is to make your own private language, adrift not just from the rest of the world but severed from any reference to persons, places, things, or thought.

A cause of SBF's failure, i think, is to be found in his hyper-utilitarian world view. Utilitarians, effective altruists, alignment theorists, all tend towards a belief in what Robert Calasso called digitability: Utility (goodness, ethics, reason, virtue) is calculable, the world is comprehensible with enough data, and moreover the world, experience, and life is reducible to computable binary quanta. The early AI researchers and behavioural scientists thought this to be the case. Alan Turing spoke prophetically when arguing that a Turing Machine could represent a brain as if it were a sequence of discrete states. The problem, of course, is that this as if has been ignored, not just by SBF, but by his former colleagues. Digitability does, however, illuminate his approach to art as just another computation. For him, life was not a social endeavour, a matter of magnitude and analogue conversation, something which can only be shared attempts to make sense of the absolute, but something that could be won or lost, and which he lost a long time ago.