In one of the essays in his collection The Trail of the Dinosaur, Arthur Koestler makes the point that many of the historical dilemmas or controversies that once seemed most insoluble were eventually overcome—not because one side or the other at some point "won" the argument—but simply because new problems emerged that made the whole previous debate seem silly.
The Crusades didn't end because Christianity "beat" Islam or vice versa. So too, Catholics and Protestants never did manage to exterminate one another in Europe's wars of religion—much as they tried. Instead, in both cases, the two sides first fought each other to a stalemate. And it seemed there was no possible exit. But then, new social circumstances arose that made the whole fight seem boring and irrelevant.
I thought in reading that essay earlier this year that the American culture wars and our current political polarization felt like very much of an impasse of that sort—a political controversy between two sides that had fought each other to a stalemate, without any obvious possibility of compromise on the horizon.
Koestler's observation was therefore freeing in a sense. It made me realize that the eventual solution to this impasse might lie not in some eventual triumph of one "side" over the other—which currently seems unlikely if not impossible (much as I would prefer my side to win); nor in some compromise between what appear increasingly to be two completely incompatible value systems.
Rather, the solution could lie precisely along the path Koestler describes: some new circumstance could arise that doesn't so much resolve the previous impasse as make it seem pointless and secondary.
It also occurred to me when I read this essay that AI might eventually prove to be that new circumstance that would eclipse our entire previous partisan culture war and cast it into the shade.
And it appears increasingly like that day may already have come. Nate Silver observes in a piece today that "we’ve probably passed a long-predicted inflection point where AI is playing a much more prominent role in American politics."
It is already a cliché of the media coverage of the data center wars that local campaigns opposing these projects bring together strange ideological bedfellows: both MAGA Republicans and DSA socialists can be found converging around local AI-skeptical protests.
But just because the relatively new problem of AI might close the book on our previous culture war, that doesn't necessarily mean it will lead to something better. After all, I for one have already come to feel that the arguments between AI skeptics and AI boosters have become something of a hopeless impasse and dilemma in their own right, where both sides seem to have ultimately irrefutable arguments in their favor, and we have a simple clash of incompatible but legitimate values.
As soon as AI has eclipsed one dilemma, then, it appears to be creating another no less hopeless one.
On the one hand, I find it hard to get behind the idea of deliberately building a technology that even its most ardent enthusiasts admit will displace many jobs, disrupt large aspects of society, and possibly destroy us all.
On the other hand, the "arms race" arguments along the lines of "if we don't build it, someone else will" may be trite—but that doesn't mean they aren't true and valid, so far as they go.
That's why I keep quoting Martin Amis's line about nuclear weaponry when it comes to AI: I'm just going on about it; I don't know what to actually do about it.
One stream of the left historically has argued that people who blame the machine or try to prevent its construction have chosen the wrong enemy. The machine could actually be our friend, they say. It could actually free humankind from drudgery, if only we could learn to socialize its proceeds instead of hoarding them.
This is the perspective Oscar Wilde takes in his "Soul of Man Under Socialism": "Man is made for something better than disturbing dirt. All work of that kind should be done by a machine." It is also the moral of Ernst Toller's play The Machine-Wreckers, about the 19th century Luddite Rebellion: "What if instead of sixteen hours you worked but eight[?] With the machine no more your enemy but your helper! What if your children, freed from drudgery, grew up in sunny schoolrooms[?]"
Ernest Crosby, in his poem "The Machines," points to the same lesson:
The devil has somehow got into the machines.
They came like the good gnomes and fairies of old, to be our willing slaves and make our lives easy.
Now that, by their help, one man can do the work of a score, why have we not plenty for all, with only enough work to keep us happy?
Who could have foreseen all the ills of our factory workers and of those who are displaced and cast aside by factory work?
The good wood and iron elves came to bless us all, but some of us have succeeded in bewitching them to our own ends and turning them against the rest of mankind.
We must break the sinister charm and win over the docile, tireless machines until they refuse to shut out a single human being from their benefits.
We must cast the devil out of the machines.
This is also where John Ganz ends up in his latest Substack post about AI, from earlier today: "[I]t’s not that AI is bad, per se; it’s that it actually belongs to all of us in common."
I tend to agree. It was our collective labor that made up the data base these models are trained on; just as it has always been collective labor that made the wealth produced by capitalism.
The question is not: is this production a bad thing?; but rather: can we share it more fairly (without completely eliminating private incentives—the absence of which, as discussed in the previous post, has tended to produce some very perverse results in historical socialist experiments)?
And if we do not share it more fairly, then what are those of us who reap the stock gains from its performance doing but living off of the collective toil of innumerable others without compensation?
As Crosby writes in another poem in the same collection: "Alas! I am eating the fruit of the forbidden tree, the tree of others' labor!"
No comments:
Post a Comment