AI is speed and speed is bad

Judging from my social media, which contains a lot of academics furiously sharing posts by other academics, some sort of dam feels like it's cracking on academia and AI. Not that academics were primed to use it (administration, well...), but there's a certain indignance and anger which feels more widespread than even a couple weeks ago.

I suspect the precipitating event is a combination of Miles Klee's very good Rolling Stone piece on the matter and the continued efforts to chip away at even the most patient academic's tolerance for meddling in what is supposed to be a slow, meticulous process. Which is good. I don't know what it amounts to in the absence of widespread, cross-field labor action, but good.

Let's go back to speed, though. When we talk about AI, I don't think the most interesting (really, appalling) thing to talk about is specific case studies of harm. It's not that those case studies are unimportant; it's incredibly important that AI makes you stupider, makes some of us psychotic or suicidal, and that it's environmentally destructive.

The bigger problem, the one we can't solve, is the way AI is the culimination of societal logics prioritizing speed and outcome over process. Even the term "AI" is a marketing driven shorthand for dozens of technologies: I shouldn't be using it, but I'm a product of my environment and I sort of like speed, too. Which is why the academics stressing that AI interferes with their process rather than focusing on it leading to sloppy work, both of which are true, is notable to me.

Those who know me know that my research on miniatures wargames production is a shell for bigger questions about process and meaning-making at work. I firmly believe that enjoying and prioritizing process leads to a host of possible exploitations at work, the classic trap of doing what you love. I also believe it's the only way out of this mess. You have to be slow, you have to be social, and those two things converge when affirmative social decisions about quality of work derive from entry into some sort of community. Craft becomes a why and a how rather than a what which exists as some sort of "authentic" sphere separate from industry and the digital.

More than plagiarism or disdain for the arts in the techbro domain of the slop factories, it's speed and efficiency which drive AI. Probably not more than the grift-oriented nature of the 21st century economy, but the grifting is also derived from an obsession with speed. Because the only way the culmination of a project is realized is through exchange value and that exchange value has to circulate far and fast.

That's all really sticky and complicated, because I just put the word authentic in quotes as a mark of disdain, indicating that authenticity is a trap. But what we've ended up with is a world without ephemeral notions like authenticity in the breakneck chase for something real, and "real" is pretty much just money. You can see the numbers in your bank account, you can buy that trip to Spain, you can eat good food. Those are real and, well, get that bag.

There are other arguments against slowness. I'm sympathetic to Sarah Sharma's work on the ways slowness (or rapidity!) for one person is undergirded by the speedy work of others, mostly minorities and women. We're basically all in this together, trapped moving ever faster, with any withdrawal based on someone moving even faster, still.

I'm extraordinarily unsympathetic to both slowness movements, like craftivists, and fully automated luxury space communists. The former ignores arguments like Sharma's entirely, the latter tends to prioritize technological fixes for more speed and eschews process for its own sake.

I just loop back to process. Maybe speed is good for you, but that "you" has to be a social "you". If you can no-shortcut your process and get somewhere you want to be, ok, but no cheating. That's incredibly unsatisfying, both to you as a reader and to me. In academia, specifically, there are no shortcuts. And that's maybe why people are still trying desperately to get in, despite fewer and worse jobs. There's still a process worth respecting deep in its bones.

At the same time the academics are getting angry, the gamers are getting even angrier. Kicked off by Larian's CEO talking about how rad AI is, then followed by a bunch of other mid-sized studios chiming in to say they agree, the Discourse is hot and heavy.

Now, I'm what you would call a gamer. I play a lot of games, digital and analog. Too many. But I'm far more cynical about AI when it comes to games: what did we expect?

By that I mean this has always been games. I got my start writing about games and labor. Nestled in my first piece which got any real notice (attention which I'm still proud of) is something which was and remains less noticed: that the games industry served and serves as a model for how work would evolve in the 21st century. Of anything I've written, that's where I will express little humility. I was right.

What is work now? You work on a project. You flit from company to company. Layoffs are so endemic that what seemed galling in 2013 (the way game companies constantly lay people off) barely seems notable. You work faster to get to your inevitable firing more quickly. Gamification has entered every bit of production, consumption, and circulation. Fan culture has infiltrated everything. Everything you buy is temporary, ready to be replaced or iterated upon the next cycle. Consumers (me included) are voracious, bottomless maws of hunger.

Sounds a lot like the games industry.

There is no place AI will embed itself more than in games. I eschew the AI inevitability arguments except here. Deep in its bones, the games industry is gangrenous, rotten, obsessed with speed so much that its feet have fallen off. And yes, I know there are various art games and even the big stuff can still move us (I have incredibly staid taste in games), but Larian was a semi-artsy throwback right up until they made a ton of money, whereupon the CEO started bleating about AI workflows.

I'm just not sure it could've gone any other way. The games industry has always, on aggregate, been at the forefront of prioritizing speed and hype. The exceptions usually prove the rule: some deliberately made game comes out, everyone goes gosh every game should be made this way, and nothing changes for the majority because the industry exists as a fully post-industrial, post-union sphere. And then that good studio closes or lays everyone off or is bought out. Or, increasingly less likely, the studio quickly pivots to the next project. Because what are you going to do? Stand still? No circulation that way and, by the way, the most played games on Steam are 8+ years old.

It's really funny how it's the people who are most secure and can take their time who choose not to via AI, isn't it?