Programming Note: Previously, I said this newsletter would be weekly with 1 or 2 main topics in each edition. Instead, I plan to send out newsletters on Tuesdays and Fridays, with 1 main story each time, along with ephemera. I will also include Song Recommendations on Tuesdays, and Movie Recommendations on Fridays. Thanks for reading these important updates.

I promise I won’t always base my missives to you on whatever I happen to be teaching or researching at a given moment, but in this case I felt compelled to chime in on an ongoing anxiety — automation and generative artificial intelligence (AI) — as I prepare to teach a class on how these forces impact work and labour markets (and how film and television take up this topic).

Other, more technically-minded writers have dug into the nitty-gritty of these AI systems, revealing what they really offer to us on a granular level. Here are some of those. My technical knowledge is sadly lacking, but I wanted to speak instead about the rhetoric being spouted by generative AI’s proponents.

Let me start by saying that the current rush of money being siphoned into these projects, from OpenAI to Google’s Bard to Meta’s own investments, is built on the failure of other tech markets in the last year or so. This is something I often harp on: these hype-storms are cyclical, and as the tech industry has begun to implode (relatively speaking), venture capitalists and other investors are always desperately looking for the next Sure Thing to plant their flag into. The metaverse idea has sputtered, crypto is no longer risk-free (not that it ever was), and the release of ChatGPT was like a rallying cry for these wealthy morons who only know how to spread cash around.

The main point, though, is that the discourse about automation has barely changed in 70 years. Desk Set, a Katharine Hepburn/Spencer Tracy comedy from 1957, is about a group of women reference librarians (researchers and fact-checkers for news media and others) who are faced with the introduction of a massive computer into their workplace, which they fear will replace them. At the time, computers took up entire rooms, operating like the existential threat of the monolith in 2001: A Space Odyssey. While the technology has changed in size and ability, the way we talk about it remains remarkably similar.

As Richard Sumner (Tracy), the “efficiency expert” engineer who brings in the computer, says to the women, “The purpose of this machine is to free the worker from the routine and repetitive tasks and to liberate his time for more important work.” Sam Altman, the CEO of OpenAI (the company behind ChatGPT), told ABC News this month, “It is going to eliminate a lot of current jobs, that's true. We can make much better ones. The reason to develop AI at all, in terms of impact on our lives and improving our lives and upside, this will be the greatest technology humanity has yet developed.” Maybe, but do you have to be in charge of it?

This has been the narrative, the myth, around machines for, well, centuries. The Industrial Revolution introduced new, awe-inspiring machines to workplaces, and anxieties likewise ran high. It’s at this time that the Luddites came out swinging.

Today, you might be called a Luddite if you aren’t on social media, or if you’re slow to adapt to new technology. But in reality, during the Industrial Revolution, British workers, especially textile workers, were not opposed to technology or its advancement, and indeed many of them happily used machines to assist in their work. Instead, these workers were demanding better wages and more work in an age of great unemployment, and British troops attacked them to force them back to work.

So, the Luddites responded by smashing the machines.

The Luddites had no problem with technological progress, they had a problem with how their bosses demanded more of them for less money, and would use machines as a wedge to get around standard labor practices. (There’s a great book by Gavin Meuller called Breaking Things at Work: The Luddites Are Right About Why You Hate Your Job, which charts this history and brings it forward to the digital economy.)

To put it simply, the problem for Luddites wasn’t technology and the threat of robots taking their jobs, the problem was, and is, capitalism and the bosses using the technology to pursue anti-labour solutions.

The current iteration of generative AI and the automated promises it makes is a classic example, then, of an unwanted utopia. The search engines we use, the way we learn, the images we see, all of this is rapidly changing based on this technology, even though the people making these decisions have no real interest in doing so responsibly, sustainably, or equitably. Labour markets are being forced to adjust, even if they don’t understand what the technology can actually do. The process, as ever, runs backward — the VCs, founders, and CEOs determine what capitalist utopia they want to enact, and the economy falls in line.

Meanwhile, other possible outcomes of technology like this, like a radical future vision of fully-automated luxury communism (machines replace most labour so that we truly be liberated from work and can focus on leisure and other activities, under a post-scarcity communist system of abundance), are rendered as hopeless pipe dreams, at least until the process rights itself.

Moreover, as Aaron Benanav and others have shown, machines have never brought about the fabled gains in productivity that they promised from the very beginning, and there’s no reason to believe that the way generative AI is implemented will be any different, because it will be according to the same financialized, capitalist logic and not according to an actual idea about how productivity is tied to value.

We must remember to historicize. We’ve been here before, and there’s no reason to forget about the lessons previously learned, or to assume that things are really new and different this time. Spoiler alert, but at the end of Desk Set, the reference librarians find out that the computer was always intended to assist them, to free up time for further research, and no one is being replaced, after all. 66 years later, this still strikes me as a properly utopian idea of how new technologies ought to be integrated into our lives, even if it never seems to work that way.

At the same time, the job of a reference librarian no longer exists, because we do have Google. It’s not the tech, doofus. It’s capital.

TL;DR: generative AI is kind of cool. Perhaps there is a way to make it work for us. What we’re seeing so far is not that, and never can be as long as the Sam Altmans of the world receive carte blanche to design the future in their image.

Share

Ephemera

society if everyone subscribed to my newsletter

• The inimitable Ed Zitron wrote for Insider about how everything Big Tech makes is getting worse, tying a tighter tech market to chasing trends and, as a result, deleterious effects on user experience across the internet. I’ve certainly felt this myself, trying to use things from Google Search to Netflix and finding the experience to somehow be worse than it ever has been (Twitter is an obvious face for this problem, but it’s happening on a much wider scale.)

• Malcolm Harris, author of the new book Palo Alto: A History of California, Capitalism, and the World (which I eagerly want to get my hands on), was a guest on The Real News podcast to talk about the fall of Silicon Valley Bank in the context of SV’s anti-labour history.

• If you’re in the mood for some unhinged humour, I suggest Conner O’Malley’s latest disturbing foray into the dark American id: “Rebranded Mickey Mouse.”

• Great list made by Letterboxd user ashley of horror movies with explicitly or coded transgender characters.

Song Recommendation: “Futures Bet” — U.S. Girls

Share Unwanted Utopias

Leave a comment