—zk

+++

I’ve been thinking a lot about why it is that my position on what Dagmar Monett Díaz calls “degenerative AI” is perceived as so extreme by folks at my institution as well as by friends, peers, and, sometimes it feels, everyone I encounter. Dear reader, my position, if you’re not already aware, is what some might call “anti-tech extremism,” but is what I’d call “reasonable.” Evidence, at this point, overwhelmingly demonstrates that gen“AI” is anathema to human flourishing—its requisite material infrastructure is exacerbating climate catastrophe, it furthers the exploitation of labor, it undermines the legitimacy of what is left of crumbling democratic institutions, it erodes sociality and destroys relationships, and, because of all of this (and more!), is fundamentally a fascist technology. Let me be clear: gen“AI” is a technology designed to enact and further the project of technocratic-fascist capitalism. This is not news. People (myself included) have been saying this for years and the evidence to support this claim is overwhelming. There is no demonstrable benefit of gen“AI” to anyone except billionaire oligarchs, who are such absolute total fascist sickos as to openly discuss the need to design disciplinary collars for their future servants when they need to retreat to their doomsday bunkers and who believe that the working class and the planet itself are reasonable sacrifices to make on the altar of “progress” or the “singularity,” or what the fuck ever word they’re using these days.

And yet, last month, as I stood in a picturesque lake surrounded by trees that would soon be ensconced in wildfire smoke—coming from fires that themselves are a result of the very climate catastrophe exacerbated by the thing about which people think my position is so extreme—I was told by an otherwise very kind fellow that “there are definitely use cases for genAI.” He went on to say that “it is amazing to be able to vibe code something that I needed for work and not need to hire anyone to do it,” (he’s a lawyer) and that the “price of tokens will go down as the tech becomes more efficient,” suggesting that my worries are just those of someone who has yet to acquiesce to the requirement of adaptation to this new economy.

Perhaps this person doesn’t care about human flourishing. Maybe they think that the billions of people in the majority world suffering undue labor exploitation and environmental harms wrought by capitalism and exacted upon them by the unsustainable consumptive practices of a much smaller group of people in the so-called global north, is “just the way it is.” Perhaps they think that those experiencing this suffering either deserve not have the material and environmental conditions to flourish or that eventually they will “develop” and thus arrive at some material state of existence that approximates the excesses of our own here in the so-called global north. But I’d like to be more generous to this person. I would wager that if you asked them if they cared about the ability of a data labeler somewhere in the majority world to live a meaningful and fulfilling life in which their needs—environmentally, nutritionally, emotionally, materially, physically, and so on—are met, they’d probably say, “well of course I care about that.” So let’s assume that they care about human flourishing, to which, as we’ve already established, gen“AI” is fundamentally anathema.

This person is also not a billionaire. But they probably have a nice house and a couple nice cars and they live in an expensive city. Let’s be clear, though: they will not be saved when the doomsday, for which Zuck and his oligarch-fascist pals have so diligently prepared, inevitably arrives. If they’re lucky, they get to wear the disciplinary collars that Zuck and his friends are designing. In other words, they’re fucked. Just like the rest of us.

And yet, this person was so excited to be telling me about using a thing that is accelerating the very calamities from which their children (and mine) will likely die (and at a far younger age than we probably will). What in the fuck could possibly be wrong with this person? How could they be so absolutely deranged? Was I prepared articulate my astonishment at this man’s derangement and thus risk turning this otherwise beautiful day into an awkward game of “who gets out of the lake first because all of the sudden the conversation has gone silent?”

I think at some point I chuckled, and said, “yeah, it’s pretty crazy, isn’t it…” and then headed to shore.

+++

Now, I think it’s probably worth noting that the evidence I have for thinking that the position I hold is perceived by others as “extreme” is entirely anecdotal. And even more so, it’s anecdotal in a way that is deeply personal. It’s about weird looks; it’s about not wanting to be somehow disinvited over to the house of folks who are my kid’s friend’s parents because I might say something that makes them uncomfortable; it’s about interacting with that family friend who has a loved one who works for a company owned by one of those tech oligarchs; it’s about losing opportunities to speak about technology at work; it’s about not being taken seriously by colleagues that I’m hoping will join or grow our union; it’s about so much more than any specific moment when someone has actually told me, out loud, that my position is too “extreme.” I find that my anxiety about all this is often about the fact that I just don’t want our social circle to end up shrinking because I’ve gone and said too much at a dinner party. And, honestly, I worry about that, especially given how heated some of the conversations I’ve been in about this stuff have gotten.

I do sense that my position—that gen“AI” should be outright eliminated, many data centers decommissioned, and some variation of decomputing enacted (along with the requisite political-economic shifts in order to end technocratic-fascist capitalism)—is indeed perceived as “extreme.” And I think others sense this, too, because, on a wonderful and far-too-short call about resisting gen“AI” in higher ed, someone posted in the chat: “why is this seemingly a fringe position we have within the University?” Indeed, why is an explicit and militant resistance to the existence and proliferation of gen“AI” within higher education seen as being crazy or fringe? Especially when the evidence—you know, the thing that academics like to rely on to substantiate the claims they make—is so fucking clear.

+++

The other type of person—and they can be the same person as the “use case guy”—that looks askance at me when I suggest that maybe this gen“AI” stuff doesn’t need to exist and that we’d be OK without it is what I’ll call “the inevitability guy.” There’s a range of this type of guy, going from: (a) “it’s here so we might as well embrace it to the max and use it to our own economic benefit” to (b) a purportedly more socially-responsible resignation that “it’s here and we can’t put our heads in the sand so we have to work with it as well as teach other people how to work with it” (“ethically” is often added to this statement).

The latter type is emblematic of a well-meaning technocrat and also tends to represent the opinions of folks I often encounter in my work as an academic at a large research university. Indeed, countless university VPs and executive managers say (and maybe believe, though the more likely scenario is that they are far more cynical than to have any kind of belief and just enjoy making a lot more money than us regular professors) that it’s possible to use gen“AI” “ethically” (which, as we know, you just can’t do). Nonetheless, they say it, and often include it in their official university policies about gen“AI” over the principled objections of faculty who actually know about this shit.

In fact, few months ago, I was drafting an email and I was having a hard time figuring out some phrasing. A colleague—and fellow active union member—suggested enthusiastically that I use ChatGPT because “it really helps get a first draft done quickly. Then you can tweak it. It’s so much easier!” I shook my head in disbelief, you know, with like the cartoon noise and shit. “You’ll eventually come over to the dark side, Zach,” they said, with a chuckle. What in the actual fuck, dude? Could this person, who has expressed their obsession with productivity and their enthusiasm for gen“AI” in helping further their productivity, ever be convinced that the very thing for which they express such enthusiasm is undermining our union’s fight for better working conditions as well as our solidarity with the international working class of which we should see ourselves as a part? Honestly, I don’t think it would be worth my time to try, especially given how clear it is that they don’t take the position of principled and militant refusal seriously.

Meanwhile, there are colleagues who identify as more “critical” but continue to have this strange (and completely empirically unfounded) faith that there are some redeeming qualities to using gen“AI” for certain tasks, or that there’s somehow a “critical literacy” about gen“AI” that students need to be given and then they should be free to make their own decisions about its use. What the fuck is critical literacy about this stuff except to understand its myriad harms and figuring how to end its fucking existence?

Alternatively, others (especially in the visual arts community) believe there are possibilities to work with the technology in ways that transcend its exploitative nature, to somehow use it against itself or to use it to expose its own ideological underpinnings. The extent to which such projects are successful is debatable, but I think it’s important to note that such projects use a range of technologies that might fall under the umbrella of “AI” but, as we know, the term “artificial intelligence” has always been bullshit and isn’t particularly descriptive when it comes to what kinds of technologies are actually being used.

Sometimes these artists and scholars adopt a kind of assemblage-theory, postphenomenological, actor-network, Deleuzian, or similarly complicated-seeming idea about the interrelationships between the human and non-human who operate in dynamic, complex systems, in which “agency” doesn’t necessarily reside with the actual people involved because of the complexities of the system itself. Such artistic practice and/or scholarship suggests that there’s a way for “AI” to be used differently, to somehow elucidate its own conditions of production and in doing so allow us to see new opportunities to bring new modes and ways of being into existence. I sympathize with these approaches, having written through such lenses myself over the years, and having made lots of artistic projects that utilize certain kinds of computing in order to try to get viewers to examine the way they themselves are seen on the terms of the capitalist firms which deploy the computational systems to which these folks are subject every single day. Over the last several years, especially since the early days of gen“AI,” however, I have come to question my own artistic endeavors as interventions that sought to catalyze some critical consciousness about computation. I wonder about their role and function, they way they are often subsumed by the neoliberal academy in service of its own propagation, and whether it’s possible to make work that does what I thought my own work was doing. (An interesting aside here is that all this self-doubt is probably why I haven’t been good enough at promoting my work and that itself might be why I worry that it is ineffectual).

I might dare to say, however, that there is something fundamentally different about gen“AI”, both the suite of technologies themselves (for now, let’s just say hyperscale corporate LLMs and their requisite technological infrastructures) as well as the economic-ideological force that has both shaped the conditions for and encouraged their adoption. (This will be the topic of another essay or post at some point, but not now).

Now… this might sound like I’m hedging a bit. So I’ll say it again just in case. I’m not hedging. Fuck genAI and every single fucking server in every single data center its existence relies upon. End ‘em all right now.

+++

Anyways, it seems to me that this idea that my position is somehow “extreme” is a function of the constraining of possibility and a shocking atrophy of the political and sociotechnical imaginary, which I’ve addressed before. When a new suite of technologies or new ideas about an existing technology are foisted upon us by Silicon Valley billionaire-oligarchs, who, have the platform sovereignty, plutocratic gravity, and oracular authority to make it appear as though those technologies or ideas are inevitable, and when the background conditions for the emergence of those technologies and their attendant politics are right, it’s nearly impossible to not see those technologies or ideas as inevitable.

I mean, how many articles and LinkedIn posts have you read (even the ones not written by “AI”) where someone starts out by saying, “as technology develops…” or “as AI evolves…” or really any statement that ascribes to technology some kind of autonomy that doesn’t involve decisions by people—as if technology just evolves on its own like a fucking monkey or fish. What leads us to accept this idea?

But maybe I’m wrong. What do I mean when I say “us,” anyway? Maybe the assumption that others feel my position is extreme comes from the “others” with whom I’m typically surrounded. I mean, there actually are graduate students and faculty out there doing incredible work on all this stuff and their positions are not dissimilar to mine. These folks care about human flourishing and the requisite environmental conditions for that flourishing across the globe, ranging from the lives and wellbeing of data labelers and other workers who have been horribly exploited to the rivers being permitted for wastewater dumping from data centers that were originally supposed to be “closed loop” facilities. There are lots of other people, too, fighting the good fight. Think about this new generation of neo-Luddites and the great organizing done by the folks keeping up stuff like NYC OFF TECH.

So maybe this whole rumination I’ve taken us on here is my fault (if you’re even still reading). I’m just in the wrong place at the wrong time and that’s why all of this is so lonely and maddening.

What if it’s just that, at this point in my life and career, I’m surrounded by mostly people who have benefitted from the status quo? And those people are the very people who would probably argue something along the lines of, “yeah but there are some good use cases.”

Should I bother trying to convince them? I’m not sure. But no matter how hard I try, I want to tell them:

Fuck your use case. Get the fuck out of here with that shit and take a good look around you. The world is fucking burning and the “it’s-gonna-be-here-really-soon-we-promise” benevolent AI God isn’t gonna do shit to fix it because it’s never coming. You’re fucked. Your kids are even more fucked. Good luck seeing them hit a ripe 50 years of age, cool guy. And Sammy Altman and all his fascist oligarch pals know that you’re fucked and your kids are fucked and they don’t give two shits because they think you and your kids are a perfectly reasonable sacrifice for that never-to-be-seen AI God that they keep saying is gonna come. But in the meanwhile, they’re happy to sell you enterprise genAI and make sure that you think there’s a good use case for it. I mean, “hey, did you hear how ethical Anthropic is? Their execs were hanging out with the Pope! Claude can’t be thaaaaat bad. Come on! Plus I vibe coded this thing for work and it only took a few minutes. Can you believe that? I mean, there’s definitely a use case for this. I get that it’s not perfect. But no innovation is ever perfect at the beginning, Zach.” Fuck you.

Clearly, me saying any of that won’t convince anyone.

+++

Maybe our generation just has to get out of the way. I was reading the interview with Koohan Paik-Mander that just came out, and I guess maybe those of us who are proud Luddites and aren’t part of gen Z just need to let the kids do their thing. I don’t know. I’m not sure how I can help right now because everything feels pretty fucking hopeless and it seems to me like the “use case” guys are everywhere, and they certainly think my opinions are a bit “too extreme.” I wonder if they’ll think climate death is “extreme.” Or if they’ll admit that fascism and the destruction of even the teeny tiny semblance of democracy we had left is a bit “too extreme.” (Don’t get it twisted, I don’t think representative “democracy” is democracy at all. But perhaps people will think that “opinion” is “extreme,” also… though if everyone were required to read Graeber’s The Democracy Project and his essay “There Never Was a West,” then I suspect this would be considered less opinion and more fact, but alas… ).

And now of course I come to the conclusion, in which I’ll tell you that everything I just said is probably counterproductive. I have to not blame working people (regardless of whether they realize that they are part of the working class) for believing in the inevitability of gen“AI” or believing in the legitimacy of a political-economic system that requires exploitation. I have to find ways not to dismiss people who seem to have taken the bait, who have embraced gen“AI” and its sycophancy to an extent that overconfidence in their own position about it seems reasonable. I have to have faith that the “use case” and “inevitability” guys can be moved by something, someone, a story or an experience, I don’t know what. But something. Because if I don’t have that faith, then I don’t have faith in people, and isn’t that what I’m actually talking about here? Having faith in people?

Gil Durán has it right in the sense that the target of our ire and of our actions has to be the billionaire technofascist oligarchs. Is this—the argument against billionaire power—where the building of broader solidarity, even with “use case” guys, starts? Or do they believe that the existence of billionaires is inevitable, too?

I have a lot of other work to do so this is where I’ll leave things. Not great, I suppose. But here we are.

Take care, everyone.

And though the “market” bounced back to some extent, suggesting that all “investor confidence” has not been lost, fears around the relationship between probabilistic automation (a more accurate term for certain domains of “AI”) and the existence of companies whose entire functions appear as objects of potential probabilistic automation has not abated.

When nearly any job appears able to be automated (and this idea of appearance is doing a lot of work for so-called “AI companies”), it begs the question that university administrators have been publicly grappling with (often in order to enhance their reputation in the eyes of the very capitalists who have thrown them into this apparent crisis in the first place), which is “what the hell are students supposed to be learning about so that they can get jobs?” This question assumes that the purpose of a university education is to get a job, which I think is pretty fucked up, but we’ll bracket that little ideological quibble for now.

We don’t need to get into the long history of a humanities education being derided as impractical, but you’ve definitely heard the stories about someone’s uncle at the Thanksgiving dinner table incredulously asking, “you’re majoring in History?” And there’s the litany of insufferably non-class-conscious disses on service industry workers who supposedly didn’t succeed in life because, after getting a degree in a field that they were deeply interested in learning about, they ended up serving delicious coffee or cocktails. God forbid. (Maybe they organized a union, too. Probably even more upsetting for that uncle.)

“The Humanities,” for its part, has long sought to fight back against the claims of impracticality, often trying to do so on the terms of the capitalist class itself, suggesting that “soft skills” and “critical thinking” are essential tools for “today’s workforce.” In other words, the stuff one learns in a humanities education will help one become an obedient and docile member of the working class, ensuring one’s willingness to have their labor exploited and making sure that the capitalist class knows that humanities graduates are just as enthusiastic about profits, quarterly earnings statements, and investor returns as their computer science counterparts. Nothing screams critical thinking like licking the boots of capitalists.

Other advocates for “The Humanities” appeal to the history of higher education, advancing the position that higher education should continue to be seen as a “public good” instead of a private one, and that a humanities education prepares students for active civic participation in a thriving democracy. This would be terrific if we had a democracy to participate in. Unfortunately, evidence suggests that isn’t the case, and perhaps it never has been.1 Plus, there’s zero chance someone like Daniela Amodei—co-founder and president of Anthropic—gives a shit about a thriving democracy. In fact, she’s benefitting from us not having one. Because if we had a true democracy—including an economic democracy2—her company wouldn’t exist (and we’d all be better for it).

However, Ms. Amodei recently told ABC News that “The things that make us human will become much more important instead of much less important”… “And what I mean by that is when we look to hire people at Anthropic today, we look for people who are great communicators, who have excellent EQ3 and people skills, who are kind and compassionate and curious and want to help other people.” Oh, really? Help other people you say? I’m sorry, but in what way is Anthropic (or any “AI” company)—from exploitation of people and planet on which its material infrastructure is built, including the mining of minerals, the shipping of those minerals, the refining of those minerals into usable materials, the fabrication and assembly of components and the attendant shipping, emissions, and water usage, to the emissions produced and water gobbled up4 by its data centers to the acceleration of dis- and mis-information to its broader second-order effects such as the centralization of wealth and power and the perpetuation of antidemocratic technocratic-authoritarian decision-making—helping people? What kind of mental gymnastics do you have to do to earnestly believe that you actually want to hire employees who are interested in “helping people”? This shit is just as delusional—though differently so—than what is labeled “AI psychosis.” Maybe this dogmatic belief in LLMs is the OG “AI psychosis.” Apropos here is the recent resignation of one of Anthropic’s “AI safety” researchers, who wrote that he has “repeatedly seen how hard it is to truly let our values govern our actions.” Shocking, I know.

Nonetheless, a number of industry folks with humanities backgrounds as well as humanities academics have fawned over the headline of a recent article in Fortune that covers Ms. Amodei’s ABC interview. The title of the article is “Anthropic cofounder says studying the humanities will be ‘more important than ever’ and reveals what the AI company looks for when hiring.” See? Students who study the humanities are prepared for real life! What an incredible turn of events!

But what makes things particularly rosy for Ms. Amodei is her brother’s prediction in his absurdly long “essay,” posted a few weeks ago that “AI could wipe out jobs across several industries, making it harder for workers to pivot.” Don’t worry, humanities students, you’ll be OK, though. Plus, thank goodness we have oligarch-intellectual soothsayers like Dario Amodei telling us about the future. What would us regular people with normal human intelligences do without his sage wisdom and oracular authority?

But if Dario Amodei’s prediction is even remotely accurate, jobs will be harder to come by. More competition and fewer jobs means workers can be exploited even more than they already are and the capitalist class can, if this prediction proves correct (meaning if oligarchs wields their “plutocratic gravity” in order to make it real), deliver larger and larger returns to shareholders and similarly unmerited salaries and bonuses to executives. If you were a completely unhinged technofascist or cyberlibertarian with billions of dollars to your name and had little to no empathy for your fellow creatures (as studies have shown is the case for the wealthy), this seems like an ideal scenario.

But none of Dario or Daniela Amodei’s assumptions slash proclamation-prognostications are guaranteed. And that’s what humanities students should understand. This shit is absolutely not inevitable. No matter what these people say.

And this is what a humanities education should teach you. It should teach you to tell Daniela and Dario Amodei and their predictions to fuck off. It should teach you to tell Ms. Amodei that her riches ($3.7 billion net worth) are unjustified and that everything she has was made on the backs of the working class and that just won’t continue to happen because the working class will be taking back control of the economy, thank you very much. And that, no, we actually don’t need probabilistic automation, or what she likes to call, “AI.” A humanities education should teach us to question the consent that has been engineered in order to produce the common sense about “innovation” and “entrepreneurship” that pervades society. It should teach us to question the supposed teleology of technological development. And it should especially teach us to question the seeming common sense around the prognosticating power of today’s oligarch intellectuals.

The value of a humanities education cannot be measured in potential career earnings. It cannot be measured because it should call into question the social function of that very measurement. And a real humanities education—one that, in Ms. Amodei’s words, engages the “things that make us human,” that cultivate the kind of compassion that she disingenuously suggests she values in her employees—would also cultivate in students an awareness of their position within the international working class, a kind of international labor solidarity that would demand the end of labor exploitation, the destruction of any and all technologies that perpetuate that exploitation, and the autonomous communal democratic control over the design and development of all technology. It demands that we recognize each other as people who have dignities and rights that are being violated at every turn by the capitalist class, and it demands that we recognize that inherent dignity in each other. How can Daniela Amodei possibly say that she values a humanities education, one that would see the workers being exploited across the globe by the various subsidiaries of the data annotation company with which Anthropic contracts, as having inherent dignity when their material everyday existence is so profoundly more precarious than hers?

The humanities should teach us to ask these questions and then to act on the answers.

There is no measure by which one could possibly conceive of the endgame of companies like Anthropic or OpenAI as even remotely aligned with a real education in the humanities. But many humanities-oriented academic units at universities across the country have caved under the systemic political-economic pressures that face nearly every university, especially large public research institutions. Humanities education has been reshaped to adhere to the demands of a capitalist class that wants to ensure that students don’t question the “workforce readiness” that those capitalists demand, despite the fact that questioning that very premise is exactly what these students should be trained to do.

I believe that when you really study the humanities, you study things about how the world works. And the last thing people in power want is educated folks who understand how the world works. They especially don’t want people who know the history of how things got to be how they are. Because soon those in power would be surrounded by folks who understand that the system is rigged against them and must be changed. And that would be a serious problem for folks like Amodei, Altman, and their techno-oligarchic ilk.

Maybe that hypothetical problem can still become a real one for them.

A couple weeks ago, I went with my family to a screening of An American Tail. Admittedly, I had never seen the film before (at which my fellow elder millennial parents expressed a certain dismay). I was, of course, struck by the rich and expressive animation. I didn’t realize how aesthetic of an experience it would be. The richness and depth of the scenery—the detailed, often dark and moody background imagery as well as the manipulations of texture and light in the foreground—and the expressiveness of the characters and the drama of the compositions, come to a crescendo in the final sequence in which the mice of New York City resist and defeat a gang of cats that have terrorized them.

The villain in An American Tail, Warren T. Rat, is a cat in rat’s clothing. He’s the head of a gang of cats who terrorize the mice of New York City. He pretends to be a rat—a fellow rodent—to gain the trust of the other rodents in the city. He doesn’t pretend to be a “good guy,” necessarily. The mice of the city are aware of his role in the exploitation of labor (one of his “guys” runs a sweatshop at which Fievel, the protagonist, is briefly detained/employed), but he also seems to be able to mitigate the impact of cats who seek to eat and generally terrorize the communities of mice. Since Warren T. Rat appears to the mice as a fellow rodent, the mice of the city, despite their misgivings about him, prefer the conditions under which they live and in which he clearly has something of a hand (or paw) to any alternative. The devil you know, right?

But our little Fievel finds out the truth. Warren T. Rat is actually Warren T. Cat, the head of the very gang of cats that has been terrorizing the mice and making their existence so precarious as to lead them to trust Warren T. “Rat” even though he takes advantage of their precarity and anxiety in order to enrich himself. In other words, Warren T. Rat/Cat creates the conditions for the precarity that he then exploits.

Sounds awfully familiar, doesn’t it?

To me, it sounds like a pretty good description of the sickos that Time put on the cover of its stupid fucking magazine. These technocratic authoritarian oligarchs are the very type of scum villainized in An American Tail.1

At the end of the film, all the mice realize the same thing Fievel discovered, that Warren T. Rat was actually Warren T. Cat. In a moment of desperation, after his fake rat nose has fallen off, Warren T. Rat/Cat says the very thing that today’s tech oligarchs say to us all the time: “Hey, hey, hey! Wait a minute. Who are you gonna believe? Me or your own eyes?”

The mice then unleash the “Giant Mouse of Minsk,” which has been described in various publications and reddit threads as “nightmare fuel.” The cats absolutely lose their shit and jump off the pier into the water. It rules. Honestly, the way that this scene is drawn and animated, that thing would scare pretty much anyone. And when conceiving of Warren T. Rat/Cat as Altman, Musk, Andreesen, Bezos, etc., it’s fucking cathartic… and inspiring. I’ll leave it at that. But think about it.

Who are you gonna believe? Me or your own eyes?

This quote kind of bounces around in the back of my head with some regularity, and I’m reminded of it pretty much every time I open LinkedIn, which is, as I’ve mentioned before, a lot more often than I’d like to admit.

When I open LinkedIn and see effusive posts about AI “revolutionizing” something or another, I often think about how, like the mice of New York City, some of us derive certain benefits from the conditions to which we are subjected, even though things could be significantly and materially better for all of us if were to explore the alternative (i.e., the ridding ourselves of the oligarch-intellectuals who create the conditions for the maintenance of the precarity they then exploit). Some folks benefit from, or, maybe more accurately, aspire to benefit from, the grift and so they themselves become second-order grifters, like some of the NYC rodents we encounter in An American Tail. (e.g., LinkedIn “thought leaders”).

What I mean, then, if we return to the American Tail metaphor (with which I should probably dispense pretty soon here), is that Warren T. Rat/Cat’s quote is powerful, but for me it’s powerful beyond how the movie intends it. Obviously once the mice realize he’s a cat, the structural conditions that Warren T. Rat/Cat himself helped produce can no longer be obscured by the material experience of the mice or rats who otherwise derived even a small benefit from those conditions.

I think what has made this quote stick in my head is that there’s no Warren T. Rat becoming Warren T. Cat moment for us. Nor can there be. Because the fucking mask is already off with today’s technocratic-fascist oligarch-intellectual capitalists. And we should have run them off the proverbial fucking pier a long time ago. So now I probably have to stop thinking of myself as a mouse in an animated New York City and move past the introductory metaphor.

But I was reminded of this “who are you gonna believe” scenario when I read a post about my institution on LinkedIn that gained some traction, especially with the “what’s the point of college now?” set, who think they are saying something new and provocative when they’re just being annoying and derivative.

The entirety of the original post reads as follows:

I am graduating soon and I have come to a harsh conclusion. Unless you are at an elite institution a Computer Science degree is becoming a bad investment. I attend Michigan State University. The pace is too slow. The curriculum is outdated. AI is banned in classrooms which is insane given where the industry is going. Classes should be more difficult with AI allowed. In a required senior level class 75 percent of the room did not have a full time job offer. Everyone wants to blame the economy but nobody wants to admit that the average candidate simply isn’t good enough. I also have no idea what the administration is doing with the money. They have cut TAs and ULAs for classes that desperately need them. We are paying more for less support. The university failed to sustain employer pipelines so I had to build them myself. At Imagine Software, we worked to bring candidates directly to employers while charging them far less than the university does. I learned more from student organizations like Imagine Software, AI Club at MSU and the Claude Builders Club than I did in any lecture hall. These groups actually care about the massive shifts happening in tech. I paid 10,000 dollars a semester for classes that did nothing for my skillset. I paid nothing to join these orgs and they gave me everything. This model is backwards. We are paying a premium for an education that is obsolete before we even graduate. I am posting this to be critical because I hope it brings change. We need to stop pretending this system works. Anyone else feel the same?

There is a lot to unpack here. I’m not gonna do it justice. But a couple important things for our purposes here: first, this student’s post is evidence that they didn’t ask themselves why they feel this way and what it is that might have led to them believing the purpose of college to be something specific in the first place? Of course they didn’t do that, because that would require not only taking a humanities course but also believing that very humanities course shouldn’t be instrumentalized in the service of the transactional paradigm of college education to which they quite obviously subscribe. I think it’s worth pointing out the obvious here—that the adoption of an idea of the purpose of higher education is of course not this particular individual student’s fault (we’ll get back to the issue of ideology later).

But there are also material and social incentives to adopting this view of higher education, and it’s not like administrators (or faculty, especially those who have adopted the ever-incentivized neoliberal-entrepreneurial subjectivity, for that matter) do a great job of countering that narrative. Even (or especially) in the humanities. Think about how many pieces about “the value of the humanities” that you’ve read that couch that value in terms of job-readiness and economism (“they learn to think critically,” we say, praying that someone out there at McKinsey or Deloitte believes us). I’m not even going to give that “humanities-graduates-are-actually-the-most-desired-in-industry” shit the dignity of a fucking link because it’s just such incredible capitalist bootlicking. Fuck that shit entirely.

Either way, I did very much appreciate this student’s critique of the administration’s priorities and the manifestation of those priorities in the administration’s choices about how they invest (or not) in the core (research and teaching) mission of the institution (see, e.g., our University’s President giving his high school pal a cushy $300k/yr job plus a $600/month car allowance while asking Departments, some of whose budgets are nearly entirely salaries, to cut 9% of their budgets). The fact that the university is a fundamentally authoritarian and antidemocratic institution hellbent on a certain kind of neoliberal optimization that turns its intended purposes completely upside down and undermines the mission and ideas for which it supposedly stands is one of many reasons for the exploding interest in unionization on the part of tenured and tenure-track faculty across the country (and certainly a core reason for my union organizing activity).2

As Peter Fleming describes with such precision in his piercing critique of western academia, however, the situation in which universities like MSU find themselves is, at least in large part, structurally determined. We can’t get more state funding without the state (similarly full of petit-bourgeois aspiring never-to-be-billionaires and capitalist bootlickers desperate for their cut of a decreasingly small slice of a pie that should belong to the fucking working class in the first place) being willing to give us that funding. So tuition goes up. We instrumentalize ourselves, turning higher education into a transactional relationship, a means to a job and nothing more. All the while reshaping ourselves into a corporatist institutional form that disincentivizes collegiality and democratic deliberation and instead produces perverse incentives that lead to faculty and students feeling burnt out, computationally surveilled, algorithmically alienated, and powerless to make any real change to a situation that feels increasingly untenable with every passing day.

But this doesn’t mean that students and faculty can’t become aware of the situation at hand and, perhaps, help illuminate that situation for others. But what does that “illumination” even look like? And is it possible? And would it do anything? Or is that sentiment not giving actual people with actual lives and material circumstances enough credit? Is it paternalistic? Either way, believing there is any possibility of a “revelation” of sorts leading to some kind of action is optimistic, and such optimism requires me to believe that the actions and words of so many faculty and administrators aren’t fundamentally cynical, the reactions of people who know they’re helpless to make real, substantive change, and so, while recognizing the hypocrisy of their commitments—e.g., teaching students to use AI “ethically”—have no choice but to go on with the charade.

For folks who study this stuff, it’s pretty obvious that what I’ve been dancing around here is the historically thorny issue of ideology. I’m not particularly interested in getting bogged down in the various interpretations of Marx’s (or Marx and Engles’) deployment of that term and what Marx or Marx and Engles together intended by it, nor the subsequent critiques of the concept of ideology (as understood by the person doing the critique), or the extent to which I might fall into one of the historical camps that pops up along the way (Althusser’s being probably the most prominent, I suppose, but I’m not an expert).3

One of my favorite books about ideology is Roland Barthes’ Mythologies. I use it to teach my students about semiotics. In Mythologies, Barthes does a great semiotic analysis of a magazine cover—describing the complex and layered relationships between signifier and signified, between sign and supersign—and I think we could do the same with the aforementioned cover of Time. On this cover, the capitalist class who exploits people and planet and is fundamentally parasitic in nature, is pictured as the very working class that they exploit, which both underscores the invisibility of the labor of which they have taken advantage, and suggests that the practice of exploitation is a legitimate form of labor itself. Meanwhile, the allusion to the history of photography and the development of the American metropolis suggests the continuities between the production of AI and the building of cities. This analogy is only true insofar as the labor and environmental exploitation required for both has cast shadows over the lives of working people all over the world, the darkness of which they will never be able to escape. In comparing the history of San Francisco with the mining required to produce the AI systems that have generated massive income inequality in contemporary San Francisco, Kate Crawford writes:

As San Francisco drew enormous wealth from the [gold] mines, it was easy for its populace to forget where it all came from. The mines were located far from the city they enriched, and this remoteness allowed city dwellers to remain ignorant of what was happening to the mountains, rivers, and laborers that fed their fortunes… Brechin suggests that we should think of the skyscrapers of San Francisco as inverted minescapes. The ores extracted from holes in the ground were sold to create the stories in the air; the deeper the extractions went, the higher the great towers of office work stretched into the sky.4

This digression, about semiotics and mining, about the visual and textual representations of the world that come to delimit the parameters of common-sense discourse, is evidence I think of why talking about the relationship between ideology and one’s material and social experience of everyday life is so complicated and also so crucial to examine. Like Stuart Hall said, it’s not like we can pull back the curtain on someone’s “false consciousness” and all of the sudden they’ll somehow see the world for what it really is—that’s not how ideology really works, because ideology has feedback loops with material circumstances and the political-economic base that determines those material circumstances.

There’s a very famous graduation speech by David Foster Wallace, called “This Is Water,” that I, probably mistakenly, reference in my courses when talking to them about the role of design in shaping the symbolic universe through which people traffic in their daily lives. I talk to students about how that symbolic universe exerts ideological force, guiding us towards particular ideas and decisions that reinscribe the hegemony of the dominant ideology into our material everyday existences. I’ve always liked Rancière’s phrase, the “distribution of the sensible” to describe this situation, although he never really meant it to be about design specifically. This idea of the “distribution of the sensible” refers to the delimiting of sensory experience that determines how we participate as political subjects.

Anyway, here’s the story David Foster Wallace tells in “This Is Water”:

There are these two young fish swimming along and they happen to meet an older fish swimming the other way, who nods at them and says “Morning, boys. How’s the water?” And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes “What the hell is water?”

This is ideology—we swim in it every single day. We are often unaware of its power to shape our ideas about what is possible. Sometimes we don’t see it. And sometimes we do, and choose to ignore it. David Foster Wallace says some amazing shit in that graduation speech, but he never said if the second fish’s question was earnest.

What I think is most salient for us here is that the student whose LinkedIn post reminded me of the “who are you gonna believe” scene, is embedded within a particular symbolic universe that, through its hegemony of ideology, circumscribes action by producing (and reproducing) communicative supports for the values they expressed in their LinkedIn post. In and of itself, this isn’t a super complicated situation to grasp. As David Foster Wallace writes just after his fish story, “The point of the fish story is merely that the most obvious, important realities are often the ones that are hardest to see and talk about. Stated as an English sentence, of course, this is just a banal platitude, but the fact is that in the day to day trenches of adult existence, banal platitudes can have a life or death importance.”

So, it follows, the most important realities need to be seen and talked about. And while the sentiment of the rest of his speech is utterly beautiful, including the numerous appeals to compassion and empathy, it’s essential to acknowledge that his capacity for such fellow-feeling is nearly impossible for anyone with the kind of wealth and power that has been accrued by the very people who determine the ideological waters in which we are forced to swim and in which all of us—even the most “critical” scholars of liberal political economy—must live the majority of our lives. In other words, the ideological water of technocratic-authoritarian-neoliberal political economy disincentivizes the kind of empathy (and accompanying class consciousness) that would bring about something better (degrowth) for all of us (and for our children and grandchildren). Imagining such a radically different reality is almost impossible because of the ability of oligarch-intellectuals to use their “plutocratic gravity,” “oracular authority,” and “platform sovereignty” to further entrench the positive feedback loop between political economy and ideological expression.

So how do we see and talk about the “most obvious and important realities”? And does seeing and talking about them really do us any good? Four summers ago, a series of posters, with stenciled lettering, lined the streets of Stockholm, declaring: “WE KNOW IT. YOU KNOW IT. SOMETHING IS WRONG.” Underneath this declaration, a pithy four-line stanza, handwritten: “Plant trees, save the bees, clean the seas, punch nazis.” I like this. It’s funny. But even if we know something is wrong, shit is pretty fucking grim right now and it’s hard to imagine being able to make substantive change. And those individual actions that they recommend, while important and worth doing, cannot alone build the class-conscious collectivism required to end the extractivist and exploitative capitalist mode of relations.

A long time ago, as the design disciplines began to reckon with the fact that design thinking and human centered design—all that shit coming out of Stanford’s d.school—were just a way of making the exploitative nature of capitalism more palatable and were effectively building new tools for white supremacy and the furthering of the corporate-colonialist project, I was thinking about how “creativity” had become weaponized in these discourses to serve the capitalist/neocolonialist class. At the time, I told my Dean that I didn’t want my students to be creative. I wanted them to have imagination. Imagination is bigger than creativity. Instead of “creativity” helping solve a problem that was never a problem to begin with, imagination asks why it is construed as a problem in the first place.

The Giant Mouse of Minsk is the manifestation of imagination. It is more than a fictional mythic figure from a kids’ movie. It is more than the massive mechanical mouse designed and constructed by the mice of New York City to scare the gang of cats who terrorize them. It is the symbolic terror of precarity that those mice felt every single day of their lives, turned on those who enacted that everyday terror.

The mice of New York City needed the imagination of Fievel to hatch the plan that would liberate them from the tyranny of feline exploitation. They also needed the community meetings facilitated by other mice like Bridget, an Irish labor-organizer-mouse.

Such a politics is only possible when it appears to become possible. This idea of appearance and what about everyday life is apparent is central to how certain ideologies maintain their stranglehold on our imaginations and how those ideologies maintain positive feedback relationships with the political-economic base. Of course, here we are dangerously close to the silly contemporary art cliché of “making the invisible, visible” (I very much miss the twitter account that perpetually and hilariously riffed on this cliché).

Nonetheless, both art and design as well as labor- and community-organizing are modes of altering that which appears possible. Organizing with others—whether against data centers or against austerity politics or for the unionization of a group of employees—is something that changes appearances. It alters in meaningful ways what one perceives about the world and what appears to be possible or not. It expands the horizon of political possibility because it requires and cultivates imagination.

And unions, especially assertive ones with more clear politics and an aesthetic edge (think the black cat of the IWW or the work of the Farm Equipment Workers Union prior to its absorption into the UAW), scare the shit out of capitalists. They can help us build our Giant Mouse of Minsk. This is why MSU’s upper administration spent tons of money on one of the most famous union-busting law firms to keep our tenure-system faculty from unionizing. They were scared. And we won our union, anyway.

Neither trade unionism nor industrial unionism can be the endgame, of course. Neither can interventionist art practice. Both are always subject to the pressures of the political-economic system under which they exist (see e.g., the onset of concessions bargaining and the decreasing militancy of labor unions beginning in the 1970s as well as the historic tensions within the kinds of contemporary art practice that seek to illuminate injustice yet themselves are part of a “market”).

My point here is that acts of working-class5 solidarity (in art and life) do two things: they scare the capitalist class and they catalyze imagination. They don’t just make us aware of the role of ideology in shaping of the contours of everyday life; they show us that those contours are mutable. Art has this potential as well. I mean, I did start writing this post after seeing an animated film with my kiddo, right?

[deep sigh]

Sorry for the meandering post without some kind of pithy conclusion. Take care of each other.

In these meetings I have been bludgeoned—I mean absolutely bludgeoned—by appeals to “teaching students how to use AI ethically” or “adopting ethical AI use guidelines” or some variation of these terms that seems to suggest an amorphous yet warm fuzzy feeling. These phrases are always deployed with an emphatic but gentle tone, suggesting that any (in this case, mine) visceral, direct, and rather unkind response to the outright absurdity of such propositions would be uncouth. If I turned those meetings into a drinking game where I took a shot of whisky every time someone said “ethical use of AI” or “using AI ethically” or some variation of this general proposition (that there is a way to do so), I’d be dead by the end of every single one of those meetings.

I have yet to find, within my own institution,2 anyone talking about “ethics” and “AI” who has actually defined the term “ethical” in the context of the use of genAI. Sure, people keep talking about teaching their students to “ethically use” AI, and they offer numerous examples of students using it to do certain things and not do others, in effect offering use cases. But a use case is not the same as having a definition of what ethical really means in the context of the use and adoption of genAI tools.

I was just in a meeting yesterday in which a rather self-congratulatory group of faculty presented about their supposedly innovative approaches to teaching students to use AI “ethically,” the development of classroom AI policies, and the development of our institution’s guidelines on genAI use. One faculty member, for example, co-developed their course’s genAI policy with their students, and, together, they created a set of “acceptable use guidelines.” The faculty member offered some examples of what those acceptable uses were and were not. Mostly, these lists centered on a differentiation between “helping” a student do their work and doing the work “for” a student.

Notably, this faculty member didn’t say if the students were asked to describe why they deemed certain uses of genAI acceptable, nor did they suggest that there was any classroom discussion about whether any use of genAI at all is acceptable. This faculty member said that, in class, they “modeled” the ethical use of genAI by using a particular genAI platform and then “employing transparent validation strategies,” to make sure the output was accurate. To me, this just sounds like doing double the amount of work that a decent researcher could do without needing to check the work of a stochastic-parrot research assistant. When I asked this faculty member if “ethical use” of genAI merely amounted to being transparent about its use and the “validation strategies” employed (whatever the fuck that means), he repeated his claim with the caveat that there are lots of different dimensions along which we might judge “ethical.” Oh, really? Are there?

Hilariously, this faculty member suggested that genAI offered some utility to students searching for case precedent in a particular domain of law in which he is an expert, despite the evidence of genAI systems consistently making up false case precedent. He made it seem as if experienced and competent attorneys had been somehow inefficient without genAI’s assistance. I feel like the history of the legal professions is a pretty legit counterclaim to this flimsy thesis. But I digress.

Anyway, this faculty member was then congratulated by subsequent presenters for “enacting their values” in their course policy. I’m sorry, but what in the actual fuck? Whose values are being “enacted” here? Do the students not value the lives of the massively exploited data laborers in the global south who labeled the training data for the LLMs those very students are now using? Do the students not value the lives of the workers being exploited by mining companies and nation states which are extracting vast amounts of minerals from the earth while destroying precious environmental resources and releasing harmful toxins into nearby communities, effectively killing and maiming people, in the process? Do the students not value those very environmental resources or those communities?3 Do they not value the nearly unfathomable amounts of fresh water and electricity consumed by the data centers that could otherwise be used by people who actually need it? The list, as you well know, dear reader, goes on and on and on.4 And some of this shit is happening right down the fucking road—Saline, Ypsilanti. I mean, what the hell kind of “values” are we talking about when the debate over acceptable use comes down to if a student is using it for “cheating”? I don’t give a fuck about cheating when people are dying and climate catastrophe is being accelerated by a bunch of technocratic authoritarian oligarch-intellectuals.

Meanwhile, this faculty member is a full professor. Highly respected in his field. And yet, nowhere to be found in this presentation was an engagement with the literature on the history of ethics and computing or any engagement with the massive body of scholarship on genAI as a material, social, and political-economic subject of intellectual and scholarly inquiry. Nor did he apparently think to talk about any of that with his students when introducing to them the idea that they’d develop some acceptable AI use policy together.

And what about the ethics of implying that the tools used by students under their new “acceptable use” policy are somehow “intelligent”? Did anyone talk to them about how unethical it is to think about “intelligence” in the same way Sam Altman does? There is a real, systemic harm in construing “intelligence” along the lines of the form of intelligence prized by our Silicon Valley oligarchs (slash sociopathic entrepreneur innovators, most of whom should be incarcerated for the harms exacted on marginalized populations who never had a fucking say in the development of any of the “innovations” under the thumb of which they suffer).

The whole history of the kind of “intelligence” privileged by Silicon Valley oligarchs is not only completely one-dimensional but it is deeply rooted in eugenics.5 How is it in any way ethical to reinforce that idea and then to map that onto the many, diverse, and wonderful intelligences (plural!) of our students?

So there I sat. Flabbergasted. But not surprised, I guess. I don’t know. The more investigative journalism and scholarship published—about the harms exacted upon the international working class and the environments on which we rely, both by genAI systems and the technocratic-authoritarian capitalist owners of their means of production—and the more clear it becomes that this situation is the classic neoliberal “privatization of gains and socialization of harms” on steroids, the more shocking these kinds of conversations are to me. At least amongst faculty members. I guess it might be different if we were in a room with the President of our University. But between colleagues, I mean, don’t we all know this genAI shit is, to be quite blunt, fucking us and it’s fucking our students and it’s fucking the world? I mean, is that not patently obvious?6

++

I think I’d be less angry if people didn’t hold onto the pretenses of teaching students how to use AI “ethically.” If they just said, “I’m committed to teaching students to use genAI and here are the reasons for this commitment,” and then enumerated reasons that don’t appeal to “ethics.”

But maybe there’s more at play here that has to do with the psychic violence of the optimized academy.

Is suggesting that there’s somehow an “ethical” way to use, or teach students to use AI, maybe just faculty members trying to soothe themselves, to rationalize the cynicism required to overcome the cognitive dissonance that is the inevitable result of having to survive under an authoritarian technocracy in which being seen as “anti-innovation” is a stain on one’s reputation from which it is impossible to recover? Perhaps the pervasiveness of the appeals to “ethical” uses of AI are part of the university’s infrastructural affect, a result of the psychic damage inflicted by the long history of technological innovation being perceived as a fix for problems—inequality, climate change, and so on—that are fundamentally non-technical in nature.

At this point, this feels like as plausible an explanation as any.

++

And this doesn’t mean that we or our students shouldn’t engage with computing. Nor does it mean we shouldn’t consider the selected application of small-scale LLMs or other more useful approaches from within the suite of AI or AI-related fields, such as Machine Learning. What I’m saying is that we shouldn’t lie to each other, our administrators, or our students when we tell them about how to—and when to—use particular technologies. And, furthermore, it means we should be explicit about the nature of the development of the technologies we are using, regardless of the utility they seem to present to us or our students in a given moment (the infamous “use case”).

The nature of such technological development is fundamentally antidemocratic and this essential quality of all technological innovation under capitalism could not be more obvious than in our current moment—a moment in which billionaires (trillionaires?) offer up the international working class and the environments on which we rely for ritual sacrifice. These oligarch-intellectuals convert, as Evgeny Morozov writes, “intellectual positions into market arbitrage while wielding (and often owning) digital megaphones to reshape the very reality their investments bet against.” Morozov goes on:

Today, it’s increasingly clear that it’s the tech oligarchs — not their algorithmically-steered platforms—who present the greater danger. Their arsenal combines three deadly implements: plutocratic gravity (fortunes so vast they distort reality’s basic physics), oracular authority (their technological visions treated as inevitable prophecy), and platform sovereignty (ownership of the digital intersections where society’s conversation unfolds).

…

Our oligarch-intellectuals begin as interpreters par excellence. They position themselves as technological mediums, passive channels for inevitable futures. Their special gift? Reading the tea leaves of technological determinism with perfect clarity … Armed with their prophetic visions, they demand specific sacrifices—from the public, the government, and their employees.

Those sacrifices include the lives of humans and nonhumans in environments across the globe.

There is nothing even remotely democratic about the kind of technological innovation the inevitability of which we are assured by the very people who have the means to make that inevitability a reality, and who materially benefit from it more than any of us ever will. And yet, we pat ourselves on the backs for using their tools “ethically”? Are you shitting me?

++

It is, additionally, astonishing to me that, although I work in what was once regarded as one of the best agriculture schools in the country, no one in these meetings brings up the fact that if you don’t have water, you can’t have agriculture. And these days, data centers are guzzling fresh water at world-historical rates and at scales that were even a few years ago completely unfathomable.

MSU was, quite literally, the first agricultural college in the United States. And while a deep dive into the history of MSU isn’t within the purview of what I’m writing here, it is impossible not to acknowledge the dissonance between the (admittedly, colonial) connection to the land our students once had (including the mandated three hours per day of manual labor in 1857) and my colleagues’ apparent resignation to the inevitability of natural resources being a sacrificial conduit for the ongoing incursion of genAI into every aspect of everyday life. The problem, of course, is that the supposedly inevitable technological advances—facilitated by the extractivist nature of planetary scale computation as well as the data centers that form its core infrastructure—aren’t gonna do shit for us if we’re all dead because we had to stop growing food and drinking water so the data centers could have the water and the land.

How can MSU “encourage all members of the university community to engage with generative AI tools responsibly, ethically, and creatively”7 when its administration and faculty should know damn well what the implications of that engagement are and how the implications of that engagement are fundamentally anathema to the original purpose of the institution?

+++

Look, Marx didn’t talk about ethics and morality8 because it should be patently fucking obvious that human and non-human suffering is not a bug in the capitalist system—it is one of its core features. It incentivizes competition. It helps create the dynamism of capital itself. The kind of mental fucking gymnastics that faculty and administrators have to do in order to suggest that there is some kind of “ethical” way to use and to teach the use of genAI is, to me, more complicated than reading all three volumes of Capital.

If we were to suggest that it could ever even be possible to adopt some kind of “ethical” use of AI at MSU, such use would have to be predicated on a communally autonomous democratic approach to technological innovation within the university community. Even then, I think, following Marx, the deployment of the term “ethical” or “ethics” becomes a distraction from the real work of creating a convivial society, full of convivial tools, the design and development criteria for which are democratically determined by everyone who is impacted by their development and use.9

It is nearly impossible to imagine what such a situation would actually look like, what kind of technologies we would design, what our everyday lives would look like, together, as a community. Or what kinds of interventions into the environment we would deem acceptable, or what kind of costs we would be willing to incur. What kinds of solidarity might emerge, and how our entire conceptions of education, progress, innovation, and a good life might change.

But I’d rather spend my energy on cultivating this imagination and engaging in the activism to bring it to fruition than doing those mental gymnastics required to soothe my psyche as I lie to myself and my students.

+++

This post ended up being way too long with too many tangents or “side-quests” as one of my fav podcasters, Molly Conger, likes to say. My apologies.

-zk

+++

OK, so we all know how dangerous it can be to just kind of scroll endlessly on LinkedIn. Because LinkedIn is filled with grifters and genAI boosters of all stripes, as well as people who, somehow, earnestly still believe that there are technological fixes to problems that are fundamentally social and political in nature. And the posts from all these different kinds of people are bad for my blood pressure. So, I should not be scrolling on LinkedIn. And yet.

The other day, as my rage at the completely naïve and/or willfully ignorant posts through which I was scrolling began to subside and my eyes were beginning to glaze over, I saw a post that immediately brought me back to reality because it really got me thinking about the question of job automation and the centrality of “job growth” to our ideas about a healthy economy.

The post in question reads as follows:

I asked my Waymo driver a question about job security, but they didn’t respond.

Just me...at a red light, talking to an empty driver’s seat in downtown San Francisco.

I watched the wheel turn left, right...smoothly, perfectly, silently. 😳

And that is when it hit me.

This car eliminated its own driver…but it created more jobs than I expected.

The Waymo app showed me:
- Remote operators watching over fleets
- Safety teams analyzing edge cases
- Engineers pushing updates daily
- CX folks answering questions like mine and monitoring the cameras

1 job gone = dozens created

The empty seat isn’t empty.

It’s filled with humans....like data scientists in Phoenix, safety monitors in Mountain View, engineers across three time zones.

My “driver” exists in 10 different forms I hadn’t imagined.

So the next time someone says “AI is taking all the jobs,” remind them...

the wheel may have no hands on it, but the future of work has more fingerprints than ever.

#AI #Waymo #FutureOfWork #Automation #Jobs #SanFrancisco #Innovation #HumansAndAI #AIinWork

I think I was curious about this post because of all the AI bullshit these days, Waymo seems sort-of benign or at least not nearly as horrifying as literally anything OpenAI is doing. So, this post is functioning like a palatable insinuation that, “hey, maybe this AI thing isn’t all bad, right?” Plus, Waymo tell us they are pretty safe,1 and if they’re also creating jobs on top of having a decent safety record, I mean, maybe those “AI is taking all the jobs” people need to chill out.

Now, my immediate reaction was, of course, dismissive. But I sat with the post, and, if we’re being generous and taking its claims seriously—and assuming the best of the author’s intentions and motivations (which we probably shouldn’t because they are a self-proclaimed “thought leader in AI”, but whatever)—then we have to do some work to actually disprove the claims they’re making, to make some kind of meaningful counterclaim, or, perhaps, to suggest that the claims they’re making don’t actually bear the significance that they’re ascribing to them.

So I’ll refrain from dunking on the person who wrote that post as we might dunk on, say, Sam Altman, who clearly doesn’t actually understand anything about his own technology or about the world that his aspirations for hyperscale data centers are about to destroy—like, say, that a lot fewer humans will exist to benefit from his supposedly human-benefitting technology if there’s no fresh water left because it’s been used to cool all of his stupid data centers.

Instead, I’d like to take the person who posted on LinkedIn at their word and try to figure out why it’s misleading to suggest something along the lines of what I think the thesis of that post is, which is, “Waymo is, on the whole, good for workers because it actually creates more jobs than it eliminates.”

There are a few approaches by which we can examine the veracity of this thesis. First, we can think about the extent to which any technological innovation under capitalism is actually good for workers. We’ll come back to this later.

But we can also examine the specific claim that supports the thesis, which is, at least in this post, namely, that, for every driver job lost to a Waymo autonomous vehicle, the company (read: Alphabet/Google) creates at least that many jobs if not several times that many. It’s possible that this post is limited in its scope to the author attempting to push back against a (valid, though she’s certainly not treating it this way) concern about the possibility of job loss due to automation.

The best rough-and-ready metric for this would be to just divide the number of Waymo employees by the number of Waymo vehicles in their “fleet.” Depending on your data source, Waymo has somewhere between 2,500 employees (Wikipedia) and 3,300 employees (LeadIQ). Waymo has, roughly, 2,000 vehicles in their fleet (source). So, yes, there are, not even counting third party contractors, more than one Waymo employee per vehicle. On this measure, the author of the post is indeed, at least partially, correct. For every “driver” Waymo has “replaced,” there are at least as many jobs currently in existence. Do jobs-in-existence equate to “jobs created”? I don’t necessarily think so. And, given these numbers, I certainly wouldn’t say “dozens.” That would require Waymo’s staff to be like 20-some thousand.

This is also just a snapshot in time and it isn’t the kind of durational data you’d need to actually figure out if Waymo is really “creating” jobs. We can make some inferences about this based on how tech companies—and capitalists, more generally—do things. And, spoiler alert, it’s not good for workers.

It’s just not economically sound for Waymo to grow in its workforce in any way that resembles “for every job lost, dozens created,” unless they’re charging riders and unconscionable amount per ride or don’t ever want to be profitable (they do). So at some point, the investment in staff (setting aside subcontractors and precarious workers in the Global South for just a moment) has to plateau. It’s not like Alphabet/Waymo is some kind of benevolent corporation that isn’t seeking to squeeze every last cent of surplus value out of its employees. However, at least as of May of this year, Waymo is not profitable. But they’ve made some proclamations about when that profitability will happen and their potential for profitability is something that a not insignificant number of people regularly speculate on. According to TechCrunch, however, at least as of April of 2025, “It’s not clear how far Waymo is from breaking even, much less profitability. Alphabet doesn’t break out Waymo’s financials in its earnings report. Instead, Waymo is included in Alphabet’s “other bets” section of its balance sheet, which in 2024 recorded an operating loss of $1.2 billion.” One source of potential revenue for Waymo’s parent company has always been the monetization of data captured through its pervasive surveillance apparatus. And, indeed, a draft of Waymo’s new privacy policy demonstrates the continuity of this practice across Alphabet/Google’s business units.

According to TechCrunch’s Rebecca Bellan, “Waymo is preparing to use data from its robotaxis, including video from interior cameras tied to rider identities, to train generative AI models…” as well as to leverage that data to further personalize ads (which is something Waymo already does).

OK, well, yeah, any company owned by Alphabet will do some sicko surveillance-for-ads-under-the-guise-of-convenience shit. Which is inherently bad for the working class because (1) Google continues to “enclose” various aspects of digital and physical-to-digital “commons”; and (2) all that data has to get stored, processed, inferences run on it, models trained/re-trained with it, and so on, which all uses a bunch of energy and water that no one has anymore and the consequences of which will be felt by working class people all over the world in the form of ongoing climate catastrophe, ranging from water shortages to increased energy bills to rolling brownouts to the aggregate (and much larger) consequences of increasing CO2 emissions in the form of increasingly volatile weather, droughts, superstorms, and so on.

Anyway, one of the “dozens” of jobs created, the author of the post suggests, is “remote operators watching over fleets.” We don’t know the ratio of remote operators to actual Waymo cars on the road, nor do we have a sense of Waymo’s internal metrics for what this ratio can or should be. But we can get a sense of Waymo’s aspirations based on how they talk about “fleet response.” They write: “As the Waymo Driver improves over time, it can solve more ambiguous scenarios independently and needs less help. But as with the rest of our operations, a helpful human is no more than a touch of a button away.” (Fleet response) Based on this as well as other Waymo corporate literature, it seems to me that many of the current employees at Waymo (and, more likely, their third party contractors both in the US and abroad) are actually serving to eliminate their own jobs. But of course when you’re an independent contractor or otherwise precarious employee doing what is effectively piecework in training automated systems, the company doesn’t care about whether your job technically exists or not. And you probably don’t have much of a choice about whether you’re doing a job the long-term viability of which you’re actually helping eliminate because you need to put food on the table.

Regardless, there will be, at least for the foreseeable future (and I’m talking like a really long time here), humans doing at least some of the driving in supposedly autonomous vehicles. The question for a company like Waymo is the extent to which you can ensure that any of those humans get paid as little as possible relative to the skills required so that you can pay your investors and executives as much as possible.

Now, as far as I can tell, many fleet response specialist jobs are through third-party contractors, meaning that these workers don’t get the benefits and perks of being official Google/Alphabet/Waymo employees. At least according to posts on Reddit, Glassdoor, and Indeed, many work as independent contractors facilitated by companies like TransDev. This is significant because we’re talking about healthcare, vision, dental, retirement, etc. The kind of safeguards, that, say, a union might fight for.

But this still doesn’t even capture what I see as a more important dimension of this conversation, which is the history of the taxi industry and the change in the nature of this type of work since the “innovation” (read: regulatory arbitrage) of Uber and Lyft in the early 2010s. It also doesn’t capture the centrality of the privatization of transit—and the privatization of everything else—to our ideas about how people should or could get around in cities and beyond.

The person whose work you MUST read when you think about the history of taxi-like transportation, especially in the Bay Area, is Veena Dubal. So I think it’s best that we see the kind of historical continuities of the neoliberal shifts in the taxi industry in the late 70s and 80s to the regulatory-arbitrage-slash-“innovation” of Uber and Lyft in the 2010s, and now Waymo and other AV-taxi systems through the lens of her writing. Prof. Dubal’s two most germane essays for our purposes here are “A Brief History of the Gig,” originally published in a 2020 issue of the excellent Logic magazine and “The Torched Waymos: Burning Effigies of Tech Oligarchy” (June 2025).

In “A Brief History of the Gig,” Prof. Dubal traces the historical continuities across the decline of unionization in the taxi industry at the onset of neoliberalism through the gigification of drivers under the thumb of Uber and Lyft in the 2010s. She writes:

The independent contractor model that underlies today’s gig economy first developed in the taxicab industry in the late 1970s, as the United States shifted towards a neoliberal conception of society in which almost everything was to be subjected to the forces of competition. Workers and households were reimagined as entrepreneurial concerns, sole proprietorships that should fend for themselves in the tumult of the great American marketplace. Embodying this logic, the taxicab industry was one of the first among US businesses to slough off the costs and liabilities—minimum wages, healthcare benefits, disability insurance, among other things—associated with direct employment.

And even though the Chauffer’s union, which had represented cabbies in the city since before the New Deal, was, in many ways, alienating its own membership through its political conservatism (and its connection with that era’s defining corporate unionism that we might call “concessions bargaining”), the gains it had won its membership included “municipal regulations on fares and restrictions on the number of taxis operating in the city” as well as ensuring that “drivers earned a living wage.” But, as Prof. Dubal writes:

Taking advantage of rank-and-file discontent with traditional unions, as well as an existing carve-out for independent contractors in labor laws, cab companies all over the US reorganized their business models in this period, claiming to “free” their drivers from the imagined restraints of employment. In San Francisco, taxi companies approached drivers in 1976 and asked if they would like to “lease” their taxis on a shift-by-shift basis, so that the companies could rid themselves of the expenses and risks associated with employing workers, including unemployment insurance, workers’ compensation, and a guaranteed commission that taxi drivers had previously been paid.

Eventually, this neoliberal-individualist and entrepreneurial ethos manifested in the further gigification of transport labor through the illegal (but somehow not prosecuted) “innovation” of Uber and Lyft.

And while the scope of Waymos’ territory is relatively small, early indications of its encroachment on the already precarious taxi-like transport market suggest looming problems for “human” drivers. A Business Insider article documents these emergent concerns:

Jason D., a 50-year-old Uber driver in Phoenix, told Business Insider it had become harder to make money in recent years because of increased competition with other drivers, lower fares, fewer tips from riders, and higher operating costs. Now, he said, the rollout of Waymo One robotaxis has made this problem even worse.

“Driverless taxis are flooding an already competitive Phoenix market and taking money from human drivers,” said Jason, who drives full time and asked that his last name not be included for fear of professional repercussions.

Meanwhile, according to the BI article, Benedikt Frey of the Oxford Internet Institute, is unsurprised:

“As the technology gets better and cheaper, drivers will feel it in their wallets,” Frey said. “We’ve seen this movie before: When Uber first showed up, it reduced traditional taxi drivers’ earnings by about 10%.”

Seems like a little bit of a problem. Doesn’t it? Or are there so many “fingerprints on the wheel” that we’re not worried? Everyone seems to be chill right now, though, because only in Phoenix can Waymo vehicles actually do airport pickups. BI again:

Nicole Moore, a part-time Lyft driver who’s also the president of the driver advocacy group Rideshare Drivers United, previously told BI that trips to and from airports were many drivers’ “bread and butter.” She said competition from robotaxis for these rides would be a concerning development for drivers.

But, this concern—even if speculative—only captures a small slice of the historical shift in the taxi-like transportation industry, and that shift reflects the centralization of wealth and power in society as a whole. And that centralization is manifest not only in the history of the privatization and gigifaction of taxi-like transport, but in the history of the tech industry in the Bay Area, ranging from Google to OpenAI, in a place where homelessness is so egregious that UN experts have excoriated the city over the issue.

After describing the various forms of opposition to Waymo being forced down the throat of an already suffering working class in San Francisco, Professor Dubal notes in her excellent essay on the burning Waymos of San Francisco:

The prioritizing of technology capital’s interests over the concerns and lives of everyday people — first responders, transportation workers, the people who rely on access to public transportation — is particularly egregious. In a microcosm, it is how people feel about neoliberal governance more broadly — the sense that their lives are treated as secondary to elite interests.

++

Overall, to me the most compelling critique of the LinkedIn post that kicked off this whole essay is as follows:

It doesn’t matter if AI is taking all the jobs or not, because the manner by which technological innovation under neoliberalism happens will only harm the working class and harm the environment and you trying to be clever by telling us, “the wheel may have no hands on it but the future of work has more fingerprints than ever” is completely fucking irrelevant because it doesn’t change the nature of technological innovation under neoliberalism, which is fundamentally antidemocratic and anti-worker. That’s why I think it’s stupid to celebrate Waymo “creating jobs” because the workers themselves should be the ones who determine the course of technological innovation in the first place.

And until we—the international working class—are in control of the means of production, and until we can democratically determine the course of development of all technological innovation in society—deciding whether we want autonomous vehicles, how they will work, what environmental costs we are willing to incur and how to distribute the burden of those costs without harming the already vulnerable, and so on, in other words, all the decisions that should be taken into account when any new technology is conceived much less produced—then we cannot ever concede that Waymo or any other technological innovation that seeks to generate profits through its exploitation of both people and planet is somehow “good for workers.”

So as the Waymos burn, remember that those Waymos and the capitalists that owned them will never genuinely seek to benefit the working class. This claim is supported by a vast and heterogenous body of empirical evidence that is political-economic, historical, and environmental in a way that no cute saying about wheels and fingerprints ever will be.

I cannot overemphasize how monumental this victory was. For more than five years, I have participated in our organizing effort, which began in earnest in 2019. For the past two years, we engaged in negotiations and arbitration with MSU administration and their expensive, union-busting outside counsel, Ogletree Deakins (one of the most famous union-busting law firms in history), to determine the composition of the bargaining unit that they would, at every turn, attempt to avoid recognizing. All the while, we continued to organize, grow our membership, and build the kind of organization that will truly give faculty the voice they deserve in our institution. Finally, just a few weeks ago, our union was recognized and the real work of bargaining our first contract—a contract that will benefit our membership in ways that reflect the priorities that we collectively and democratically articulate—can begin. If you are interested in some background on our effort, my brilliant colleague, Karthik Durvasula, has written a terrific and detailed piece.

I began helping organize our union for a range of reasons, each of which has taken precedence in my mind at different points throughout the campaign: I am committed to the democratic governance of my institution which cannot be done through a faculty senate that is merely “advisory” to the administration; I believe that unionizing is a demonstration of solidarity with other members of the working class, and in particular those workers on campus who are already unionized; I believe a union can offer concrete protections through contractual provisions that the institution of tenure, under constant attack and at risk of disappearing outright, fundamentally cannot; I believe that unionizing is itself an act of solidarity that undermines the false meritocracy that is, in spite of universities being full of very smart people who should know better, part and parcel of the neoliberal academy; I believe that the solidarity and collectivity fostered through the process of unionizing is a gesture at something better than the society we have now, that, perhaps, it can help us understand certain aspects of what an anticapitalist society might look like.

The history of trade and industrial unionism, particularly in the US, is defined in part (at least in my very-much-not-a-historian brain) by a struggle over what the function of unions are and what their role is within organizations and society more broadly. These tensions are evident in books like The Long Deep Grudge: A Story of Big Capital, Radical Labor, and Class War in the American Heartland by Toni Glipin, as well as Jon Melrod’s excellent Fighting Times: Organizing on the Front Lines of the Class War.

The questions of the purpose of a union, how we mobilize around that purpose, how that purpose is decided, who decides it, and what are the edges of acceptable or possible purposes in the first place, are questions that we as a union have had to grapple with. And, as academics, it is hard for us not to engage some of these more “philosophical” questions about unions. But these questions also have impacts in how we organize on the ground because they say something to us about what our members value. When we arrive at the bargaining table, we must have a very sophisticated understanding of what our members think we are, what we value, and how what we value shapes the priorities for which we will fight in negotiations. What I think is crucial for the purpose of this essay is that there is a particular material reality that gets produced by the expression of shared purpose and the normative values which made that purpose acceptable or desirable. I’ll come back to this key idea in a bit.

++

A few years ago, one of the things that we on the UTSF Organizing Committee had the opportunity to do—thanks to our affiliation with MEA/NEA—was to attend one of the last trainings Jane McAlevey did before she passed away. This training was facilitated by the UC-Berkeley Labor Center, and, not only was it inspiring for me, but also ensured that now I get regular updates from the Labor Center on all the important work they’re doing.

As I was celebrating our recognition, I got an email from the UC-Berkeley Labor Center about a new report they released, titled, “Labor’s AI Values.” Most of my research these days centers on genAI and often uses higher education as a case study for understanding the impacts of genAI in society, so this report seemed quite germane to my research interests—interests that have taken on a more explicit class analysis and political-economic flavor over the past couple of years.

The report is short (12 pages), so it’s worth reading in its entirety. But there are a few things that stood out to me, and I’d like to talk a little about those things, why they stood out, and what I think this report indicates from a broader historical perspective, especially with regard to the intersecting histories of the labor movement and technological innovation under capitalism.

But before I talk about the “Labor’s AI Values” report, I want to take a brief detour through the pages of Toni Gilpin’s compelling and comprehensive history of labor organizing at the International Harvester corporation. Titled The Long Deep Grudge: A Story of Big Capital, Radical Labor, and Class Warfare in the American Heartland, Gilpin’s book has been, for me, almost impossible to put down.

The Long Deep Grudge is ostensibly about the history of unionization efforts at the International Harvester Corporation, which began its life as the McCormick Harvesting Machine Company, in Chicago, Illinois in the 1840s. But the book is about so much more than that, because it also traces the radical labor history of the city of Chicago and demonstrates the central place it should occupy in the lives of working people across the country to this very day. I can’t recommend the book enough, especially for us Midwesterners who grew up perhaps attending the auto show at McCormick Place in Chicago in the 1980s and 90s. There’s something deeply disturbing about that for me now that I’ve been exposed to Gilpin’s work. But my childhood memories of seeing (union-made) automobiles at an exhibition hall named after one of the most tyrannical anti-labor industrialist families in the history of the world will have to wait for another day.

One of the most profound insights of this book for me is how the ideological tenor of union organizing reflects the horizon of political possibility in society. What I mean by this is best expressed by Gilpin in the introduction to the book. Gilpin uses the origins of the United Farm Equipment Workers of America (FE)—its radical roots, fomented by anarchists and the IWW (the International Workers of the World, more commonly known as the “Wobblies”) and solidified by organizers who were members of the Communist Party—and its subsequent merger under the umbrella of the much more centrist UAW, to articulate a compelling, inspiring, and also sobering history of the labor movement in the US. Gilpin writes:

In late 1979, as the Chrysler Corporation teetered near bankruptcy, the UAW consented to stiff wage and benefit cuts and the age of concessions bargaining was born… The collapse of the labor movement meant there has been little to stop the already rich from getting even richer by raking back more of what the working class has earned. Once again, 1979 marks a key demarcation point: Before that year, as the US economy grew, so too did workers’ wages and benefits, as would be expected in any reasonable equitable society. Since then, however, as national economic productivity has continued its upward trajectory, only the fortunate few have reaped the benefits: incomes for those at the top increased exponentially, as have corporate profits, but the bottom 90 percent of all wage earners have seen their compensation stagnate or decline.

This 90 percent includes myself and many of my colleagues (notably, our president and provost sit well beyond this 90% percent threshold). Of course, the curve gets awfully steep once you get past that last 10%, and the top 5% of income earners in the US earn somewhere north of $350k depending on what source of data you’re using (for the record, our university president and provost are also above this threshold). And then after that 5% mark things get super crazy, but you get the idea.

Anyway, Gilpin goes on to write:

Labor became tethered to the premise of an ever-expanding American economy; when the economy faltered, union leaders were left without a framework to clearly assess who was doing what to whom, and they lacked the vocabulary to articulate a challenge to this crisis that might, at least, have allowed workers to channel their fear and anger. Instead, those feelings have festered into cynicism and despair and the demoralizing sense of personal, rather than systemic, failure. Unions seemingly did all they could possibly do within the constraints of American capitalism, ratifying what might be characterized as the labor addendum to Margaret Thatcher’s dictum ‘There Is No Alternative.’ Within its brief lifespan, however, the FE endeavored to counter that notion, strenuously insisting that there was, in fact, a contrasting conception of the proper role for unions in America.

The idea that the economy can grow forever without limits—a growth enabled by, and fostering of, technological innovation—is the most dangerous fairy tale ever peddled by the capitalist class and its promotional apparatus. The fairy tale of infinite economic growth both producing and powered by technological innovation has been thoroughly, empirically, debunked. There is irrefutable evidence, for example, that technology will not “solve” climate change (Hickel, 2020), and, furthermore, that technological innovation under neoliberalism has served the extraction of profit at enormous human and environmental costs (Blakeley, 2023). And yet, the rhetoric of economic “growth” maintains its grip on the public imagination. But even “green growth,” championed by the most seemingly “progressive” of politicians, has been deemed empirically baseless (Hickel, 2020, Ahmed, 2018). Nonetheless, the myth of the possibility—and necessity—of continued economic growth, as well as the role of technological innovation in that growth, is so powerful that many people, ranging from laypeople to economists to politicians, will go to extraordinary lengths to justify its logic. Including, at least since 1979, “big labor.”

But it is precisely this idea—that infinite economic growth is possible and that it can be both fueled by and productive of technological innovation under some variant of capitalist political economy—that is itself anti-worker to its very core. This is because it is a lie, and the capitalist class is notoriously good at lying to workers. So, for anyone in the labor movement to take this premise as anything but suspect is… not great.

++

So now, given the historical context of organized labor in the US since the onset of neoliberalism—which coincides with the concession-oriented and more “pro-business” turn in the UAW that Gilpin describes—I think we can glean some important insight from the “Labor’s AI Values” report, especially with regard to the relationship between labor and technological innovation.

One of the first things I noticed is that there weren’t statements from labor organizations that offered some technological context for their comments. So, as far as I could tell, they were most often referring to “AI” without qualifying this term in any way. This is important, because I worry that generative AI has come to dominate the popular imagination in a way that keeps us from appreciating the differences between genAI and other types of “AI” research and application, including algorithmic systems of inference and recommendation or Machine Learning.1 I will continue to refer to generative AI here, specifically, because of its cultural prominence and because it is being pushed by the biggest AI companies towards enterprise applications, meaning efforts integrate genAI (specifically) into the workplace at scale, in spite of a number of reports about its lack of utility even for the corporations that have sought to embrace it.

In the statements from the various labor organizations presented and synthesized in the “Labor’s AI Values” report, they often say “workers should have the right to” something-or-other, or that “the right to” something-or-other is “important” or “essential.” But how will workers get that right? Is that possible, in this day and age, to negotiate such rights into our contracts? Do we have that power? The answer to that question probably varies quite a bit, but I wonder how many of our unions really have the power to get written into our contracts “the right to be protected from any potentially discriminatory impacts of [AI in hiring, promoting, and firing of employees]” (p.4) or “the right to participate in the design, development, and procurement of technology from the outset” (p.6), especially given how many third-party vendors most employers use for purchasing and maintaining technology. These two specific assertions of rights fall within the sections titled “Laying down the rules for responsible tech use” and “Centering workers,” respectively. Does the working class currently have the power to “lay down” any kind of rules about “responsible tech use” or to force the capitalist class to center workers more than they center their shareholders?

Meanwhile, under “Improving jobs and livelihoods,” the National Writers Union is quoted as saying that “human creativity is ‘the work of millions of human lives’ and therefore must be protected from exploitation.” In this same section, the report states that “Many groups affirm that workers have the right to a ‘just transition’ as digital technologies transform the labor market (UNI, CWA, NWU, IATSE). Employers should provide notice prior to layoffs, retraining, priority bidding for comparable roles, severance pay, and partial wage replacement (CWA).” And, “other groups look to governments to provide social security, lifelong learning, and retraining so that workers can take advantage of new opportunities created through AI.” I am skeptical that there will be new opportunities created by any genAI technologies developed under capitalism that do not conflict with the value of human creativity being protected from exploitation. Especially when that creativity requires things like food, shelter, and medical care, not to mention a decent internet connection. I am, furthermore, skeptical that a “just transition” is possible under anything resembling the current sociotechnical and political-economic systems under which we live. I don’t think I need to provide any evidence to substantiate that claim [gestures around and looks despairingly at the world historical wealth inequality that is worse even than the “gilded age,” as in, you know, the era that prompted the radical activity that resulted in the unionization of International Harvester].

There were also some quotes from certain labor organizations that I felt were particularly naïve and/or reflected the relatively limited political horizon of unions in the US. Or at least they really needed some explanation. For example, the UNI Global Union is concerned with “ensuring that AI ‘serves people and the planet’ and is compatible with ‘fundamental human rights.’” Now, given what we know about the extent to which genAI: (1) undermines the dignity of workers and all people, really; (2) is, at best, incompatible with, and is, more likely, catastrophic to the life of the planet; and, (3) has only undermined—not served—human rights… I feel like that’s one of those ones where you’re like, “come on, really?”

So, while the report does a good job of summarizing a number of organization’s aspirational ideals, there are certain aspects of it, legible through what is implicit in some statements or absent in others, that I think speak volumes. The aspirational ideals expressed by the labor organizations surveyed articulate a political horizon that, at the very least, accepts some key premises of capitalism that folks like the Farm Equipment Workers of America (FE) back in the early 1900s weren’t readily willing to accept. In describing the Communist Party’s influence on the FE, Gilpin writes: “for the grudge match between this powerful company [International Harvester] and this radical union hinged on something more fundamental than how much workers should get paid: it came to be a struggle for control over the very nature of work itself.” (Emphasis mine)

For me, this is what is so profound about books like Gilpin’s or like essays such as Missé and Martel’s “For Democratic Governance of Universities: The Case for Administrative Abolition,” which advocates for the abolition of all university administrators (Presidents, Provosts, Deans, etc.) and shows how universities can be democratically managed without the interference and budgetary impacts of the highly-compensated administrative class. Such writing suggests that the parameters of political possibility for us as subjects of capitalism are not as narrow as we might think. And that’s the kind of political education that the FE, at its radical height, was able to offer its members. We should be struggling over the very nature of work itself, and if we’re not doing that, we’re doing it wrong. We’ve already ceded too much ground.

It feels like the fact that labor organizations across a number of industries don’t put forward more radical positions about AI—positions that more explicitly reflect the material realities that have been empirically documented by researchers and investigative journalists—demonstrates that the aforementioned fairy tales about infinite economic growth and entrepreneurial technological innovation have been accepted as fact by the leadership of these labor organizations. This tacit acceptance has pernicious consequences. In other words, we’re still living the legacy of the very process that brought the FE under the UAW and produced the contemporary, cynical, individualist, anti-labor conditions under which we live and that Gilpin describes in her book’s introduction. Now, of course, it is not only possible but quite likely that the rank-and-file of many of these unions hold significantly more radical positions than those expressed by their national leadership. But without concrete evidence that a labor organization can, at the very least, enforce contractually or through other means the values they hold about AI indicates that this report is worth examining in its historical context with regard to the relationship between organized labor and technological innovation under capitalism.

I also think that these assertions of “AI values” by the labor organizations surveyed in the report reflect the dangers of industrial unionism in the US, dangers the Wobblies saw more than a hundred years ago. The fact that there is only ONE mention of “ghost workers” in the Global South was shocking to me. Not only should labor organizations be aware of the exploitation of folks in the Global South that makes AI possible (and that has also made capitalism itself possible), but we should be militant about our solidarity with those folks. They are not only the most exploited when it comes to the production of AI tech (ranging from folks doing the tagging of data sets to folks mining the minerals that end up as capacitors or other components in servers powering genAI) but also the most vulnerable when it comes to the impacts of climate change that data centers for AI are themselves exacerbating exponentially.

++

“First, power. Then, cooperation.”
—David Brody, labor historian

First, power. Then, cooperation. This is the mantra by which I want to organize. And worker power means more than just the power of a particular trade or industrial union. It means worker power. International worker power. This is why the Wobblies were seen as such a threat, why the Haymarket affair was so crucial in the history of the FE and why the labor movement today needs to recuperate its own radical history. It needs more than concessionary stances, and it needs more than a demand of rights. It needs to take its rights from the capitalist class. And that needs to happen in every sector of society, including higher education. This means having a clear-eyed view of generative AI from beyond the purview of the recent concessionary history of collective bargaining and the acceptance of infinite growth via technological innovation under some variation of capitalist political economy that came to dominate labor in the 1970s. It means saying that we understand that the empirical evidence points to the fundamentally deleterious nature of genAI for both labor and the environment (which are inextricably interwoven, especially if we begin from a standpoint of radical international labor solidarity). And not only do we understand these basic facts about genAI, but that we refuse to use genAI and we refuse to be integrated into its pernicious feedback loops.

Since the earliest days of the industrial revolution, as demonstrated so clearly by Brian Merchant in his Blood in the Machine, innovation under capitalism has been fundamentally antidemocratic in character. By this, I mean technological innovations are foisted onto the working class by entrepreneurs who are aspiring or existing members of the capitalist class. The working class—ranging from information technology workers, to folks in factories assembling servers, to the folks working in refineries that yield the raw materials for the components that go into those servers, to the miners working in the mines that harvest the minerals sent to the refineries—has no say in the design criteria for the technologies to which they become subject, that they use, or that they build. The international working class doesn’t get to say if they think ChatGPT should even exist in the first place. They have no voice in whether Mark Zuckerberg’s stupid AI glasses are actually something that should exist or if the hundreds of billions of dollars invested into Meta’s AI projects, money which was made on the backs of the international working class, should instead be used to feed people across the world who need food. And it’s fucked up that the working class doesn’t have a say in any of this stuff. And unions should be at the forefront of saying as much. And not only that, but unions should be at the forefront of the kind of struggle demanded by the urgency of the situation in which we have found ourselves today—an urgency that demands a reckoning with the role of unions themselves in society.

First, power. Then, cooperation. Maybe.

++

I’d of course be remiss if I didn’t mention another sobering statistic with which Toni Gilpin opens her book. She writes that “by 2018, the percentage of the workforce within the ranks of organized labor had plunged to 10.5 percent.” In other words, even if unions were to attempt to enforce the values they espouse, or enact values that were even more in conflict with the historically concessionary approach of unions to business, how much impact could those of us with the benefit of being organized actually have?

To address this, I’d probably need to write a follow-up of some sort about the history of the concept of convenience, the Protestant work ethic, the myth of meritocratic entrepreneurialism, and the rise of the petit bourgeois in the US. Will I write that follow-up? Dunno. Things are bad.

Hi everyone — my name is Zach, and I work at Michigan State University and the opinions I express in this space are my own and not those of my employer.

I’m going to limit my comments to talking specifically about generative AI because it’s at the forefront of everyone’s minds right now and, unfortunately, has become synonymous with AI research writ large. Now, I’ll offer some points of departure for conversation, specifically with regard to the guiding questions listed in the agenda—namely: “Where do you see liberal arts making the biggest difference in how we use AI?” and “What’s one risk or opportunity you think people aren’t talking about enough?”

1. Liberal arts education wakes us up to the world around us.1 It teaches us to understand the world as it is and it fosters in us an ethical imagination2 for the purpose of changing that world. This is why fascists don’t want students to study it. It’s why they want colleges to become technical schools. Anything that allows us to get a handle on how the world actually works and who benefits from how the world works is thus constituted as “woke.”

2. Liberal arts education helps students see interconnections between peoples, histories, and ideas, and is thus ideally suited to systems thinking.

3. It follows then that the difference liberal arts education can make with regard to the use of genAI is that the liberal arts should lay bare the systems underlying and motivating its existence and use.

4. Such a project would leverage the mountain of empirical evidence to demonstrate that there is no meaningful societal benefit for the use of generative AI. Full stop. A serious liberal arts examination of genAI as a system would suggest the following:

   1. Generative AI is environmentally deleterious and there are no grounds to believe this will get better.

   2. Generative AI’s core capabilities and functionalities cater to the kind of authoritarian technocracy to which we have seen fascists—both contemporary and historical—aspire.3

   3. GenAI, like all other innovation under capitalism, must have a “business case.” Often, this means that any real social benefit is subsumed by the augmentation of surplus value, and the benefits that accrue to the working class are both secondary and disproportionately small to those that accrue to the capitalist class. The benefits that accrue to the working class often serve to engineer the consent required for the maintenance of capitalist hegemony.

   4. Generative AI is, like every other innovation under capitalism, fundamentally antidemocratic in the sense that those who are most impacted by it and those who are forced to use it have had no say in its design and development. This is, for the capitalist class, a feature, not a bug.

5. It thus follows that, looking at the second question on our agenda, the opportunity people are not talking enough about is twofold. I will call this opportunity “negation towards conviviality”:

   1. First, negation: by this I mean stopping the use and proliferation of genAI platforms. Totally and completely. Indeed, the Luddites of the mid 19th century were not “anti-technology” as history has painted them, but instead were the most sophisticated analysts of technology of their time, because they understood better than anyone the implications of the antidemocratic nature of innovation under capitalism. And they responded in kind. Brian Merchant’s Blood in the Machine is the best book on the subject right now. To end genAI as we know it today is thus to open a space of transformative possibility.

   2. Second, conviviality: the transformative possibility opened by the negation I just described is a possibility for direct democratic determination of all technological innovation in society. Everyone who uses or is impacted by any tool—ranging from computing to transportation—should have a say in that tool’s design and development. Such tools, in the parlance of philosopher Ivan Illich, would thus be called, “convivial tools.”4

6. The liberal arts also can help us avoid the trap of pessimism when presented with such a proposal—I mean, who could imagine the entire world stopping using chatGPT tomorrow much less the working class having a say in the development of all technologies to which they are subject?

   1. The liberal arts would teach us realizing this imaginary scenario requires the social bonds that capitalism has sought to destroy—a willingness to be in solidarity with one another and to participate in collective action.

   2. The liberal arts would teach us that perhaps a starting point toward this solidarity is labor organizing. It is the first step to a different world because once you start organizing, your idea of what’s possible, your political horizon, immediately expands.

   3. But the liberal arts would also teach us that we need more than labor unions. We need the kind of solidarity that cannot come through electoral politics, because representative democracy is a sham. It’s not actual democracy. Which is why it works so well for the capitalist class and not for the rest of us. Political theorist Camila Vergara has astutely demonstrated this in her excellent book Systemic Corruption, and David Graeber’s history of the concept of democracy is a good companion for her seminal work.

Indeed, Graeber wrote in another classic of his, Utopia of Rules, that “the ultimate, hidden truth of the world is that it is something that we make, and could just as easily make differently.” A better description of what the liberal arts should teach us you’d be hard pressed to find.

Subscribe now

A bit of context: a course about genAI

This semester (Fall 2025), I’m teaching a “special topics” course on generative AI. This course was open to students in our Graphic Design BFA program as well as our UX degree program. It’s ostensibly a “studio” course, meaning that its focus is on the production of designed artifacts of some sort. Part of how I see my role in the curriculum of both the Graphic Design and User Experience degree programs is to address emerging technologies and to think with students about their role as users of those technologies as well as designers of the interfaces to those technologies if not some functional aspects of the technologies themselves. This is, more or less, how I sometimes describe my area of expertise and the kind of research I do—I study the politics of technology, particularly as enacted through the interfaces to digital products and services (the title of my first book is Interfaces and Us). But over the last six years or so, the confluence of a couple different events—getting tenure, my involvement in labor organizing in the higher ed sector, and the emergence of deep learning as a mode both of analysis and production of scholarly activity—the focus of my research has pivoted to the intersection of AI and higher education.

Over the last several years, the discourse on generative AI within our institution as well as within the academic programs in which I participate as a core or affiliated faculty member has trended noticeably (if I’m being generous) overly-optimistic if not completely naïve. Now, if I’m not being generous, I’d call it willful ignorance or, perhaps worse, a deep cynicism that is reflected on the surface by a strangely enthusiastic resignation to the supposed inevitability of genAI by trying to demonstrate its ability to “enhance learning” or “be a partner in creativity” or some other absolute total and complete bullshit. And any apparent naïveté is, of course, not naïveté at all, but instead a function of the hegemony of technocratic-authoritarian capitalist political economy.

This commitment to a political-economic orientation to understanding the impact of various technologies on everyday life is central to the unique approach I take to my research. My expertise is defined by my ability to synthesize methods of analysis from different fields (e.g., Science and Technology Studies, Design Studies, Cultural Studies, Political Economy) in order to address the implications of various technologies on the people who become subject to those technologies, either through using them or through participating in their design, development, or production. Having worked as a designer in advertising and tech, I also have first-hand experience of attempts to both produce certain effects on users and shape societal discourse on those effects.

So I felt uniquely well-suited to teaching a class on genAI for designers. And I thought this semester might be a good one for this class, since genAI has been at the forefront of so much conversation, both within the industry (you can read my thoughts on that if you’d like) and within academia (you can hear me talk about my thoughts on that as well).

The structure of the course is relatively simple—it looks at five facets of genAI: its functionality (how does a genAI system actually work?); its origins in the rocks and minerals of the earth and the material requirements to produce an actual server; the labor, both to harvest and process the raw materials for servers but also the data labor of tagging, training, and content-moderation; the environmental requirements of data centers; and, the ideological and financial dimension (i.e., who really benefits from the use and proliferation of genAI?). The course seeks to help students find robust and empirically rich ways of understanding the impact of genAI on those subject to its use and production.

Our first two classes of the semester were interesting and energizing for me, in part because so many of the students came to the course without the kind of “rose-colored glasses” that I had—unfairly—expected them to have about genAI. These students seem to have a sense that something is amiss, and they can point to a number of the ills that genAI is producing. This, I think, is a good start, and it is going to give us some more space to think about the production of communicative design work that can push back on the dominant narratives about genAI within the university and beyond.

The question and my response

But at the end of our second class, a student came up to me and asked “so, is this course more of a ‘negative’ view of genAI?” This got me thinking.

I kind of demurred when answering this question, and said that it was more about “consequences.” But I’d like to explicitly address this idea that there is a “negative” and thus a “positive” view of genAI, because I worry about the “two-sides-ing” of something so consequential.

An important caveat (with which I began our semester and which I mention with some regularity in any conversation about genAI in which I get involved): I’m talking specifically about genAI here and not all “AI” research, which is much broader and includes stuff like machine learning and systems of algorithmic inference. Furthermore, I’m talking specifically about genAI as it exists today and not in some idealized fictional world that doesn’t have the profit motive and juicing quarterly returns as its primary objective.

Now back to this idea that there is some kind of “negative” and “positive” debate about genAI. The short version of the story is that there are not “two sides” to any “debate” on genAI. There’s not even really a “debate” worth engaging in when we actually do the kind of research that as research-academics we should be expected to do. There are only two sides if you listen to someone like Sam Altman who has a vested interest (a very large one, given his net worth) in the existence and ongoing development of the technology. I think it’s naïve (or again, perhaps deeply cynical if not nihilistic) to give him—or anyone else with a substantial material and financial stake in the continued development of genAI—any credence in any argument about its social benefit. No serious scholar would listen to his arguments without at the very least pointing out a massive conflict of interest.

It also seems to me that, in my many conversations with computer scientists and engineers over the years, that there is a relatively frequent (and very patriarchal-slash-white-male-testosterone-infused) smugness with which many technical experts address those of us who might be considered “non-technical.” Altman and his ilk are absolute paragons of this type of behavior, thinking that they know what’s best for the world even as they understand less and less of how the actual world works. This is, of course, evidenced by their myriad statements on the kinds of things that they think AGI (Artificial General Intelligence) can achieve for humanity in spite of the preponderance of empirical evidence, much of which is written by actual scientists, that the rush to create AGI has been and will continue to be catastrophic for the environment and for the vast majority of humans on this planet.

Now this isn’t completely uniform behavior across the CS and engineering fields of course. There are plenty of reasonable folks who would respect my, or any other scholar’s, expertise or experience. One of my best friends is head of engineering for a division of a major entertainment platform and he’s fully aware of the havoc he’s wrought on the world and is certainly more than welcoming of my critiques of his employer. He also majored in History.

In other words, CS and Engineering folks are rarely taught something that is basic knowledge in the humanities and social sciences—all technologies are political and they emerge from within a certain political-economic situation and tend to reflect the asymmetries of power and privilege of the political-economy from within which they emerge. Technology is never neutral and never could be. This is why the vast majority of technological innovation under capitalist political economy—particular the neoliberal and increasingly neoliberal-authoritarian variety we experience today—serve the interests of the shrinking, centralizing, oligarchic capitalist class and continue to exploit the working class. Of course, that exploitation has to be justified and/or concealed, and the trojan horse of “convenience” tends to do a very good job of this, especially when immiserated working class folks are constantly trying to optimize their own lives because they are treated like computers by the world around them—a world constructed by the very people invested in ensuring genAI’s widespread adoption and use.

My point here is that this course is an attempt to get a handle on what is actually happening in the world—it is a material analysis of genAI as it exists in the world today and it looks through a serious material and financial lens at the situation as it is, not as those with power and privilege want us to believe it to be. This is why I think it’s a mistake to suggest that there is a “debate,” or that there’s a “positive” and a “negative” side to the situation we find ourselves in with regard to the computational infrastructure of everyday life. And that’s why I’d suggest that we avoid such language when talking about that situation.

+++

PS—I thought I’d spend some time putting some links in this message but I really appreciated what Ed Zitron said about how absurd it is that the burden of proof—precisely because of the asymmetries of power and privilege in society that I just talked about—is on those of us construed as “skeptics” of genAI instead of on the “booster.” So maybe I’ll add some links later. Maybe not. But when corporate promotional rhetoric masquerades as “fact” or when someone like Sam Altman can make absurd proclamations without them being interrogated in the slightest by the press, I think I’ll just say what I’m gonna say and add some footnotes later.