Socialists and AI: No other choice?

Thus [the bourgeois economist] saves himself from all further puzzling of the brain, and what is more, implicitly declares his opponent to be stupid enough to contend against, not the capitalistic employment of machinery, but machinery itself.

—Karl Marx

As Marxists, we cannot disentangle Artificial Intelligence (AI) from capitalist political economy.

AI is not, in fact, “intelligence” but rather a system of sophisticated algorithms that do not generate new ideas but can only spit out what it has been fed. How tech companies developed these algorithms was through “data scraping,” taking large amounts of personal and private data to teach its machines.

AI is also not “artificial.” AI is powered and built by a large force of workers from the Global South, including Kenya, Venezuela, the Philippines, and India, among others.

Moderators for digital platforms work 18 to 20 hours a day sometimes, at a rate of $1.46-$3.74 per day, looking at incredibly violent and disturbing text passages and images with no breaks.

Of course, these digital companies offer no support to these workers, with many workers suffering from post-traumatic stress disorder and deep psychological damage.

Many attempts to organize workers at these digital sweatshops have been crushed. Daniel Montag, a worker in Kenya, told Time that workers at a digital sweatshop subcontracted by Meta were all fired after attempts at unionization.

Reading these reports is eerily reminiscent of the piecemeal sweatshops that Karl Marx described in Capital: worked to the bone under confined conditions for pennies.

This is coupled with the intensive fossil fuel and water use that AI data centers require, with AI data center needs exceeding the already heavy burden of other digital services such as Netflix and Zoom.

Digital infrastructure is not immaterial: the virtuality of the digital merely mystifies the grueling labor and extraction that enriches tech companies.

However, the moral arguments used to reject generative AI unfortunately often fall flat in the imperial core. Many ask, “Which major industry isn’t fueled by the blood of workers in the Global South?” Or, “Is it possible to be morally pure in a capitalist economy?” Of course not. But, if such arguments are not compelling, I offer here a few different arguments about why organizers on the Left should reject AI as it is being developed under capitalism.

Community

Writing, artistic expression, learning, and thinking are social acts. When we write or create art, we do so with the goal of reaching an audience intellectually or emotionally.

When we read and think, we learn from writers and process what society, past and present, has communicated.

As I often teach in my own writing courses, a key aspect to becoming a good writer is awareness of your audience and taking the opinions of others into account.

For organizers, the social aspect is particularly important. As historical materialists, our writing and thinking should be continually developing with and in community. Generative AI takes away this community aspect, and the only interlocutor is the faceless user interface of Claude or ChatGPT.

What potential organizing opportunities do we miss when we outsource community dialogue to machines? And why would we want to outsource idea-generation to machines that aggregate reactionary views?

As studies have shown, AI algorithms commonly make decisions based on racial stereotypes.  We should not risk reproducing these same reactionary tendencies in our own organizing by trusting machines that are often designed for surveillance and military applications.

Also, nothing in the past five years has suggested that those in charge of these large AI companies, such as Mark Zuckerberg and Elon Musk, are interested in fixing the racist and reactionary tendencies of these systems.

The social effects of AI are horrifying: AI psychosis, cognitive decline, heightened surveillance, and militarization. When the U.S. military horrifically bombed a girls school in Iran, strong evidence suggests the target was chosen by an AI algorithm.

When organizers use AI, whether for writing or graphic design or anything else, we are feeding tech companies that actively seek to thwart our aims.

Feeding these tech companies our data also gives the surveillance state more ammunition to use against us. Our thoughts and ideas should be ours, not shaped by AI or given to tech companies as another data point to further suppress and surveil socialist organizing.

Also, AI art is sloppy and ugly, and AI writing tends to be poorly written and formulaic. When restaurants use AI art, does the food not look like it’s infested with worms?

Relying on AI hurts our ethos as organizers. Earlier this year, Vijay Prashad and Tings Chak were caught using AI to write an article on China and Uyghurs. The deeply problematic politics of their argument aside, the AI article contained several hallucinated sources that detracted from whatever credibility the article and the publication, Monthly Review, had.

If our goal is to reach the wider community, we should not give the public easy grounds to dismiss us. And if we believe in the power of labor and creativity, our materials should reflect that.

We can learn how to write and draw together, or at the very least pay competent writers and artists for our materials.

This belief in community and labor should be centered in our thinking on AI, which leads me to the threat of AI and automation to labor.

Labor

Taste is ultimately subjective, and some people do genuinely love AI slop when it generates fingerless hands or worm-infested hamburgers. So, when discussing AI with the broader public, our arguments should center on labor.

The dream of tech companies is to depress the value of labor power, so the capitalist class can extract more surplus value.

We are beginning to see AI used in multiple industries: automated grading and teaching to replace educators, digitally generated art and content to replace artists, and vibes coding to replace software developers and engineers.

AI automation is not only targeting jobs of the professional managerial class either. Many companies are building AI-enabled robots to begin replacing jobs in factories and oil refineries. Even gig work is under attack by automation through companies such as Waymo and Starship Technologies.

The proliferation and development of AI poses a grave threat to workers. According to one tracking tool, roughly 265,000 jobs have already been cut due to AI.

The specter of AI is used in the way Marx describes automation in Capital, Volume I:

The instrument of labour, when it takes the form of a machine, immediately becomes a competitor of the workman himself. The self-expansion of capital by means of machinery is thenceforward directly proportional to the number of the workpeople, whose means of livelihood have been destroyed by that machinery. The whole system of capitalist production is based on the fact that the workman sells his labour-power as a commodity. Division of labour specialises this labour-power, by reducing it to skill in handling a particular tool. So soon as the handling of this tool becomes the work of a machine, then, with the use-value, the exchange-value too, of the workman’s labour-power vanishes; the workman becomes unsaleable, like paper money thrown out of currency by legal enactment. That portion of the working-class, thus by machinery rendered superfluous, i.e., no longer immediately necessary for the self-expansion of capital, either goes to the wall in the unequal contest of the old handicrafts and manufactures with machinery, or else floods all the more easily accessible branches of industry, swamps the labour-market, and sinks the price of labour-power below its value.

In other words, machinery and automation takes away the labor power of the worker, and the only commodity she has—labor—is rendered unnecessary. Workers who are not laid off now must work more as their workday intensifies to compete against machines.

Take, for example, an educator. A university begins to lay off instructors as tasks such as grading and curriculum development become automated by AI, and the remaining instructors take on more classes, more students, and ultimately more work to maintain their employment. The quality of education and instruction will decrease across the board for both students and instructors, but the university only sees short term value gains.

In a capitalist political economy, where the goal is to extract as much value as possible, technology will always trend (with notable exceptions) toward automation, deskilling, and the suppression of the value of labor power.

Spoiler alert for Park Chan Wook’s No Other Choice

In Park Chan Wook’s newest film, No Other Choice, the main character, Yoo Man Su (Lee Byung Hun), is laid off from his manufacturing job at a paper factory. A new position at a paper factory opens up, and Man Su decides to eliminate his competition, literally. He kills every other unemployed worker competing for his position, and when Man Su gets the job at the end of the film, he walks into a factory where he is the only person. All other workers have been eliminated, and the only thing for Man Su to do is watch machines make paper. No community, no solidarity. Barely even any work.

Park Chan Wook’s vision of automation grows ever closer as tech companies, large corporations, and governments promote and fund AI in service of value extraction, war, and surveillance.

AI is another weapon aimed directly at the global working class, whose precarious conditions of daily living are already close to unbearable.

Is the time we save outsourcing our creativity and cognition to automated algorithms worth it?

Opinions expressed in signed articles do not necessarily represent the views of the editors or the Tempest Collective. For more information, see “About Tempest Collective.”
Featured Image credit: Tang Yau Hoong; modified by Tempest.