“If we really want to make ethical AI, then we might want to consider this perspective: there is no ethical computation under capitalism.”
This is the “mic drop” line at the end of the recent Philosophy Tube video, Here’s What Ethical AI Really Means. It is the defining example of someone saying something that is clearly supposed to be brilliant but sounds like a dad joke or a mid-2000s Reddit post, but then if you tell them that, they would say it's on purpose, but it absolutely wasn't.
This kind of nonsense “activism” and/or “education” harms the prospect of meaningful change.
Computation, Production, and Consumption
The phrase “there is no ethical computation under capitalism” and the phrase it derives from, “there is no ethical consumption under capitalism,” both originate from a critical perspective on capitalism, but they address different things.
“There is no ethical consumption under capitalism” critiques the consumer side of capitalism. It suggests that due to the pervasive nature of capitalist production methods, which often involve exploitation, environmental harm, and unethical labor practices, it's nearly impossible for consumers to make purchases that are completely free from these issues. The critique here is that every act of consumption in a capitalist system is complicit in some form of exploitation or harm, regardless of the consumer's intentions or choices.
“There is no ethical computation under capitalism” extends a similar critique to the realm of technology and AI. It implies that computational technologies, including AI, are inherently unethical under capitalism because they are developed and deployed within the same exploitative and profit-driven framework. This suggests that the systemic issues of capitalism unavoidably taint the processes, objectives, and impacts of technological development.
At first glance, it might seem that “ethical computation” rectifies some of “ethical consumption’s” issues in that computation is some form of production. However, the simple shifting of “what act” does not designate “for whom.”
I have fallen for this trap myself; years ago, I regularly asserted that it should be, “There is no ethical production under capitalism.” But this undermines the importance of the labor of the subordinate class – the one that ultimately does the actual work of production in service of the ruling class. The labor itself creates value and thus is the subordinate class’s main weapon against the ruling class.
However, if all production is unethical under capitalism, then all of that labor, which the subordinate class has to engage in, is unethical. All labor and production necessary to build power for that class is, too. Similarly, if all consumption within capitalism is unethical, including doing the necessary things to exist, like eating and owning a car to get to work (which is also unethical), what is the point of anything?
Thus, “there is no ethical computation under capitalism” gives us a similar kind of impasse. The fact is, technology can be very good. It has unquestionably made people’s lives better, in part by making certain jobs easier. There is very likely less slave labor due to technological developments. There is, of course, a dark side. New forms of exploitation arise, jobs that once supported people disappear, and many things become less valuable. There is plenty of nuance.
Ethics is an abstract concept without any objective basis. While many argue against the idea of relativity in ethics and morals, it is impossible to separate these things. However, “no ethical computation” completely flattens the ethics of computational endeavors, leaving no room for individual or collective agency to influence or reform these practices for the better. It conflates an ethical viewpoint with an objective process.
When viewed scientifically, capitalism is essentially the socialization of production while retaining the previous (private/feudal) mode of appropriation. This creates multiple relationships to the means of production, and one of them retains control and the end product, creating classes of people with differing material interests. This not only creates social conflict but also results in mathematical untenability.
I think a shift away from discussing these activities through an ethical lens is necessary. Ethical considerations are either irrelevant or too abstract to be meaningfully applied to these areas, especially given the complexities and systemic nature of global capitalism. Further, “bad ethics” is not what is wrong with capitalism (as I have repeatedly demonstrated in my work).
Sloganizing Everything
Relying on overly simplistic slogans like “there is no ethical computation under capitalism” reduces complex socio-economic and technological issues to black-and-white statements. This fails to capture the nuanced realities of how technology interacts with and is shaped by capitalist systems.
This approach tends to frame issues in a binary manner — good vs. bad, ethical vs. unethical — without acknowledging the spectrum of possibilities and the multifaceted nature of technological impacts in society. It suggests a world where any action within a capitalist framework is inherently unethical, which is a statement that is bound to fuel nihilism. Why do anything if it’s all bad? If things I must do make me bad, why should I try to be good?
This sense of futility or cynicism discourages engagement with anything positive, whether that be simply helping in one’s community or building some form of power to attempt to force progress and change. It can also stifle more nuanced discussions about navigating the structures for better outcomes now, at least for as many people as possible.
As people become more nihilistic, they also become more polarized. Rather than having any objective realization of interest, the paralyzed subject can only dictate and argue their preferred ideals. Constructive dialogue with those who may not fully align with one’s viewpoint becomes an impossibility because the discussion really isn’t about finding beneficial actions to take. This can alienate potential allies and hinder the formation of broad coalitions that will be necessary to make fundamental changes.
I think the mindset of “yes, we know these slogans don’t quite say the critique, but the awareness the slogan creates starts the conversation, and that’s important(!)” is fundamentally counterproductive. It’s not just that they “don’t quite say it,” it’s that they don’t say it. Often, the sloganized version of a critique leads to an understanding that’s fundamentally at odds with the actual critique (see: “abolish the family”).
This misdirection results in obscuring the real issues and thus efforts based on it are poisoned from the start, regardless of how good the intention is. And that’s assuming efforts even emerge at all; these slogans often lead to superficial engagement at best. While they might spark initial interest, they seldom encourage deeper investigation or understanding, leading to a buzzword-level awareness and nothing more.
Mic Drop
The “no ethical computation in capitalism” line comes to us at the very end of the video essay. It is the “mic drop” moment, the thing we’re supposed to walk away thinking “woah” about. However, after a brief dramatic black screen with Mr. Robot-inspired music, Abagail Thorne returns to say that “it feels a bit weird to do the Nebula ad after that emotional ending, but like I said, the people who make this show work really hard, and if you wanna help us then you can!”
What follows is an ad for Nebula, a content platform “owned partly by the creators,” but also by CuriosityStream, which has a USAID member on its board. Thorne goes on to say, “We're trying not to be uber-capitalist.”
This is why I think it’s so interesting the critique that gets support from people within the broader structure, made by personalities that get pushed by media outlets (and, in Thorne’s case, hired by Disney), paints all action as bad, rather than dichotomizing for whom.
I would argue that all consumption in capitalism is as ethical as it can be. We have nothing to do with how something is produced, and therefore, our choices as to what to consume don’t matter in any collective context. But the dichotomy isn’t “consumption vs. production;” writing this article was the production of culture, just as writing Here’s What Ethical AI Really Means is.
“For whom” should be a class question, first and foremost (and class should be based on qualitative relationship to power via means of production). While one might think that means “who is producing,” we should understand that production is often done by the class that isn’t truly benefitting from it. So, whether we must ask questions about what is “ethical” to produce (or if we instead do what I advocate for and ask questions based on material interest), the question isn’t “who is producing” but rather “who benefits from the producing.”
Computation is a form of production, and the subordinate class (either as a whole or individual members) can absolutely be the primary beneficiary of it. It can (and should) be wielded against the interests of the ruling capitalist class, and I think that is a good case for what someone might call “ethical computation in capitalism.” No, this isn’t easy by any stretch, but trying to figure out a way for that to happen is better than the paralysis of “all action is bad.”
Again, I understand that isn’t “the point” of such a slogan, but I do not care. When a slogan undermines the actual critique of the ruling order it represents, then the ruling order benefits from the slogan, no matter how much discussion it generates.
For more on how subversive thought is subordinated to the capitalist structure (as well as a lot more talk on the phrase “there is no ethical consumption under capitalism”), watch my documentary Marx for Sale (or buy it as a pamphlet).
Conclusion
The continued use of slogans like “there is no ethical computation under capitalism” ultimately undermines meaningful critiques of capitalist structures. While catchy, these slogans distort the critique and lead to nihilism and paralysis.
Slogans lend themselves toward deterring the interrogation of who benefits. The material interests of different classes and the potential for leveraging technology and other means of production are not discussions of morality or ethics.
As I put in the credits for Marx For Sale, one of my current projects is a documentary on AI. I’m thankful Ms. Thorne articulated her argument a couple of months later (interestingly enough), fleshing out the limitations of “leftist” critique of AI so I don’t have to.
There’s a lot more to talk about than slogans, but this is a potent example of why I don’t frame my critique through them.
Have you written any essays (or books) on blah blah leftism? Like, why identity politics doesn't help the people it purports to help, and actually hurts them? (I'm not sure you think that but i think you might.)
I'm looking for material for 2 different audiences. The first would be material for liberals who *may* be open to class politics. I'm particularly looking for shorter stuff for these people. I don't know about you, but I have a very hard time getting people to read stuff I send them. I don't understand that, but it is.
So I'm looking for something basic, written in simple language, for a liberal, rainbow sock-wearing, provocative-shirt wearing person who has absolutely no idea that they're being misled, that it's a united system ("uniparty") versus the masses, etc. A video would also work but it would need to have closed captioning.
2nd one is for me. And I like books, and essays. Can you point me to a work of yours you'd recommend about identity politics or synthetic leftism? If it teaches about socialism it would need to be intro-level. I haven't read any of the classics, literally I'm just at the start of my voyage on this ship. So I don't understand the sects, and the acronyms, etc.
Thank you in advance.