Against Artificial Intelligence

There’s a lot of talk lately about how artificial intelligence is entering a new phase. This article from Futurism says it was a “banner year” for the artificial intelligence “industry.” (It’s an “industry”? Gross). It proclaims that AI researchers are now actively trying to create computers with consciousness. Another article from a few months ago summarized what’s called “generative AI”—or AI that can be creative. And I’m sure you’ve heard of the ChatGPT thing, which mimics human language. This article says that by 2025—two years from now!—90% of online content could be produced by AI like ChatGPT.

We’ve been hearing this kind of thing for a long time—that AI will achieve “consciousness” (whatever that means) and that it will be able to be creative (whatever that means). And that it will be able to produce written language as good or better than humans can. It’s always been bullshit, but maybe it’s closer to being real now. But the key point is—there is no real desire for any of this. Nobody really wants this. It is being pushed from the top down. There is no mass need or desire for it.

The elites have invested billions of dollars into the AI industry. There are all kinds of NGOs, think tanks, initiatives, programs, departments at universities, endowed chairs, and on and on, all with the goal of producing AI. The media is made up of people who went to good schools and love the idea of Smart People Being In Charge. And nothing fits this better than Smart Scientists making AI that will run society. That’s why we see so many breathless hype pieces about how AI is coming and how it will improve things.

The same media that loves to talk about how important “democracy” is, also loves to hype AI. The same elite liberals who are behind thinktanks with names like Defending Democracy or whatever will also pour millions into AI research. But nothing is less democratic than AI—having a small group of scientists at elite institutions working on machines that will take jobs from the mass of people. If there was a popular vote about AI, I would bet anything that the majority would be against it—would want it to stop altogether. AI is not something that the mass of people want. Elites have already invested heavily in it, and they are protecting their investment by hyping it—and their media lackeys are all too eager to push their propaganda.

Even the occasional piece warning about the dangers of AI that makes it into elite media—like this recent one in the Wall Street Journal—totally misses the point. It makes an argument that without “consciousness” AI will be sociopathic. This is fair enough I guess, but it starts from this flawed premise that is never questioned—“We’re building machines that are smarter than us and giving them control over our world.” Who is “we”? Are “we” building these machines? I’m not! You’re not! Who is? A small group of elite scientists and technologists, funded by the richest people in the world, who want machines to take over everything because it will eliminate workers from the equation. The elites love to talk about the “problems” of AI—like putting consciousness into it—but they never talk about who is building AI, and why they’re doing it. Even in critical pieces like this one, the premise is that “we” are all building AI together somehow. This is so obviously stupid and untrue. AI is never voted on, it has nothing to do with the mass of people. It is a top down thing that’s happening. “We” (whoever that is), has nothing to do with it.

The ethical need for AI is barely ever articulated. AI is usually just pushed on us as some self-evident good, or as something unstoppable that we all just have to accept. But sometimes, when the elite media bothers to defend the need for AI, they’ll say things like it will be able to solve the most difficult problems. This of course means climate change. That is the existential threat that justifies every measure to solve. But this is stupid, because climate change is not some problem that we are too stupid to solve. It isn’t something that a machine that’s smarter than us will be able to solve, like there’s some complex formula to solve it that we just aren’t seeing. Climate change is a political, social, and economic problem. It isn’t a computational one. It isn’t even that complex. It’s just about political will. How can a smart machine help with that? It can’t. So even here, AI is useless.

Then there’s the creativity thing. This piece about “Generative AI”—AI that can be creative—is interesting because it openly states what is driving all of this: “The dream is that generative AI brings the marginal cost of creation and knowledge work down towards zero, generating vast labor productivity and economic value—and commensurate market cap.”

Translation—AI is being produced because the capitalist class wants value created without labor costs. When machines do the work, you don’t have to pay them. With human labor, value is created, but you have to pay the laborers for the value they create. Capitalists don’t like this. They want the value that labor produces, but they don’t want to pay for it. This has always been the dream of capitalism—and they think that with AI, they will be able to achieve it.

This is interesting in a lot of ways. It shows how deeply stupid the rich people who control the world are—they think that their companies will get more valuable if workers are eliminated. That’s what the line about “market cap” means—that the value of companies will vastly increase once workers are replaced by machines. These idiots think that machines will create the same value that workers do, and they won’t have to pay them. But who will buy their products or services? If workers everywhere are out of a job, then there won’t be a base of customers to give value to companies. This is one of the basic contradictions of capitalism, and of course the capitalist class always has been and always will be blind to it. They are incapable of learning anything that goes against their basic motivation of increasing profit by any means necessary, and viewing workers as anything other than things that they unfortunately have to pay to help achieve profits, is impossible for them.

But even more deeply, this belief in “generative AI”—machines that can be creative—shows a total lack of understanding of art, beauty, imagination, and anything else that relates to creativity. The word “creativity” is always used in these discussions, but that is too broad of a word. Creativity can relate to too many things. The more important question would be less about creativity and more about imagination. Imagination has a more robust philosophical and aesthetic history than “creativity” does. Imagining is a specific thing in aesthetic philosophy—creating is a much more general term that means so many things that it really ends up meaning nothing.

Imagination is the interplay between the understanding and sensation. Imagination is the only way that the senses in sensation make contact with the understanding—that is the very essence of art.

Machines have no senses, no sensations, and so even if they are able to gain “understanding” of anything—which I doubt is possible—it won’t be able to produce imagination, which is the whole basis of art. AI starts from nowhere—just lines of computer code—and so it will lead to nowhere. Imagination has to use the raw materials of sensation and understanding to combine and lead to new visions and styles and forms of beauty and truth. AI has none of these raw materials, and so it won’t be able to lead anywhere. Imagination is a specific thing, and it is impossible for machines, and always will be. We never hear about imagination when AI is discussed, just creativity—because creativity can mean anything, so it ends up meaning nothing, and serves their propaganda purposes better. Empty words like creativity are always preferred in propaganda to substantive ones like imagination.

And one of the most important things about art is the audience that it is made for. Art isn’t complete until it has been received by the audience. This is sometimes called reception theory. Art doesn’t fully mean what it means unless and until it is received well. But who is art created by AI for? What needs is it speaking to? What worlds is it imagining? It starts from lines of computer code, so it has nothing to offer human audiences, who are beings of sensation and understanding. Art is about subjectivity—it is not objective, but intersubjective. Two subjectivities reaching out across space and time—that is what art is. AI will never have subjectivity in a full sense—even if it achieves a kind of understanding, it won’t have sensation, and so won’t be a true subject, and so it can’t reach out to another subjectivity.

Here we see the same contradiction that we see at the core of capitalism itself—capitalists try to enrich themselves by making their customer base poorer, and it just destroys the economy. In the same way, AI is destroying subjectivity while seeking to promote it.

The fact that billionaires have wasted so much money on AI, and are just are so committed now, I think is a genuine sign for hope. They keep pushing interest in this kind of thing because they’re protecting their investment. They’ve dumped so much money into it, they can’t stop now. It’s all propaganda. There is some hope that, if they do actually succeed in making AI, it will end up being bad for them. There’s always this talk about how AI will replace low level workers, but executives are really more easily replaced. What do corporate executives actually do? They make decisions about how to maximize profits. Can’t machines do this better? That’s what machines are really best at—finding out how to increase numbers. The hope, then, is that the corporate class spends a fortune creating AI that will end up making them obsolete, not everyone else. Remember, rich people are very stupid, in a deep way—they never think about why they’re doing what they’re doing, or what the meaning of their actions is. It wouldn’t be surprising at all if they ended up putting themselves out of a job. Here’s hoping.