We have all known people in our own family, work, and social circles who are “smart stupid.” They may be very skilled, even brilliant, at some pursuits. Often, those are technical skills, such as programming, surgery, restaurant supply management, or other domains. But in other spheres, they’re complete morons.
They can be highly credulous. They follow twisted paths of logic to bizarre conclusions. They fervently believe in things without any evidential basis. They malign people who may be actual experts in the topic at hand. They trivialize the likely consequences of rash actions that follow on their beliefs.
Now, to some extent, we’re all smart-stupid. We’re smarter in some areas than others. On the other had, genuinely smart-stupid people insist on ignoring their weaknesses, making bold declarations on subjects where they should be aware of their deficiencies. Unfortunately, they sometimes have the power to get people to follow their misguided diktats.
Without a doubt, Elon Musk is one of those people. Nowhere can you see these deficiencies more than his zealous pursuit of AI. Here’s something he posted yesterday on X, the same day that the current regime unveiled its “Liberation Day” tariffs. (More on the connection of AI to the tariffs later.)
This sentiment doesn’t come from nowhere. Musk is part of a community in the technology world, including many tech executives, who believe that AI is more than just a good thing. It’s more than an inevitability. It’s actually the chief duty as human beings to create our coming AI overlords. As an added bonus, we might be able to use related technology to achieve “immortality” by uploading a copy of our minds into machines, achieving the “Singularity.” *
There are different flavors of this doctrine, from the Rationalists, who want to build a benevolent AI overlord to usher in a golden age, to the Zizians, a recent addition to the roster of American cults that believes, among other things, that it’s a good thing to train ourselves to become vegan psychopaths to prepare for the AI tyranny to come.
Before getting into their “thinking,” I have to point out the dynamic that leads people to believe in such nonsense. Smart-stupid individuals like to pile one supposition on top of another, higher and higher into the ether of ridiculousness. At no time do they stop to question whether any one of their conclusions makes sense. Instead, they swiftly move to place another piece of the house of cards, and another, and another…
How do people convince themselves of these intellectual castles in the air? Possible explanations include simple arrogance, social isolation (such as spending way too much time in social media echo chambers), Dunning-Kruger-like inability to see their own ignorance and inability, poor education in the humanities (“If I can master Python, Java, Rust, I can figure out anything!”), regular drug or alcohol abuse, living in the clouds of lofty privilege, and other possible factors. The end result: a zealous belief in absurdities, and a siege mentality against anyone who dares point out their ridiculousness.
Part of the backstory of AI fetishism lies in a science fiction story, “I Have No Mouth And I Must Scream,” written by Harlan Ellison, the enfant terrible of the SF community in the 60s and 70s. If the title isn’t a clear enough of a hint for you, Ellison’s story is one of the most depressing things you might read in this genre. A sadistic machine intelligence has killed most of humanity, except for a handful whom it keeps alive simply to torture. There’s no escape from this hell, even by suicide, since the godlike AI simply resurrects them for the next round of inventive cruelty.
That’s it. That’s the story.
You might think, “Gosh, that’s a horrible way to spend some pleasant time in my favorite reading chair, except if it’s a useful warning against trusting AI too much.” Others have a different take. In fact, they make a very perverse reading of Ellison’s story into a philosophy of life.
But first, let me caveat what I’m about to say. A lot of people who are fervent believers in AI, who want AI to take over many core functions of our lives, are misguided, but not as weird and malicious as we’re about to get. As more benign as their arguments might sound, I don’t think that AI can or should take over as many important decisions as these enthusiasts would argue. I don’t want AIs to govern when to pull the plug on very sick people, who gets into universities, which job candidates are deserving consideration, or whether a criminal defendant is innocent or guilty. I still want human beings to ultimately turn the switch left or right on those decisions. AI should be a servant, not a master. We don’t need or want tech oligarchs to usher in the AI oligarchs.
Now, on to the reductio ab absurdum of this thinking.
In the 2010s, many Internet forums were abuzz with discussion of Roko’s Basilisk, a thought experiment based on Ellison’s story. The argument went like this: Suppose an all-knowing, all-powerful AI is an inevitability, or at least a high probability. A machine intelligence that smart will know, who helped create it, and who impeded it. It will then use its powers to punish the people who stood in its way. Therefore, everyone has a duty to do their utmost to create the AI god-emperor, to avoid its wrath.
Now, on the face of it, this is the type of argument that, on its face, is not worth taking seriously. It’s a lot like Descartes’ “demon deceiver,” or its modern off-shoot, the universe as a simulation, which lead exactly nowhere. (Am I not going to get out of bed, because I think the bed, the bedroom, my house, and everything else is unreal? David Hume, among others, would disagree.) A cute idea maybe to bat around in a dorm room discussion, but that’s about it.
Except other people took it much more seriously. For the full story, I recommend the recent episodes of the podcast Behind The Bastards about the Zizians. The host, Robert Evans, has done a heroic (and no doubt painful) job of sifting through the blog posts of the Zizian’s founder, news stories about the Zizians, and other resources to paint a picture of how an overwritten piece of Harry Potter fan fiction led this group to deciding the best techniques for bringing about the AI-dominated future was becoming effective psychopaths, trying to put the evil half of their brain to sleep while keeping the good part awake, living on a decrepit tugboat in San Francisco harbor, and murdering the owner of the land on which they were squatting. It was the ultimate example of smart-stupid people (many of them, including their founder, were Silicon Valley tech professionals) following one silly proposition after another down the road to destruction. It’s a horrifying listen, but I recommend it.
Very little stood in the way of the Zizians traveling down the fast track to destruction. They built an isolated community of true believers. None of them had any education to speak of in domains they were exploring: for example, they seem to have had no idea that “upgrading” themselves through sleep deprivation was a very, very bad idea. Nor did they base their notions on any deep thinkers about morality and ethics. They baldly declared they wanted to become “Sith Lords,” embracing the worldview of Star Wars villains, or they wanted to emulate Ricky Gervais’ character in the UK version of The Office. These bits of popular culture were their guideposts, not centuries of political philosophy and history.
As I said earlier, there are less poisonous versions of AI fetishism than where the Zizians took it. But even belief in a more benevolent AI overlord is still a belief in an AI overlord. It’s no less silly and unjustified than the Zizian nightmare.
Which brings us back to Elon Musk. In the past, Musk has expressed that he finds Roko’s Basilisk an interesting thought experiment, worth considering. (It even helped him get a date with Grimes.) Does that make him a Zizian? No, but he definitely moves in the same circles where people take this kind of nonsense seriously, the thought problems that lead nowhere, except sometimes leading people to do stupid things.
And he certainly also has an unjustified faith in AI. His tweet on April 2nd is the most succinct statement possible. We’re just organic creatures whose role is just to give birth to the great machine intelligences. The fact that he can make a great deal of money off AI can’t hurt, either.
What’s the basis of his view of AI’s inevitability, and his weird idea that being superseded by machines is somehow glorious? The same porridge of popular culture that gave birth to the Zizians: science fiction stories, Internet memes, lowbrow space opera movies, blog posts, tweets, and so on.
In light of this, There are other statements he has made that are worth pondering. Infamously, he recently said that empathy is a “bug” in Western civilization. He promotes, through NeuraLink, the idea that by implanting devices in our brains, we’ll become happier and more productive as cyborgs than we ever would have been with just brain tissue. (This is one of the tenets of transhumanism, another movement that takes way too seriously something that science fiction writers like to write about. To solve human problems, transhumanists believe we should become less and less human.) He thinks that AI can take over many functions of government.
All of these smack of a wide-eyed, unquestioning, and ultimately unwise aspects of techno-Futurism, none of which is informed by a working knowledge of, well, how the world works. ** Or even how AI works.
Yesterday, Trump unveiled the new tariffs. It took no time for observers to figure out how they were calculated:
- The formula was based on trade imbalances, not actual tariffs imposed on the United States.
- The formula, bizarrely, was based on Internet domains, not actual countries. That means we’re imposing tariffs on islands with no human inhabitants (but there are some protectionist penguins lurking there, apparently), and even an island where the only inhabitants are US military personnel.
- The source of the formula was, in all likelihood, someone asking a chatbot, “What is the easiest way to impose tariffs?”
a
We don’t yet know who that someone was. But who, just who, might have an unjustified belief in the ability of AI to replace government experts who know something about tariffs, the health of trade between the US and other countries, and oh maybe, how the world economy isn’t organized around Internet domains? If it wasn’t Musk directly involved, it’d be amazing if his influence wasn’t behind this world historic moment of abject stupidity.
As I said, none of this AI fetishism is based even on a clear understanding of AI. We don’t know yet if AI is capable of “general intelligence.” Whether AI can ever have agency is also highly doubtful. (Sleep easily tonight. No current incarnation of AI is going to “wake up” tomorrow, Skynet-style, and decide to start hurling nuclear weapons indiscriminately.) We give AI a lot of information, get it to start building up patterns and conclusions, and set it on a task. A person who doesn’t know Jacques Merde about economics asks it to design tariffs, it’s going to give the result we have.
There’s a lot of techno-futurism that’s pretty harmless. If people want to experiment with putting machinery in their bodies, go for it, as long as you’re not harming anyone else. (That’s just basic utilitarianism. Oh, the humanities!) But if you want to start inflicting that techno-futurism on other people, based on absolutely no expertise in economics, health care, science, the effective functioning of government, international relations, history, demographics, the environment, climate, manufacturing, agriculture, law, transportation, intelligence sources and methods, or any of a thousand other possibly relevant topics, then you’re a dangerous idiot.
* When I was in college, I had a classic dorm room discussion about the science fiction concept of copying our minds. I’ll admit, I thought at first it really was a kind of immortality. Then my friends pointed out that it really wasn’t. If I die and someone moves the copy of my brain into another body, I, the person talking right now, am still dead. It’s as deficient a form of immortality as saying that John Wayne was still alive as long as Rich Little could do impressions of him. Ditto for “uploading” our brains into some electronically-generated pseudo-reality. You really want to tell people who believe that’s a good thing to spend more time outdoors, preferably with other people.
** For another quintessential smart-stupid guy who knows Jacques Merde about anything, check out the collected works of Curtis Yarvin, the computer programmer who, from his cubicle, came up with a cunning plan for re-inventing societal organization, turning the nation into a collection of corporate-run city-states. Who loves Curtis Yarvin, the guy who declared that “Democracy is done” without ever cracking open a book of history, political philosophy, or political science? Why, of course, JD Vance, Peter Thiel, Marc Andreesen, and other people who want to demolish the status quo! Behind The Bastards also has a good series covering him.
