This is the third part (paid) of a five-part essay on the psychology of AI nerds. They are intended to be read in order (introduction included in part 1). The fourth part—IV. Legibility—comes out tomorrow. Here are the previous parts:
AI Nerds Are People Who Like Everything — I. Overexistence (free)
AI Nerds Want Out And They’re Taking Us With Them — II. Escapism (paid)
III. Apoliticality
AI nerds don’t think the world can fix them—so far, as we saw earlier in parts 1 and 2, their experience has been pretty consistent on this point—and so they reject this planet we call home and also any instrumental mechanism that promises, in some way or other, that it can change to accommodate them.
That’s why AI nerds seem to hate politics—or rather, technical people do, for this is a broader sentiment. That’s why they dismiss the saying that “everything is politics.” They kill two birds with one stone: they shut down any potential hope before it aggravates their social suffering and, simultaneously, devote all their time to yet another form of escape: fixation and obsession.
I remember my time as an undergraduate at the aerospace engineering faculty; I had never known then, or since, a group of people so detached from the real world yet so concerned with making it a better place. In my specialty, aerospace vehicles, we used to dream of squeezing more speed out of the engine by tweaking alloys in the turbine blades or swapping in a sliver of carbon composite or something equally nerdy, but seldom would we wonder, aloud or otherwise, why Airbus needed faster Eurofighter Typhoons in the first place.
Then I’d switch on the news at home and realize the warplanes on which we participated, albeit only indirectly as oblivious students without a clue of how the world works, had been deployed in Syria, Yemen, and Afghanistan.
Five years immersed in that environment, drunk on that cocktail of candor, naivete, and blind predisposition, led me to a haunting epiphany: the people who end up in politics are the worst suited for the task (almost a platitude to those versed in being a voting adult). You’ve heard that “power corrupts,” but the truth is closer to “power selects for the corruptible.” People pure enough to do it well (well for the people, not themselves), who enjoy the selfless pleasure of a good deed, who blithely believe that good deeds exist isolated from second-order consequences, who are careless about membership, optics, protocol, and other bureaucracies that slow down the math, would never be politicians by choice.
AI nerds (whom I separate from the CEOs they work for) tend to display “Light Triad” energy—they believe people are mostly good, worth treating with dignity, and not to be used as instrumental goals—so they don’t fail by punching down, like your typical political party operative, but by not punching at all. They don't possess enough cynicism, cunning, or callousness to withstand the ladder-climbing dynamics that permeate political life, so they give way to those who understand that bombing countries is a “necessary evil.”
5 DAYS LEFT: I’m running a Halloween sale for free subscribers at a 33% discount until November 3rd (Monday). I probably won’t do another one until 2026, so make sure to get yours at the reduced price of $80/year (vs $120/year). You can also opt for a standard monthly sub.
The AI circles I lurk online remind me of college; autistic savants worried about one thing and one thing only: how to fit more datacenters in the world, more GPUs in the datacenter, more model parameters in the GPU, and more data tokens in the model—which is a convoluted way to say that they like to hit the proverbial anvil with the hammer, and care about little else. Their lives are a perpetual pursuit of solutions to an infinite sequence of simple but hard problems. Or rather, they used to be. Something fundamental radically changed a few years ago.
Keep reading with a 7-day free trial
Subscribe to The Algorithmic Bridge to keep reading this post and get 7 days of free access to the full post archives.