Be the Anti-LLM

Eric Weinstein’s Call for Human Originality

In my corner of the internet, Eric Weinstein often gets pummeled with criticism for his specific brand of smug self-aggrandizement. Even putting aside his theory of everything, he makes himself an easy target for ridicule when he keeps complaining on podcasts that the government never calls him to ask for his advice. “I love it,” writes Richard Hanania, mocking Weinstein. “It’s like he’s an ambassador from the realm of pure logic the US government must deal with as an equal.”

Still, I’m a sucker for Eric’s unique capacity for coining new terms and reframing conversations. I don’t always agree with him, but I still appreciate his ability to help me view issues from a different perspective.

In a recent conversation with Steven Bartlett, Eric performed one of his signature reframings that really caught my attention. Bartlett observed that AI has achieved a level of intelligence that threatens white-collar jobs—a comment that’s usually accepted at face value. But Eric rejected the premise entirely, arguing not that large language models (LLMs) aren’t that intelligent, but that humans aren’t that intelligent.

Here's Eric’s argument, edited for clarity:

You and I are two chatbots for the most part. More or less we’re LLMs. More or less we don’t do a single intelligent thing all day long. And the reason LLMs are able to mimic us is that we don’t realize that intelligence is a last resort for us. You know, if you think about greetings, more or less what do we all day long are LLM interactions. “Hey buddy, how are you?” “Good, good. Things have been really busy. How about you?” That’s an entirely scripted conversation.

I love this observation because it sets up a challenge for humans to differentiate ourselves from AI chatbots: Don’t live your life shackled to a script. Instead, work toward being original at every turn. In a world of chatbots, be the anti-LLM. In this way, you’ll not only set yourself up to be less easily replaceable in the job market, but you’ll also be—at a fundamental level—less of an automaton.

Speaking for himself, Eric says, “That’s why I want to do podcasting outside of the LLM. … I want to talk about things that I’ve never explored, where I don’t have something…ready.”

The world could very quickly become much livelier if we all took up this challenge. I’ve been ranting for years about how modern writers have failed to develop a voice. Instead, nearly all modern fiction writers default to writing in bland understatement.

Online, it’s no different. If you spend any amount of time on a microblogging platform like X, you’ll notice that even the edgiest people (with a few exceptions) write in the same style of snarky understatement. Often, a person’s entire online style and personality will be built around using no capital letters or punctuation.

But the challenge to not act like an LLM goes beyond writing style. It also directly applies to lifestyle. “What are they doing at Burning Man?” Eric asks. “They’re trying to live luxuriously under oppression—simultaneously luxurious and as dirty and disgusting as you’ll ever be. Hopefully they’re having tons of eye-opening, mind-bending experiences, chasing some way of getting out of the LLM.”

There is a lot of talk these days about the death of culture. As philosopher Mark Fisher observed, to be in the 21st century is to have 20th century culture brought to us by 21st century technology. (I.e., we still listen to and emulate the Beatles, even though we now stream their songs on Spotify.)

Ideally, new technology should help us break free from this cultural death spiral. That’s what technology has done historically—revolutionizing and revitalizing the visual arts with the camera, developing the new art of cinema with the video camera, developing a plethora of new music genres with the creation of the steel-string guitar, the synthesizer, etc.

Perhaps AI will end up giving us tools that help us push culture forward once again. Or perhaps it will inadvertently nudge us into pushing culture forward out of sheer boredom with AI slop.

As Eric notes: “Do you know what AI is bad at? Things where there isn’t much data.”

That’s what we humans are still good at. We—embodying the spirit of the anti-LLM—can still turn our minds toward unexplored horizons.

For what it’s worth, this view fits in perfectly with Eric’s ultimate aspiration for humanity, which is to become multi-planetary as soon as possible—before we destroy ourselves with our fancy new technologies.