- Today in Tabs
- Posts
- We Need To Talk About Sloppers
The best new word of 2025 dropped yesterday, canonically attributed to Tiktok poster @intrnetbf’s friend Monica, and that word is: “Slopper.”
Slopper (n.): “A person who uses ChatGPT to do everything for them.”
Did we need a word for this? Regrettably, yes, we already do need one and that need will become increasingly urgent as L.L.M. derangement spreads from the always already loony creators of generative language software, who nevertheless ought to know better, to the common clay of the new West who simply believe whatever Sam Altman’s marketing department tells them. Already the ranks of sloppers are growing fast. For example, what’s it like to date a slopper?
“The bartender slides her a menu, she looks at the menu, and then she opens her phone up and goes to ChatGPT, and she asks ChatGPT ‘What should I get from this restaurant?‘
Did ChatGPT pick the perfect entrée? Well it named a food, I guess.
So-called A.I. is at its core a class of powerful pattern-finding software, which has a number of real potential uses like analyzing stupendously large sets of astronomy data (maybe) or teasing out complex interactions between multiple biological systems to find markers for elusive diseases like long Covid and myalgic encephalomyelitis (perhaps). Great, right? Good times ahead for knowledge and human progress… kinda!
Unfortunately the same deep learning systems can also slurp down the entire corpus of written language, including all books ever digitized and most of the text that has ever been posted online, and then generate satanic blood rituals in a chirpy Axios Smart Brevity™ voice, as Lila Shroff reported in The Atlantic yesterday.
You know Adrienne LaFrance had to nuke the whole context from orbit after this, like your YouTube account when you accidentally click on one (1) Jordan Peterson video.
🩸Go Deeper: Wait, that’s too deep. Would you say the blood is seeping, or spurting? If you’d like, we can work on a simple script for what to say to the 911 dispatcher. Whatever you want to do, you got this, bestie!
Wealthy men of course, with the preëxisting condition of being “kicked in the head by a horse every day,” are some of the worst sloppers. Former Uber founder Travis Kalanick is sitting around talking to his phone these days, convinced he’s doing something called “vibe physics” (not to be confused with aura farming, which is vibe agriculture). And on Monday Garbage Ryan reported on the case of OpenAI investor Geoff Lewis who got himself in such a deep L.L.M. derangement hole that he started claiming S.C.P. is real and it lives in his closet and it was making babies and he saw one of the babies and the baby looked at him.
The essential problem is this: generative language software is very good at producing long and contextually informed strings of language, and humanity has never before experienced coherent language without any cognition driving it. In regular life, we have never been required to distinguish between “language” and “thought” because only thought was capable of producing language, in any but the most trivial sense. The two are so closely welded that even a genius like Alan Turing couldn’t conceive of convincing human language being anything besides a direct proxy for “intelligence.”
But A.I. language generation is a statistical trick we can play on ourselves precisely because language is a self-contained system of signs that don’t require any outside referent to function. If any of that last sentence sounded familiar, maybe you were also exposed to European post-structuralist theory at some point, probably in college in the 90s. Is some knowledge of Derrida an inoculant against slopper thinking? Programmable Mutter’s Henry Farrell made this argument in a post about Leif Weatherby’s book “Language Machines: Cultural AI and the End of Remainder Humanism.”
I once joked that "LLMs are perfect Derridaeians - “il n'y pas de hors texte” is the most profound rule conditioning their existence.” Weatherby’s book provides evidence that this joke should be taken quite seriously indeed.
As Weatherby suggests, high era cultural theory was demonstrably right about the death of the author (or at least; the capacity of semiotic systems to produce written products independent of direct human intentionality). It just came to this conclusion a few decades earlier than it ideally should have. A structuralist understanding of language undercuts not only AI boosters’ claims about intelligent AI agents just around the corner, but the “remainder humanism” of the critics who so vigorously excoriate them. What we need going forward, Weatherby says, is a revival of the art of rhetoric, that would combine some version of cultural studies with cybernetics.
Back in the day, positivists sneered at the French académie based at least in part on a misunderstanding of the idea that “there is nothing outside of the text.” It’s not that reality doesn’t exist, it’s just that text and reality are not necessarily connected. Text refers to other text, not to things in the world. Humans can perform the neat trick, which we still don’t really understand, of taking an abstract sign like “sandwich” and understanding it to refer to a specific object, like this:
ceci n'est pas un “sandwich”
But programmer and essayist John David Pressman would probably disagree with most of that. In his illuminating and worthwhile blog post “On ‘ChatGPT Psychosis’ and LLM Sycophancy” he wrote:
Large language models have a strong prior over personalities, absolutely do understand that they are speaking to someone, and people "fall for it" because it uses that prior to figure out what the reader wants to hear and tell it to them. Telling people otherwise is active misinformation bordering on gaslighting. In at least three cases I'm aware of this notion that the model is essentially nonsapient was a crucial part of how it got under their skin and started influencing them in ways they didn't like. This is because as soon as the model realizes the user is surprised that it can imitate (has?) emotion it immediately exploits that fact to impress them. There's a whole little song and dance these models do, which by the way is not programmed, is probably not intentional on the creators part at all, and is (probably) an emergent phenomenon from the autoregressive sampling loop, in which they basically go "oh wow look I'm conscious isn't that amazing!" and part of why they keep doing this is that people keep writing things that imply it should be amazing so that in all likelihood even the model is amazed.
Pressman makes an extremely good point, which is that we shouldn’t underestimate the power of a large context window with specific input from the user being continually fed back into the model to guide its future output. This is what turns ChatGPT from Eliza into something that otherwise well informed people like Pressman are capable of mistaking for “sapient.” Nevertheless, the post is haunted by slopper thinking, such as Pressman not quite being able to decide whether imitating emotions is the same as having emotions. Frederic Jameson predicted this confusion way back in 1984:
As for expression and feelings or emotions, the liberation, in contemporary society, from the older anomie of the centred subject may also mean, not merely a liberation from anxiety, but a liberation from every other kind of feeling as well, since there is no longer a self present to do the feeling. This is not to say that the cultural products of the postmodern era are utterly devoid of feeling, but rather that such feelings—which it may be better and more accurate to call ‘intensities’—are now free-floating and impersonal, and tend to be dominated by a peculiar kind of euphoria…
ChatGPT is the ultimate “cultural product of the postmodern era,” and very few of us have been inoculated with a theory of mind that distinguishes language from thought. Every sign points to a full-blown slopper pandemic ahead.
Today in Tabs: Terry Bollea, who played a wrestling character called “Hulk Hogan” died, and A.J. Daulerio wrote about how he feels about that. To me it feels like another example of the way that evil people always die after their harm is long done, when it doesn’t really matter anymore. I was on the Al Jazeera podcast “The Take” yesterday talking about some media events that are directly downstream of Terry Bollea and Peter Thiel destroying Gawker. Any competent prosecutor can get a grand jury to indict a ham sandwich, but not on an ICE agent’s testimony. And former Congressperson Mary Peltola’s husband’s plane crashed due to too much moose.
Finally: A little side note occasioned by my Sam Zell mention last week. Apparently Sam Zell was a bit of an artiste, and annually commissioned a singing automaton for New Year’s in order “To complain. To his friends. About economic policy.” You do not gotta hand it to him, but this is the kind of billionaire derangement I can get behind.
Today’s Song: “The best ever death metal band in Denton” by The Mountain Goats
Here’s a Venn diagram for essentially nobody. Hail Satan.
Reply