August 2026

The Decadence Preservation Society

By in ai, the geist of the zeit, the statusphere

Anastasia Berg and Jon Baskin:

Even more important is to create and protect spaces that allow people to read, write and discuss ideas without the encroachments of AI or any other digital technology.

…In response not only to the evangelists of posthumanism but also to a more pervasive despair about the ability of our words and actions to make things happen, these spaces must double as training grounds for the guardians of human language, thinking and imagination. For the “radical” task before us is not to hope that something resembling our concepts and practices will reemerge in the distant future; it is to do whatever remains in our power to preserve them in the present.

Like most reflections on AI from the writer’s guild, I don’t find this one very thought-provoking. What I end up thinking is that they have identified a real question but chosen the wrong scale for it. AI will force us to decide which activities are valuable only for their results and which derive some of their value from the human performance itself. A concept becoming difficult to apply is not the same thing as a concept becoming unintelligible. AI may make it harder to know whether Fred wrote his paper, whether Susan knows what she’s talking about, or whether some award-winning novel owes half its prose to Claude. That can obviously produce serious problems of verification and attribution. But the very fact that we care about those questions suggests that merit, authenticity and fraud remain perfectly intelligible. Indeed, fraud may become more salient when imitation becomes easier. Counterfeit money doesn’t make the concept of genuine currency unintelligible.

Their line about unassisted work becoming a mere “hobbyist’s predilection,” like using a typewriter, makes the confusion clear. It assumes that people undertake difficult activities primarily because there is no cheaper way of obtaining the output. But as I just said recently, a forklift has not made a 500-pound deadlift unintelligible. It has merely made obvious that moving 500 pounds from A to B and deadlifting 500 pounds are different activities.

But my main impression is that I’m once again a bit surprised at how uniform the reaction to the strange new tool has been among the supposedly-creative: pull up the drawbridge and ready the boiling oil. Where’s the curiosity, the experimentation? How psychologically conservative, even reactionary, of them! The stereotypical artist used to be the person who horrified respectable people by saying, What happens if we use this new thing in a way nobody intended? Photography, collage, tape manipulation, synthesizers, sampling, cut-ups, video, photocopiers, digital editing — the creative temperament was supposed to be promiscuous toward tools. Give the artist a machine and the artist immediately misuses it. Now the machine arrives and a surprising number of literary people become the church ladies: Absolutely not. That infernal contraption will alter the traditional relationship between Man and Paragraph!

In a funny coincidence, Arts & Letters Daily linked to this article, in the achingly-progressive Guardian, on the same day as the one above. The authors argue that contemporary people possess the costumes but not the spiritual substance of fin-de-siècle decadence: the luxury, excess, artifice, bodily display are there, but the authentic ethos has vanished. Which is, amusingly, structurally identical to every conservative lament about lost forms: we have Christmas without Christianity, universities without learning, manners without morality, architecture without craftsmanship. Except here, apparently, we have decadence without decadence. The old decadent said: Bourgeois morality is suffocating; propriety is fraudulent; good taste is merely convention; let us cultivate artifice, perversity, intoxication, useless beauty. The new nostalgic decadent says: Young people are practicing decadence incorrectly! There used to be standards! Aubrey Beardsley has somehow become Norman Rockwell.

I mean, I too think of scrolling, “looksmaxxing,” and self-quantification as dubious pleasures, but it’s simply hilarious how we’ve moved from “Away with Victorian moralism! Follow sensation!” to “No, no; not those sensations!” The interesting question is what an embodied human being can do with the artificial thing. If Wilde had somehow been handed Claude, I doubt his first reaction would have been, “Alas, this device possesses no phenomenological interiority; lock it in the cupboard.” I suspect you’d lose him for three days. That’s the missing decadent response: “What deliciously improper thing can I do with this?”

Instead, at almost precisely the place where their historical decadents would seem to recommend promiscuous experimentation with artifice, the authors become guardians of traditional authenticity. AI artifice is disqualified in advance because it isn’t backed by the correct metaphysics of embodiment. Again, how can you not laugh in delight? The self-appointed heirs of “the first duty in life is to be as artificial as possible” encounter perhaps the most powerful engine of artificiality ever devised and again reply, “No, no, not like that!” They make much of the decadent ideal that pleasure should be useless, an end in itself, opposed to Victorian utility, productivity and profit. Yet by the end they have turned decadence itself into a social-improvement program. They advocate for a “New Decadence” to resist optimization, restore embodied pleasure, revitalize culture and “force fresh blood through today’s complacent, anemic culture.” Poor des Esseintes. Even decadence has KPIs.

Fin-de-siècle fetishism aside, we’ve been living under the Baby Boomer ethos for a half-century now. David Brooks was writing about “bourgeois bohemians” around the turn of the century. That template of rebellious hedonism has long been a stale cliché in itself. But it really begins to feel its age when change arrives from the direction of the future, aided by technological developments that we have long thought of as progressive. If the kids today began to show signs of a return to sobriety, religion, monogamy, and so forth, that could easily be understood as a lamentable return to reaction. You could say, “Ah, the pendulum has swung; the squares are back,” and remain secure in the belief that you still understand the map. But what happens when you become the person saying, “That’s not a real relationship, that’s not authentic intimacy, you’re supposed to want human connection, this technology is degrading something essential”? You may possibly be correct, but you can no longer pretend to be Dionysian.

For people whose cultural identity was formed around novelty, liberation, experimentation, transgression, technological optimism, or being on “the right side of history,” AI can produce an almost metaphysical embarrassment. It threatens to reveal that progress was not a destination at which they arrived. It was a train they happened to ride for a while. And trains keep going.

Claudemarking, Claudewashing and Nudereading

By in ai, jests japes jokes jollies, the feeling of absurdity, writing

Join me, if you will, as I bow my head and give thanks to the trickster god who loves us and wants us to be amused. What’s that? You’re agnostic? Well, let’s see…

I wonder how long it will be until some enterprising coder—oh, really? That was fast.

Let’s sum this up. Claude is beginning to watermark text, and others seem likely to follow. I don’t know if this has an official name yet, so let’s call it “Claudemarking.” If you copy and paste text that has been “processed” by an LLM, it will bear the devil’s clanker’s mark, which others will be able to detect. But the human entrepreneurial spirit being what it is, this will immediately be accompanied by means for removing said clanker’s mark. Let’s call this “Claudewashing” (which will quickly become an epithet for those suspected of evading justice). Inevitably, an online service will appear with a name like “HumanizeMyClaude.ai” and a landing page showing a smiling undergraduate giving a thumbs-up. Philosophically, this is funny because if a mark can be stripped, absence proves nothing. If a mark survives editing, presence proves only that the text passed through Claude at some point. That leaves us exactly where we were: arguing about what “authorship” means, except now with cryptographic confetti scattered over the argument.

So we already have Pangram; soon we’ll have Claudemarking. I’m usually not one for predictions, but I’m going to take a stab at the likely next innovation in digital prophylactics: SynthScreen. I envision this being a new version of that wearable tech that has never quite caught on, like smart glasses. SynthScreen would have a full Iron Man HUD for prose. You’re reading an essay and little red brackets suddenly snap around a sentence:

SYNTHETIC-LIKELIHOOD EVENT DETECTED
Triadic construction: 87%
Em-dash residue: elevated
Metaphorical symmetry: abnormal
Recommended action: discontinue reading and unsubscribe immediately

Naturally there would be different protection levels. Normal web browsing requires only the glasses. High-risk environments — corporate LinkedIn, university DEI statements, Substack during a discourse outbreak — require full Narrative-95 filtration. Then comes the social etiquette. Someone sends you an article and you ask, “Was this SynthScreened?” They say no. “You mean you just sent me raw text?” “They told me it was human-written.” “Did you see the test results?” “No, but I trust them.” “Jesus Christ, man!” Offices establish designated clean-reading zones. Libraries put signs on the door: THIS FACILITY MAINTAINS A LOW-SYNTHETIC ENVIRONMENT.

Eventually, all this fussy hypochondria around reading as a biosecurity procedure produces a counterculture that glamorizes risk and danger. A new (old) practice emerges. Let’s call this “nudereading.” People. Text. No intermediaries. You just read and use your judgment, like they did in the Before Times. Underground book clubs meet without scanners. A dissident magazine announces that it accepts submissions in whatever state they arrive and refuses all provenance testing. I’m picturing scenes like Dead Poets Society, where the students slip away to a cave in the night to read and chant text with no heed of provenance. The elders are scandalized. “You don’t know where that text has been!” “Do you have any idea how many models may have touched it?” “Your cousin’s friend’s co-worker nuderead one newsletter in 2029 and never recovered!” School assemblies feature reformed clankertext addicts describing how it began innocently with a few Claude sonnets and ended with them reading anonymous machine-generated erotica in unlicensed prompt dens. Clankertext culture takes on the most subversive connotations, like gay bathhouse culture of past decades. Soon, somebody, probably the Atlantic, publishes a stern public-health article entitled “The Dangerous Romanticization of Unprotected Reading.”

Then you get the full subculture apparatus: clandestine reading rooms, coded symbols, pseudonyms, bars where the house rule is that nobody asks whether the prose is human or machine. “Don’t provenance me” becomes an assertion of privacy. People wear pendants shaped like em dashes. There are digital zines with manifestos about textual autonomy and the right to unreadability of origin. Meanwhile clankertext itself develops an aesthetic of deliberate excess. Why pretend to be human? Let the machine write six-page sentences, hallucinated metaphors, impossible registers, prose no sane MFA workshop would permit. What began as stigma becomes style. “Human-passing” is now the insult. Young writers sneer at clean, tasteful, provenance-certified prose as “meatsack realism.” You get the idea.

Anyway, the “effort-to-result” ratio of all this bot-sniffing reminds me of nothing so much as watching a dog chase its tail. I think I previously referred to this farce as an arms race, but if so, I retract that. An arms race at least suggests two sides accumulating genuinely greater capabilities. Here a lot of the motion is just circular, like the dog and his tail. We devise a technique for identifying synthetic prose; therefore we create a technique for making synthetic prose harder to identify; therefore we devise a technique for identifying prose that has been deliberately made harder to identify, and round and round we go…until somebody finally publishes a piece in the Atlantic titled, “How I Learned to Stop Worrying and Love Nudereading,” culminating in a revelation: “I had spent years trying to determine whether the intelligence behind a sentence was artificial. It had never occurred to me to ask whether the sentence was intelligent.” Cue the podcast appearances, and an eventual book titled Raw: Reclaiming Reading in the Age of Synthetic Text.

So, then, as I was saying: Would you like to join me in a prayer of thanks?

Prose Makes the Man

By in aphorisms, foolosophy

It is wisest to write in ‘I-natural.’ But many write in ‘I-sharp.’

— Paul Valéry, The Collected Works of Paul Valéry, Vol. 14

At first glance, Valéry seems to be recommending modesty: lower the temperature, stop preening, write from the self without making the self the whole cosmos. Sound advice, as far as it goes. But “I-natural” has its own dangers. For many of us, the natural self is not Montaigne in a dressing gown but a man in pajama pants, Crocs, and a stained T-shirt, muttering about errands while scratching himself where no guest should have to look. Rawness is not the same as honesty. The unfiltered self may be genuine, but so is bad breath. There is a kind of writing that mistakes lack of preparation for candor, as if the reader were morally obliged to sit through every undigested thought because it emerged sincerely from the author’s head. This is slovenliness with a clean conscience. The slovenly “I” forgets that anyone else has entered the room.

But Valéry’s “I-sharp” is not innocent either. The word first suggests the self pitched too high: the ordinary ego inflated into drama, portent, grievance, revelation. One did not simply go grocery shopping; one passed through the fluorescent nave of late capitalism and beheld, among the bananas, the spiritual exhaustion of the West. But “sharp” also points in the other direction. Couldn’t we say there is a good sharpness? A sharp-dressed man has not necessarily falsified himself. He has made himself presentable. Likewise, a writer may dress his thoughts for company without becoming a fraud. Indeed, it can seem rude not to. To invite the reader into one’s “natural” thoughts without arranging them first is less a confession than a breach of hospitality. Style, at its best, is not vanity but consideration: the combed sentence, polished syntax, and offered chair of prose. The trouble begins when style stops serving the guest and becomes the host’s mirror. The sharp writer remembers he has an audience, but too often he wants that audience staring at the cut of his coat.

The real “I-natural,” then, may not be the naked self at all, but the socialized self: the self that has acquired tact. Some sport a romantic suspicion of manners, as if every rule of civilization were an enemy of truth, and every unfiltered impulse carried the authority of a mountain spring. Yet good manners are a kind of contrived naturalness. Children have to be taught not to interrupt, not to monopolize, not to announce every whim as it passes through them. Small talk, so often despised as empty, is one of civilization’s little mercies. We ask about the weather not because we are amateur meteorologists, but because weather gives two strangers a neutral patch of common ground on which to stand until conversation can move under its own power. Writing has similar courtesies: proportion, clarity, pacing, restraint, an occasional joke placed where the reader might need one. These artifices do not betray naturalness; they create it. The slovenly writer says, “Here I am; endure me.” The sharp writer says, “Here I am; admire me.” The natural writer says, “Come in; I’ve made a path.”

Why Y?

By in ai, foolosophy, technology, the geist of the zeit

Gregory Conti:

AI means the death of authorship—and of the inspiration that work which you know reflects human achievement, and human competence, provides to all who read it. We will increasingly become mere “prompters,” at best “collaborators” with the machine. As sci-fi as it sounds, we are already well on our way to becoming mere interchangeable vessels for ChatGPT’s accomplishments and insights. The awe that we feel in reflecting on an Einstein or a Dostoevsky, we will feel toward no one again; Claude cannot be someone’s hero. The whole concept of a “life’s work” that belongs to a great progenitor of ideas is under threat; soon we may have less reason to care who “wrote” a paper with an AI’s help or who inaugurated a theory by prompting Claude than we do about who was riding Secretariat when he won his triple crown.

Conti is reflecting on what AI means for human achievement in the big scheme of things. I don’t have strong feelings about the majority of his points, but this one stands out like a sore, dumb thumb to me.

The crucial word is interchangeable. Give the same powerful tool to ten different people and they do not become interchangeable unless the only thing they contribute is pressing Enter. They bring different questions, accumulated knowledge, obsessions, standards of evidence, aesthetic preferences, capacities for judgment, willingness to reject attractive nonsense, and, perhaps most importantly, different ideas of what is worth doing in the first place. AI can supply an enormous amount downstream of those choices without making the choices cease to matter.

In fact, Conti’s own essay inadvertently demonstrates this. Claude apparently got close to an argument he was already developing, and he found that “phenomenally deflating.” But why was that particular argument significant to him? Why had he chosen those primary sources? Why was he interested in the question? Why does he accept some conclusions and reject others? Claude did not spontaneously decide one morning that technological succession was a threat to human flourishing and pitch an essay to Compact. Gregory Conti did. The machine’s competence can potentially alter how much credit he deserves for particular inferential moves, but it doesn’t turn him into a fungible nozzle through which Claude squirts ideas.

The machine, unlike Secretariat, doesn’t want to run anywhere. Secretariat was an organism pursuing the same immediate end as Ron Turcotte: get around the track ahead of the other horses. Claude has no intellectual project corresponding to Conti’s. It doesn’t care whether AI destroys authorship, whether the argument is published, whether it is true, whether it contradicts something it said yesterday, or whether anybody reads it. It can perform cognitive labor inside a project, sometimes astonishing amounts of it, but the project still has to come from somewhere.

“Claude cannot be someone’s hero” is therefore true and almost completely beside the point. Neither can a forklift. Yet the existence of forklifts has not destroyed our ability to admire someone who deadlifts 1,000 pounds. A forklift can make the human achievement look absurd if the only thing that matters is “how many pounds moved vertically?” But that was never what the human achievement meant. We admire a human being’s ability and desire to do it.

Cars didn’t destroy running. Calculators didn’t destroy mathematical virtuosity. Cameras didn’t eliminate painting. Recorded music didn’t make live musicians pointless. Mechanical advantage doesn’t automatically annihilate human accomplishment because human accomplishment has never consisted simply in maximizing output irrespective of the means. By Conti’s logic, chess should have become meaningless sometime around Deep Blue, because there was “one fewer arena in which a person, through their diligence and hard work and intellect, can shine.” Instead, millions of amateurs remain devoted to the game, and Magnus Carlsen remains famous precisely because nobody gives a damn that Stockfish could annihilate him.

A person who spends twenty years reading, thinking, developing peculiar preoccupations, deciding what questions matter, interrogating an AI, recognizing which of its suggestions are fertile, rejecting ninety-nine others, connecting one result to something he read fifteen years ago, and shaping all of that into a coherent body of work isn’t an “interchangeable vessel.” He’s doing something much closer to what authors have always done: imposing the form of one mind upon heterogeneous material. Conti sees a real problem — machine contribution can become large enough that conventional human credit becomes fraudulent — and expresses it as the much grander and far less defensible proposition that machine capability renders human individuality irrelevant. Those aren’t remotely the same claim.

AI attacks the scarcity of competence much more directly than it attacks the scarcity of salience. Producing a competent essay, competent illustration, competent computer program, competent critique — these things may become astonishingly cheap. Deciding that this obscure conjunction is interesting, that that apparently trivial experience deserves attention, that everybody else is asking the wrong question, that a ridiculous analogy is actually illuminating: those are different faculties. Weirdness is one of the sources from which importance originates. Dostoevsky is not Dostoevsky because he was extraordinarily proficient at generating grammatical Russian prose. Plenty of people could do that. What mattered was the fantastically peculiar configuration of obsessions that made The Brothers Karamazov seem necessary to him in the first place.

This is also why the “interchangeable vessels” line strikes me as almost exactly backwards. The more abundant generic intellectual competence becomes, the more valuable non-interchangeability may become. If everybody has access to something capable of producing a polished 2,000-word essay on Augustine, then “can produce polished prose about Augustine” stops distinguishing anyone. What remains distinctive is: Why Augustine? Why this passage? Why connect it to something your warehouse supervisor said yesterday? Why does that remind you of an obscure Tolstoy diary entry? Why do you find the conventional interpretation unsatisfactory? Why make a joke about forklifts halfway through?

There was a smart-ass kid in my Philosophy 101 class who argued that science could answer all philosophical questions better than philosophy could. My professor’s quip was that science couldn’t answer the question “Why?” Decades ago, when LLMs were still science fiction, that fault line was already obvious to anyone who looked. Science, technology, AI may quickly and effectively answer the question “Given X, how do I achieve Y?” They may have less to say in response to “Why Y?”

If I had to speculate on future trends, I’d suggest that we are probably headed toward a world in which competence is cheap and personality is expensive. And, amusingly, that may make the intellectual oddball rather more valuable, not less. The person who can write the 400th respectable essay about “AI and the Future of Humanity” is in trouble. The lunatic who suddenly realizes that prednisone, French symbolist decadence, and the ability to easily put on one’s socks belong in the same sonnet still has a job.

Most of these big-picture overviews of AI deal in ventose generalizations rather than the interesting particulars. This makes the genre frustrating and nearly useless to me. A sentence like Conti’s “we are already well on our way to becoming mere interchangeable vessels for ChatGPT’s accomplishments and insights” is nearly evidence-proof. What observation would establish that people had become “interchangeable vessels”? What observation would disprove it? The phrase creates an atmosphere rather than identifying a mechanism.

And because so much AI criticism operates at that altitude, it has developed a remarkably standardized inventory of anxieties: loss of authorship, loss of authenticity, deskilling, homogenization, dependence, replacement, alienation, loss of meaning, death of creativity. The nouns change places, but you can often predict the next paragraph. There is something genuinely funny about essays warning that AI will produce generic thinking while themselves reproducing a highly recognizable genre template.

Homophonia

By in editorial vigilantism

Joel Miller:

If you click the link — and you should; it’s a good, interesting post — you’ll notice he tried to spoil my fun by fixing it, but luckily, I had gotten the original version in my inbox, hence the screenshot.

At any rate, it got me thinking. I wish there were a sort of interactive feature on big platforms where a reader could append, say, a red asterisk to a grammatical or spelling error, similar to how Twitter allows “Community Notes” to be appended to inaccurate tweets. Obviously you don’t want readers to have editing privileges, but it would be nice to be able to highlight “poured,” and have a little pop-up menu appear with an option for “red asterisk.” Then the next time he logs into his dashboard, he sees a little notification: “UPDATE: 2 red asterisks on Too Much To Read?”

That would solve the awkward social problem perfectly. At present, correcting a typo requires an escalation of intimacy. You have to either leave a comment or email a stranger and thereby transform yourself from “satisfied reader” into “person who felt sufficiently exercised by your misuse of a homophone to locate your contact information.” The medium makes the correction seem far more important to you than it actually is. (There is something insufferably pedantic about the disproportion between offense and urge to intervene. You have nothing substantive to quarrel with. You enjoyed the essay. You think well of the writer. So naturally the one contribution you feel uniquely equipped to make is: “Admirable discussion of five centuries of reading practices. I regret to inform you that you have used the wrong verb. TWICE.” It’s the copyeditor’s version of Columbo: “Just one more thing, Mr. Miller…”)

I would keep the asterisks invisible to ordinary readers, too. That distinguishes the feature from Community Notes. Community Notes exists partly to warn the public; the red asterisk exists to discretely tell the proprietor that his fly is down. No public humiliation, no comment-thread argument about whether the Oxford comma is mandatory, no deputized grammar police decorating an article like a Christmas tree.

There could even be an author’s equivalent of “resolve”: correct the error and click Fixed, whereupon the people who starred it receive no notification whatsoever, because that would be deranged. The asterisk disappears and civilization advances by one micron.

I’m just giving brilliant ideas away for free because that’s how much I care about the digital commons.

Unhappy Endings

By in books, foolosophy, history

I recently read Patrick Leigh Fermor’s trilogy documenting his travels as a young man walking across Europe in 1934, from Holland to Constantinople. This is from The Broken Road, the third book, describing his encounter with what sounds to me like the Bulgarian Elmer Fudd. (The initial sentence, beginning on the previous page, starts, “an old man all in brown sat reflected…”) Most of the narrative throughout the trilogy is carried by Fermor’s descriptive density rather than excitement or adventure, but this was my favorite moment of genuine humor. It’s wonderfully dry because Fermor never steps outside the narrative to explicitly tell us that the man is ridiculous. He simply keeps adding details until the scene becomes a cartoon. A gun with a report like an exploding bomb; the hunter hidden by a cloud of smoke; zero ducks to show for it.

The gun itself sounds less like something designed to kill ducks than something designed by ACME Co. Another tremendous bang and cloud of smoke, and again the ducks sail on “undepleted” (with one, perhaps, pausing to rearrange his bill). The hunter’s response is not to reconsider his marksmanship, or his rickety weapon, but to shake his fist and call the ducks pimps. “Oooh, you scwewy pezewengi!” Fermor’s deadpan assent to this odd insult makes it even funnier. He could have written, “he cursed the ducks as pimps,” which would merely report eccentricity. Instead he momentarily lets the old man’s world govern the prose. The ducks have behaved without honor. They have evaded justice. Therefore: pimps. Why not?

This passage, from earlier in the book, is more representative of the general poignancy the reader feels throughout the trilogy, knowing with hindsight what “Europe, 1934” means. The Russian exiles were, in one sense, right. The Soviet regime was not eternal. It really was “rife with the germs of its own dissolution.” But a “temporary” madness can last longer than everyone who remembers the world before it. History did eventually turn the wheel, but these men were no longer there to be carried home.

Fermor’s best touch is “sudden giveaway silences.” The sighs express nostalgia; the silences betray doubt. Some part of them knows that Grand Duke Cyril and the double eagle are ceremonial props in a story they tell one another because the alternative is admitting that exile is permanent. Hope continues in speech, then falters whenever the room goes quiet. And “autumnal sadness filled the long room” is wonderful because the season is historical as well as emotional. These men are the scattered surviving leaves of a political and social world whose winter has already arrived.

The best sentence of the whole trilogy for me, though, is this one: “Nearly all the people in this book, as it turned out, were attached to trails of powder which were already invisibly burning, to explode during the next decade and a half, in unhappy endings.”

“As it turned out” is especially brutal in its mildness. No trumpet blast, no retrospective prophecy. Just the belated knowledge that history had already begun its work while everyone was eating, talking, flirting, offering beds, discussing history, and going about ordinary life. And “invisibly burning” avoids the smug consolation of inevitability. Fermor was walking through a Europe with Hitler and Stalin in power, but the people he met did not experience themselves as characters in the prologue to disaster. Their futures were hidden from them, as ours are hidden from us. Only afterward does the route appear strewn with powder trails.

The final phrase is almost painfully restrained. He could have named executions, exile, war, occupation, dispossession, suicide, disappearance. Instead he uses the language of stories — “unhappy endings” — as though the euphemism were the only way to hold all that accumulated knowledge without crushing the sentence. The reader almost leans forward to grab at Fermor’s lapels. “No, not ‘nearly all.’ Which ones? What happened to Nadejda? To Rosa? To Lise and Annie? Did anyone get across a border? Did someone marry and disappear into Switzerland?” Instead, “unhappy endings” leaves them where Fermor himself largely had to leave them: walking away from him into an approaching darkness.

One is only sometimes warned, when these processes begin, of their crucial importance: that certain poems, paintings, kinds of music, books or ideas are going to change everything, or that one is going to fall in love or become friends for life; the many lengthening strands, in fact, which, plaited together, compose a lifetime. One should be able to detect the muffled bang of the starter’s gun.

Here the “muffled bang of the starter’s gun” announces beginnings. A book, a friendship, a love affair, an idea enters your life almost casually, and only afterward do you discover that the course has been set. You wish retrospectively that there had been some audible report: Pay attention. This is one of the moments from which the rest of your life will proceed. Later, the powder-trail sentence gives the image its terrible inverse. Those people too have already crossed an invisible threshold, but this time toward catastrophe. The fuse is burning while they are still talking, dancing, drinking, flirting, making plans. Nobody hears the match being struck.

We imagine a life as a sequence of recognizable turning points because retrospect supplies the punctuation. While living it, though, the crucial moments usually arrive disguised as ordinary afternoons. The conversation that becomes a lifelong friendship does not feel different from ten forgotten conversations. The person whose loss will divide your life into before and after is, at first, simply someone sitting across a table.

But we were talking about endings. And so, after all the extravagant precision of Fermor’s descriptive prose, he reaches the thing that matters most and can give us only a few restrained syllables of judgment: unhappy endings. The great raconteur suddenly has almost nothing to say.

Ink Libel

By in ai, books, extraordinary popular delusions, media/propaganda, the madness of crowds

Pulping our humanity!

Horrific destruction!

A CULTURAL CRIME AGAINST HUMANITY!

Wait, what’s that?

But contrary to many news reports, the old books bought from these sellers do not appear to be “rare” in the typical sense of the word. According to Becker and the online discussions I read, the books in these orders are almost entirely nonfiction, were mostly published from the 1970s to the 1990s, and were written in a variety of languages. Becker wrote in a blog post that he has received orders for “stuff like The Insider’s Guide to Metro Denver from 1995 or How to Use Corel WordPerfect 1991.” These belong to a category of books that you could call rare because a small number of copies remain, but they may also be out-of-date, undesirable to readers, and not valuable per se. In some cases, booksellers have noted, the cost of shipping exceeds the books’ resale value.

 

Oh.

Boy, do we look foolish! If only some idiot blogger had kept his head while the rest of us were losing ours!

Well, mistakes were certainly made — and amplified, and magnified — but extremism in defense of The 1995 Insider’s Guide to Metro Denver is no vice, as I believe Erasmus might have said. At any rate, I’m sure we’ve all learned just enough to save face without forcing us to question our priors — so, no harm, no foul; all’s well that ends well!

I don’t want to let the Atlantic fully off the hook, though:

Regardless, the idea of an AI behemoth destroying books has sent a bolt through the collective psyche. Book destruction is historically linked to authoritarian regimes and attempts to control public discourse, as well as ancient tragedies—the burning of the Library of Alexandria comes to mind. The encroachment of AI, particularly in the realm of books, is now everyday news: Last month it emerged that a top-selling Kindle book may have been partly written by AI, and grandparents are buying children’s books filled with AI-generated slop.

Even if a literal book burning is not under way, the process of ingesting millions of books, stripping them of authorship, and blending them into a homogenous “intelligence” branded with the name of a chatbot certainly feels destructive. It masks the hard work and collaboration that goes into knowledge creation, undermines incentives for authors to write books, prevents experts from finding one another, and gives AI companies tremendous power over what information people can access. Given this scale of cultural damage, the physical destruction of books seems almost quaint.

“Regardless” here is doing the work of an emergency generator after the factual power supply has failed. The current it supplies is too weak to illuminate anything, but strong enough to animate a corpse. Yes, a “bolt” did indeed shoot through the collective psyche: the mad scientists of viral outrage sent it there. They had stitched together a monster from misshapen rumors, decaying conspiracies, rotting scraps of metaphor, and adjectives scavenged from dead headlines; once electrified, it lumbered off through the global village, bellowing about antique books, pulped humanity, and crimes against mankind.

This is an embarrassing effort to salvage a virtue out of a moral panic. If I may take some creative liberties in paraphrasing, “We investigated the burning of Alexandria and found that Alexandria probably isn’t burning. But the fact that people imagined it burning reveals a deeper fire.” No, what it reveals is that your audience consists of gullible cretins who should be ashamed of themselves. I’d like to see a Venn diagram of the people forever warning us about “misinformation” and the people who instantly, uncritically believed that AI was devouring “ultra-rare” books. I suspect it would actually be a perfect circle.

(Misinformation, n. The stupid things other people believe. The corresponding belief among one’s friends is called context, raising awareness, or an important conversation.)

I’m talking about you, 404media. I’m talking about you, Ted “The Honest Broker” Gioia. I’m talking about you, Futurism. I’m talking about you, TechRadar. And you, Ryan Fleck. You, Guardian. You, Hyperallergic. Oh, there were so many of you. And let us not forget the supporting cast among the Bandar-log: the minor imbeciles, volunteer panic merchants, ambulatory rumor mills, and unpaid interns of hysteria sharing their infectious outrage on Substack Notes, TikTok, Facebook Reels, and sundry other sewage dispensers. Your appalling ignorance is rivaled only by your overweening self-righteousness. You squeezed out fake tears over library books that had not been checked out since the Reagan administration. You rent your garments over worthless monographs that had been abandoned at Goodwill and remained unbought. Were you fooled, or were you doing the fooling? Did you mistake wishful thinking for corroboration, or merely trust that your readers would? (Take your time. Neither answer helps.) The slightest ounce of skepticism or simple patience would have sufficed; instead, you mistook your priors for evidence, your indignation for inquiry, your metaphors for facts, and one another’s repetition for independent confirmation. You stand amidst a mess of fabrications and aver that your heart, at least, was in the right place, even if your brain was AWOL. You embroidered rumor, embellished speculation, and passed the resulting monstrosity down the line with the solemnity of monks preserving the Gospels, only you produced something more like The Protocols of the Techbros of Silicon Valley. The online environment is, incredibly, an even stupider, uglier place today because of your malign efforts.

In a just world, you would cancel your internet service, surrender your passwords, and sew your mouths shut. If you possessed even the vestigial stump of integrity, you would renounce all “hot takes” and wander for years like so many Ancient Mariners of epistemic disgrace, buttonholing wedding guests, commuters, grocery clerks, and innocent diners, forcing each in turn to hear the tale of your sordid imbecility, your appalling groupthink. You would conclude by saying, plainly and without qualification: “I was wrong because I was too stupid to question a story telling me exactly what I wanted to hear.” But this is not a just world. So tomorrow you will return, chastened by nothing, confidence undiminished, to pronounce upon some fresh catastrophe you first encountered nine minutes earlier. I see you, though, and I know what you are.

The Anxiety of Artificial Influence

By in ai, jests japes jokes jollies, the feeling of absurdity, writing

In the future, the author will be issued a transparent typewriter in a bare room. Every keystroke will be time-stamped. Wastebaskets will be searched for unauthorized phrases. A court-appointed observer will certify that the writer stared helplessly at the wall for the requisite number of hours before producing each paragraph. Any sentence arriving too fluently will be flagged for enhanced interrogation.

Publishers will require a complete chain of literary custody: notebooks, deleted drafts, browser history, coffee receipts, testimony from spouses who heard the author muttering in the shower. “We loved the novel,” the editor will say, “but unfortunately there is an unexplained twenty-three-minute interval during the composition of Chapter Nine.”

If any anomaly appears, the author will lie immobilized in an fMRI machine while a panel of forensic neurologists watches each sentence form in real time. Any suspiciously rapid activation in the metaphor cortex will trigger an alarm. “Subject appears to have generated the rain worried the windows without the customary forty minutes of anguish. Possible neural autocomplete.”

But even that will not settle matters. The chip may have been removed before the scan. The LLM may have planted ideas months earlier. The author may have absorbed machine-generated phrasing without conscious knowledge. Soon publishers will require a full cognitive provenance extending back to childhood: school essays, dreams, overheard conversations, every book ever read, and sworn testimony from the author’s mother that he was already making up stories before ChatGPT existed.

Finally, writers will have to be bred in controlled facilities, raised without internet access, and released into sterile composition chambers under the supervision of the Authors Guild and three independent neurologists. Until then, only one author can meet the standard:

Catching Crumbs

By in ai, battling personal entropy, foolosophy, writing

You can strength-train your brain with AI just as well as with other humans, and often far better.

— Venkatesh Rao

On August 7, 2024, I published a post called “I’m Bored.” It was the only thing I posted that month, the longest unintended silence in the history of my blog. One reader emailed weeks later to ask whether everything was all right. Had I suffered some personal trouble?

Nothing had happened. That was the trouble. I was bored with the state of online writing, bored with the people arguing on social media, bored with watching the same empty controversies reappear in slightly altered costumes. My reading had gone stale too. Books passed before my eyes without striking sparks. Work was consuming most of my time and energy, and by the time I got home, writing required an effort of will I could rarely muster. I had not dramatically renounced the life of the mind and I had no intention of officially quitting writing. I had simply stopped finding much pleasure in it. I wasn’t depressed, though — it was more like being in suspended animation, waiting for… I knew not what.

In retrospect, I was intellectually overfed and undertrained. I consumed words constantly, online and off, but converted almost none of them into useful work. Reading had become grazing: intake without exertion, stimulation without adaptation. I was taking in too many empty intellectual calories with no way to work them off, and the result was not nourishment but sluggishness.

One year and one day after that nadir, I started a writing streak that has continued uninterrupted to the present. I have published one post a day ever since. In fact, I quickly made a conscious decision to limit myself to one post a day. That’s how fast the ideas and words are coming. What turned the trickle into a firehose? Work is still busy, after all. And I still train hard in the gym. The only thing that has changed is my discovery of AI.

Before going further, I should specify what I mean by “AI,” a term now broad enough to conceal almost every distinction worth making. It can refer to image generators, recommendation systems, medical diagnostics, workplace surveillance, customer-service bots, deepfakes, autonomous weapons, plagiarism mills, or software designed to eliminate somebody’s job. People regularly begin by objecting to one of these things, move without warning through three or four others, mention data centers drinking water, and conclude that asking a language model to help consider an essay is intellectually fraudulent. Rather than stages of one argument, these are several different arguments bundled together under a convenient abbreviation, which makes most casual discussions counterproductive.

I am not offering a general defense of everything called “artificial intelligence.” My subject is one particular use of large language models: sustained conversation with them as intellectual companions, sounding boards, research assistants, dogsbodies, and training partners. Whatever objections apply elsewhere are not relevant here.

Even within this narrower category, “using AI” tells us very little. A person may know nothing about Heidegger and instruct a model to produce eight hundred words giving the impression that he does. Or he may paste in a paragraph from Heidegger and say: Explain this to me as if I were five, because I cannot tell what the hell he is getting at. Surely he is not saying X? Would it be fairer to say Y? What distinction am I still missing? Both people are “using AI,” but one is asking it to conceal his ignorance while the other is asking it to help him locate his ignorance precisely.

My own use resembles the second. I don’t ask for finished thoughts to present as mine. I begin with a thought of my own — a quotation, an irritation, an analogy, a half-remembered anecdote, a joke that might contain the germ of an argument — and offer it up for discussion. The model responds. I object, qualify, extend, or redirect. It returns the baton. The conversation may continue intermittently through the workday until, by the time I get home, the stray thought has acquired structure and weight. Before this became possible, such thoughts occurred just as often. Most of them died unattended.

A passing idea is perishable. It must arrive when one has time, energy, access to the right books, and perhaps a willing interlocutor. Miss the window and the excitement drains away. You may scribble a few words in a notebook, but when you return ten hours or ten days later, you find a cold memorandum left by an earlier version of yourself: “High Sparrow versus Daenerys — religious tyrant/secular liberator.” Whatever made the comparison seem urgent has vanished.

A notebook can refrigerate a thought. A conversational model can reheat it. Better still, it can keep the thought in motion while you are temporarily occupied elsewhere. As I said above, it’s like passing a baton back and forth. During a break at work, I mention something that has occurred to me. When I next look, the thought has been considered, connected to other ideas, perhaps challenged at its weakest point. I carry it a little farther and hand it back. The exchange preserves continuity across a fragmented day. I do not have to recreate the original moment of inspiration from scratch after spending eight hours moving boxes and pallets.

This has shown me just how many crumbs of thought used to fall to the floor. I had assumed that my low output reflected a shortage of material. In fact, material was everywhere. My problem was I lacked a way to gather and process it before it was swept up or crushed into dust.

This is why I like Rao’s strength-training comparison. Regular use changes one’s working capacity. A single interesting conversation may produce an idea, but daily conversation trains the ability to recognize what is profound in a passing observation, distinguish a premise from a punch line, identify the weak point in an analogy, and carry a thought beyond its first attractive formulation. Each exchange is a repetition. Over time, the repetitions accumulate and the mental muscles grow.

The process becomes self-reinforcing. More thoughts survive long enough to become posts. Finished posts increase confidence and appetite. Greater appetite makes one more alert to usable material. Frequent discussion builds the stamina required to follow an argument through several turns rather than abandoning it at the first difficulty. Reading becomes more nourishing because it once again has somewhere to go. Instead of consuming words passively, one reads with the expectation that they may soon be tested, disputed, connected, or put under load.

The conventional fear is that AI will make people mentally fat and lazy by doing their thinking for them. That can certainly happen. A person can use a model as a forklift, moving intellectual weight without exerting himself. But I had become mentally sluggish through passive consumption before I began using AI. The critics who say that “AI does your thinking for you” have usually failed to ask what the user is actually doing. There is a dramatic difference between issuing the command “Write me an essay about Heidegger while I take a nap” and spending a day exchanging arguments, objections, analogies, jokes, revisions, memories, and quotations until a finished thought emerges. One use eliminates exertion while the other increases its volume and intensity.

The same is true in a physical gym. A barbell is an artificial contrivance. So are plates, racks, benches, cables, and machines. They allow resistance to be measured, movements to be repeated, weaknesses to be isolated, and loads to be increased gradually. No one concludes that the resulting strength is counterfeit because it was developed in a climate-controlled room with specialized equipment.

Well, almost no one. A purist — and I have met such — might insist that authentic strength should be acquired “naturally,” by pitching hay, digging ditches, hauling stones, and shoveling mulch for eight hours a day. He might regard the cable pulley as a technological shortcut that performs the labor on the lifter’s behalf. He would be wrong. The equipment does not provide the strength. It organizes the resistance. So does a language model. To say that AI thinks for me because I think with it is like saying that a barbell lifts for me because it supplies the weight. The barbell provides the resistance. I still have to perform the repetition.

The comparison also clarifies why the best model of LLM interaction is not dictation but Socratic dialogue. The machine should not be treated as an oracle handing down judgments. Its first answer is the opening move in a cross-examination. What do you mean by that? What in the text supports it? Are there competing interpretations? Did your last answer contradict the first? Explain the point without the specialized vocabulary. Give me an example. Now give me a counterexample. Surely this implication cannot follow. Where have I gone wrong?

A model may answer confidently and still be mistaken. Fluency can disguise error, especially in difficult or contested subjects. No one should ask a question, receive an answer, and go confidently on his way as though he had consulted holy writ. But a confident mistake that survives one question may have trouble surviving five follow-ups, a demand for evidence, a competing interpretation, and an attempt to restate the point in ordinary language. Conversation does not guarantee truth, but it does give error more opportunities to expose itself.

But the possibility of confident bullshitting is not a defect unique to machines anyway. Human teachers, scholars, friends, and authors also speak confidently, contradict themselves, misunderstand questions, repeat fashionable errors, and defend bad first answers out of pride. The remedy has always been the same: further questions.

Socrates did not improve minds by delivering immaculate lectures. He worried at assertions until their hidden assumptions appeared. He made people define their terms, confront contradictions, and discover that they did not know what they thought they knew. That should be the standard of LLM interaction. Naturally, like Socrates, the technology has inspired a sizable constituency eager to make it drink hemlock.

Books offer depth without responsiveness. Human beings offer responsiveness with finite patience. The LLM occupies the territory between them. That territory has altered my intellectual life. It has not made every thought profound, every post necessary, or every argument correct. Producing more writing is not necessarily the same thing as producing better writing. A daily habit can generate its own quota of weeds. But after the sterility of August 2024, I am not inclined to sneer at abundance. The important change is not merely that more words have appeared on my blog. It is that reading, questioning, and writing have again become parts of one continuous activity.

Strangely, I find that my conversations with an LLM have indeed changed my writing, but not in the way one might expect. For some reason, which I won’t pretend to fully understand, I find myself much more willing to elaborate and elucidate my thinking when brainstorming with an “interlocutor from nowhere.” The LLM’s absence of a personality somehow makes me more inclined to spell out my thoughts rather than condense them into a rhetorical nesting doll. Probably this is because with an LLM, the more fully you state the thought, the more useful the reply tends to be. It may then be simple self-interest: you get out what you put in, so make the extra effort. In any event, I’ve noticed that this has made my writing longer and more overtly explanatory than ever before, and the practice has followed me here.

Previously, I tended to summarize my thoughts in extremely condensed fashion. If the reader didn’t get all the allusions, the implications, or the brilliant inside jokes, I didn’t care. I admit this owed to both impatience and indifference. It’s up to you to keep up; if you can’t, that’s not my problem. Writing to an LLM has somehow broken down this recalcitrance. Maybe it’s a temporary change. Maybe when the novelty of abundance wears off, I’ll return to speaking in koans and whacking you with a bamboo stick. I won’t pretend to know; I’ll just enjoy the new experience.

Two years ago, I thought I was bored because there was nothing left to say. In reality, thoughts were occurring constantly. They had nowhere to go, and most died before I could give them form. AI gave my thinking a place to train. Arguments over whether conversations with an LLM are “real” conversations strike me as the worst kind of sterile scholasticism. Those so inclined can waste their energy delineating an LLM’s absence of a psychological center, or soul, as if it matters, but the rest of us will shrug and get on with the important business of harvesting productive thoughts.

An Ode to Prednisone

By in battling personal entropy, poetry

Mark Sandman once sang of a cure for pain;
Then prednisone rose singing through my veins.
Decrepitude, age, inflammation fade;
The failing flesh stands startled and remade.

From gravity’s old grip, both flesh and mind
Rise, sloughing off the years I dragged behind.
Sing, O Muse! I vault, caper, shout, and roll,
While borrowed fire goes roaring through my soul.

I crow with Peter Pan, afloat and free;
The earth below has lost its claim on me!
To pause the clock, to scorn the costs, I cry:
No matter if I never sleep—I fly!

But time returns; the buried ache resumes;
The god departs, and age reclaims its rooms.

Obiter Scripta, no. 166

By in aphorisms, books, obiter scripta

Such is my turn of mind that I don’t expect to learn everything, from A to Z, from books; but only to get seeds of thought that I can cultivate within myself, in my private garden. I can make something only when I set out from very little and that very little bears fruit in me; were I to take larger quantities, I’d produce nothing—what is more, I do not understand what has already come to fruition.

— Paul Valéry, The Collected Works of Paul Valéry, Vol. 14

My shelves are a kind of private seed bank: a library not principally of finished thoughts to be learned, but of dormant thoughts waiting for the conditions in which they can grow. I am the reverse Johnny Appleseed. I do not spread seeds of thought across the republic; I hoard them in climate-controlled conditions. I am not merely inattentive to the author’s whole argument; I am prowling through books with sacks, secateurs, and dubious property rights. Teach me? Author, feed me.

The Smallholder Renaissance

By in ai, foolosophy, socmed, the geist of the zeit

We are all at the beginning of a vast migration into endless intellectual hinterlands that we didn’t know existed, at least not in any accessible way, until AI opened them up. And in the process, we are becoming not merely intellectual homesteaders, subsisting in sod houses out on the frontier, but farmsteaders — running complex and expansive intellectual operations with automation. As with the settlement of the American West in the 19th century, the railroads may follow, with more industrialized agriculture in its wake, but for now, we’re in for a period of AI-assisted smallholder intellectual culture.

— Venkatesh Rao

I am not accustomed to finding myself on the technological frontier. Usually, by the time I arrive, the frontier has parking regulations and a Walmart.

I still use bookmarks. I use a browser rather than apps. I read writers who have been dead for a century and maintain a personal website in an age when most people regard owning a website as an eccentricity somewhere between repairing your own shoes and corresponding in Latin. I have never felt any particular obligation to keep up with the future. Most new technologies strike me as ingenious solutions to problems I was content to leave unsolved. And yet I find myself unusually excited by large language models, for many of the same reasons Rao details.

What makes the technology appealing is not that it feels futuristic. It feels like a return (or RETVRN, as the reactionary kids like to say). My schematic history of the internet has three stages. The early web was the classical world. The social-media era was the Dark Ages. The arrival of LLMs marks the beginning of a Renaissance.

The early web was exuberant, decentralized, amateurish, ugly, learned, obsessive, and alive. People built places. They made personal websites devoted to forgotten poets, regional ghost stories, guitar pedals, local architecture, obscure religious sects, or every movie in which a particular character actor had appeared. Pages were linked to other pages because someone had found them interesting, not because a platform had calculated their likelihood of prolonging engagement. You followed paths laid down by individual curiosity.

Most of these places were badly designed. That was part of their charm. Their proprietors painted the walls whatever color they pleased, installed blinking signs, tiled the background with tiny images, and arranged the furniture according to private systems comprehensible only to themselves. Visiting the early web was like wandering through a settlement of owner-built cabins and eccentric little villas. You never knew what might be behind the next door.

The web was not free of stupidity, quarrels, fraud, pornography, vanity, or political mania. The classical world was not free of sewage. But its dominant sensation was expansion. Each link opened onto another little territory. Online life felt larger than ordinary life because people were continually making new rooms in it.

Then came social media, or “socmed,” a nod to Orwell’s “INGSOC.” Facebook promised a global village and delivered a brutalist-style housing project.

Instead of building your own place, you were assigned an apartment inside an enormous standardized block. You could decorate the interior with photographs, slogans, and approved widgets, but the architecture belonged to someone else. The corridors, elevators, sightlines, locks, and exits were controlled by the administration. Your “page” was not a piece of property. It was a unit allotted to you under terms that could change without notice. Even the shift in vocabulary revealed the change. On the old web, you built a site. On social media, you maintained a profile. One was a place you made; the other was a dossier about you.

The metaphor extended beyond surveillance. Life in the socmed block came with thin walls and vigilant neighbors. Every remark could be overheard, screenshotted, reported, stripped of context, and brought before the building committee. The authorities did not need to police every room because the tenants volunteered for auxiliary duty. Everyone became both resident and informant. Snitches were showered with riches rather than laced with stitches. We were herded into a few gigantic courtyards where everyone reacted to the same controversies, learned the same slang, performed the same approved identities, and watched the same handful of people denounce one another. The global village became a company town, then a prison yard.

The early web had encouraged eccentricity because most people were difficult to find. Socmed discouraged it because everyone was permanently exposed to the crowd. An odd private interest had to compete with the urgency of whatever the entire building was shouting about that morning. The socmed era felt paradoxically smaller even as its platforms contained more people, more information, and more activity than the old web ever had. The landscape had been compressed into a feed.

LLMs have given me back the sensation that the internet opens outward. The change is not technical decentralization. My solitary cabin remains connected by a power line to an enormous corporate data center. The frontier depends on centralized utilities, and the owners of those utilities may eventually behave like railroad barons. Rao’s warning is built into his metaphor. The railroads may arrive. The land may be parceled, standardized, consolidated, and planted with whatever crop produces the highest return. But for the moment, the experience is one of private exploration.

Socmed constantly asks: What is everyone talking about? An LLM asks: What are you thinking about? That difference is enormous. Two people can use the same model and travel in opposite directions. One asks it to write marketing copy. Another studies Akkadian verbs. Someone constructs a private concordance of Jules Renard. Someone else spends an afternoon producing heroic art venerating Willis Carrier, then veers into Nietzsche, social conformity, and the spiritual significance of swearing under a barbell.

The shared technology does not require a shared culture. It can support millions of private cultures. The operation does not have to scale. It need not attract subscribers, generate revenue, establish a brand, or enter the common conversation. A line of inquiry may justify itself because it enriches one morning. A discussion may be worth having even if it never becomes content. Under platform logic, private thought is merely unpublished inventory. Under smallholder logic, it is part of maintaining the soil.

The LLM also supplies something the old web could not: responsiveness. The early web contained endless strange settlements, but you still had to find the right one. Somewhere, perhaps, a retired professor had built a page explaining the obscure passage you were trying to understand. Somewhere, a hobbyist had documented the half-remembered event you wanted to investigate. Somewhere, another eccentric had noticed the connection that had just occurred to you. Theoretically, you could find these people. In practice, the friction often killed the inquiry. Now the trail talks back, so to speak.

For a somewhat cerebral and solitary person, this is exhilarating. Ideas can be discussed without worrying about giving offense, monopolizing someone’s time, testing the patience of a friend, or discovering that the other person has no desire to spend forty minutes distinguishing two possible interpretations of a sentence by Wittgenstein. Even the long-suffering Lady of the House has to beg off sometimes: “No thinky stuff tonight, darling; I have a headache!” A human companion may be willing to discuss your obscure fixation once. The interlocutor from nowhere will still be there for the ninth round.

This is what makes the newest technology feel like a Renaissance. It restores older habits under new conditions: private scholarship, amateur investigation, correspondence, commentary, commonplace books, eccentric syntheses, the small republic of letters. It gives the intellectual hermit some of the advantages of the city without requiring him to become an urban intellectual. Or, to put it less grandly, the future finally arrived in a form attractive to a conservative temperament. I had expected the future to appear as one more demand that I keep up. Instead, it appeared as permission to wander off.

Rao’s agricultural metaphor is especially apt because LLMs do not merely open new territory. They increase the amount of territory one person can cultivate. A notebook preserves a thought. A search engine locates information. A word processor records finished sentences. A conversational model can help gather, sort, test, connect, preserve, and develop material while it is still fresh.

This makes the LLM something like the Norman Borlaug of the blank page. Its innovations have introduced high-yield varieties of prose, intensive cognitive irrigation, mechanized threshing, and synthetic fertilizer for exhausted intellectual soil. A plot that once supported one stunted observation every six months can now produce three essays, a satire, seventeen aphorisms, and a prolonged inquiry into the timely appearance of seasonal candy.

Uncounted millions have been saved from idea-starvation. Does anyone hail the machine as the savior of the humanities? Naturally not. The literary environmentalists stand at the edge of the bumper crop and complain that the wheat has too many em dashes. They warn of textual monoculture, depleted semantic topsoil, dependence on proprietary seed stock, genetically modified phrases, and the extinction of venerable heirloom sentences cultivated by hand since the eighteenth century. “Yes, humanity has produced more writing than ever,” they concede, “but at what cost?”

These objections are not all foolish. The new abundance includes vast quantities of rubbish. Mechanization can increase the yield of weeds as easily as wheat. The coming railroads may favor industrial plantations over smallholders. A tool that enlarges private intellectual life can also be used to automate spam, propaganda, fraud, and empty production on a scale previously impossible.

For now, though, before the acceptable uses harden and the land is fully surveyed, there is room to build strange little operations. One can cultivate a few acres of private intellectual ground without submitting a business plan. One can follow an obscure thought until it disappears over the horizon. One can produce something useful without first asking whether the crowd will approve.

The early web gave the solitary mind territory. Social media gave it a crowd. LLMs give it infrastructure. The classical ruins are still there. The housing blocks remain occupied. The railroads are coming. But beyond them, for the moment, the land is open.

Crush the Soul

By in battling personal entropy

Some people pray or meditate to get right with themselves and the world. Others pick up iron and put it down again. You break through your usual self-imposed limits, feel a little pain and fear in the moment, but come back together a little stronger on the other side of it.

The purpose of powerlifting as a spiritual discipline is to crush the soul. The soul is naturally vain, lazy, talkative, and full of excellent reasons not to do another rep. Other disciplines seek to elevate the soul or purify it. We place it under several hundred pounds until it stops talking nonsense.

The gym has its own severe monastic order. We rise at appointed hours, observe fixed rituals, submit to repetitive labor, mortify the flesh, and periodically gather around a man undergoing an ordeal to offer prayers in solidarity. The air smells of sweat and liniment instead of incense, the bells are iron plates striking the floor, and the monks’ chants are in guttural Anglo-Saxon oaths rather than Latin. Our spiritual director carries a stopwatch and answers pleas for mercy by saying, “Three more.”

You do not emerge purified exactly. You emerge sweaty, trembling, and briefly unable to form a sentence. But the scattered pieces come back together on their own, and the self that returns is harder to frighten than the one that went under the weight.

If You Think This Has a Happy Ending, You Haven’t Been Paying Attention

By in bread and circuses

Ed West:

The final series of Game of Thrones came out two years ago and ended with great disappointment, in fact so much so that millions of people demanded that it be remade (which perhaps says something about how mundane our problems were before 2020).

Yet although all those people who called their daughters Khaleesi are probably regretting it now, the show’s finale at least told a certain truth. The attraction of Game of Thrones, compared to most historical fiction, is that alt-history allows its creators to tell an almost purer truth about history and human nature. As amoral as some of the characters are in Westeros, they behave as people would have done in such a situation, devoid of modern idealism or comforts.

Daenerys, the queen over the water, is the only true idealist among the contenders for the throne, and campaigns to end slavery in Essos before liberating her homeland. And yet when she arrives in Westeros she turns out to be far more murderous and oppressive than what went before. Likewise the ancien regime, terrible that it was, had nothing on the horrors unleashed by modernity, which contained far greater lies than the world of chivalry.

In his newsletter, West linked to this article of his from 2021. As I mentioned before, I recently watched the entire Game of Thrones series as a favor to a friend who wanted someone to discuss it with. I loved it. There were a few choices I felt were wrong turns, and the later seasons definitely moved things along at “plot speed” rather than the slower, detailed pace of earlier seasons, but overall, I thought it was very well done. The ending in particular surprised me by how completely it worked. It made perfect sense to me.

That’s why I was amused to be reminded just how unpopular the ending had been. When the finale aired, I remember seeing people complain about it on Twitter, but I didn’t really pay attention. I wasn’t watching the show, and I assumed the iron law of social media was in effect: the discourse is always driven by the loudest and stupidest people. So, again, I went into the series open-minded, knowing little, half-expecting it to conclude in a clown-car pileup, only to discover that the ending was not merely defensible but, in its broad conception, brilliant.

Seven years late? Doesn’t matter. Let me try to talk sense into the sorts of simpletons who sign online petitions and name their daughters after fictional mass-murdering tyrants. And let’s begin by boldly seizing the dragon by both horns.

Those who felt betrayed by Daenerys’s mass destruction of King’s Landing by dragonfire seem to feel that this was some sort of perversion of her character. I would claim that it was the inevitable denouement of it. Ever since she walked into fire and came out unharmed with three young dragons, she had been a messianic liberator who, crucially, believed in her own mythology. A common complaint was that her “turn” felt forced, that we hadn’t been shown enough cracks appearing in her character to accept her transformation into a monster. But the idea that she needed more psychological deterioration quietly smuggles in the premise that atrocity must be preceded by visible mental collapse. The eyes must go wild, the hair must come loose, the villain must begin talking to herself in mirrors. Otherwise viewers feel cheated because the person still looks recognizably human. But Daenerys did not need to become a different person. She only needed to encounter circumstances in which her existing convictions produced different results.

She had always believed that she possessed exceptional moral authority, that resistance to her was usually proof of corruption, and that spectacular violence could cleanse the world and establish justice. In Essos, the story often placed obvious monsters, like cartoonish slaveholders, opposite her, so those beliefs looked heroic. They flattered the complacent assumptions of the viewer. In Westeros, she confronted divided loyalties, a rival claimant, political indifference, personal loss, and a population that did not greet her as its ordained savior. The moral framework remained the same. Only the material fed into it changed. She does not wake up one morning and decide, “I am evil now.” She concludes that mercy has failed, love is unavailable, fear will work, and history has authorized her to act. Plenty of real atrocities have been committed by people who remained coherent, purposeful, affectionate toward intimates, and convinced of their own righteousness. Insanity is almost a comforting explanation because it quarantines evil inside a visibly defective mind.

The contrast with the High Sparrow provides all the clues anyone should have needed. Both embody fanatical certainty armed with power.

The High Sparrow arrives already marked as dangerous in a way that a modern viewer would instantly recognize. He is old, male, religious, sexually censorious, hostile to pleasure, dressed in rags, surrounded by chanting zealots. A Savonarola figure. Modern viewers possess an entire vocabulary for him before he finishes his first speech: fanatic, fundamentalist, theocrat, hypocrite, patriarch. His humility is presumed fraudulent; his moral certainty is understood as a lust for power; every punishment carried out in the name of virtue confirms the diagnosis.

Daenerys receives the opposite treatment. She is young, beautiful, sexually liberated, multicultural, and opposed to slavery. She speaks of breaking chains rather than enforcing commandments. Her armies advance beneath the banners of emancipation, accompanied by dragons and triumphant music. The High Sparrow says that sinners must submit to the gods; Daenerys says that oppressors must submit to history. A Napoleon figure. Modern audiences recognized the High Sparrow as a fanatic because he spoke the obsolete language of salvation; they mistook Daenerys for a hero because she spoke the modern language of liberation.

But the political structure is clearly similar. Both possess a moral authority that cannot be negotiated with. Both divide the world into the redeemed and the guilty. Both interpret resistance not as disagreement but as proof of corruption. Both make suffering purgative. The High Sparrow strips Cersei naked and marches her through the streets for the salvation of her soul; Daenerys burns men alive for the liberation of mankind. He is medieval enough for viewers to fear. She uses the language in which modern power prefers to complacently describe itself.

Daenerys is not merely a secular version of the High Sparrow, though. She is the more formidable historical type: the liberator who becomes an emperor without ever ceasing to think of herself as a liberator. That’s the Napoleon comparison. The revolutionary promise travels abroad on horseback (or dragonback) bringing universal emancipation to peoples who somehow keep having to be conquered before they can enjoy it. Her dynastic claim, personal destiny, and humanitarian mission fuse into one argument: she ought to rule because history, blood, suffering, and justice all point toward the same person.

Her final violence therefore does not require a descent into cartoon madness. She does not abandon her ideals; she draws their final conclusion. The people have failed to welcome their liberation; rivals have failed to acknowledge her destiny; mercy has failed to produce obedience. What remains is fear and overwhelming firepower. The idealist and the tyrant were always the same woman.

I had no trouble recognizing this because I am temperamentally conservative. Varys was my favorite character in the show. He knows that when institutions collapse, “freedom, equality, peace” do not automatically follow. Starvation, armies on the march, opportunists, zealots, burned cities, and strongmen follow. He is temperamentally conservative in the anti-collapse sense. He has no illusions about rulers, but that makes him more afraid of believers with armies.

Thus, the ending was not a betrayal of Daenerys but a revelation of the audience: viewers who instantly recognized tyranny when presented in stereotypical guise failed to recognize it when it flattered their unthinking modern prejudices. I’m tempted to conclude that much of their anger was simply compensation for their embarrassment at realizing how easily they can be manipulated by a pretty face and some stirring rhetoric. I guess I was right after all: the discourse is driven by the loudest and stupidest people.

Be My Signal, End This Noise

By in ai, environment

I love the smell of napalm deployed against the Smugstack consensus. Smells like victory for common sense. But what I really enjoy here is seeing how they’ve already got it all set up to run each post through a one-click authenticity detector. It’s like the Amazon Prime option for self-righteousness. “Scan for AI text.” Oh, you mean use the same black-box AI software that depends on “data centers drinking up all the water”? Or does the software only use holy water when performing moral hygiene?

I kid, of course, because data centers house computing, storage, and network infrastructure generally. Email, streaming, online shopping, search, social media, business software, photo backups, and AI all draw on the same broad industrial system. They’re not giant chatbot terrariums with private aquifers. In other words, before you jump on the trend to complain about data centers and environmental destruction, try deleting your Netflix, your TikTok, your Hulu, your Spotify, your Disney+, your Instagram and your billion-and-one selfies first. Let’s see how serious you are about this! Oh…

1 2