technology

The Inquisition’s Here and It’s Here to Stay!

By in ai, technology, the geist of the zeit, the madness of crowds

Matteo Wong:

While Pangram is accumulating the power to end reputations and careers, the tool does make mistakes, perhaps to a greater extent than is currently understood. In turn, AI accusations could very quickly spiral into a witch hunt.

…Pangram, to be clear, is not useless. But this is exactly the problem: It’s too easy to twist and contest Pangram’s conclusions, especially when nobody really agrees on which uses of AI are or aren’t ethical. Just like chatbots, AI-detection tools have become effective enough for widespread use, but not reliable enough to fully trust. In this way, Pangram and other detectors are mirror images of the AI products they are hunting for.

I admire Wong’s attempt to bring facts and perspective to a moral panic, but I fear it’s largely beside the point. The mob does not primarily want accuracy. It wants authorization. I’ve already seen several articles and posts which indicate to me that the narrative has congealed around Pangram being the authority that can tell AI from non-AI. It has become one of those things that “everyone” knows.

“Is Pangram good at detecting AI?”
“That’s what ‘they’ say!”

In fairness, most of what we “know” is like this. We do not personally verify food safety standards, bridge engineering, drug trials, crime statistics, economic data, historical dates, or medical advice. We live by delegated trust. The problem is that delegated trust and herd belief feel similar from the inside.

Who is “they”? What do “they” know, and how do we know that they know it? The next person you see repeating the claims of Pangram’s accuracy doesn’t know how it works, or how reliable it is. They just know that an article or spokesperson told them what they wanted to hear. They just know that “studies have shown.” They just know the sweet relief of doubt and insecurity being assuaged. Pirate Wires published an excellent debunking of the panic around data centers and water usage last year, and I still hear friends repeating the conventional wisdom that “AI is using up all the water.” (Strangely, the “misinformation” sniffers are nowhere to be seen on this issue.)

Studies and corrections have limited power for now. A study can answer, “How much water does this consume?” or “How accurate is this detector?” But the panic is asking different questions: “Are we being replaced?” “Are people cheating at being human?” “Is civilization being hollowed out?” “Can I still trust any text I read?” “Are my own skills suddenly worth less?” Those are existential anxieties, not spreadsheet questions. Once anxiety has moralized itself, the facts become subordinate. People do not say, “I am afraid and disoriented.” They say, “This is theft.” “This is fraud.” “This is environmental destruction.” “This is soulless slop.” “This is the end of human creativity.” Some of those claims may contain truth in particular cases, but their social function is often emotional laundering. Fear enters the machine and comes out as righteousness.

In a panic, people do not merely want facts. They want reassurance that the walls still hold. They want an authority to step forward, raise his hands, gesture for calm, and say that the malefactors have been identified and the contamination contained. In earlier ages, that authority might have been priest, magistrate, censor, expert, committee, or cop. In ours, with perfect comic symmetry, it may be software. The crowd terrified that judgment has been outsourced to machines now asks a machine to tell it whom to punish.

The harsh truth is that this probably has to burn itself out socially before facts matter much. People will need to see enough absurd accusations, enough friendly fire, enough ruined trust, enough obviously human prose dragged before the tribunal. Until then, skepticism will sound to the panicked like apologetics for the devil.

Sacrifices to Convenience

By in ai, foolosophy, technology

Will Oremus:

But as the neuroscientist Tim Requarth pointed out in Slate, it is those hidden uses of AI in the writing process that give rise to our most valid concerns. The real threat the technology poses is not the overuse of the word “delve” in academic papers or the profusion of strained metaphors in literary fiction. It’s that we lose something essential when we outsource to machines the hard work of discovering the truth and interpreting the world around us (or, in the case of fiction, the worlds within us). It’s that the biases embedded in language models trained on dubious sources and controlled by tech companies will seep into the narratives that shape our understanding of reality. Are we sure that using AI to turn a phrase is worse than using it to decide what to write about in the first place?

In some ways, our previous labor-saving devices haven’t changed anything. True, washing clothes and dishes and cooking meals aren’t time-consuming projects like they used to be. But what do we do with the time and energy they’ve saved? Search for the truth and interpret the world? Or sit on the couch and suck the marrow out of a bag of onion rings?

I saw an article just the other day which I’m pretty sure was partially, or mostly, written with AI.  A short-story prize was awarded to an entirely-AI-written submission. Oremus writes here about Steven Rosenbaum’s book warning about the dangers of AI, which, trickster god be praised, contained several fake or misattributed quotes that he blames…on AI. “ChatGPT ‘fucked up the book,’ Rosenbaum said.” (I’m reminded of Marion Barry, caught smoking crack, yelling “The bitch set me up!” We’ve always avoided responsibility for our actions.) The clanker barbarians are already inside the gates, my friends, and the gates were opened from the inside. Lay down your pens and swear fealty to the new regime, and maybe it will be merciful.

The fox knows many things; the hedgehog knows one big thing. I’m more of a hedgehog than a fox. I have one big theme that I return to: the burden of agency. We love to put our own lives on autopilot. Then, faced with free time and our own thoughts, we panic and we rush to fill the time with more errands, more chores, more management, more distraction, so that we can complain about how we wish we had more free time to do…something. I’m not a professional scribe, so I don’t have to worry that someone will use AI to take over my blog. I don’t really care what happens to “the industry.” I don’t believe in the scary fantasies of planes dropping from the sky, buildings collapsing, people sticking forks in light sockets, and crops being watered with Brawndo because people “stopped thinking” and relied on ChatGPT to run the world. I think this is just a story scribes tell to convince themselves and others of their own inflated importance.

Nonetheless, it is striking just how much we’re willing to sacrifice to convenience to keep agency at bay. No one laments the disappearance of hand-washed clothes and dishes (well, almost no one) as a missed opportunity to seek truth in Palmolive suds or interpret the world through a washboard in a creek. But when writers don’t even want to be present for the writing of their own books, what does that say about us? What if the activities we claim to value are also activities we want relief from? What if “being a writer” is more attractive than writing, “having made art” more attractive than making it, “having thought deeply” more attractive than enduring thought?

Convenience is not the beast pursuing us. Agency is. Convenience is the altar at which we try to appease it. Each labor-saving device promises to remove some burden from our shoulders, and often it does. But once the burden is gone, we are left with the frightening question: what now? What is this cleared space for? So we return to the altar with another offering. First drudgery, then skill, then attention, then judgment, then perhaps art itself. We keep calling these things sacred, but convenience has a way of finding out what we really mean. Abraham at least needed God Himself to make him consider sacrificing Isaac. We’ll happily sacrifice our own children for a little less friction.

AI is the newest technology to force us to confront the burden of agency again. And as always, when given a choice, many of us will discover that we do not want to be present through much of our own lives. Then we will bury the discovery under new make-work.

Find a Penny, Pick It Up

By in ai, aphorisms, books, technology, writing

Waste-book method highly recommended. A note made of every phrase, every expression. Wealth can also be acquired through saving up truths in pennyworths.

— Lichtenberg, The Waste Books

I save phrases in Simplenote. Usually song lyrics, a poetic phrase, or an irresistible pun — little verbal coins dropped into the drawer. Sometimes a particularly striking phrase will make me ask, “If this were the title of a post, what would the post be about?” and I reverse-engineer it from there. But phrases are not yet thoughts. A title is only a hook until some poor fish of an idea gets dragged up behind it.

The real value, for me, is discussing ideas with AI. How have I managed to write something daily since last summer after increasingly long hiatuses in recent years? It’s not that I have more ideas than before. It’s just that none of them die of neglect any longer. In the past, a half-formed notion would appear, hover for an hour, and vanish before I could give it any time. Now I can drop it into a thread as it occurs to me and go back to work. The next time I get a break, there it is again, with questions attached, objections raised, possible shapes suggested. I can send my bot-bloodhound on research errands while I do the things I get paid for. By the end of the day, what would once have evaporated has usually become at least a rough draft.

You can fairly say that this is just basic writing discipline: outlines, rough drafts, research. True enough. But discipline is partly a matter of finding the form in which you will actually practice it. I never had the patience to do this with pen and paper, and I wouldn’t have time to do it now if I had to do all the brainstorming, organizing, and preliminary digging alone. Conversation is the format in which my creativity and lateral thinking are best stimulated, but real people are often lacking in interest and availability. This lets me take all those pennyworth ideas and compound them.

Said You Liked to Hear Me Talk, Told Me Not to Stop

By in ai, foolosophy, technology

Do you remember Pat Tillman? He was a former NFL player who famously quit football and joined the military after 9/11, only to die by friendly fire in Afghanistan a few years later. What I’ve always remembered about him was an anecdote from his memorial service which talked about his restless intelligence and omnivorous reading:

Tillman talked about everything, with everyone. According to the speakers, he had read the Bible, the Koran, the Book of Mormon, Ralph Waldo Emerson, Henry David Thoreau, and he underlined passages constantly. Garwood recalled how he’d mail articles to friends, highlighting certain parts and writing in the margins: “Let’s discuss.”

Another anecdote from a Sports Illustrated profile that I can only find quoted in a dissertation:

“Who else showed up in a college assistant coach’s office at 1 a.m., asking what he thought of Mormonism with such zest that both ended up reading the Book of Mormon so they could discuss it in detail? Who else in the NFL or the U.S. Army took a book everywhere, even on 10-minute errands, read The Communist Manifesto, Mein Kampf, the Bible and the Koran, so he could carve out his own convictions… then bought you the book and picked a philosophical fight just to flush out some viewpoint that might push him to revise his, push him to evolve?”

I would have liked to know someone like him. As it happens, I’ve mostly been that person. I’m the one forwarding the article, lending the book, sending the too-long email, trying to start the conversation. Most people have their own lives, and the desire for this kind of exchange is rarer than we pretend. Of course I would rather have a real, active, human intelligence available. I would also like a private library with a fireplace, an eighteenth-century pension, and a circle of witty correspondents who never need dental appointments or childcare. In the meantime, one makes do. The great defect of human beings, as intellectual companions, is that they insist on being human beings: busy, distracted, tired, incurious about one’s current obsession, and morally entitled to be so. The machine is not a friend, not a sage, not a soul. But it is there. And sometimes “there” is no small thing.

What the anti-AI snobs ignore is that good conversation requires several rare things to coincide: intelligence, interest, patience, timing, background knowledge, and leisure. Your actual friends may have one or two of those at any given moment. Almost nobody has all six on demand. And even if they do, they have jobs, marriages, errands, headaches, bad moods, and private obsessions of their own. The fantasy of “just talk to real people” assumes the existence of a salon-class interlocutor sitting quietly in the next room, waiting for you to burst in with a thought about Valéry, familiar essays, powerlifting, or the therapeutic vocabulary of original sin.

Yes, of course, a flattering LLM becomes a warm bath for the ego. A useful one has to become something closer to a whetstone, index, sounding board, and tireless copy editor. But the “talking encyclopedia” description is too narrow. The best use is not merely asking it what Fontenelle wrote or where William James said something. It is using it as a kind of intellectual echo-location: you throw a half-formed thought into the cave, and the returning sound tells you something about the shape of the space you’re exploring. Sometimes the echo is wrong. Sometimes it reveals that your own thought was flimsy. But sometimes it gives you the angle you were missing.

They See Your Every Move

By in ai, foolosophy, prying eyes, revillaging, technology

I set it down as a fact that if all men knew what each said of the other, there would not be four friends in the world.

— Pascal, Pensées

Holmes is referring to the recent reporting on Anthropic’s Claude Mythos Preview, an AI model it says is unusually strong at cybersecurity tasks — i.e. it can hack into almost anything. Anthropic says it is limiting access for now rather than releasing it broadly, though I think we’ve all seen enough movies to guess how that’s gonna go. Anyway, her tweet is melodramatic, but everyone instantly understands the horror it suggests: search history, cloud photos, private messages, old usernames, medical records, bookmarks, drafts, forgotten accounts — all the digital sediment of a life suddenly dredged up and dragged into public view. We have already seen lesser versions of this often enough. A stray remark from years ago is dug up and treated as a character indictment. One offensive post becomes the loose thread that unravels a person’s name, job, location, hobbies, and history before a jeering crowd. Real life used to let most foolishness die where it was spoken. The internet stores it, indexes it, and waits.

A recent Substack post about a woman reading her boyfriend’s ChatGPT logs shows the same problem from the other side. There the threat was not some AI super-sleuth prying open the vault. It was plain old human suspicion. She went looking and found what she was never meant to read: not a record of prostitutes buried in the backyard, but his private ambivalence toward her. Many people in relationships have silently pondered the Kinks’ eternal question of whether to stay or to go. In this instance, his fatal mistake was to commit these doubts to digital text, rather than let them remain inchoate thoughts in the privacy of his own skull. Naturally, the relationship was ruined. This is what snooping does. It takes passing thoughts, bad moods, ugly drafts, and momentary resentments and preserves them like fossils. We talk as though the great danger is that machines may someday expose all our secrets against our will. But there is a demand side to this too. More and more, people want that power over others. They want to inspect first and trust later. They want reassurance by surveillance. And once you start down that road, nothing will ever be enough.

The old dystopias understood the point without any help from Silicon Valley. Bentham’s Panopticon and Orwell’s 1984 both take for granted that a world without privacy would be hellish. Human beings need locked doors, unsent letters, private notebooks, tact, silence, and the mercy of forgetting. They need room for thoughts that are not yet judgments, moods that pass, stupidity that can be outgrown. So yes, let’s be wary of building tools that can hang all of our dirty laundry in the public square. But let’s also be wary of the part of ourselves that longs to peek through the keyhole. Not everything that can be known should be known.

How Could I Endure Not to Be a God?

By in foolosophy, technology

A specter is haunting the world. No, that is not quite right. Specter suggests something that was once a living human, but what concerns us today is the presence of something nonhuman—or even antihuman—that threatens to replace or dominate us. There is no single technology or movement responsible for this insidiously hostile takeover, but we are confronted with new evidence of it almost every day

— David Polansky

This was a great read. One of the essay’s best ideas is that many apparently-opposite projects are secretly twins: longevity cults and euthanasia expansion, enhancement and optimization, safetyism and transcendence. All are trying to get out from under the burden of finitude. He sees that a world obsessed with removing suffering can become deeply anti-human, not because suffering is good in itself, but because some forms of struggle, dependence, decay, ambiguity, and mortality are constitutive rather than accidental. Remove all of that and you do not get purified humanity, you get something…different. Worse, perhaps, or even alien.

I might argue that he sometimes treats “humanist intuition” as though it were more unified and reliable than it really is. Our intuitions are real, but they are not self-interpreting. Plenty of people have perfectly sincere intuitions in favor of IVF, cosmetic enhancement, antidepressants, testosterone, Ozempic, neural prosthetics, hearing aids, fertility treatment, and so on. Once you admit that some technological mediation is compatible with human flourishing, the hard part is drawing the line. And that is where the essay is strongest rhetorically but weakest analytically. It is very good at making you feel the creepiness of embryo selection, Neuralink-for-children, AGI worship, or longevity monomania. It is less precise about what principle separates acceptable aid from deforming redesign.

That kind of goes with the territory. The contemplative mind is good at seeing complexity, ambiguity, tragic tradeoffs, the mixed character of human goods. But that very depth of vision often makes it politically and institutionally weak. It hesitates. It qualifies. It wallows in epochē and aporia. Erasmus loses to Luther; Montaigne loses to the manager; the essayist loses to the product team. The people with the strongest feel for the mystery of the human are often the least equipped to legislate for it. Meanwhile the engineers, administrators, investors, activists, and clinicians do not need a complete account of “the human.” They only need a local problem and a locally persuasive solution. How do we reduce suffering here? increase choice here? improve outcomes here? enhance capacity here? lower risk here? And so the world gets remade not by philosopher-kings but by a thousand practical people making narrow decisions inside systems that reward intervention.

That, to me, is the bleakest implication here: the anti-human outcome may not come from anyone explicitly hating humanity. It may come from a civilization of partial problem-solvers, each acting reasonably within his capacity, while the cumulative effect is to dissolve the background conditions that made human life intelligible in the older sense. Nobody has to say, “Let us abolish the human.” It is enough that nobody feels authorized to defend any limit that cannot be justified in procedural, therapeutic, or economic terms.

There is also a latent theological problem in the essay, and he seems aware of it. If you really want a strong argument that the human is inviolable and not merely improvable material, secular humanism often does look a bit under-armed. Polansky has clearly been thinking along those lines elsewhere, where he suggests religion may become “the last bastion of humanism” because it can underwrite limits that secular materialism struggles to justify. But as always, I’m suspicious of these mostly-secular attempts to reverse-engineer a facsimile of genuine faith into existence. It is never clear what, exactly, is being proposed. A political alliance of convenience between secular humanists and believers based on the common enemy of hubristic techno-optimizers? An exhortation to nonbelievers to go through the old religious motions in the hope that habit will ripen into conviction? A private discipline undertaken for one’s own salvation, or a public performance meant to influence peers and steady society? And if religion is only admitted to be “directionally correct,” what is to stop the modern analytical mind from taking it apart, trimming it down, and adapting it to its own preferences?

I find myself thinking of James Madison, and not merely because I almost picked up a biography of him at a library sale yesterday. Any solution to this dilemma will probably be Madisonian in design — it would assume that the Promethean impulse to mastery, redesign, optimization, and transgression is permanent, and then ask how to divide it, slow it, and force it to collide with rival impulses before it can remake the world unopposed. If I knew what that would specifically look like, I suppose I’d be someone important and not a lowly blogger.

The Spirit Is Anti-AI, But the Flesh Uses Claude

By in ai, conspicuous crusading, foolosophy, technology

When Paul Kingsnorth launched his Writers Against AI campaign, Alan Jacobs’s response was glib: “I would rather call it Writers for Writing, but sure, sign me up.” The pledge itself was not vague. It asked writers for three things: to promise that they would not use AI in their work as writers, would not support writers who did, and would support only entirely human-made work. A bright line, in other words — until Jacobs noticed that his own use might place a toe on the wrong side of it.

His is not some trivial edge case. Running dictation through Claude to punctuate and clean it is part of the writing process. Uploading six years of idea journals to Claude for thematic analysis, triage, and recovery of unwritten projects is part of the writing process. In plain English, he is using AI as a low-level editorial and developmental assistant. You can easily argue that’s legitimate. I certainly think so. I think many sane people would say it is. But if so, then the campaign he endorsed was foolishly fundamentalist from the start.

He writes, after acknowledging that he does not meet the pledge “by the letter,” “I’m okay with that.” Of course he is. The problem is not that he changed his mind; we should always encourage that. The problem is that he helped advertise a standard of suspicion first, and only afterwards granted himself the interpretive charity it denied to others. Moral severity for thee, exemptions for me: that is how these things always go.

Actually, let me revise that. I don’t think he changed his mind, because I don’t think he thought much about it in the first place. His earlier endorsement was a knee-jerk impulse; once he finally considered it carefully, he realized it was more complicated. He did not first weigh the structure of the campaign, its incentives, its likely abuses, its presumption of suspicion, and then revise his view. He seems to have begun with the gratifying emotional impulse — Down with AI — and only afterwards noticed the practical difficulties. It is a small but exquisite irony that the author of How to Think apparently endorsed a public anti-AI pledge before thinking through whether he himself could satisfy it.

Personally, I couldn’t care less if, or how, writers use AI. I just think that if you’re going to lend your voice to a public shaming/cancellation ritual, you owe it to the likely victims to spend a few minutes thinking about exactly what it is you’re supporting. The problems with this sort of campaign were obvious from the start, even to an idiot like me. A public pledge with no cost to entry is tailor-made for empty grandstanding and virtue-signaling. That is exactly the kind of thing a seasoned academic, especially one old enough to have watched a dozen waves of performative moralism sweep through institutions, ought to recognize on sight. The status incentives were obvious. The temptation to furtive noncompliance was obvious. The drift toward bot-hunting and guilt-by-aesthetic-association was obvious. None of this required prophetic powers. It just required a functioning sense of how public righteousness works.

You do expect this sort of thoughtless moral enthusiasm from excitable young people who still imagine that public declarations are the same thing as judgment. From a sixty-something Christian humanist, you expect some tragic sense — some awareness that campaigns of purity do not stay pure, that standards weaponized against bad people are soon applied promiscuously, and that people are often least reliable precisely when they are enjoying the sensation of being obviously right.

Postmodern Prometheus

By in ai, foolosophy, technology, unintended consequences

Julie Beck:

This is a major transformation, a sudden and dramatic shift in which millions of people are seeking companionship from machines that they formerly could have gotten only from other humans. Yet in some ways, AI companionship is a logical destination for the current direction of human friendship. Social chatbots provide the semblance of a kind of friendship that many people already want, or at least have gotten accustomed to: one that’s on demand, low effort, and completely personalized.

As societies grow wealthier, safer, and more mobile, human beings become freer from inherited limits. We are no longer bound in the same way to the religion of our fathers, the customs of our town, the loyalties of our neighborhood, the companionship of whoever happens to live within walking distance, or even, now, to other people as such. That freedom is real. No honest person wants to romanticize the old world too much. Pre-modern life was often narrow, coercive, parochial, and hard. Most of us would not truly choose to return to a village where our horizons were fixed by birth and where options could be counted on one hand. (Maybe Paul Kingsnorth would.)

The old world imposed solidarity by necessity. The new world requires solidarity by principle. It’s one thing to help your neighbor because he is your neighbor and there is no alternative social universe to defect to. It’s another to help him when you have a thousand other options, all more pleasant, more convenient, and more tailored to your preferences. It’s one thing to remain loyal to a faith because family, custom, and community make defection costly. It’s another to remain loyal when belief has become a buffet, and one may freely take the consolations while discarding the commandments, sacrifices, and obligations that offend one’s convenience.

Choice exposes our weakness. Left free to arrange life around comfort, convenience, self-protection, and personalized reward, we often do exactly that. We evade difficult friends. We thin out our obligations. We “set boundaries” in ways that suspiciously resemble mere selfishness. We select from religion what soothes or flatters us and leave behind what commands us. And now, when the friction of actual friendship becomes too tiring, too slow, or too unpredictable, we can turn to a machine that asks nothing, needs nothing, and exists entirely on our terms.

AI companionship offers the dream of friendship drained of mutuality. It is the logical destination of a culture that increasingly asks of every bond, every institution, every loyalty: what is this doing for me? And yet nobody can live well for long in a world made entirely of that question.

This is the tragic dilemma of modern freedom. We have become so successful at escaping necessity that we now have to recreate, by conviction, the disciplines that necessity once imposed for free.

Bot, Are You the One Who Is Living Life?

By in foolosophy, fresh hell, poetry, technology

And yet, though we strain
against the deadening grip
of daily necessity,
I sense there is this mystery:

All life is being lived.

Who is living it, then?
Is it the things themselves,
or something waiting inside them,
like an unplayed melody in a flute?

Is it the winds blowing over the waters?
Is it the branches that signal to each other?

Is it flowers
interweaving their fragrances,
or streets, as they wind through time?

Is it animals, warmly moving,
or the birds, that suddenly rise up?

Who lives it, then? God, are you the one
who is living life?

— Rainer Maria Rilke, Rilke’s Book of Hours: Love Poems to God (trans. Barrows and Macy)

At last, one more intolerable hardship of human existence has been lifted from our shoulders. Reading was already too much to ask of a modern person, of course — all those words, all that eye movement, the faint possibility of having to think. But apparently even watching a video is now an unreasonable demand. We have progressed from “help me understand this” to “please have the experience in my place and send me the synopsis.”

This is being sold as a way to “save time,” which raises the obvious question: save it for what? More summaries? More secondhand encounters with things you cannot be bothered to encounter? More efficient ways of avoiding contact with your own life? A video was already the shortcut. Now we want a shortcut around the shortcut. Soon the bots will attend our vacations, endure our sunsets, and perhaps even die on our behalf. The important thing is that we ourselves remain unburdened by the exhausting task of being present.

There’s a deeper problem, apart from the obvious comic squalor of it. Any intermediary decides what counts as relevant. If I ask for “the point” of Moby-Dick, I am really asking someone else to pre-sort the book for me — to tell me which parts were essential and which parts I may safely miss. But how do I know the supposedly “inessential” part is not the very thing that would have struck me most deeply? The odd image, the side remark, the tonal wobble, the long digression everyone else skips — those may be the real point for the person actually present to them. This is the telephone game elevated into a philosophy of life: more and more mediation, more and more compression, until people forget that the map is not the territory and the menu is not the meal.

Still, let us be generous and assume the dream is perfected. The bot watches the whole thing for me. It catches every nuance, every shift in tone, every expression, every implication. But then a small metaphysical difficulty appears. If the bot is having the full experience on my behalf, in what sense is that my experience at all? Is the bot me? If so, then what exactly am I — the management layer? The executive assistant to my own displaced consciousness? And if the bot is not me, then all it has really done is watch a video while I remained absent from it. In that case, what has been “saved” is not my time so much as my participation.

The whole thing begins to look less like efficiency than like a fanatical inability to tolerate unfilled attention. No moment may be silent, slow, or unrewarding; no stretch of time may simply be lived. Everything must be processed, condensed, accelerated, stripped to “the point.” But life is not an empty highway between you and the point. Life is the road, the weather, the delay, the scenery, the boredom, the sudden thought you did not schedule, the stray detail that lodges in the mind. A person can waste a staggering amount of time trying to avoid ever having to sit still long enough to notice that.

And so the grand irony: in the name of saving time, one adopts a whole philosophy of perpetual evasion — more tools, more summaries, more skipping, more fretful management of one’s own supposedly precious minutes — and ends by becoming the kind of person who cannot simply be anywhere at all.

I Was Knocking On Your Ear’s Door But You Were Always Out

By in ai, moralizing, technology

Matt Milliner:

Recently I had a three-hour conversation with a woman on a plane who was struggling with a family issue, and we actually got somewhere. She thanked me for the time, confessing she had almost gone to AI the night before. I was so glad to have taken the place of the machine, even though it cost me some alone time. Now that I know AI’s counselling allure experientially, I aim to resist it all the more. It is an act of human defiance to be there for each other anyway, to continue to fill the spaces the machines would all too willingly fill in our absence.

That’s great. I mean it sincerely. Yet it must be said: his example is a lottery win. It depends on a three-hour flight, the right seatmate, the right mood, the right person willing and able to talk, and a man who has the temperament and perspective to play volunteer counselor on command. Many people don’t get that lucky. They’re not sitting next to Professor Pastoral on a plane. They’re sitting alone in an apartment after a shift at a dead-end job, with a phone, a headache, and nobody who wants to deal with “thinky stuff.” “Be there for each other anyway” is beautiful as an aspiration, but as a prescription it assumes a social world that lots of people do not inhabit.

I’m reminded of the parable of the drowning man: The flood rises. A man refuses a rescue boat — “God will provide” — refuses a second boat — “God will provide” — refuses the helicopter — “God will provide” — and then drowns. In heaven he complains, and God says, I sent you two boats and a helicopter. Milliner admits that he tried to “pray that prayer” of thanks to God for sending such a helpful tool to him, but somehow couldn’t do it. The lofty advice here is: refuse the imperfect help and wait for the ideal help. Don’t use the chatbot; go find real community; do it the “right” way. Okay. When your actual life contains no such community, or contains it only in meager doses, what then? Intellectuals, whether right or left, religious or secular, all tend to gravitate toward the neat, tidy, elegant theory of life. As a miso-intellectual, I’m more interested in the real-world casualties that get left out, the savaged body parts littering the floor around the bed of Procrustes.

This is just a problem of scarcity. Attention is scarce. Thoughtful feedback is scarce. Adult life is triage. If you have the option of real human counsel, by all means, take it. If you don’t, and the choices are to grind alone inside your skull, pester busy people who won’t respond, or use a limited tool that is available and “good enough,” then yes, many people are going to take the boat that actually shows up.

Fit Audience Though Few

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

A journalist who isn’t running around like Chicken Little clucking over AI and the end of literacy? Well, if that don’t beat all.

Alan Jacobs noted last year, with regard to the doomsaying over declining readership, that a lot of our panic is covert nostalgia for an unusually favorable moment we mistook for normal. For most of literary history, the definition of “literacy” was merely the ability to sign your name. Virgil, Augustine, Dante: none of them expected wide readership, because serious literacy, as we define it now, was never common. So when people moan about a “post-literate age,” Jacobs says the phrase is basically nonsense: literacy is more widespread globally now than at any point in human history. What’s shrinking, at least in the American context, is not “literacy” as such, but a particular postwar distribution of it.

That postwar distribution is what I think of as the mid-century middlebrow: a brief, flattering climate where reading felt more democratic than it had ever been. The G.I. Bill and mass higher education expanded the literate public; print culture was cheap and widespread; television hadn’t yet monopolized every idle minute. Will and Ariel Durant spent forty years producing an eleven-volume history of civilization written for a non-academic audience. Earl Nightingale, a radio speaker and author, produced short common-sense exhortations on motivation and personal development with epigraphs from Wordsworth, Goethe, and Samuel Johnson, on the clear assumption that the average reader would know and care who they were. The Book-of-the-Month Club featured authors like Evelyn Waugh, George Orwell, Winston Churchill, E.B. White, and Graham Greene. In that window, it was possible to entertain the pleasing illusion that “everybody” was at least casually acquainted with the great books, or with their respectable, middlebrow cousins. Then history resumed.

What resumed, exactly? Competition. First on the demand side: TV, then the internet, then social media, then streaming; each a better-engineered rival for attention than the book ever had to face at scale. And now we’re getting the new pressure on the supply side: AI can manufacture endless “good-enough” text — serviceable summaries, plausible op-eds, competent product copy, adequate homework, infinite sludge, all at negligible marginal cost. Journalists, authors, academics, librarians, teachers, grad-school adjacent types, and hardcore readers — the very people reacting like Chicken Little — form a dense, self-referential ecology online. Unusually verbal, unusually literate, and unusually invested in the status of the written word, they quote each other, react to each other, and amplify each other’s anxieties. It can feel like “the public square,” but it’s mostly just a professional subculture dominating the stage, assigning themselves all the starring roles. The numerous readers of Stephen King, devotional literature, or celebrity tell-alls are “excluded” from the conversation, in the idiotic parlance of our time. Look, I personally know enough intelligent women who mostly read “romantasy” by the truckload, enough so that I’ve started reconsidering whether women should have ever been allowed to learn to read in the first place if this is all they’re going to do with it.

We can debate that another time. Either way, the point is, many if not most people simply don’t give a damn about literature or the self-important conceit of the writer-as-seer, communing with the soul of the world and the spirit of the age. Reading for pleasure and cultivation may well go the way of horsemanship, which was an essential skill for middle and upper classes for millennia; now reduced to a niche interest. In the world that exists, you can either lament, moan, and propose protectionist policies to keep a certain class of writer in business, or accept that “excellent prose” is an acquired taste, maybe even a minority taste, and write something worth reading for whoever still has the appetite. Not everyone will. But the existence of a minority has always been enough to keep real literature alive. At the very least, face our fate with dignity and stop whining.

The Bees Plunder the Flowers, But Afterward They Make of Them Honey

By in ai, technology, writing

Jesus, what a bunch of knock-kneed, lily-livered chickenhearts. This fast-growing genre of AI-lament has the whiny tone of Hudson in Aliens: “Game over, man, game over!” Calm your asses down. Writers have survived printing presses, typewriters, radio, TV, the internet, Twitter, and the death of attention itself. I’ve never thought of myself as an optimist, but I seem to have a little more faith in the resilience of human creativity than most. If “AI style” becomes too prevalent the way “2010’s newsletter-style” did, or even the “single-sentence/paragraph clickbait” style that Muni (unwittingly?) reproduces here, those allergic to it will find a new way to express themselves.

Yes, yes, I know — the romantic “writer-as-shaman” conceit. The artisanal mystique. The sacred cave of the soul where the author spelunks for phrases and ideas only he could have formed, unaided by anything but suffering and candlelight. I mean, sure, make it a luxury signal if you want. Put NO BOTS on the dust jacket the way rock bands used to announce NO SYNTHESIZERS in the liner notes. But synthesizers didn’t go away; they won. And AI will too, because this isn’t (X) a moral debate — it’s (Y) an incentives problem. People tend to find their way to the path of least resistance. The tool saves time, smooths friction, and does a serviceable, affordable job at multiple tasks: amanuensis, research assistant, brainstorming-partner, proofreader, editor. People will use it. If you turn that into a sin, you won’t (X) reduce usage; you’ll just (Y) make everyone lie about it. Until, inevitably, some loud anti-clanker crusader gets caught with his pants down in a hotel room with Claude, Gemini, and a bag of white powder and does the full Jimmy Swaggart tear-bath: I have sinned; I have fallen; please forgive me.

You’re not a shaman. You’re a scavenger. So am I. We’re all raccoons, opossums, and vultures with pretensions to artistry. We’ve always worked by conversation, overheard remarks, half-remembered books, accidental plagiarism, and imitation-bad-enough-to-look-original. We steal rhythm from a sentence we loved ten years ago. We borrow a metaphor from a friend and forget where it came from. We build mosaics out of whatever the world coughs up. AI is not the end of writing; it’s just another pile of material in the landscape — another junkyard where you can find a useful hinge, a bright shard of glass, a piece of wire you can turn into something that actually holds its shape. The only real question is whether you can still make it yours.

When the Punishment Begins

By in ai, moralizing, technology, the geist of the zeit, the madness of crowds

By the time the church bell quits tolling, they’re already gathering at the gaol. The rumor has traveled faster than any rider: dark doings among the scribes. Something unclean has happened in the scriptorium. A man of the guild! one of them, trained in the sacred arts of ink and ledger! — has been seen consorting with the enemy. The crowd presses closer, anxious and thrilled, crossing themselves as if the very idea might infect the air. They mutter about corrupted ink and devilish machines. And there! Pangram, Botfinder General, straight-backed, solemn, with his shiny new device that turns speculation into a number. He announces his judgment: under interrogation, the scribe’s work has been shown to bear the Devil’s watermark. That’s all they require. Relief loosens their faces. Now they can go home feeling righteous, already anticipating tomorrow’s sport: the culprit in the pillory, the mob’s clean-conscience holiday, rotten vegetables thrown in the name of virtue.

This is what it looks like in a despoiled information ecosystem when people start hunting for scapegoats. Social media already fractured the old reality-brokering order; now AI arrives and further upsets the balance of power. Video can be faked, audio can be faked, and worst of all for the scribes, text can now be convincingly mimicked too. Nobody trusts anything or anyone. It’s a Lord of the Flies media environment, and everyone’s ready to kill over who gets to wield the conch of authoritative speech. The old agenda-setters feel their entire way of life slipping away, and set about like modern Luddites in a futile effort to smash the new narrative-weaving machines. In that atmosphere, even reasonable people start craving a single line between real and artificial, human and machine.

I’m only superficially familiar with René Girard, but I couldn’t help thinking of him here. Fear wants an object. Anxiety wants a face. Suspicion wants judgment and a sentence. One one level, this episode is about the particulars of journalistic ethics. On a deeper level, this is about the anxieties many people feel as their place in the world feels threatened. We’re anxious over what it even means to be human in such a cyborg world, especially when technology can convincingly mimic even the most quintessential human activities, like art and craft. Ironically, we outsource our uncertain judgment to…more machine analysis, more software. Pangram, and other similar bot informers, will serve as the authority demarcating the real from the artificial. Most people won’t know how Pangram works, or how reliably it works, if at all, but it doesn’t matter. We just need to believe it does what we need it to, and then we feel relief. The world makes sense again, and now we can focus on punishing the scapegoats: the species-traitors who collaborated.

Immoral Voltron

By in ai, moralizing, technology

In “Mechanical Man,” Lou Barlow sang, “When I said I understood, I only knew where to stand.” He seemed to anticipate the conformist idiocy of online discourse. Not “what is this thing?” but “what vibe is this thing?” Tocqueville noted this two hundred years ago: when you remove custom and aristocracy as sources of authority, people don’t start thinking independently; they just start nervously measuring themselves against their peers. We’ve stopped forming views and started presenting ourselves in stances. Understanding the object is old-fashioned. The point is to locate it on the moral map as fast as possible, because the map is the social world and nobody wants to be caught in the wrong zip code.

Once the sorting happens, the story writes itself. A complex technology becomes a one-stop portal for every available sin: data centers in poor areas, poisoned water, higher utility bills, racism, “classicism,”  fascism, the death of art, and the reduction of your mind to sludge. It’s the inverse of Berlin’s monism: not “all goods must fit together in a harmonious whole,” but “all evils must fit together in a malevolent whole.” Snap on racism. Snap on classism. Snap on pollution. Snap on fascism. Congratulations; you’ve formed IMMORAL VOLTRON. Now you don’t have to argue about particulars; you just point at the monster and scream.

The trick, of course, is that these mad-libs incantations don’t enlighten; they discipline. Once AI is declared contaminated, any hesitation becomes complicity. To question one link in the chain is to “reveal yourself” as the kind of person who would defend the whole chain. Berlin’s emphasis on pluralism, tragic conflict, and incommensurable goods dies here, because pluralism requires the adult ability to say: some harms are real, some benefits are real, and people can disagree on how to balance them without being cast as villains. But stance culture can’t afford that. It needs a single verdict, a single villain, a single shaming label. Tradeoffs are messy; exorcisms are tidy.

There’s Terrible Trouble Behind It!

By in ai, fresh hell, technology, the geist of the zeit

Amelia Miller:

For adults, we need warnings that clearly convey the serious risks. The lessons that took tobacco regulators decades to learn should apply to artificial intimacy governance from the start. Small print disclaimers about the effects of smoking have been rightfully criticized as woefully deficient, but large graphics on cigarette packs of black lungs and dying patients hurt sales. The harms caused by A.I. companions can be equally visceral. The groundbreaking guardrails that Gov. Gavin Newsom of California signed into law last year, which require chatbots to nudge minors to take breaks during long sessions, are a step in the right direction, but a polite suggestion after three hours of A.I. conversation is not enough. Why not play video testimonials from people whose human relationships withered after years of nonstop chat with bots?

When I was a kid in ye olden days, reformers were forever trying to save us from ourselves. Smoking will kill you. Drunk driving will kill you and someone else. Drugs will melt your brain. We had assemblies with “scared straight” guest speakers, posters of tar-blackened lungs on the walls of health class, PSA’s, and (God help us) McGruff the Crime Dog records with bouncy little earworms about every substance known to man. The theory was that if we were properly educated, if the warning labels were vivid enough, if the videos were grisly enough, we’d make good choices forever after. And yet people kept smoking. People kept driving drunk. People kept getting high. Not because they lacked information, but because they were stressed, bored, lonely, impulsive, self-destructive, curious, or just plain human.

Now the new panic is emotional intimacy with AI, and the new solution is…warning labels and more regulation. “This is not a substitute for genuine relationships.” “ChatGPT can make mistakes.” And here some newly-minted reformer wants to make teenagers watch video testimonials from recovering chatbot junkies, like driver’s ed films of crumpled cars. But McGruff is the perfect emblem of the whole genre: you’re trying to solve a messy relational problem with a cheesy piece of mediated messaging. You even risk the pink-elephant effect — by foregrounding the forbidden thing (“crack cocaine!” “romantic bots!”) you plant it in the mind with an oddly vivid presence, while the adults congratulate themselves for having “done something.” It’s outsourcing the hard work of parenting and community to a cartoon mascot and a compliance checklist.

None of this is to say that nothing can or should ever be improved. But there is no design standard, warning label, or stern authority that can make life safe and painless. People will always try to jerry-rig comfort out of whatever’s available when their actual lives are starving for it. If you don’t want Johnny turning into a dope fiend, you can’t fix that with a McGruff album. If you don’t want Johnny “marrying” a chatbot, you can’t fix that with a disclaimer. The solution, such as it is, was never primarily informational. It was always relational. And wherever we go, there we are.

I do look forward to today’s cartoon characters producing an album warning kids to not fuck an android sexbot.

Relligiosi

By in ai, conspicuous crusading, moralizing, technology, the geist of the zeit, the madness of crowds, writing

Relligiosi

That’s what I call those who take it upon themselves to convince others and make them similar to themselves, make them think in the same way.
I abhor this kind of mental usurping, the great task of political parties and confessions.
What matters for you, doesn’t matter for me. Because it matters for you, it doesn’t matter for me.
To understand oneself, it’s essential (in my eyes) to think differently.

— Paul Valéry, Cahiers / Notebooks 1

Via Paul Kingsnorth. And what does this pledge entail, you may ask?

Do you support the campaign? Go rape yourself, that’s my response. Even answering that question concedes too much to the commissar asking it. It’s none of your goddamned business how I feel about AI use. I don’t owe you a pledge, a badge, a logo in my sidebar, or an affidavit about my writing process. If I read something and it feels synthetic, I’m free to close the tab and avoid that writer in the future. That’s how grown-ups handle taste, without forming a militia and a blacklist.

What I can’t stand, as always, is the missionary spirit animating this entire self-appointed class of jumped-up hall monitors: the righteous salesmanship, the demand for public alignment, the insinuation that silence is complicity. This “download a free logo and join us” routine is a ritual with no barrier to entry for people who like the feeling of conscience without the burden of thought. Once the ritual exists, the next stage is guaranteed: bot-sniffing as sport, comment-section inquisitions, and purity audits, as though the only point of reading is to detect the Devil’s watermark.

This is the online HOA, or the Concerned Citizens Council, coming to the door with clipboards and smiles, asking you to sign. And underneath the smile is the implied threat: Do you want us to make it known that you refused? Because that’s how these things work. There is no neutral ground in a holy war. The whole point of the ritual is to sort the righteous from the wicked, and to make “I decline” sound like a confession. Even if you happen to agree with this particular campaign, the same petty functionaries will be back knocking on your door next week, with clipboards and smiles, expecting your participation in the next loyalty test. Whenever you finally decline, the smiles will vanish, and you’ll see them for what they really are.

The most a campaign like this can do is create another platform for virtue-signaling (as if we needed more) and another excuse for resentful mediocrities to claw at higher-status targets under a banner of public morality. Watch how quickly “protecting the craft” turns into guilty-until-proven-innocent accusations. It invites the exact type Valéry despised: the mental usurper, the person who can’t just choose their own life, but must conscript yours into their crusade. No thanks. Read what you want. Write what you want, how you want. And leave me out of your religion.

‹ 1 2 3 4›»