It's easy to get depressed about the state of serious book coverage in the US these days, but as long as Parul Sehgal is writing regularly, all is not lost: www.nytimes.com/2026/04/03/m...

Patrick Radden Keefe is one of the few truly greats in our industry:

I think it's all to the good that people are working out their own ways to use (or not) this new technology. But I will say that one of the three people whose experiments with LLMs convinced me to give them another try last fall is one of the best reporters and writers on staff at TNY.

Yeah, sorry, no: if you can’t come up with 1000 interesting words about a book you’ve been assigned to review, you should probably not be in the business of book reviewing. At the very least you should refuse the assignment. open.substack.com/pub/samleith...

A lot of AI panic, understandably, is about what *other* people are doing with the tech: how it’s being misused/abused by students, companies, scammers, etc. A smaller but growing concern (e.g. in E. Klein’s and M. O’Rourke’s recent NYT pieces) is about how LLMs affect their primary users.…

Melville, getting—early, in “Notes for January”—into my year-long sequence titled “Notes for General Circulation”: ‘God keep me from every completing anything.’

One of the more durable fault lines on the left (broadly conceived) is between people who think the goal of politics is to secure power for them and theirs, and those who think the goal of politics is to find some near approximation of equality for everyone. It’s often hard to spot the fissure…

PSA: if you have MUBI, don‘t miss the opportunity to see this.

try and imagine a bigger W as an academic

Favorite part of No Kings in Manhattan yesterday was watching a 70ish yo woman impatiently but charmingly take a megaphone from a teenager to demonstrate the right rhythm for the “shut it down” chant. (She then, of course, gave it back and let the teenager take over.)

William Butler Yeats gazing in the direction of Bethlehem:

Happy hundredth to Frank O'Hara, but also happy 100 to Gene Shalit, who is still alive and who, I learned from some brain-breaking feed synchronicity, was born two days before FO'H.

"another golden stairs" really could've been a classic dead song. easy to imagine these words in deadhead circulation, on shirts & stickers, referenced as wisdom, unpacked at the dead scholars caucus, etc.

Oh no, RIP Tracy Kidder. 20 years ago I lucked into one of the more remarkable events of my life, a party with him, Paul Farmer, and several other astonishingly talented human beings. Tracy was of course a brilliant writer; he also seemed to me a total mensch. Here‘s a photo I took of him and Paul.

We therefore learn from this learned commentary made by a Babylonian scribe over 2000 years ago, that in the Akkadian language, the word shitti means "shit"

A simple test for whether you should put an LLM in charge of a process, whether that process is organizing photos or picking targets for a bombing run: When things go wrong (eventually they will go wrong) will a computer saying “You’re right. I shouldn’t have done that“ seem a satisfactory response?

Strange to think that a century from now people are going to collect these as curiosities. They’ll marvel at the feckless absurdity of a populace that brought this man to power not once but twice, and they’ll be right, and then in their next breath they’ll tell themselves “We would never…”

I’ve said it before, I’m sure I’ll say it again: there just might be a lesson here for anyone who wants it. I’m convinced anyone wants it.

Not to make light of a bad situation, but any copy editor past or present will recognize this caption's headline as its own kind of tortu(r)ous ga(u)ntlet.

"In the midst of what is going on now, it is hard, when you sit down at a desk, to feel confident that morning after morning spent fiddling with words and rhythms is a justified activity – especially as there is never any certainty that the whole thing won’t have to be scrapped." — T.S. Eliot, 1942

I genuinely love this recipe but its first four words are such a classic, and classically insane, NYT Cooking misdirection.

After watching lawyers—people who work in a field where ~90% of prose is written by someone other than the person credited for it—get het up about a Claude-assisted article, I’m once again insisting that anti-AI arguments are not going to be helped by naive conceptions of authorship and originality.

OH. MY. GAWD.

Calvin Tomkins died on Friday at the age of 100. For nearly 70 years, he filled the pages of The New Yorker with portraits of modern artists, from Marcel Duchamp to, just recently, Tala Madani and Rashid Johnson. Read David Remnick’s remembrance. newyorkermag.visitlink.me/T5eFd6

I think Trump is genuinely scared of nuclear weapons, just like he genuinely hates windmills, but every once in a while I’m reminded that all it would take to start a nuclear war is a rerun of the same neurochemical cascade that prompted him to run for president after Obama mocked him at the WHCA.

Anyway I mentioned it in a reply to that last thread but one thing that’s been striking is how anti-AI negative polarization has led people to loud assertions about authorship and originality that would have been laughed out of a Viennese parlor room for their naïveté a hundred years ago.

It‘s *very* easy to see how LLM generated art might be the, ahem, natural terminus for certain strains of late-20th c. avant-garde poetry. I’m also surprised that people haven’t seen LLM output in the context of "il n'y a pas de hors-texte" style Theory. LLMs are semiotic wildernesses/wonderlands.

One fact that goes unmentioned in this piece: publishers, especially corporate publishers, generally find ways to justify selling books that readers want to buy. Is there anything in the history of publishing that gives us reason to think this time will be different? www.nytimes.com/2026/03/19/b...

meta being called meta is so incredible. it’d be like if microsoft was called zune or nasa was called challenger

I’m sure I could find some Cheever story that evokes this feeling, but the book that comes most immediately to mind reading this sentence is Salter’s LIGHT YEARS. A favorite of mine, I hasten to add, even if I sometimes think of it as a gorgeous airless coffin.

Trump has been, I have no hesitation saying, a greater catastrophe, but the Iraq War remains the defining political event of my life, in the sense that it shaped my politics more than anything else. Still vividly remember being at Jimmy’s in Hyde Park when Bush announced the first attacks on TV.

In case you made the mistake, as I did, of reading Ross Douthat’s latest exercise in racist propaganda—oh, I know, he’s just asking questions!—here are two pieces I‘ve written that might help you get your day back on-kilter: www.theguardian.com/news/2021/ap... www.nybooks.com/articles/202...

Horrible and heartbreaking in several dimensions.

Just published from Columbia UP! Congrats @johannawinant.bsky.social !! cup.columbia.edu/book/lyric-l...

I unironically think that the latter stuff is what’s going to make AI revolutionary in a minor key. (If the overhype doesn’t kill us or bankrupt the economy first, a big if.) LLMs are extremely effective translators between human language and computer language.

I know this is naive (’twas ever thus) and that I should be thankful for any counterpressure we can get to this abominable war of aggression, but it really rattles something deep that increasing gas prices are likely to have more influence on the course and conduct of the war than dead schoolgirls.

Whether LLMs are conscious ought to be a ridiculous Q to anyone who's spent more than a few minutes learning how they work, but/and it sure doesn't help matters that Anthropic baked this nonsense into Claude's training. On this front at least, ChatGPT and Gemini's canned responses are much better.

we are at the stake, And bay’d about with many enemies, And some that smile have in their hearts, I fear, Millions of mischiefs.

This isn’t hard. It’s bad to make guns easily available. It’s bad to put AI in charge of targeting decisions. Guns and AI can both kill but they can’t murder—they can’t bear moral responsibility—because they’re things, not people.

The emerging “Claude killed the kids” frame on this story shows just how easily anti-AI counterhype can end up doing the government’s work for it. If you use a computer to choose targets—esp. but not only if you give it bad data to work with—and it chooses poorly, that‘s on you, not the computer!

Looking for a #longread for your Saturday afternoon? Check out this story, in which I answer all your burning questions about how you can monkeywrench carbon capitalism just by living in a treehouse. (Including the most common q of all: how do treesitters go to the bathroom up there?)

Not today, Satan.

J. H. Prynne, on reading

Solidarity today with everyone who has an eighth-grader headed to a public high school in NYC… 😬

A real sign of basic LLM illiteracy that people seem to think a Claude instance would have any privileged insight into *how* it "thinks," let alone how any other Claude instance might have been used. Instances "know" what they've been trained on, and what data they've been fed. That's it.

Heaven have mercy on us all—Presbyterians and Pagans alike—for we are all somehow dreadfully cracked about the head, and sadly need mending

The use of AI to identify targets (without subsequent specific human confirmation) should be considered a war crime. And by that I don't mean "we should make a rule" I mean that straightforward application of the existing rules would tell you that it is a war crime.

page one of Peter O’Leary’s <Phosphorescence of Thought>

I'm long separated from my limited experience here, but shooting down 3 friendly F-15s sounds like more than a deconfliction/comms problem, it sounds like an autonomous air-defense systems problem. Good thing we're not, say, recklessly rushing to replace more humans with AI in lethal decision loops

ubi solitudinem faciunt, pacem appellant.