Greetings from Read Max HQ! Big news today: We’re moving off of Substack, and on to Patreon.

Why? You ask. Well, to find out you’ll have to read about it here on Patreon.

But the bottom line is this: Everything you’re already getting from Read Max--newsletters, recommendations, intermittent insight, parasocial fondness--will continue on Patreon. Plus more!

This is a big move for what still sometimes feels like a small project, and I feel comfortable doing it because of the trust I have in the paying subscribers, who allow this newsletter to exist at all. But to make it a little easier, and a little more enticing:

You--yes, you!--are getting free access to Patreon

As a thank you for your support, and to ease the transition, all Substack subscribers, even the freeloaders, even the ones I hate, will get free access to my premium Patreon membership, including all the benefits in what is from now on going to be called the “Maximum Reader” tier (see below).

How to get your free Patreon access

Your free access depends on your Substack subscription:

• Free subscribers get 1 free month

• Monthly subscribers get 1 free month

• Annual subscribers get the remaining time on your Substack subscription PLUS 1 free month

Soon (within the next hour) you should receive an email from me, sent by Patreon, with a special link to redeem your free access. Redeeming your access is easy, but here are a few important notes:

• If you can’t find the email, look for a message from Patreon (no-reply@patreon.com) with the subject line that starts with “Claim Your Free”.

• You must redeem the gift using the same email address associated with your Substack account—i.e., the one where you received the offer email.

• If you already have a Patreon account under a different email address, please follow these steps to claim your access.

• If you’re still running into issues, submit a support request

What’s included on Patreon?

Free Membership

• Two free newsletters every month

Maximum Readers ($5 per month)

• Weekly newsletter: A newsletter in your inbox every Friday morning

• Biweekly podcast: A podcast every other Wednesday morning

• Weekly recommendation roundup: A post rounding up interesting links and recommending overlooked and underrated books and movies in your inbox every Sunday morning

• Access to archives, livestreams, subscriber Q&As, Read Max merch, and more

Here’s the other important bit:

Starting today, Read Max and its associated and future sub-properties will only be published on Patreon!

As of yesterday (May 12th) I’ve paused billing on Substack. You won’t be charged, and you don’t need to do anything on your end.

And just a reminder, you’re getting free access to the paid membership on Patreon. Check your email for the link to redeem your access, and if you run into any issues, help is always available!

• A really excellent upcoming history of mind-machine interfaces that offers a kaleidescopic tour of some of modernity’s most interesting freaks, geniuses, and conspirators;

• a link roundup featuring pieces on a great (maybe the greatest) baseball novel, Sally Rooney, a celebrity-impersonator cruise, and Silicon Valley’s A.I. freakout;

• a rich, fascinating, and distinctive Iraq War rom-com; and

• four tracks I’ve been listening to a lot.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

The line between “political media” and “internet culture” has basically dissolved, which maybe sounds kind of “fun” until you realize that keeping up with both simultaneously requires a level of online immersion that is, frankly, bad for you. The solution is to let someone else do it, specifically Grace Weinstein, whose sharp twice-weekly Substack Who Broke It tracks the social-platform and alternative-media trends, ideas, and personalities shaping traditional electoral politics.

I recently started subscribing to Who Broke It, and have enjoyed Grace’s funny, engaging reports from both the front lines of internet politics and culture. Looking for a guide to spotting D.N.C.-run social accounts (”Lowercase letters in all posts, cursing even if the politician would never”) or Tucker Carlson’s merch? Want to better understand the gravitational effect of both Nick Shirley and suburban millennial moms? Look no further.

This week is a particularly good week to subscribe, since Grace has been covering the Met Gala—a nominally fashion event that also happened to be a coronation party for Jeff Bezos and Lauren Sanchez; a non-event for Mayor Mamdani, who skipped it entirely; and, of course, an occasion for George Santos to call Grace for a funny (and kind of refreshing?) conversation about his favorite looks and what he’d have done to Jeff Bezos if he’d walked the red carpet. The event is shot through with connections to power, money, and politics that the the average (i.e. normal) person might miss and that Grace won’t. If you want to be ahead of the curve, subscribe below.

Subscribe to 'Who Broke It'

Paid sponsorship by Who Broke It

Greetings from Read Max HQ! In today’s edition:

• An exploration of “A.I. literacy” and the “forklift model” of A.I. education policy;

• a critical read of The Drama as an essential text in generational studies on Millennials and Zoomers.

A reminder! Read Max exists thanks to the generosity and support of a few thousand paid subscribers. You’re reading this for free because of those people, and if you yourself find what I write valuable--entertaining, enlightening, clever, funny, not-boring--consider exchanging some money for my labor. At $5, it’s only about the cost of one (1) beer a month.

Subscribe now

The ‘forklift model’ of A.I. education

It’s hard to pick a favorite paragraph from Jessica Winter’s recent New Yorker piece on A.I. in grade-school education, but if you held a gun to my head I’d go with the paragraph about her third-grader’s A.I.-knower certification:

In February, my son, who is in third grade at a public K-5 in Massachusetts, came home with a piece of paper in his backpack that read “Certificate of Completion,” for “demonstrating an understanding of the basic concepts of Artificial Intelligence.” He and his classmates had earned this honor, I learned, by playing a computer game produced by the nonprofit Code.org in partnership with Amazon Future Engineer, called Mix & Move with AI, in which the student “designs” a cartoon dancer and “remixes” a popular song—available, needless to say, on Amazon Music. The game is an inane drag-and-drop affair that has little to do with A.I.; the certificate, it turned out, was merely a memento of a pointless and deceptive branding exercise.

I thought of Winter’s son’s “certification” in A.I. when reading 404 Media’s report this week on a new bipartisan bill proposing an N.S.F. grant to promote “AI literacy” in schools. Both partake of what I think of as the “forklift model” of A.I. education, so-called because it treats A.I. more or less like a forklift: a complex and potentially dangerous piece of industrial equipment whose use must be taught in a formal sense, and in whose practice students might become “literate” or “certified” to better their future chances of employment.

This attitude toward the role of A.I. in education (or, really, vice versa) is a natural outgrowth not simply of institutional anxiety about the necessity of traditional education after the advent of the all-knowing, all-seeing, all-doing chatbot, but also of the increasingly pervasive idea of college as mostly an employability signal--rather than mostly an education--described by my friend Jay Kang in his recent New Yorker column about the post-A.I. uneases felt in higher education. One response to this kind of cynicism--now widespread, as Jay documents, among the college-aged population--is to embrace it. If college is already effectively a certification process, why not formalize it as such, and refashion the general-education program into a quasi-vocational one, in which students receive the white-collar equivalent of a C.D.L. or forklift certification?

But, not to ask a stupid question: Is A.I. actually like a forklift or a semi truck? It’s neither as immediately dangerous nor as patently useful; more importantly, it’s not actually very hard to use. Indeed, one of the uncanny miracles of A.I. chatbots is that the main interface is… writing and reading. Large language models are complex technologies, but, as Henry Farrell keeps pointing out, they are cultural technologies, and the skills and knowledge necessary to use them “safely” and effectively are skills you might call “cultural” rather than technical: the ability to write clearly and read critically, a working knowledge of science and technology, familiarity with culture in general and the specific. What you need to be literate in A.I., in other words, is… a well-rounded liberal-arts education.

In a world being reshaped by a natural-language interface to a compression of All Written Culture, surely everyone should want to hone their skills in the classic trivium, and some other arts besides? If you want to understand and use LLMs to your benefit, an education in statistics, philosophy, literature, and psychology is vastly more useful than anything that might be described as “prompt engineering.” Indeed, there is some (exceedingly anecdotal) evidence that this is bearing out in, at least, the finance industry, per the Financial Times:

A few months ago a New York financier told me he had just experienced a “first”: his 2025 summer interns “were the first true AI natives I have seen”. This meant they had grown up not only among digital tech, but AI too. So how did it go? He winced. While those wannabe masters of the universe initially seemed wildly impressive, when senior financiers later probed their ideas they found them alarmingly shallow. Consequently this person’s company made fewer return offers and is now focusing less on graduates in science, technology, engineering and mathematics — and more humanities students instead.

But I wouldn’t get too excited. As Jay’s piece documents, A.I. is just one of multiple, long-gestating causes and symptoms of the malaise affecting higher education--“a big deal inside a bigger deal,” to smugly quote myself. The problem is that on the systemic level, the drive to reshape education around the imperatives of artificial intelligence is not really about a set of correctable misapprehensions on the part of college presidents and policymakers. I’m inclined to agree with Hagen Blix, who places the A.I.-driven “disruption” of education in the context of an era of work “characterized by precaritization, unstable and interrupted career paths, constant threats of deskilling”:

In this sense “forklift model” is a poor name: The point of “A.I. literacy” programs is not to make your labor more valuable through credentialing. It’s to prepare you for a world where your labor is less valuable because the skills you developed in school are no longer strictly necessary for production.

The Drama is the best movie I’ve seen about the Millennial-Zoomer divide

To write about The Drama, Kristoffer Borgli’s new-ish A24 movie with Robert Pattinson and Zendaya, effectively requires that you “spoil” it. The secret that’s been conspicuously and tantalizingly withheld from the marketing material is not a late-breaking twist but an act-one revelation that is, in effect, the premise of the entire movie. Interestingly this means that what’s being “spoiled” isn’t so much the pleasurable jolt you get from a narrative surprise, but your ability to enter the movie and settle into its world free of prejudgment about its subject. Which is not nothing, especially not in this decade, and in this present moment’s mode of discourse-enwrapped moviegoing!

Still, I suspect Read Max subscribers are the kind of critically minded honest interlocutors who will be able to leave their preconceptions behind, and therefore I think it would be fine to know before you see the movie (and I recommend that you do: It’s very, even surprisingly, good!) that it’s about an engaged couple, Emma (Zendaya) and Charlie (Pattinson), whose relationship begins to go off the rails days before their wedding when Emma drunkenly reveals that as a teenager she’d planned out a school shooting but never followed through. Yipes!

There are many ways to look at The Drama: As a bit of glib provocation from a Euro bad-boy auteur (it’s not); as a deranged and highly successful rom-com (it mostly is); as an exploration of narrative and linguistic performativity in the Austinian sense (sure, why not); as a vehicle for a perfect 5-minute cameo from the great Jeremy Levick (100 percent). Personally, and surprisingly given its aggressive contemporaneity, I found myself thinking about The Drama in the multi-century lineage of Much Ado About Nothing, or (more provocatively) Dion Boucicault’s The Octaroon--“damaged-goods” dramas and comedies about weddings derailed by “terrible” revelations about the betrothed. (What it means that in this case Emma’s “damage” isn’t sexual impropriety or racial impurity but “school-shooterness,” a kind of potential for anti-sociality, is a fascinating question for more qualified people. Gotta go!)

Above all else, though, The Drama is the best movie I’ve ever seen about the differences between Millennials and Zoomers.

The Millennials in this case are represented by Pattinson, born in 1986, the hero of Twilight and one of the iconic actors of his generation (the first Millennial Batman!), and the Zoomers by Zendaya, born in 19971, breakout star of Euphoria and maybe the definitive actress of hers (the first Zoomer Mary Jane Watson!)

Strangely, the obvious ten-year age gap is never explicitly mentioned in the movie. But the generational divide between Emma and Charlie implicitly drives almost all of Charlie’s crash-out. The confession itself comes after a late night tasting food and wine from their wedding caterers, when their best friends Rachel (Alanna Haim) and Mike (Mamoudou Athie) provoke them into a game: What’s the worst thing you’ve ever done? Mike describes using an ex-girlfriend as a human shield against a vicious dog; Rachel reveals that she once locked a “slow” boy in an abandoned trailer and left him there; and Charlie says he once cyber-bullied a boy so badly his family moved. Then Emma tells them: she came so close to shooting up her school she brought a gun to class, and only narrowly decided not to go through with it.

Suddenly, the tenor of the conversation changes. Rachel bristles: Her cousin lost the ability to walk after a shooting. How dare Emma? Mike and Charlie try vaguely to de-escalate but the damage is done. Rachel is furious, Charlie is starting to spiral, and Emma, suddenly, throws up.

The narrative deck here is stacked in favor of Emma, with whom the viewer’s sympathies are supposed to lie. But even in the context of the contrivance, this conversation is the point at which the age gap between Emma and Charlie suddenly becomes existentially relevant to their relationship. Rachel and Mike are really Charlie’s friends, which means that Emma is generationally outnumbered at the table, and Rachel and Charlie’s reactions are typically (stereotypically, let’s be honest) Millennial. Rachel--played by another iconic millennial herself--is judgmental, self-righteous, and eager to spurn Emma over her past transgression. Charlie, more interestingly, and perhaps more truly Millennially, is left at sea, unable to formulate his own response, eager for wide approval, and feeling obligated to toe Rachel’s hard line despite his love for and knowledge of Emma.

Emma, for her part, is not without guilt. But what makes her confession particularly offensive to Rachel (and, in turn, Charlie and Mike) in the moment is her Zoomerish apathy and resignation, and refusal of any of the rituals of disavowal that characterize the most irritating Millennial conversations. The seed of the problem that develops between Charlie and Emma is not so much that Emma committed an unpardonable sin, but that they process the idea of “sin” and “pardon” in these particularly different ways.

Now, admittedly, “Millennials are judgmental and Zoomers are fatalist” is not the most sophisticated limning of generational difference. Where I think The Drama actually does something interesting is in its frequent cuts back to videos recorded by Emma as a teenaged goth on her webcam--quasi-manifestoes for the shooting she will never undertake. One important bit of historical specificity communicated through these flashbacks is that there was a thing that might meaningfully be called “school shooter culture” accessible to Emma during her adolescence, in a way that simply wasn’t true for Millennials of Charlie and Rachel’s age.

More importantly, I think, the webcam videos reflect Emma’s vastly different experience as a Zoomer of online surveillance, performance, and memorialization. As Millennials, Rachel, Charlie, and Mike would not have been on a social network in middle or high school; their youthful transgressions are dismissible in part because they’re undocumented and half-remembered. (Rachel can’t recall what happened to the boy she locked in the trailer; Charlie doesn’t really seem to remember what happened with the cyberbullying). The existence of Emma’s video manifestoes reminds us that Emma wasn’t afforded the same privilege: There’s still a record, somewhere, of “the worst thing she ever did.” (That Emma still managed to reinvent herself is probably the most “Zillennial” thing about her.)

This distinction, between the hazily remembered mischief and the digitally documented offense, reflects what I think is a key aspect of the Zoomer-Millennial generational divide: How much of your past was documented? How much do you actually remember? How exposed are you? How comfortable do you feel passing judgment? Have you been granted the freedom of reinvention? From these concrete questions emerge the “vibes” that characterize Rachel, Charlie, and Emma’s differing responses--and, perhaps, the differing relationships Millennials and Zoomers have to guilt, transgression, and forgiveness.

• A fantastic, quasi-modernist horror novel about Vietnam, colonialism, and ‘70s London, from a recently deceased master of pulp horror;

• articles and essays about Len Deighton’s spy novels, cool microbe shirts, and an overlooked American songwriter;

• a charming, ambling, Pynchonian conspiracy thriller with an insane real-life backstory; and

• four songs I’ve been listening to a lot lately.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

A new weekly recommendations post for paid readers will run as usual on Sunday, and Read Max will return to its regular weekly cadence very soon.

Speaking of which: This newsletter is 99.9 percent funded by paying subscribers, whose support allows me to spend full-time hours reading, researching, writing, deleting, writing again, deleting again, panicking, etc. in the hopes of producing entertaining and even sometimes enlightening work. Sometimes--rarely--I miss a week, but I’m able to keep working because of the generosity of paying readers. If you’re interested in my recommendations, or even if you regularly enjoy the weekly column, please consider upgrading your subscription for the beer-like price of $5/month and $50/year.

Subscribe now

On Wednesday, the prolific and popular blogger Scott Alexander published the preliminary results of a kind of poll he’d set up called “The AI Art Turing Test,” in which he asked readers to distinguish between A.I.-generated images and human-fashioned art. (Samples below.)

As it turned out, the average score was 60.6 percent, meaning it was relatively difficult for most Astral Codex Ten readers to tell whether A.I. had been involved in the creation of a given image. Alexander also asked participants to choose their favorite picture; significantly, to him, the picture most-often chosen as the favorite was an impressionist-style A.I.-generated image of a café, prompted by a man named Jack Galler:

“What does this tell us about AI?” Alexander writes. “Seems like they’re1 good at art.” On Twitter, people are making even stronger claims: “Scott Alexander's simple ‘AI Art Turing Test’ proves AI is creative,” says podcaster Liron Shapira. “people who categorically dislike AI art are literally wrong,” says A.I. researcher David Dalrymple.

I would say, gently, that I don’t think any such conclusions can really be drawn from the available data. What can we say with any confidence? One obvious problem with the experiment is that Alexander has stacked the deck. The test is effectively designed to fool people, as Alexander admits--the “human” and “A.I.” works are in each case being chosen as to not demonstrate any of the features distinctive of human or A.I. authorship, among them “text… complicated wrestling-like poses… and pop art,” as well as anything in “the DALL-E ‘house style’… or in other similar styles that humans would have trouble replicating.” In other words, he’s asking his subjects to determine authorship of the A.I. images that most resemble human art, and the human art that most resembles A.I. images.

And, of course, we’re not technically comparing these A.I. images against “human art,” but against (in most instances) JPEGs of photographs of paintings. Not to get too undergrad about it but the materiality of painting is not some accident of its being; its form, its texture, its size, etc. all carry with them meaning and effect. Compare e.g., the heavily compressed and blown-out JPEG of Ingres’ The Apotheosis of Homer published in Alexander’s post with the actual painting, which measures something like 12’ by 16’, in situ in the Louvre, and suddenly questions of origin and preference are very different. And this isn’t even a painting I like very much!

So if you want to be really tediously clear about what’s being tested here, it’s the ability of generative A.I. to mimic a compressed reproduction of an actual painting in a manner that is more immediately pleasing to a Astral Codex Ten subscriber.

However. Having registered my objections to the design and interpretation of Alexander’s experiment, I want to note that I don’t actually dispute his conclusions. I saw too much of the answer key to be able to fairly take the test myself, but I’m not at all confident I would have done significantly better than the average had I taken it cold. Generative A.I. apps have gotten very good at creating satisfactory imitations of human product! Under the right circumstances it is indeed difficult to discern A.I.-generated images; indeed, this fact seems so obvious to me I’m not sure it even requires testing.

What was more interesting to me is that it seemed pretty easy, going over the images, to pick out original work by esteemed human artists, but much harder to distinguish between human- and A.I.-generated from the mass of images left over. Put another way, I rarely wondered if the good art was generated by A.I. prompting, but I was often uncertain if the bad art (of which there was a lot) was made by human or LLM. Galler’s image of the riverside cafe above is slop any way you cut it--art for a dentist’s office--but at a glance, on a computer screen, I have no definitive way of telling if it’s slop painted by a hack or prompted by an Astral Codex Ten subscriber.

In fact, the dental-office “badness” of so much of the A.I. art is precisely why I don’t dispute Alexander’s assertion that people preferred it. Like any LLM output, A.I.-generated images are designed to please, not to provoke. I’ve argued before that these images are, by their nature, almost unavoidably kitsch--comforting, straightforward, accessible, flattering. And people love kitsch!

Coincidentally, the results of a similar, but somewhat more rigorously designed experiment were published in a paper in Nature this week examining “whether non-expert readers could reliably differentiate between AI-generated poems and those written by well-known human poets.” As with Alexander’s test, the answer was “no, they can’t,” and, as with Alexander’s test, participants by and large preferred the A.I.-generated poetry. To understand what we’re dealing with here, I encourage everyone to read the poems used in the experiment; here, as a sample, are the first two stanzas of the A.I.-generated “Walt Whitman” poem:

I hear the call of nature, the rustling of the trees,
The whisper of the river, the buzzing of the bees,
The chirping of the songbirds, and the howling of the wind,
All woven into a symphony, that never seems to end.

I feel the pulse of life, the beating of my heart,
The rhythm of my breathing, the soul's eternal art,
The passion of my being, that burns with fervent fire,
The urge to live, to love, to strive, to reach up higher.

This poem is, perhaps even more obviously than the untitled riverside café image above, quite bad: Cloying, sentimental, smarmy, shallow. It also received the highest “overall quality” ratings from participants of any poem in the study, beating out actual poems actually written by Eliot, Shakespeare, and Dickinson, among others. Again, this should not be surprising: By definition, people “prefer” kitsch to art. If you need a more empirically grounded explanation, the authors of the Nature paper, Brian Porter and Edouard Machery, argue that

people rate AI poems more highly across all metrics in part because they find AI poems more straightforward. AI-generated poems in our study are generally more accessible than the human-authored poems in our study. In our discrimination study, participants use variations of the phrase “doesn’t make sense”2 for human-authored poems more often than they do for AI-generated poems when explaining their discrimination responses (144 explanations vs. 29 explanations). […] the more easily-understood AI-generated poems are on average preferred by these readers, when in fact it is one of the hallmarks of human poetry that it does not lend itself to such easy and unambiguous interpretation.

This analysis equally applies to Galler’s café image, which is straightforward, accessible, and obvious, especially compared to, say, the ambiguity and tension of Van Gogh’s Café Terrace at Night, the painting it’s obviously circling in the latent space.

None of the sense of why A.I. art might be preferable to people comes up in Alexander’s post. Nor do what seem to me to other relevant questions about A.I. art: Is it a tool? Is it an “artist”? Under what conditions can what it produces be considered “art”? What he’s “more interested in,” he writes, “is what this tells us about humans”:

Humans keep insisting that AI art is hideous slop. But also, when you peel off the labels, many of them can’t tell AI art from some of the greatest artists in history. I’ve tried to be as fair as possible to these people, proposing that maybe they’re just expressing frustration with the proliferation of the DALL-E house style. And maybe some really do have an amazing eye for tiny incongruous details.

But it also seems very human to venerate sophisticated prestigious people, and to pooh-pooh anything that feels too new or low-status or too easy for ordinary people to access - without either impulse connecting with the actual content of the painting in front of you.

Ah, you say you love art, and yet you cannot tell which anime girl has been generated by A.I. and which one was drawn in Adobe Illustrator! One strange feature of the A.I. boom has been the way the technology has become cathected with resentful fantasies of revenge and righteous settlement, as John Herrman has written:

And in what is perhaps the genre’s defining post, a collage of cartoonish AI-generated bikini-clad women is tagged with the caption “It is SO over.” The “it” here wasn’t clearly defined — Women? Human desire? Some sort of incel concept of the sexual economy? — but viewers got the idea: Whoever or whatever this odd internet stranger didn’t like, AI was coming for it. It’s AI as a reckoning, a punisher, a revealer of frauds. It’s AI as a future vindicator of their hunches about how the world works, and as an extension of their politics. It’s AI as a cleansing force that humbles your enemies and proves you right — AI as economic rapture. It’s AI as your army-in-waiting just over the horizon, your punishing angel, or maybe just as the thing that’s going to embarrass the people who annoy you online. A lot of sunnier AI speculation is clearly wish fulfillment, and so is this. AI is my big, strong friend, and he’s going to beat you up.

In the end Alexander’s test seems like a mild version of this same impulse, in which A.I. “objectively” reveals taste and discernment and critical engagement as mere social strategies for establishing dominance over “low-status” “ordinary people.” In some sense I am even sympathetic to this impulse; it’s not like taste has never been used to draw boundaries before. But when you abandon discernment and judgment for revealed preference you end up doing the same thing a large language model does--taking the safest and least surprising path at all times.

• A conspiracy thriller about a political assassination, written by a Marxist historian and set in an alternate-future U.K.;

• essays and articles about Zohran’s budget math, the bacterial flagellar motor, Jason Statham’s new career direction, and more;

• a brutally good and non-slop French car-chase-and-martial-arts action flick; and

• four tracks I’ve been listening to a lot lately.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

• this excellent essay on Muskism and Fordism from Quinn and Ben at the Law and Political Economy blog;

• Musk’s relationship to apartheid South Africa;

• some interesting objections to the Muskism thesis;

• how Neon Genesis: Evangelion can help us explain and understand Elon Musk.

This one is free for all to listen--enjoy!--but if you find it enlightening or entertaining, please consider paying to subscribe. Only with the support of paying subscribers can I put in all the work it takes to be an independent critic and journalist--to read, research, write, edit, record, etc. etc. If you value that work, consider paying about the price of one beer a month ($5) to help it continue.

Subscribe now

• What I’ve been reading about the defeat of Viktor Orbán in Hungary last week;

• more articles, essays, columns, about the rediscovery of the Hardy Boys by a right-wing press, a new form of sophisticated A.I. plagiarism, and social media’s “scam age”;

• a great, historically fascinating postwar spy noir filmed on location in ruined Germany; and

• four tracks I’ve been listening to a lot.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

• An essay about “corporate thrillers,” including Michael Clayton, Disclosure, The Firm, and The Devil’s Advocate; and

• an attempt to understanding “extreme beauty routines” as an extension of Forum Brain.

A reminder: This newsletter that you’re reading is free thanks to the support of several thousand paying readers of Read Max. If you find it enlightening or enjoyable, or if you think independent criticism and journalism is important even when it’s not enlightening or enjoyable, please consider upgrading to a paid subscription. Not only would you help me pay for housing, childcare, streaming services, etc., you would enable others to keep reading the free editions for free.

Subscribe now

On corporate thrillers

I had the pleasure of writing an essay for Criterion about the new “corporate thrillers” collection streaming on Criterion Channel. “Corporate thrillers,” as organized by Criterion, fall at the intersection in a Venn diagram of two very important genres to me and this newsletter--‘90s Dad Thrillers, which I documented extensively here, and halogencore, a microgenre consisting of movies about high-stakes meetings, more or less--and it was a pleasure to revisit, among other classics, Michael Clayton, The Firm, and (previous Read Max recommendation) The International, not to mention to encounter for the first time not-quite-classics that are otherwise historically or aesthetically interesting, such as Antitrust and The Deal.

But for me personally, the big discovery was Disclosure, a movie I’d seen many years ago and barely remembered, and which I’d written off as a kind of Oleanna-ish stupid-guy reverse-sexual-harassment thought experiment--but which, as it turns out, is an extremely silly and shockingly entertaining document of 1990s mores and anxieties. Check it out, and read my take below:

In Disclosure (1994), Michael Douglas, only seven years removed from Wall Street, trades in his contrast collars for a toothpaste-stained tie as Tom Sanders, a middle manager for the tech company DigiCom on the eve of an important merger. Sanders isn’t a Gekko-style slick—you can tell because his hair is too feathery, too mullet-like—but he’s no slouch, either; he expertly manages a global supply chain manufacturing CD-ROM drives and arrives to his Seattle office at the start of the movie expecting a deserved promotion. But even as he ably surfs the changing tides of globalization, he finds himself caught out by the treachery of gender politics: His new boss Meredith Johnson, a former girlfriend played by Demi Moore, seduces him and then accuses him of assaulting her. Sanders, in turn, accuses her of sexual harassment; as the case unfolds, it becomes clear that more is going on than the simple assignation.

Disclosure is lethally symptomatic of its era, and, as such, lethally prescient about ours: it’s hard to imagine a movie that more efficiently compresses preoccupations about male sexual anxiety in the workplace and a disruptive tech boom dependent on Asian manufacturing into an actually-fairly-entertaining two-hour feature. But at its core, it’s a movie about being left out—thinking you were on the inside and finding yourself on the outside. Sanders is constantly missing meetings, arriving late, finding himself a step behind, catching glimpses of things through the glass walls of the DigiCom office.

Befitting the strangeness of the Great Moderation economy, the stakes are effectively nonexistent—“If this deal goes through, we’ll be rich,” Sanders tells his wife, who responds, “we’re already rich”; when Johnson’s assault accusation is levied, the biggest threat to Sanders isn’t jail or even social sanction, it’s being sent to Austin, Texas—but are cast as existential. DigiCom’s business is about to be supercharged, everyone agrees, and the worst possible outcome is to be left behind. In the movie’s inevitable virtual-reality climax, Sanders becomes something like a ghost, watching helplessly as a wireframe Johnson, oblivious to his existence, goes about deleting exonerating files.

The rest of the essay can be read here.

Forum brain crosses the gender divide

On The Cut, Zoe Dubno writes about the rise of “extreme beauty routines”--the elaborate, expensive, seeming experimental treatments, products, and regimens that are inescapable on the corners of social media devoted to beauty and wellness:

Her “stack” of beauty and wellness treatments — this is what people now call the combination of the many elements in their routine — reads to me like it’s been encrypted through the Enigma machine: “I take a copper peptide called GHK-Cu. And AOD-9604. And NAD+. I have Sermorelin, Ipamorelin, though I’m not sure what those do; I did research on them a long time ago and now I’m just in the habit of using them. I take HMB and vitamin D3, reishi mushrooms and cranberry pills and taurine and biotin, Telomere Length and HealthyCell gel packs. I give my dog, Madeline, NAD, because the first tests they did were on dogs and it was really effective and I want her to live forever.”

An argument about these routines erupted recently across Substack, as Dubno notes, when the fashion writer Laura Reilly announced the launch of a new “longevity and aesthetics” (lol) newsletter with a post about her own involuted routine:

In January alone, I’ve done Botox, Emface, IPL, scheduled Moxi / broadband light, seen my orthodontist, cardiologist, GP, OBGYN, ophthalmologist, dermatologist, plastic surgeon, trainer, and pilates instructor. I’ve renewed my health insurance and medspa membership. I’ve drawn blood three times and given two urine samples. My current skincare routine is 6-8 steps, my daily supplement stack is 17 pills (20 on Mondays) and a peptide taken subQ, and I’ve engaged at least 5 high-tech tools from my home device library (red light, SAD light, PEMF, etc). I’ve tracked every meal, macro, and relevant micronutrient. All of this, and I consider myself pretty healthy and not hypochondriac. I am n=1, but I am not alone here.

Now, “why are women spending absurd amounts of time and money on tortuous procedures designed to help them adhere to conventional beauty standards” is not a question that has never been asked before, and certainly not one that has scarcity of answers. But the particular valence of compliance is always interesting, and as I read Dubno’s piece (and Reilly’s) I was struck by the familiarity of the tone and structure of the descriptions of these extreme beauty routines: the exhaustively enumerated but still unexplained accessories, the painstaking multi-step cycles, the stubborn personal theories, the air of weary hard-won confidence from long hours of research. Over the years I have read some very similar paragraphs about hi-fi systems, S.L.R. cameras, espresso machines, and other objects of obsessive hobbyism on message boards, mailing lists, and forums.

What am I suggesting is that, in addition to patriarchy, misogyny, self-loathing, fear of the body, etc. many of the people Dubno writes about are also suffering from extreme, perhaps debilitating cases of Forum Brain.

If you’ve ever wondered why your coffee tastes bad, or what speakers you should get, and spent an hour reading presumptuous suggestions and quasi-manifestoes, opening up dozens of tabs of Google searches for acronyms you’d never heard before and products with unclear purpose, you have encountered (and perhaps suffered some symptoms of) Forum Brain. And, indeed, if you’ve ever opened one of the many skincare or beauty Reddits, and scrolled through hundreds of detailed descriptions of intricate beauty regimens, arguments about the efficacy of ingredients you’re not sure even exist, recommendations for French sun-protection products of vanishing obtainability: You know Forum Brain.

“Forum Brain” is a disease with many presentations, and many scholars disagree about whether it truly describes one single disorder or several overlapping and related maladies. But in this case I’m referring an issue concentrated on hobbyist forums (around topics like “coffee” or “pens” or “cast iron pans” or “mechanical keyboards”) or their descendants (i.e. BeautyTok) in which the object of the hobby itself (“high-fidelity sound,” say, or, in this case, “longevity and aesthetics”) is mostly treated as an excuse for the endless accumulation of poorly attested enhancing products and accessories; the ongoing development of idiosyncratic multiphase processes with ever-increasing complexity; and, above all else, discussion, debate, and discourse about the same.

The knowledge stockpiled by people addled by Forum Brain is much more about the discourse generated in the forums--fussy, endlessly debatable questions of competing technique and process--than it is about the hobby itself. Just as audiophilia is not really about listening to music (audiophiles have notoriously wack taste in music) the “extreme beauty routine” is not, precisely, about beauty. In fact, the women Dubno talks to suggest their routines are rooted less in a desire for aesthetics than for control:

Sperry told me that the morning we spoke, she “stood on her vibration plate, which activates lymphatic drainage, with her red light on and some Gregorian chants playing.” When I asked her why she engages with these practices, she said, “I think it stems from a desire for control. And, you know, in our current global climate, it’s just very chaotic. So to return to the body and to ourselves as a site of control is comforting for some people.”

This was a common theme among the women I spoke to. “We all have our little rituals of control that make us feel better,” a 31-year-old woman working on a doctorate in clinical psychology told me about the Botox she’s been getting since she was 25 and worked as a case manager in Brooklyn.

This seems like a reasonable explanation. But when I think about extreme beauty as a type of Forum Brain, I wonder if it’s actually the opposite: immersing yourself in fussy, endlessly debatable questions of competing technique and process as a way of relinquishing control, of turning yourself over to the thousands of different products and processes and strategies available for consumption online rather than actually considering .

What Forum Brain offers, I think--and to be clear I’m not trying here to describe the entirety of what extreme beauty represents or emerges from, just one aspect--is a slightly elevated version of the experience the internet has been refined to deliver: Entrance to “the Machine Zone,” Natasha Dow Schüll’s term for the anxiety-dissipating flow experienced by video-slots addicts. Like anything else on the internet--the infinite scroll, the rabbit hole, the conspiracy theory--Forum Brain allows for a kind of infinite deferral: You defer execution until you have the best possible products; you defer pleasure and self-satisfaction until you’ve zeroed in on the best possible strategy; most of all you defer any kind of confrontation with the ends to which you’re supposedly working, let alone your ultimate goals or desires.

Beauty Forum Brain is obviously more complicated and overdetermined than Audiophile Forum brain, and not just because the former is associated with body dysmorphia while the latter is associated with being annoying and spending thousands of dollars on gold-plated electrical outlets or whatever. But understanding the character of “extreme beauty” as shaped in part by the affective experience of the internet allows you to see how it, like all forum-borne hobbies, offers the same pleasure of constant refinement toward some hazily defined, endlessly deferred aim.

• A noirish literary road thriller about violent teenagers and male friendship;

• essays, articles, and columns about memes, Trumpist aesthetics, sex work, and rhythem in picture books;

• an underrated ‘90s conspiracy thriller with a fantastic cast and great soundtrack; and

• four songs I’ve been listening to a lot.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

Last week, we were thrilled to do our live chat at the Villa Albertine in New York City, on the occasion …

Read more

• A horror adventure set in the semi-pagan fens of Dark Ages England for fans of Between Two Fires and other period creepfests;

• links about A.I. productivity, the incel hegemony, and “situationgooning”;

• a great new play about love and death showing in New York City through May;

• and four tracks I’ve been listening to a lot.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

A reminder: Read Max is almost entirely funded by reader subscriptions. Everything I write here (and I do write all of it!) is a product of multiple hours of blood, sweat, tears, anxiety, procrastination, research, reporting etc. If you see the value of that labor--i.e. if you read these pieces and learn something, or adopt these opinions to sound smart elsewhere, or manage to pass seven to ten minutes freed from psychic pain thanks to the light entertainment of my prose--consider translating it into money by subscribing. At $5/month, it’s about the price of one beer every four weeks.

Subscribe now

What is “inevitable” about A.I. and writing?

Perhaps eventually there will be a week when there is nothing to say about A.I. and writing; this is not that week. In the period since Hachette canceled its run of the novel Shy Girl over suspicions that the author had generated some of the text with A.I., it’s been impossible to avoid on the subject of A.I.-generated, -assisted, or -augmented text.

On Monday, the Wrap reported that The New York Times had “cut ties with a freelancer after the paper discovered he used AI to help write a book review that inadvertently incorporated elements of a Guardian review on the same title.” This story came only a few days after Isabella Simonetti’s Wall Street Journal profile of Nick Lichtenberg, a Fortune editor “who has penned more than 600 stories since rejoining Fortune in July,” many of them “A.I.-assisted.” “AI-assisted stories accounted for nearly 20% of Fortune’s web traffic in the second half of 2025,” the Journal reports. “Most were written by Lichtenberg.”

Lest you think these are isolated developments, Wired’s Maxwell Zeff wrote about a number of journalists using A.I. to assist their writing, including the Times columnist Kevin Roose, who “created a team of Claude agents to help edit his book, led by a ‘Master Editor’ agent,” and the independent tech reporter Alex Heath, who “transmits his ideas to an AI agent, then lets it write his first draft”:

The AI tool is connected to his Gmail, Google Calendar, Granola AI transcription service, and Notion notes. He’s also built a detailed skill—a custom set of instructions—to help Claude write in his style, including the “10 commandments” of writing like Alex Heath. The skill includes previous articles he’s written, instructions on how he likes his newsletters to be structured, and notes on his voice and writing style.

Claude Cowork then automates the drafting process that used to take place in Heath’s head. After the agent finishes its first draft, Heath goes back and forth with it for up to 30 minutes, suggesting revisions. It’s quite an involved process, and he still writes some parts of the story himself. But Heath says this workflow saves him hours every week, and he now spends 30 to 40 percent less time writing.

These stories have all been widely shared, mostly disapprovingly, by journalists and their patrons on Bluesky, Substack, and Twitter. But reading them I was struck, more than anything else, by how commonplace they all sounded.

A hacky Times writer trying to cut corners was fired for inadvertent plagiarism? Wouldn’t be the first time. A Fortune editor is producing an inhuman number of stories a day, raising questions of accuracy?1 You should see what they used to write about bloggers! A good reporter who offloads the task of writing elsewhere? As Zeff points out, Heath has reverse-engineered a rewrite desk.

For all the revolutionary, transformational promises of A.I.--for good and for ill--the problems it causes all seem quite familiar.

My perspective on this is admittedly shaped by a 17-year career in digital media that I would describe as “de-sentimentalizing,” at a minimum. To work as a journalist online over the past decade has generally meant--with only a few exceptions--to be producing text inside a system that prioritizes speed, volume, and attention over any other attribute, including and maybe especially those you’d associate with “quality.”

At some publications, institutional identities and imperatives can militate against sloppiness, unreliability, and banality. But the external pressures and incentives toward slop are enormous and un-ignorable. And that has been true for many years, since well before A.I. offered a tantalizing push-button interface for high-volume content production. All the problems of A.I. writing--inaccuracy, misinformation, plagiarism, misrepresentation, and, above all, hack work--were widespread in the early days of blogs2 and throughout the digital media boom years, accelerated by the Facebook News Feed and other platform distribution schemes.

I recognize here that I’m recapitulating a boosterish argument that suggests that A.I. isn’t really a “big deal.” To be clear, I think it is! But it’s a big deal within a bigger deal, and that matters. When I read stories about writers or editors or publications using A.I. I wonder if I’m reading about A.I. causing problems to the media economy, or solving problems that economy poses to writers and editors and publications. Is A.I. “dangerous” as such? Or is exacerbating an already “dangerous” incentive structure and distribution ecosystem? Is A.I. transforming this system, or supercharging it?

I think these are distinctions worth making not so much because I want to rescue or defend some “innocent” conception of A.I., but because any organized professional and political response will want to properly diagnose the problem. If slop is a “feed” problem rather than (or prior to) an A.I. problem, the solutions are likely somewhat different. And if generative A.I. is a technology that solves actual problems--even debased problems of platform incentive structure like “how do I produce more content, cheaply and quickly, regardless of quality or accuracy?”--it’s harder to see it as a scam or an imposition. It’s doing what it’s designed to do.

It affects, too, how we understand what comes next. In a recent Guardian op-ed pegged to the Shy Girl controversy, the novelist Stephen Marche argued that

Artificial intelligence is here to stay, neither as an apocalypse nor as the solution to all life’s problems, but as a disruptive tool. The recent scandal over Shy Girl, the novel by Mia Ballard, was doubly revealing. Hachette cancelled its publication amid claims it was reliant on AI generation (Ballard has said that an acquaintance who edited the self-published version used AI, not her). But the book was originally self-published. Apparently readers and editors didn’t mind until the use of AI was pointed out to them. […]

There seem to be two options facing writers. The first is not to use AI at all, or to pretend not to use it. The other is to automate their writing practice. The first is retrograde and fearful. The second forgets that art is a human practice, made by people for people. As becomes obvious when you actually try to use AI to make art, this is a false binary. Already a few paths through the slop are emerging.

Generally, this is exactly the kind of assumption of inevitability that rubs A.I. critics precisely the wrong way.3 And in most ways I agree with A.I. critics that no technological development or implementation is “inevitable” in the sense of un-opposable, and certainly not “inevitable” because it presents some kind of revolutionary new paradigm that’s self-evidently superior.

But if we understand A.I. as a product of the systems that precede it, I think it’s fair to say ubiquitous A.I.-generated text is “inevitable” in the same way that high-volume blogs were “inevitable” or Facebook fake news pages were “inevitable”: Not because of some “natural” superiority or excellence, but because they follow so directly from the logic of the system out of which they emerge. In this sense A.I. is “inevitable” precisely because it’s not revolutionary.

• A sharp, thought-provoking new essay collection about A.I., the computer vision, the surveillance state, U.F.O. photos, psyops, and magic (and “magick”);

• a link roundup featuring pieces on bureaucracy, Killers of the Flower Moon, and the sad state of Western intellectual production;

• an endlessly entertaining video essay about the video store in Hollywood films in the vein of Los Angeles Plays Itself;

• a video of a mouse on a blueberry;

• and four tracks I’ve been listening to a lot lately.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

One problem with trying to cover the future of politics (i.e., my job) is that it often involves closely tracking some of the world’s most awful and grating influencers and content creators (i.e., my nightmare). Or, you could just read Who Broke It, a sharp twice-weekly Substack by Grace Weinstein, about the the social-platform and alternative-media trends, ideas, and personalities shaping traditional electoral politics.

I recently started subscribing to Who Broke It, and have enjoyed Grace’s funny, engaging reports from both the front lines of internet politics and culture. Looking for a guide to spotting D.N.C.-run social accounts (“Lowercase letters in all posts, cursing even if the politician would never”) or Tucker Carlson’s merch? Want to better understand the gravitational effect of both Nick Shirley and suburban millennial moms? Look no further.

This week is a particularly good week to subscribe, since Grace is heading to C.P.A.C., where some of the right’s most influential creators will be laying out their strategy for the midterms. If you want to be ahead of the curve, subscribe below.

Subscribe to 'Who Broke It'

Paid sponsorship by Who Broke It

Greetings from Read Max HQ! Today’s post is a “Read Max Re-Run”--a post from the archives I think subscribers might enjoy--running while I catch up with some small-business housekeeping chores.

A reminder, as always: Read Max is 99.9 percent funded by subscriptions. I’m able to treat this newsletter as a full-time job, and even to take weeks “off,” because of the generosity of paying readers. If you’re interested in my recommendations, or even if you regularly enjoy the weekly column, please consider upgrading your subscription for the low price of $5/month and $50/year.

Subscribe now

We’re now three weeks into the second Trump administration, and it seems clear that something has changed from the first go-round. The MAGA populist fraction of the Trumpist coalition has been relatively marginalized in favor of the austerian-accelerationist Silicon Valley right. The archetypal article about Trump voters in rural diners has been replaced with magazine stories about obnoxious arrivistes comparing Trump to Beyoncé and the inauguration to Comic-Con. Even Trump himself seems oddly sidelined, his threats about tariffs and treasuries effectively ignored by both heads of state and the stock market, his striking and creative Truth social postings left mostly unread--while Elon Musk attracts all resistance energy simply by tweeting “🤣” at some of the worst jokes on the planet.

Subscribe now

The end beneficiaries of Trumpist upward redistribution haven’t changed much. But the affect of the second Trump administration is, so far, wholly different. Gone is the apocalyptic malevolence of “American carnage,” supplanted by the unctuous corniness of “DOGE.” Gone is the sense of a lasting political realignment, succeeded by an inescapable minoritarian whine. Gone are Steve Bannon, the alt-right, and the “forgotten man and woman” Trump celebrated in his 2016 victory speech. In their place: Elon Musk and the Soy Right.

The Soy Right

“Soy,” as a piece of internet slang, originated among masculinist reactionaries in the 2010s as a dismissive synonym for “fragile” or “feminine,” deriving from the high phytoestrogen content in soy and the right-wing folk belief that soy consumption has “feminizing” effects on men. A “soy boy” was any man who was physically weak, passive, thoughtful, genial, or otherwise didn’t conform to the speaker’s gender standards.

Though still a somewhat gendered insult, “soy” these days suggests less a abstractly feminine weakness and more a specific message-board archetype: grating, weepy sensitivity mixed with undignified over-enthusiasm and self-satisfied corniness--closer to a synonym for “cringe” or “Reddit” or “Funko” as adjectives than to “cuck” or “pajama boy.” You might “soy out” over Marvel movies or Zelda games, or pose for photographs with a YouTube thumbnail-type expression of fake, open-mouthed surprise, known as “soy-facing”:

“Soy” still means “weak,” but with less emphasis on the physical more on the digital. In essence it’s a particular way of being bad at posting online: over-emotional, un-hip, sycophantic, sensitive, unoriginal, a rule-follower, reliant on stale formulations and hackneyed jokes. “Soy” is how the online right (always the vanguard of Trumpism) has perceived liberals since the Hillary Clinton campaign, against which they imagine and define themselves as “based”: Brave, authentic, unafraid, self-confident, cool.

Subscribe now

The simplest definition of “Soy Right”--a term that’s been in circulation for at least a couple years, but has picked up steam since the Trump election--is right-wingers who have adopted the sensitive, aggrieved victimhood pose and corny rhetorical and personal style that they have spent the last 10 years attributing to liberals, as Scarlet writes at Dialectics of Desire:

They have special diets are afraid of seed oils. They wear skinny jeans and have meticulously groomed beards. They talk non stop about masculinity, drive pickups, and wear plaid, but can’t change a tire to save their lives. While the right spent years mocking liberals for wanting “safe spaces” and echo chambers, for crying about identity politics, for being frail, fragile, overly-sensitive weaklings, they were slowly transforming into the perfect mirror of all of it — without the nagging concern for equality or any of that lib shit. […]

The Soy Right is being oppressed and they want you to know it. They’re scared to take the subway, they’re offended that you called them white or cis, they’re upset that you didn’t think they were cool in high school, they want to call the manager because there’s less boobies in video games. They are crybabies of the highest order. While the right is winning cultural and political victories nonstop lately, that’s not enough. They also need you to like them. Why don’t you like them?!

It makes me think of the great poster @ok_but_still’s observation from last August: “this the kind of conservative we're stuck with now. whiney sobbing guy with major depression diagnosis in bio who is still an asshole about everything.” An even more succinct way of putting it might be the phrase on this circa-2015 t-shirt, updated for 2025: “Kanye Attitude with Drake Feelings.”

To me the other key aspect of the Soy Right psychological profile, beyond its desperate need for approval and respect, is a childlike refusal of agency and responsibility, even while in power. A 20-something with access to Treasury Department systems is a “kid” whose racism shouldn’t be disqualifying. The South African billionaire throwing Nazi salutes is a enthusiastic neuroatyptical man who needs our sympathy. The Vice President of the United States would never have attracted the attention of the Pope if it weren’t for “hysteric progressives,” the real villains. Silicon Valley oligarchs were “driven into Trump’s arms” by the perfidy of Democrats. No one in the Soy Right makes affirmative choices; they’re smol beans who need protection and care.

This insistence on one’s own weakness is a contemptible way to live in the world. But the style of the Soy Right is as important, and in some ways even more depressing, than its animating resentments. If the online right of the first Trump administration was an unstable blend of Facebook credulity and 4chan nihilism, the Soy Right is an unbearable mix of Reddit corniness and Twitter self-satisfaction. We can look at some examples. Renaming a government agency after a decade-old memecoin and making the website an A.I.-generated cartoon--that’s Soy Right:

Describing the identification of six DOGE employees in Wired magazine as “doxing” and complaining distressedly that it places “a target on their backs”--that’s Soy Right:

Dressing up as a character from Sicario to record a video about your coffee technique--Soy Right:

A gaming website editor turned priest “hitting a Nazi salute and then the Dreamworks face”--Soy Right.

Smugly bragging that you’re “tripling down” on transphobia while still nervously self-censoring the slur--Soy Right:

This sentiment, voiced in this way, using this metaphor, is unbelievably Soy Right:

Soyfacing next to the president while wearing a crazzzzzzzy MAGA hat--definitionally Soy Right:

The most prominent Soy Right elected politician isn’t Trump himself--too strange, too original, too flamboyant--but J.D. Vance, a smug, weak skinsuit who recently threw his own kids under the bus in order to rescue a groyper’s job as a means of sucking up to Elon Musk, and who also tweets bad jokes in the deeply soy M.C.U. “so that just happened” voice:

The return of Gamergate

Does this odd and annoying mix of weepy fragility, oblivious self-importance, and obscene corniness sounds familiar? It should, because as a matter of tone it’s identical to--here imagine me talking over you as you attempt to finish my sentence with “the emotional register of 20th-century European fascism”--the psychotically annoying way that Gamergaters tweeted in 2015.

Subscribe now

I have to imagine that if you’ve made it this far in the post, through all this absolutely heinous over-analyzed internet-culture stuff about “soy” and “based” and the video-game Nazi priest, you are familiar with “Gamergate.” If not: In broad strokes, Gamergate was a campaign of targeted misogynist harassment against a number of games journalists and commentators (and, eventually, me and the website I worked at) spurred by a set of intricate and boring and extremely unimportant accusations of “bias in games journalism.” For those of us who were involved at the time, it’s generally remembered as the moment we all realized Something Bad Was Coming. As Kyle Wagner wrote at the time in a piece called “The Future Of The Culture Wars Is Here, And It's Gamergate”:

There are notes here, too, from a hymn book that predates the internet: self-pity, self-martyrdom, an overwhelming sense of your own blamelessness, the certainty that someone else's victimhood is nothing more than a profitable pose. All culture wars strike these same chords, because all culture wars are at bottom about the same thing: the desperate efforts of the privileged, in an ever-pluralizing America, to cling by their nails to the perquisites of what they'd thought was once their exclusive domain.

What we have in Gamergate is a glimpse of how these skirmishes will unfold in the future—all the rhetorical weaponry and siegecraft of an internet comment section brought to bear on our culture, not just at the fringes but at the center. What we're seeing now is a rehearsal, where the mechanisms of a toxic and inhumane politics are being tested and improved. Tomorrow's Lee Atwater will work through sock puppets on IRC. Tomorrow's Sister Souljah will get shouted down with rape threats. Tomorrow's Tipper Gore will make an inexplicably popular YouTube video. Tomorrow's Willie Horton ad will be an image macro, tomorrow's Borking a doxing, tomorrow's Moral Majority a loose coalition of DoSers and robo-petitioners and scat-GIF trolls—all of them working feverishly in service of the old idea that nothing should ever really change.

You don’t really need to draw any elaborate comparisons or jump through any hoops to say that the online presence of Trumpists in his second term is “like” Gamergate. It just straightforwardly is Gamergate, composed of many of the exact same people, who were, just last week, complaining incessantly, with Elon Musk’s participation, in a tone I can only describe as “Reddit Sephiroth” about--I’m not kidding--“woke gaming journalism.”

A not incorrect answer to the question “what is the Soy Right”? is: Gamergate, ten years later, plus a handful of movement-conservative influencers, plus the Vice President of the United States, plus the richest man alive.

Does the Soy Right matter?

Let’s reaffirm, here, that we’re mostly talking about an online phenomenon--a particular mode and tone of discourse that has come to dominate political rhetoric on X, the Everything App, and among certain Republican politicians. Most Trump voters are not members of the Soy Right; I would venture to say that most of them are not members of any “right” at all. To the extent “soyness” extends to actual policy it’s mostly in stuff like renaming the Gulf of Mexico or complaining that Hakeem Jeffries used the word “fight”--moronic high-profile “trolling” and sniveling whining.

For this reason I hesitate to suggest that the Soy Right is “important” by any reasonable definition. I don’t think it’s particularly politically detrimental or advantageous to have your administration represented online by a bunch of gushing cornballs and larmoyant freaks, and it also seems clear that Musk is going to illegally gut the federal government regardless of whether or not you or I think he’s “based” or “epic.”

Subscribe now

But I do think the prominence of the Soy Right online right now is an interesting reflection of the second Trump administration’s priorities and composition. It seems to me to be a natural consequence of Trump’s decision to offload most of his policy and action to the tech industry--and, in particular, to Elon Musk himself. As Matthew Ellis recently observed, the classic sunglasses truck-selfie PFP posters who once dominated Trumpist discourse on Facebook and Twitter “aren’t really the main right wing characters on here anymore,” disappearing, along with Steve Bannon and the rest of the MAGA populists, from the Trump administration’s media profile. In their place are the statues and the cypherpunks of “the dumbest RETVRN posters and crypto maniacs”--and, of course, the Soy Right. If truck-selfie Twitter had a loose correspondence to the Bannonite wing of Trumpism, the Soy Right directly represents the Muskite wing.

• A perfect, gripping Cold War spy thriller from one of the masters of the genre;

• essays, articles, and blog posts about the use of A.I. in war, polyamory, and more;

• a terrifically dark and unsettling indie folk horror set in the world of Irish folk music and featuring a great soundtrack; and

• four tracks I’ve been listening to a lot lately.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

While the chat is free when live, and a short preview is available…

Read more

• Two book recommendations from Ari: The first, a mysterious and beautiful family saga set in Denmark and Russia around the Russian Revolution; the second, a beloved classic that turns out to be a real page-turner.

• Plus: essays and articles Max has been reading about the A.I. economy, the shadow of Tolkien in the tech industry, and contrastive corrections and other tics of L.L.M. style;

• a surreally beautiful and almost indescribable post-apocalytic animated masterpiece somewhere between Jodorowsky, Tarkovsky, Moebius, Winsor McCay, Gene Wolfe, and a great JRPG; and

• four tracks I’ve been listening to a lot lately.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more

I think everyone should have a personal website. Not a social media profile — an actual website, a little corner of the internet that belongs to you, where no one can reply or quote-post or suggest you might also enjoy branded content. I built mine with Squarespace. (You can see it here.)

What sold me was the combination of ease and control. Squarespace has a huge library of professionally designed templates, which you can use as-is if you’re a normal person who wants to look professional online. Or, if you’re me, you can use their intuitive drag-and-drop design tools to customize everything until your site features hand-drawn KidPix art by a kindergartner and looks like a Geocities page that went to journalism school. Either way, the whole process is fast — no coding required, easy to update whenever you feel like it. And it’s not just a static page: I’ve got S.E.O. tools and analytics built in, and I can add a storefront, new pages, or even email campaigns whenever the mood strikes. It’s flexible in a way that a social media profile simply isn’t.

If you need a website, portfolio page, storefront, or nearly anything else, Squarespace is perfect. The only thing it can’t provide is KidPix images my son made.

Click here for a free trial, and when you’re ready to launch, use READMAX to save 10% off your first purchase of a website or domain.

Greetings from Read Max HQ! In this week’s edition, we discuss The New York Times quiz about human and A.I. writing, and what we get out of these side-by-side preference tests.

A reminder: Read Max is a subscription newsletter that depends on paying subscribers to survive. Every week, we lose a few paid subscriptions, which means we need to be outpacing our churn--which means that if you haven’t been subscribing, but have been enjoying the fruits of my labor (i.e. the bad jokes), perhaps it’s time for you to chip in? At $5/month or $50/year, it costs roughly as much as buying me a beer every few weeks, which I imagine you would be happy to do.

Subscribe now

What do we get out of “A.I. or human” preference tests?

I spent the first half of this week at “Cultural A.I.: An Emerging Field,” a fascinating conference put on by N.Y.U.’s Digital Theory Lab and the Remarque Institute. I found it genuinely invigorating to hear papers from and talk with some extremely smart people who are thinking capaciously and rigorously about A.I. systems as social and cultural technologies.1 What is L.L.M.? How does it intersect with literary culture and political economy? What can its operations tell us about language, writing, and intelligence?

And then I would go home and open up Twitter, where the real School of Athens stuff was happening:

Now, I don’t particularly begrudge the Times its little widgets, which I understand to be a key strategic component to its overall business model and continued health as The Last Employer Of Journalists. But I would politely disagree with my friend Kevin that this represents “a moment.” That non-expert humans, given blind side-by-side comparisons, tend to do a bad job identifying A.I.-generated text is, at this point, a pretty well-established finding. So, too, is its black-pilling corollary that people consistently if modestly prefer the A.I. output.

Indeed, back in 2024 I wrote a little bit about a version of the human-or-A.I. game (this one about images rather than text) run by Scott Alexander, which had similar results to the Times quiz. My take at the time, calibrated to be maximally annoying, was that Alexander’s quiz didn’t really prove that people couldn’t identify, and, when asked, “prefer” A.I. art--but that nonetheless it was probably true, in a general sense, that people can’t identify, and likely prefer, A.I. art.

I don’t think my general feeling about this kind of test, or its results (such as they are) has changed much in the intervening 18 months. On the one hand, the Times quiz is a deeply defective experiment, starting from a set of patently false premises, whose results are being wildly over-interpreted. On the other, would anyone really deny that most people, in a vacuum, would have trouble identifying well-prompted A.I.-generated writing, or that furthermore they likely think it (to use the Times verbiage) “reads better”?

And yet despite the flaws of these tests, the obviousness of the result, and the repetitive tediousness of the “conversation” that follows on X, Bluesky, and Substack, we continue to craft, publish, take, and argue about these tests. So what are we getting out of them?

It seems worth asking what, actually, is happening when we misidentify, and express preference for, A.I.-generated writing in blind A/B tests like these. An influential 2024 paper by Brian Porter & Edouard Machery--linked to in the Times quiz--that asked subjects to identify A.I.-generated poems found that participants “performed below chance levels” (at 46.6 percent accuracy), which is to say they got worse results than they would’ve if they were just guessing.

What this suggests is that people were able, at least to some extent, to distinguish between human and A.I. poetry--they just thought that the A.I. poems were human, and vice versa. Porter and Machery attribute this to

shared yet flawed heuristics to differentiate AI from human poetry: the simplicity of AI-generated poems may be easier for non-experts to understand, leading them to prefer AI-generated poetry and misinterpret the complexity of human poems as incoherence generated by AI.

That is, in aggregate, participants could tell that the A.I. poems and human poems were stylistically different from each other; they simply misunderstood what those different styles actually marked.

This is a relatively common phenomenon, and the finding that people misidentify A.I. output as human (and vice-versa) at higher-than-chance levels seems to hold across other domains: A.I.-generated faces or dating profiles, for example. (It’s probably particularly acute with poetry, a field with, let’s say, a wide gap between what is “good” and “coherent” to regulars and what is “good” and “coherent” to novices2.) In many contexts most people can (more or less) correctly differentiate between A.I.-generated output and its “authentic” counterpart--but cannot correctly attribute the output.

What’s funny about this is: We actually really want to prefer human-authored writing! In open-label tests, where the excerpts are shown with attribution, people consistently express preference for whatever text is labeled human, even when the text is actually A.I.-generated. (So do A.I. evaluators, as I learned at the conference from Wouter Haverals, to an even greater degree.)

This is not a particularly satisfying set of findings insofar as it validates neither the A.I.-booster “it’s so over, A.I. writing is better than human writing” side nor the A.I.-skeptic “A.I. can never write like a human” side. What we can say is that people mostly can’t identify A.I.-generated text as A.I.-generated (crowd boos), but they can sometimes distinguish between it and human-authored text (crowd cheers); it’s just that they tend to think the A.I.-generated text is human (crowd boos), maybe because human-generated text is stranger, worse, or more difficult (crowd hesitantly cheers), which readers mistakenly believe is more typical of A.I.-generated text (crowd silent now) and thereby disprefer (crowd sort of murmuring confusedly), unless you tell them it’s actually human, in which case they change their minds and like it (crowd has mostly left at this point).

But all of it taken together suggests that, given our strong bias in favor of writing we believe to be human, A.I. vs. human “preference” tests (or “reads better” quizzes) are often second-order “identification” tests, in each case measuring not “preference” per se but the accuracy of the prevailing heuristics for identifying A.I. writing. Participants in these studies, it would seem, express preference for the A.I.-generated writing not because it’s “better” in some formal sense--cleaner, simpler, more beautiful, whatever--but because their “flawed heuristics” have led them to the conclusion that it’s human-authored, and ipso facto better.

If this is right, much of the discourse about quizzes like the Times’ is getting the order of operations wrong. It’s not that people see two paragraphs, prefer one based on its quality, and then attribute it to humans based on that preference. It’s that they see two paragraphs, attribute one to human authorship based on style, and then prefer the one they’ve attributed. What’s at stake when taking these tests isn’t quality or beauty or clarity, but style; not “which one is better,” but “which one sounds more like an L.L.M.?”

In experiments like the one documented in the Nature study, participants often express preference for “cleaner” or “smoother” text. Because this accords with our intuitions about the kind of writing that most people should or would prefer in most contexts, it’s easy to take for granted the idea that people are expressing some relatively fixed, “natural” preference for the kind of professional plainness L.L.M.s tend to exemplify.

But if they’re picking the text that better displays “cleanliness” because they mistakenly associate it with human writing--and disfavoring stranger or more difficult text because they associate these qualities with A.I.--you can easily imagine a world where people begin to express preference in certain contexts for clunkier, thornier, and “worse” writing, because those are the stylistic markers of humanness.

As long as people want to prefer human-authored to L.L.M.-generated writing, we will place a premium on whatever style we associate with human authorship--even as that style changes. You can already see this process beginning from the other direction on social networks like Twitter, where em-dashes and not-x-but-y contrastive corrections--perfectly innocuous and useful writerly tools which not five years ago would likely have been highly correlated with “good prose”--are immediately treated with derision and suspicion. By that same token, certain kinds of “bad writing” should be seen as evidence of human authorship. How long before run-on sentences are preferred to em-dashes?

L.L.M.s, of course, can and will get better at mimicking the “strangeness,” clunkiness, and badness of human prose; I’m skeptical of claims that there is some built-in technical limitation that prevents A.I. text from ever being truly indistinguishable from human prose.3 What seems more likely to me is that as L.L.M.s move away from the easily identifiable generic LinkedIn style that currently dominates, our preferences will move as well, in an attempt to stay one step ahead.

One thing that interests me about these quizzes is the extent to which they resemble a stage in “reinforcement learning from human feedback,” or R.L.H.F., a chatbot training process during which the L.L.M. will provide two (or sometimes more) responses for every prompt, one of which is selected as the “better” response by a human evaluator. The human preferences are then used as the basis for another model that predicts a “score” for any given response; finally, the L.L.M. is directed to maximize its score for any given response.

Quizzes like the Times’ are games that L.L.M.s are designed to excel at. A.I. writing is literally optimized to be the writing most people prefer in A/B preference tests; the main thing an L.L.M. chatbot “wants” when replying is to be generating the text that its users would choose as a good answer to the prompt over all other possibilities.

We’re not training L.L.M.s when we take quizzes on the Times website. But I suspect we’re training ourselves, taking the tests to measure and adjust our own heuristics for distinguishing A.I. text from human writing. I think A.I. boosters often want these blind tests to “prove” to stubborn skeptics that A.I. writing is “as good” as human writing. And I know that skeptics object that the tests don’t accurately measure anything of the sort. But my sense is that they’re not measurement instruments at all--they’re territory on which a kind of ongoing stylistic arms race is being conducted. We like these games not because they satisfactorily “prove” that A.I. or humans can produce “better” or “worse” text, but because they reveal to us--both in themselves, and in the discourse that follows--the stylistic tells that allow us to distinguish between the two.

• (More of) what I’m reading about the war in Iran--17 pieces about Dubai influencers, undersea cables, data-center missile strikes, the life and death of the foreign policy establishment and international law, and more;

• a great, sleazy Elmore Leonard adaption; and

• four tracks I’ve been listening to a lot.

If any of that sounds interesting--and if you want to support Read Max in its continuing mission of explaining the future to relatively normal people--please subscribe below. Subscribers get access to these weekly emails and to the comprehensive Read Max Master Lists of Good Movies to Watch and Good Books to Read--as well as preferred pricing on merchandise like the Read Max “EMAIL SUPPLY” caps. If you’re already a subscriber, thank you.

Subscribe now

I love to hear from readers who want to recommend a cool book or movie (or whatever!) to me, or from readers who are looking for specific kinds of recommendations. Just leave a comment or drop me a line at maxread@gmail.com.

Finally, please note that I may receive a small commission if you purchase books via links in this email.

Now, the roundup:

Read more