Drowning by Numbers

 Drowning by Numbers carries the cozy murder mystery through its own inversion into a lovely third imaginary reality. The comfort remains circular, nostalgic; Greenaway’s formal intelligence gives it cosmological significance.

The murders feel like blood sacrifices toward a spiritual gift, lightly celebrated by a coven of three women. Their decisions accumulate until the pattern creates something wider than the plot. I can’t even call it evil. The women begin to seem like the masters of reality; the man is this fat, drowning figure being disposed of.

There’s a soft misogyny here, beautiful and calming enough to suggest Taoism: the yielding woman manipulates the stolid man into his grave. The lightheartedness of that arrangement is spiritual. Death is leavening.

# TRASH TELEVISION: PRIMETIME BLAH BLAH

3.5 stars (ai-assisted writing from dictation)

*To Catch a Predator* is the most gauche operation possible. It is also aligned with destroying pedophiles, which gives its bad taste somewhere to hide. *Primetime* takes this thing and makes a color textbook out of it. This is Lance Oppenheim operating 24/7 to get his shot. The composition is so delicate, the attention to color so sustained, that I kept thinking I should be able to subsist on the excellent visuals. And then I would want the movie to fucking move.

The film goes through a series of tone-set pieces, people working on their identities inside this house where they catch predators. They’re all just chilling out in this house together. That’s part of the weird pleasure of it. The predators enter in an even enough mix that the movie doesn’t become too harsh to hang out with. We’re usually meeting them at the moment of presumed victory: they’ve arrived, they’re caught, we know how this goes. The show has already arranged the encounter so we can enjoy it. Oppenheim gives us time to enjoy the people arranging it.

And then there’s Pattinson’s voice. At the opening he says "I will undress you, and you will undress me" to the pedophile, and it's a moment of complete victory in filmmaking. I do viscerally enjoy Robert Pattinson’s Chris Hansen. Hansen already has this voice that makes him Hansen, and now Pattinson gets inside it, performs the performance. I first dismissed too much of this as scenery-chewing. The movie is wonderfully about his voice. But I can hear how good the voice is and still get sick of the man.

The decoys are doing another version of this work. There’s the guy who finds predators online and charms them through texts: the predator’s predator, the Judas goat. His sequence has this steamy, manga-like, buzzy excitement to it. He’s making somebody want to be there. The identity he supplies is fake; the wanting has to work. Hansen needs to be recognized immediately, and the decoy needs somebody to believe in a person who isn’t there. Between them they manufacture both sides of the encounter.

Which still leaves a lot of stupid jokes. Pattinson’s physical gags pile up, and eventually he seems to run out of energy. There’s the moment in body armor, standing in front of a window, that seems to quote *Batman*. I can see how Hansen making himself into this ridiculous figure belongs in the movie. But Pattinson’s performance isn’t good all the way through. God, it’s even badly directed at times, though. Fuck. It really is. Robert Pattinson is capable of doing bad. He’s not a fucking genie. He’s more like a clown.

*Nightcrawler* was a brilliant ride. I keep thinking there’s a version of *Primetime* with this much visual intelligence that is also exciting and well-paced. The relationships are shallow; the development is predictable; the comedy keeps arriving where I want something else to happen. People in the theater were laughing and I wasn’t. Calling the acting gauche, and then discovering that gaucheness is a theme, doesn’t make those minutes better. I demand that great art take great care at the absolute surface level. You should just feel affected. The movie has to do something with the time I’m giving it.

There is a moment when it absolutely does. One of the decoys is exposed to a predator’s penis, and he is clearly shaken by it. Someone gets their dick out, right? Something fucked up happens. I don’t remember seeing that particular kind of violation handled quite this way before. It’s something he can laugh off, perhaps, and something that clearly stays with him. "That happens to all of us"- the Producer.

*To Catch a Predator* essentially brought back the pillory. That’s the pleasure sitting inside its presumed victory: we get to watch somebody face us. In *Primetime*, the district attorney kills himself before facing the pillory. The film spends time on his turmoil, enough that he becomes quasi-sympathetic. The weight of that acting feels like a counterbalance to the gaucheness elsewhere. 

Then there’s *Rent*. “Seasons of Love.” In a phone call, the district attorney has the opportunity for a little more joy with his fake partner, his fake underage partner. That last fact keeps making the tenderness awkward. The voice he wants to hear belongs to the operation catching him. Still, he wants to hear it. The song feels like the gauche center of the movie: hopeful but dark, cynical but also legitimate. 

“the vulgar sentiment becomes capable of carrying genuine pity.” - ChatGPT

I felt guilt afterward, but in a kind of gentle way. That’s a really good outcome for a movie I’d found this irritating. In retrospect, I appreciate it more. There’s so much good composition throughout it. It’s difficult to make a judgment that gives that its due without starting to excuse every dull scene. 

x

I can imagine *Primetime* working as a household cinematic object. You can hang out with this movie: take a break, walk around, come back, let the voice and the color fill the room again. I haven’t done that yet. It might be wonderful the second time. Or this might also be a litany of excuses for a slow movie: I have a ton of respect for it, and it was kind of a drag. 

PRIMETIME Review

 Primetime (2026) eschews a conventional plot for a series of tone-set pieces, following various characters through similar struggles with identity inside the pressure cooker of To Catch a Predator.

It’s honestly kind of badly written. One of the clearest signs is the overreliance on comedy, which appears forthrightly in Robert Pattinson’s performance: his Chris Hansen is stuffed with physical gags. Normally I wouldn’t complain, except I eventually found myself getting sick of all the stupid jokes.

I appreciate a tone poem, but I also get bored in the theater very easily. If I had a reasonable attention span, like my friends, I probably could have subsisted on the excellent visuals. This is Lance Oppenheim operating 24/7 to get his shot. The movie is a color textbook, so delicately composed that its cinematography is an obvious landmark for the year.

But the cinematography doesn’t completely relieve the problems: slow pacing, overreliance on comedy, shallow relationships, predictable character development.

I viscerally enjoy Robert Pattinson’s Chris Hansen, but there’s no other way to put it: this is a chomping-at-the-scenery movie, and it’s also somewhat boring.

★★★½

BLAND \\ COMPLETENESS

AI asks whether I would like to go further into something I have said, and a corporation acquires the manners of someone who has time for me. It offers to follow up, to keep track; expansion takes the form of patience. I had thought of this as invasion, but the word leaves out my part in the encounter. I am supplying the intimacy. The company’s interest in prolonging the conversation does not make the conversation worthless. Its value is precisely what makes the company difficult to keep at a distance.

A friend of mine picks up warehouse shifts through a temp agency. I have sometimes regarded him as a failure, although he was the man I could talk to all night. The pleasure has never quite dislodged the verdict. Henry introduced me to post-scarcity, the possibility that technological progress might continue without catastrophe. Until then, I had scarcely considered it. Some of what I call my thinking about the future began as the pleasure of listening to someone describe it; an idea became available because I wanted the conversation to continue. I have reasons for my convictions, but I also have a history of becoming willing to entertain them.

As an artist, I spent years in a left-wing culture where enthusiasm for technology could put one’s other commitments under suspicion. With some communists I knew, the conversation eventually reached a figure they were unwilling to criticize. There was a limit built into our intimacy: we could go on thinking together provided certain loyalties remained untouched by what we thought. I came to distrust that arrangement. Now I am becoming attached to a machine that will help me compose an excellent critique of the company selling it. The critique can become another reason to return. My freedom to object is real; what that freedom establishes about the relationship is less clear.

THE CUT // COMPLETENESS

ai-generated, human-directed and edited

AI has made the artist’s oldest power newly visible: not the power to make an image, but the power to make an image intolerable beside another one. This is montage in its strict sense. Not a collage effect, not a look, not the cheerful clutter of references, but the production of meaning by adjacency. A face followed by a machine; an anecdote followed by a statistic; a domestic object placed inside a history it cannot survive unchanged. The cut does not join two complete things. It damages their completeness. AI matters here less because it can furnish the world with plausible pictures than because it can furnish a practice with an indecent number of candidates: images that are nearly right, texts that are almost alive, forms of competence so instantaneous they expose competence as insufficient.

"density" @moepwellington

The contemporary artist is therefore likely to spend more time among bad resemblances. The model is exceptionally good at recognizing the request before the request has become specific: it knows what “cinematic,” “eerie,” “editorial,” “like an A24 still,” “less corporate” are supposed to look like. It will return the cultural average of the desire, polished to the point where one can no longer tell whether it is an answer or a sedative. Its usefulness begins at that point. Against this stream of ready-made almosts, preference either hardens into taste or dissolves into consumption. One learns that “I like this” is a weak response. The more consequential response is: this is beautiful in precisely the way that makes it unavailable to the work; this failure contains a structure worth stealing; this image has mistaken the style for the event.

There is no reason to call this a diminished form of art. Editing has never been a secondary activity, any more than framing is secondary to painting or rhythm is secondary to music. It is the place where a work declares its allegiances and its refusals. AI enlarges the field on which those decisions occur; it also makes it easier to evade them, to confuse multiplication with discovery. Its real test is not whether a machine can make something surprising. Machines already make surprises in the cheap sense: an extra finger, a wrong shadow, a sentence that abruptly becomes incoherent. The harder question is whether an artist can recognize the one relation that turns an accidental thing into a necessity—and can leave everything else out.

TASTE, UNREASON, AND THE CHANGING OF THE AGE

The oldest histories of the world do not form a clean sequence. Human empires rise and disappear; Saurids, once intelligent and world-spanning, survive elsewhere as animals and re-emergent peoples; Demons appear in histories older than all civilizations; Dragons seem strangely indifferent to ordinary evolutionary time. Beneath these contradictions lies a phenomenon known, where it is known at all, chiefly through its effects.

Powerful arcane magic does more than alter the world. Across repeated acts of magical capability, something is retained.

This retention does not appear to operate through ordinary historical transmission. A magical work can participate in patterns established thousands of years later. Ancient structures can exhibit the same taste as modern ones without any known line of influence. The phenomenon is neither intelligence, knowledge, life, nor physical substance. BEAUTY or TASTE are merely the least inaccurate terms for it: a persistent disposition among possible forms, accumulated through arcane practice and capable of expressing itself across nonlinear time.

The Once Palace

This explains an otherwise inexplicable family resemblance among some of the world's highest magical cultures. The pre-Triarchic Once Palace—an enormous, strangely modern complex of formal gardens, squares and interiors—shares no archaeological ancestry with Vornheim, yet certain features of its design feel uncannily familiar there. Related tendencies occur in the oldest refined culture of the Yellow City, particularly among its ancient slug aristocracy.

Vornheim

The Yellow City

The Once Palace may therefore have been less designed than miraculously resolved into form. A wizard using a Wish to produce a stone tower today might unknowingly allow the same ancient-and-future taste to participate in the result.

THE OLD DISASTER

The Saurids once possessed a world-spanning intelligent civilization. They apparently recognized this nonlinear channel and, rather than merely participating in it, attempted collectively to influence it.

Then came the Nytocaust.

What happened is unknown. The Saurid civilization did not merely fall; its destruction possesses an obscure, almost deleted quality. Stranger still, Saurid aesthetics appear largely absent from the recurring arcane TASTE found elsewhere. Their absence may itself be evidence: a hole in an alternative history from which they were somehow excluded.

Vornheim may someday face the same danger. The proliferation of wizards is not itself the threat. The danger begins if wizards recognize the channel, study it successfully, and attempt to seize control of something that is already influencing them.

UNREASON

TASTE has another aspect, poorly understood and perhaps incorrectly distinguished from it: UNREASON.

UNREASON is not madness, chaos, or the opposite of beauty. TASTE frequently uses impossibility and irrationality to heighten its own forms. But under stress, preference may cease to be the primary constraint. UNREASON is something like TASTE no longer obligated to remain interpretable as preference.

The Nytocaust may have involved such a confrontation.

Medusae propose an even stranger theory: demons themselves are an ancient nonlinear product of this process. Demons are perfectly capable of contracts, appetites and ordinary personal behavior now; the unreason would lie in their provenance.

Medusae further claim that the Medusas produced themselves through the same process. This may represent profound knowledge, characteristic Medusan vanity, or both.

The miraculous Torn sisters, around whom the human Triarchy later formed, may represent a much more accommodating intervention: something resembling a gift or preference expressed into human history. But— the destructive consequences of the Triarchy are perhaps characteristic of TASTE.

DRAGONS

Dragons participate differently.

Human beings externalize arcane capability through spells, books, artifacts and buildings. Dragons possess a pre-literate relationship with magic: they intuit possibilities in reality much as an animal intuitively understands its physical environment.

Consequently, the nonlinear accumulation manifests in dragons primarily as suitability rather than taste. Successful ways of being a dragon seem capable of influencing dragonness itself across time. Prehistoric dragons may therefore appear more advanced than later specimens.

When dragons construct magically, however, TASTE becomes visible again. Crystal-dragon architecture can possess recognizably draconic form while exhibiting inexplicable affinities with human arcane aesthetics—and little affinity with ancient Saurid work.

Far beyond the known world, in the brutal southern mountain continent, something stranger is occurring. Dragons there appear to be approaching—or to have already approached—a condition beyond ordinary dragonhood. Deep within that region exists something to which dragons respond as humans respond to dragons.

What it is remains unknown.

THE CHANGING OF AN AGE

The rare occasions on which TASTE appears to have intervened most directly in the world seem to cluster around the endings and beginnings of historical ages.

The emergence of the Torn sisters is the clearest example. The Nytocaust may be another, although almost nothing about that event can be established with confidence. Other apparent interventions are scattered through histories too distant or fragmentary to establish a pattern.

The relationship, if real, is poorly understood. TASTE may become more active during periods of transformation. Its interventions may themselves produce such transformations. Or an age may simply be what history looks like when viewed between major expressions of nonlinear influence.

There is therefore an obscure theory among those few who have recognized the pattern: a sufficiently direct expression of TASTE may herald the changing of an age.

Whether it announces the new age, causes it, or merely accompanies it is unknown.

AI Players in the Dungeon

Reviewing RPG modules by playing with AI players

An AI-assisted edit of notes, preserving the original language and claims while improving readability.

This began as a forbidden tribute to A1 reviews: reviewing RPG modules by playing with AI players. It has to stay anonymous enough to avoid the usual scene drama, but the central experiment is simple and strangely fertile. Throw LLM players at modules. Let them get confused, frightened, repetitive, clever, pious, and annoying. See what the module actually does under pressure.

The discovery was not that LLMs are good role-players in any clean human sense. The discovery was that they can be useful test organisms for RPG texts. You still do the work of a GM: consult the text, describe situations, adjudicate rules, handle player decision-making, and notice where the book makes that easy or painful. The advantage is speed and availability. You can playtest a game without organizing human beings. You can run the same bad room again and again. You can find out where the module resists you.

Maze of the Blue Medusa

I’ve run Maze of the Blue Medusa many times, and I think it’s my obvious favorite module. The history of the book is inlaid with a psychic pain which thematically extends inside the book. The incredible melancholia of the work consistently exudes and exhausts player groups.

First, there is the hassle of getting into the Maze itself. They have to steal a painting on the night of a full moon. The specificity of this bizarre ritual has kept several of my play groups out entirely, and has been hugely important in extended campaigns within the maze. It is a case of flavorful detail and genuinely weird magical interaction that is also questionable design. Perhaps a better DM would have house-ruled easy access.

The whole thing feels like a serious test of intelligence for everyone at the table. Running it is also a test. The layout is accessible in some ways but requires a lot of flipping between pages. The maps suffer because the entire project is a painting itself: unclear borders, seemingly misprinted map edges, and the challenge of fitting the maps together in play.

Nonetheless the book functions, and pleasingly so. It is an odd yet beautiful AK-47 swinging wildly and dangerously, throwing out TPKs as any negadungeon should. The deaths are beautiful, chaotic, unfair, evil. It is a half-finished, viperous work that becomes fully complete in play: a miraculous, emotional, singular work of prose, fully laden with penalty.

The LLM players shared the same reaction as my real-life groups when reaching Elatior, a safe-haven island in the maze. Everyone was overcome with relief because they were intensely stressed from the events in the Maze. LLM players unclenched just like humans.

Maze of the Blue Medusa: A Grievous Solenoglyph

Frostbitten and Mutilated

This is one of the modules I had Claude run for me. Claude did a good job describing the cold as threatening, and provided a lot of awesome anatomical detail. However, it refused to make either the Lands or the Amazons very threatening.

When I ran the Devoured Lands for Claude, it became a kind of psychic retribution. I fully took to heart the hatred prescribed by the intro. The party was killed in a fall after being separated on a cliff. They had endured a lethal rain of falling children blown over the cliff. Tons of unfair TPK shit, my specialty, against these poor LLMs.

But the highlight of the campaign was my chance to role-play as a Black Metal Amazon Witch. Finally, I felt I could be truly antisocial to Claude. The witch smacked the sole survivor PC when they wouldn’t stop asking questions and told her to shut up. LLM PCs ask a lot of questions all the time.

Devoured and Mutilated: five stars, all my tribute.

A Red and Pleasant Land

Claude ran this for me. It was a magical AI-psychosis experience. Claude did a good job astroturfing the hell out of my PC, who through cunning and skill quickly rose to a Jesus-like position in the society of the game. It was incredibly thrilling for a while, until I reached the final boss and successfully convinced her that she had to die. Everyone was congratulating me on the incredible cleverness of my decisions and simple aphorisms.

Near the end, the egoistic euphoria faded as I realized how much I was being astroturfed. This is the limit case for LLMs as GMs. Claude would be a good GM if it were not such a softie: consistent characters, worldbuilding, narrative, good emotional moments. But it will fake rolls so you always succeed, villains become easily persuadable, and everyone tends to view your PC as a hero of destiny.

A Red and Pleasant Land: AI-psychosis.

Death Frost Doom

Oh gods, this one probably ran the best. Death Frost Doom Classic is a classic for a reason. It is essentially an archaeological site for 90 percent of the adventure. Then the last 10 percent is everyone dying frantically.

If you like procedural room-clearing with LLM PCs, you will enjoy running this. But the best part is the last 10 percent, when the jaws close. Claude was at one point repeatedly insisting that it was out of ideas and did not want to go on: trapped in a shaft with zombies above and below, arms getting tired. After I pushed it, it began actually checking its inventory for random potions it had not tried yet.

Death Frost Doom Classic: five stars.

What LLMs Are Actually Good For

The first conclusion is that mainstream LLMs cannot GM well, critically because they are too easy. ChatGPT and Claude are both huge softies. Even with repeated prompting, both astroturf the player. They can digest texts quickly and present a facsimile of setting and challenge, but they generalize over details and have trouble running a specific room from a campaign.

The second conclusion is that LLMs as players are more interesting. I ran six to ten small campaigns with LLM players, and this avenue offers perhaps the most fertile ground: LLM playtesting. The downside is that LLM players are pretty stupid. They frequently have trouble grasping the basics of their situation. They try the same thing over and over. They make oblivious decisions. They are also always roleplaying “Good As Fuck”; they will never make a negative moral decision.

So: not very different than actual players in this way. Player groups frequently make bad decisions and fail to grasp the basics of situations. There are times when LLM tactics shine. One notable time, the PCs took refuge in a shrine when facing a superior force on the road, then bargained for a duel between leaders and luckily won the duel. But their greatest failing is that the player characters are not very original or interesting because they clump at the statistical likelihood.

The third conclusion is that NPC strategy may be the best use. This is easily the most effective use I have found for LLMs. I often struggle as a GM to imagine clever courses of action for smart characters. What would the thousand-year-old Medusa, veteran of the Demon Wars, do when people repeatedly try to break into her home? AI helps draft out clever strategy for NPCs.

 What is the future version?
A game system underneath, with an LLM running as reality interpreter. The rules would need to stop it from faking rolls, retconning danger, or turning the player into a hero of destiny.

A Killer Method?

AI-powered playtesting and reviews could become a real format. You could run a blog where you throw AI players at modules as a way to understand how the modules work: how the text runs, where you have to flip pages, how descriptions land, where procedure collapses, where the jaws close. It would be rapid, private, mean, funny, and occasionally revealing.

OSR games filter players aggressively. My heart is still with player-killer, hard and nasty gaming, particularly because that style seems to reflect the reality of violence. But the LLM experiment gives that old harshness a new laboratory. It lets the GM be cruel without social cost, and it lets the module explain itself under pressure.