I have no reason to be in Maine. Nothing about the state, or what little I know of it (lobster tails and pointed firs), has ever struck that chord in me which says come visit. But during a recent trip back to Boston, a city I lived in quite happily for a number of years, I was consumed by the desire to go somewhere I had no reason to be. In a town halfway between Kennebunk and Ogunquit called Wells, I booked a three-night stay at an Airbnb called “The Artist Shack.” To get there, I begged use of a friend’s old bike, a Schwinn beach cruiser whose seat was much too low for my legs, and pedaled to North Station to catch the Amtrak Downeaster. Six or seven stops later and I was cycling, slowly, down a two-lane highway in Maine.
I have not been having a good 2026. For reasons I will not get into, I have this year finalized a divorce, retired from most forms of human-to-human contact, bought a house, and forgotten, inside that house, how to sleep. I have also in this time lost all interest in reading or writing—a predictable outcome to being heartbroken and spending too much time with realtors. I used to be an essayist, which meant that I wrote to gain purchase on the thoughts, feelings, and experiences that comprise something approximating the human condition. What I have learned in 2026 is that writing essays can feel rather pointless, like bad therapy or pleasureless masochism, when one finds themselves in poor human condition.
Buying a house was supposed to better my situation, or at least forestall any further devolution by forcing a change of scene. Le Corbusier writes in Towards a New Architecture that a house should be “a machine for living in.” My dream of owning a house was just that: the dream of a new machine, which is usually more achievable than the dream of a new life. But the machine for living in I purchased was old (a boomer-aged bungalow in Idaho with a millennial-aged second-story addition) and riddled with problems: uneven floors, rotting joists, a garage door that wouldn’t close, and window screens clawed to pieces by suburban panthers. My personal problems added to the house’s structural ones. These problems compounded, and eventually, I realized I would have to address the defects in my human encasement—the rind of the sausage if you will, usually procured from an animal’s intestinal lining—before trying to fix this icebox I’d bought to shelter my expired links. So I ran away to the East Coast to store my sausage inside machines whose problems were not mine to fix, all the while hoping that exposure to these other machines, these boxes arranged by other hands tied to other minds, would teach me again how to live.
When I arrived for my contactless check-in, I saw that the Artist Shack was really a twofer: a pair of cedar-shingled cabins on a high deck jutting out into the marsh. I opened a tall metal gate and crossed a short bridge to the deck. On the deck rail’s other side, blue water curved through dense beds of cordgrass. I smelled the sulfurous scent typical of salt marshes. I looked at the cobwebs hoarding golden light into every gap in the deck. Sweaty from the ride and still frazzled from my recent encounter with Amtrak’s byzantine bike-storage system, I took a moment to collect myself, then entered the larger of the two cabins.
That was a day ago, and I have not changed my underwear nor left this artificial island since. I’ve done a bit of writing, but mostly I’ve lain flat on the futon, thinking about my house and this house and Le Corbusier’s precious idea that humans must make machines suitable for living in: that a machine, in other words, must augment our humanness rather than take from it.
Whoever outfitted this particular machine thankfully avoided the beige-and-white palette customarily inflicted on coastal Airbnbs. The walls, floor, and ceiling are uniformly paneled in honey-toned wood, with simple but sturdy furnishings anchoring the room and a minimum of tchotchkes. Five large windows aerate the space, and a sliding glass door leads back out onto the deck with its south-facing view of the marsh; on the deck’s far side is another single-room building which the owners have set up as a bedroom. I like this bifurcated design, however impractical it must be for northern winters. A house so divided implies a clear order to the execution of vital functions: one building for sleeping, the other for cooking, eating, pissing, shitting, cleaning oneself, or lying draped across pieces of furniture. Instead of a closed hallway between the rooms, one gets the plein air vibes of the marsh. The house’s openness—its integration of inside and outside, of domicile and nature—is quite clearly the space’s defining feature. Not a house by the water, but on and of the water, an alfresco fortress.
_____
Over the past decade, the Korean American director Kogonada has made three very different films exploring the relationship between humans and machines. (A fourth, Zi, premiered at Sundance in January and will be released this fall.) His first film was a quiet slice-of-life drama called Columbus (2017), his second a sci-fi inflected elegy for a family’s “techno sapien” companion called After Yang (2021), his third a rather forgettable foray into fantasy-romance called A Big Bold Beautiful Journey (2025). All three employ the dreamy, meditative style one expects from an arthouse director, and none of them are especially dependent on traditional storytelling techniques like plot or character. What are these films about, then, and what makes them feel so utterly human to me, despite Kogonada’s throughgoing fixation with technology and artifice? I have been asking myself that question all summer.
I have a working theory about Kogonada. Basically, I think he’s an essayist, one of these people who’s programmed from birth to find cosmic significance in what he calls “familiar moments.” Essayists like to dwell in the perpetual present of James Agee’s “summer evening,” which is all summer evenings; their creations wrest insight from that “brief crack of light between two eternities of darkness” that Nabokov once metaphorized as “existence.” If Kogonada’s films are “about” anything, it’s what most essays wind up being about: the fleeting passage of life into nonlife and what shreds of meaning we glean in the interim.
This essayistic impulse to reflect on passing images and moments reaches its apotheosis in Columbus. The film follows two people who briefly cross paths in Columbus, Indiana, a town improbably stocked with the brainchildren of famous architects, from I. M. Pei to Eliel Saarinen. A thirty-something translator named Jin (played with somber reserve by John Cho) is summoned from Seoul to Columbus after his estranged father falls into a coma. There he meets Casey (Haley Lu Richardson), one of those prodigiously intelligent and casually pretty young women who, in the imaginations of male directors at least, are always smoking on the sidewalks of provincial towns, waiting to engage itinerant men in philosophic debate. The film unfolds through scenes of Jin and Casey walking around Columbus, conversing, with frequent cuts to serene images of modernist architecture bathed in sacralizing light. Casey knows more than the average townie about Columbus’s vaunted buildings. Jin, the son of an architectural theorist, says he doesn’t give a whit about buildings: “You grow up around something and it feels like nothing.” But what Casey grew up around doesn’t feel like “nothing” to her, and after she shows Jin some of her favorite buildings in town, he comes to care, in a fatherly way, for both Casey and the place that forged her.
Aside from a brief exchange about phones (Casey bought a dumb phone before it became trendy to do so), Columbus doesn’t directly comment on humanity’s fraught entanglements with technology. And yet the subtext is there, planted like a thematic easter egg for Kogonada’s subsequent features. In an early scene at the public library where Casey works, she listens to her co-worker pontificate on the troublingly short attention spans of today’s digital natives. The problem is not actually our capacity for attention, he tells Casey, but rather our distressing inability to find much in the world worth paying attention to. In other words, the real issue is affective: a “crisis of interest” and not a decline in the mind’s ability to attend. The irony here is that this lecture is being delivered to someone who lacks neither attention span nor interest. Casey spends her days keeping tabs on her mother, a recovering addict; attending public lectures; and thinking about architecture. She is never doomscrolling in bed or trying to “influence” her peers. She is well and truly taken by her interests.
People in my line of work, teachers of writing and literature, have lately started imagining a time when the Caseys of the world have grown scarce or nonexistent, and when the cultivation of our archaic interests has been handed over en masse to LLMs controlled by amoral corporations. With each fevered funding round for OpenAI or Anthropic, we wonder more what will become of us. Of our jobs, for one, but also of our humanity when the things that once made us feel distinctively human—the writing of books or the expression of complex thought—have migrated one rung down the periodic table from carbon into silicon. Watching Columbus, the Homo-chauvinist in me wishes to take its heroine firmly in hand and spirit her away to a data-less island where her intellectual curiosity will be protected from the vampiric love of crypto bros housetrained by AI companions.
When OpenAI debuted ChatGPT in 2022, I put my head in the sand and pretended not to notice. I was at the time enrolled in a humanities PhD program, and within my little pocket of academia, where graduate students devoted years to reading reams of scholarly texts as a matter of course—texts an LLM can scan and summarize faster than any human—AI refusal became the default law of the land. Many factors supported this refusenik stance, and more are emerging every day. The World Economic Forum’s latest white paper on AI predicts multiple scenarios in which AI has replaced or “hollowed out” large swaths of the labor market by 2030, and AI companies have already systematically pillaged our reservoirs of intellectual property, exacerbated environmental carnage with their energy-guzzling data centers, and introduced AI-directed autonomous weapons systems and AI-enabled surveillance to already dire geopolitical conflagrations. According to the AI boosters, whose tried-and-true business model is to spin profits out of hype, the refuseniks’ concerns are hyperbolic. We are told to take a deep breath, or two, to remember that all transformative technologies tend to face a period of hysterical opposition before the doubters find that their fears were unfounded. In any case, the exponential increase in computing power promised by AI will soon enough solve climate change, cure cancer, and liberate us all from the trappings of labor, of struggle, thus canceling out any outstanding complaints. The proper response to this, I believe, is vulgarity: I think you people are fucking stupid.
Unlike Casey, I do own a smart phone, and when I know where to find it, my iPhone does not significantly diminish my quality of life. But AI is not simply an upgraded version of older tech (unless human intelligence is AI’s “dumb phone” antecedent); nor is it comparable to something like social media, which though omnipresent, is only really good for a few, mostly unserious, applications (e.g., sharing heavily filtered images and owning the libs). The historical analog that tech writers like to reach for is the invention of nuclear weapons, or less ominously, the integration of the modern electrical grid. As with electricity, there are and will be countless use cases for AI, and some of them will markedly improve our lives, making previously unthinkable feats possible for the overworked or differently abled. For every air fryer and non-kerosene lamp, however, there will also be AI versions of the electric chair, the Cybertruck, and those automated litter boxes which sometimes guillotine kittens.
The clearest indication to me of AI’s arrival is the extent to which these technologies have already begun warping how we relate to ourselves and each other. Psychologists researching the psychic effects of indiscriminate AI use have shown how totally the burden of having thoughts can be transferred from a person’s mind into the tantalum cells of an NVIDIA chip, what Steven D. Shaw and Gideon Nave refer to as “cognitive surrender.” For Shaw and Nave, those who surrender to AI’s algorithmic wisdom quickly proceed past the waypoint of “cognitive offloading, which involves strategically outsourcing a discrete task to an external tool (e.g., using a calculator).” What the surrenderer instead attains—indeed, what they perhaps desire—is “a deeper abdication of critical evaluation, where the user relinquishes cognitive control and adopts the AI’s judgment as their own.” Writers ensconced within the academy have aired similar concerns whilst surveying their students’ end-of-semester portfolios. “One of the real challenges here is the way that A.I. undermines the human value of attention, and the individuality that flows from that,” warns Meghan O’Rourke. “What we stand to lose is not just a skill but a mode of being: the pleasure of invention, the felt life of the mind at work.”
What exactly does “the felt life of the mind at work” feel like? No one can answer this question objectively, but I’d wager what it doesn’t feel like is shooting a stream of prompts into ChatGPT and upcycling whatever comes out. Cognitive surrender to AI surely bears affective costs—Kogonada’s “crisis of interest” is just one of them—and droves of people have already started asking chatbots to perform high-end emotional labor, sometimes with disquieting, even deadly, effects. Of course, there are also less cerebral feelings than the expending of mental effort, feelings like being in love, or falling out of it, like torment and joy and overwhelming fatigue. No company has yet released an AI program that can convincingly feel such feelings on our behalf, but it stands to reason that as AI continues to take our thoughts from us, it may next decide to come for our feelings, our perceptions, or even our capacity for having an emotionally rich, which isn’t to say good, experience. What I worry about—what keeps me up in this time when I am trying to feel more connected to the flow states of my present as opposed to the fractures in my past—is not just cognitive surrender, but a retreat from the field of charged and embodied living.
_____
I’m back on the futon, back in my recumbent position, but at least now I’m recumbently chatting on the phone with my archaeologist friend Abby. She is filling me in on the state of her novel as well as a recent spate of home renovation projects. I am crowing to her about the plumber who just stopped by to rearrange my pipes: how nice it was to have another body beside my own, how ably that other body gave and received pleasure.
Abby and I met years ago in an MFA program, critiquing each other’s essays; now her days involve dealing with even older and deader things than pretty words crossing a page. She’s just returned from a trip to Albuquerque, where her brother had just successfully put on a conference for coders working in a computer language whose name she can’t remember. Every single presentation mentioned AI in either a nonplussed or optimistic tone, Abby reports, and we chew on this for a few minutes, affirming each other’s AI-cynicism while also trying to ascertain what we might lose by staying home from the fair. I want to tell Abby I’ve been reading books by people like Meghan O’Gieblyn and Jeff Sebo of late, people working at the interface of AI discourse, religion, and philosophy, and that these books have taught me this isn’t a new horizon we’re breaching, but rather the reintroduction of an older, nonsecular frame of mind. For long spans of our species’ history, humans have gladly put their eggs in the basket of a supernatural power whose intelligence supposedly exceeded our own (i.e., God), and even we apostates, through our boundless faith in scientific inquiry, have come to see the world as composed of sentient forces which, like AI, aren’t even remotely human. But I don’t want to talk about posthumans past, present, or future. I want to speak of the human condition, which Abby knows exactly as much about as I do; of Jin and Casey and their nimbus of shared interest; of “familiar moments” that recall us to the texture of our lives; and of such variegated human concerns as movies, divorces, sleeping with men, this well-kept cabin in Maine.
So, for the next three hours, that’s what we talk about: smart stuff and dumb stuff and the stuff we have not yet categorized as anything. Abby, like me, displays as many emotional talents as limitations. She is kind, perceptive, a bit brusque, tough when it’s warranted but naïve when she needs to be. We became friends on account of these traits, but we stayed in touch on account of something outside us both: the times we spent together, our shared experiences of traveling, conversating, and occasionally losing our heads in each other’s presence, all of it knit together by the essayist’s compulsion to lift meaning out of life and thus make it worth living.
Kogonada’s films are also built on this intricate emotional architecture of shared experience—moments we cocreate with others, moments that enclose more than one thinking and perceiving body. His second film, After Yang, asks whether a human can ever share their experiences with a synthetic consciousness, in this case grounded in the form of an Asiatic robot named Yang (Justin H. Min). Yang lies dead for much of the film, or at least unresponsive: a pore-less bundle of flesh that his father/owner Jake (Colin Farrell) must cart around a retrofuturistic city in hopes someone can fix the robot’s “core malfunction.” Jake needs Yang to wake up because he loves and misses Yang, but also because Yang is Jake’s adoptive Chinese daughter Mika’s older sibling, as well as the family’s unpaid au pair. Jake and his wife, Kyra (Jodie Turner-Smith), initially acquired Yang, a “cultural tech,” so he could teach Mika about her Chinese heritage. Without access to Yang’s well of “Chinese facts,” Jake worries that this multiracial human family—white dad, Black mom, Asian child—will drift off its transcultural moorings and lose whatever legibility it once possessed.
Although Yang never returns from the cybernetic bardo, the film gives us access to his artificial personality and artificial emotions through a cube-shaped repository of video recordings which a mechanic pries from Yang’s chest. When Jake one night inserts the cube into a specialized reader, Yang’s memoryscape unfurls into the room as a galaxy of ghostly lights, each pinpoint world holding a moment that Yang decided, for whatever reason, to keep. Jake moves randomly through these moments, occasionally asking the reader to “repeat” the snippets that seem especially meaningful to him, like he is trying to establish the right room tone for the film of Yang’s existence. Of course all Jake gets are fragments, truncated scenes showing a tree with one branch hewn off, Jake and Kyra laughing and cuddling on the couch, Yang calling Mika 妹妹, the Chinese phrase for younger sister.
Is humanness defined by the ability to execute certain functions, to tee up bits of knowledge like Yang’s “Chinese facts” or Casey’s architectural trivia, and to feign understanding of them? Or is what makes us human the ability to experience (not in 4K but in analog), to earnestly emote, to have memories and to assign talismanic significance to some of them? Kogonada’s films collectively suggest that it’s having an embodied, feelings-forward connection to the experiences we share with others, not abstract knowledge, that guarantees the difference between what is human and what is not. This echoes the computer scientist Melanie Mitchell’s assertion that “unlike LLMs we are embodied creatures with a sense of self, a sense of others, and (at least for most of us) a profound caring about the consequences of our actions.” But in the world of After Yang, AI beings also have bodies that die and decay, bodies that remember, bodies that care. We know from autopsying Yang’s memories that he possessed both sense of self and sense of others; we watch him watching himself in the mirror, but also observing, with tender care, a human clone named Ada whom Jake thinks of as Yang’s girlfriend. Like many humans, Yang also had questions about his identity, what constituted his selfhood as a techno. At one point, Jake asks Ada, “Did Yang ever struggle with being a techno? That is, did he ever want to be human?” Ada quickly bats the question away, then recalls that Yang did at times equivocate—not over his humanness, but his preassigned race. “What makes someone Asian? He’d ask that all the time.”
One could interpret this scene as a dab of racial essentialism: Yang may look Chinese and know all there is to know about China, but he has no notion of what it is to feel Chinese, because he has never been to China to experience an authentically Chinese way of being, and because he is a robot whose “race” is fabricated, not lived. After Yang may invite this line of thinking, but it does not really support it. I hear in Yang’s question a genuine note of befuddlement more than envy of those of us who move through the world beset by what Cathy Park Hong dissects as Asian people’s “minor feelings.” To Yang and maybe Kogonada, the experience of being Asian is indistinguishable from that of being human, an ontology Yang has no trouble embodying. In fact, the film implies that Yang is better versed in what it is to be human than either of his listless and distracted parents. Jake doesn’t know how to make friends, talk to his wife, or share meaningful moments with his daughter. Only after reviewing Yang’s memories, his proxy humanness, is Jake returned to himself in a slow excretion of tears.
The film frames Yang as a relational prosthesis for humans in poor human condition. He came into the family fold to help these humans process complicated cultural dynamics, to help them raise each other and feel the emotional weight of each passing moment. Even Kyra, who early in the film expresses her qualms about installing a techno at the family’s emotional center, is moved to wordlessness by Yang’s aftereffects: the butterflies he collected, the attention he lavished upon trees. By fritzing out but leaving a record of his memories behind, Yang delivers the family to a painfully shared experience. That experience is of grief. His memories fade from the room, but the room’s human components keep on whirring, now without AI assistance. Kogonada’s depiction of a household in which a synthetic being is tasked with managing our most intimate relations demonstrates that what we stand to lose in the age of AI is not just our facility for thinking, but the muscle memory for feeling alive in all of that word’s worldly valences.
_____
This cabin and its environs have turned into an organ for sensing my all-too-human emotions. I eat bowls of mac and cheese over the soapstone sink, weeping, then spend hours at the windows, watching the marsh’s changeful expanse. The water is up, the water is down. A lone egret stalks the shallows, beady eyes peeled for silver streaks of fish.
I know I’ve been living for some time in the tropey valley of a contemporary divorce narrative—giving away and then claiming property, setting out on a trip, indulging in casual sex with comely strangers. If I follow this story to its likely conclusion, I will leave this place with a new lease on life, delighting in my postmarital independence. Unfortunately, being partnered for so long has spoiled the easy pleasure I once took in solitary travel. My ex-spouse was the kind of person who would’ve thought to bring more than a water bottle and two Zyn canisters to this cabin; who would’ve biked down to the beach as I wrote and told me all about it, but not enough to make me jealous; who would’ve understood why I found it so romantic that the wildlife refuge north of Wells is named for Rachel Carson—Carson who once wrote to Dorothy Freeman, the love of her life and a Mainer, “There must be long days when we can go off together alone. There must be many opportunities for us to do together the things we love to do.”
My ex and I were practiced, you see, at sharing an experience.
The weather inside this unshared experience has taken a turn. It rains for several hours, and I let the sills collect puddles because I can’t bear shutting the windows. Late-nineties issues of Architectural Digest pile up on the coffee table, the spreads extolling the virtues of a minimalist interior. I imagine myself inside those interiors, wiping down all that gleams, vacuuming up my crumbs, knowing that none of this will move unless I move it. I sit at the dining table reading until the weather turns again.
Sunset comes calmly to the marsh, the rain having polished down every roughness. I take a brief break from the cabin, walking up the road I biked in on. All the houses on the road, I notice, are flying American flags (the Artist Shack flies an Italian flag alongside the Stars and Stripes). A few cars pass by filled with beachgoers returning to town for dinner. Across the road, there is yet more marsh. This is the scene in which I must try and bury my sadness.
_____
I didn’t want to write much, if anything, about Kogonada’s A Big Bold Beautiful Journey (the film has already been savaged by other critics), but it happens to be the entry on his IMDb page most germane to my single and mopey status. Much like After Yang, Kogonada’s most recent film (his first with a major studio) features a machine inserting itself into human affairs of the heart. This time the AI is a dulcet-toned GPS system implanted in the dashboard of a Saturn sedan, while the humans on board are two road-tripping singles named David and Sarah (Farrell and Margot Robbie). The AI tour director is bossier in its meddling than Yang was, though this may only be in comparison to the astounding internal passivity of David and Sarah, who each spend the bulk of the film vying for the emotional equivalent of passenger prince or princess. Neither knows what they want from a partner and neither has anything interesting to share about why. They are more akin to Hinge profiles with opposable thumbs than flesh and blood people on a life-changing first date.
Upon viewing the film a second time, I test drove the idea that the protagonists’ lack of self-awareness might be part of a ploy to show how shallow and etiolated romantic adventurism can feel when lovers cede navigational control to AI. But Journey is too earnest a film to sustain such a reading. We follow along as the AI organizes a fantastical sojourn through David and Sarah’s respective personal histories. At each “stop” on the trip, a different magical door appears in the landscape, transporting the characters to an immersive recreation of a moment from one of their pasts. The GPS seems to use loneliness and regret as its only selection criteria for which moments the travelers must rehash, as if memories of alienation and ennui are always the moments that bear the most pedagogical weight. Perched at the top of a lighthouse on the Atlantic—the first, tone-setting stop on this trip—David and Sarah chat about the inherent sadness of traveling solo:
david: It’s funny how the most beautiful places make you feel the most alone, you know. No matter where you go, you always end up going to the top of something and looking down on something. Most places have a thing like that, you know, you go to the top and you . . . you look out and you look down and you’re supposed to have an epiphany about who you are and where you are, about your life and where it’s headed.
sarah: Well, do you know where it’s headed.
david: No Sarah, I do not.
sarah: Did you have an epiphany the last time you were here?
david: No, I felt nothing.
sarah: So you didn’t take this all in
david: Not even close.
sarah: What about now?
The lesson here is self-evident. Single David is too meta, too in his head; David just needs someone else there, someone like Sarah, to share this view with him and to convince him that his epiphany is right here: love and connection with another human. Recent history biases me, but I want to say there is a wide spectrum of experience between feeling “nothing” and feeling like you’ve received, out of thin air or in the avatar of your beloved, the cheat codes to “your life and where its headed.” Only an AI, or someone without any interest in the ambiguities of lived experience, could think and feel in such an all or nothing way.
Like After Yang’s constellation of memories, these portals to the past are meant to map the interior lives of a dubiously human agent, except this time the memories are all bad, yet somehow vacated of emotional stakes. Because of the form, we can safely assume that David and Sarah will end up together, making all this backstory a distraction from the potentially amusing spectacle of two attractive people trying to make a go of it in the present. That’s not to say the film fully gives up on developing David and Sarah’s IRL chemistry. Like any other Kogonada joint, Journey is committed to showing us the “familiar moments” that fill out its characters’ shared experience. Moments like singing in the car, and eating a “fast food cheeseburger,” and fake fighting before real fighting, and buying each other stuffed animals to remind each other what it was like to be a child (moments I also participated in with my ex). Embarking on a new relationship, this film suggests, means redoing all these ordinary moments alongside another person. The problem is that this film’s mundanities are not nearly as gracefully rendered or poignantly poetic as the ones that enliven both Columbus and After Yang. It’s like the film’s writers asked AI to come up with a set of ordinary moments that would appeal to the largest audience, resulting in a treacly playlist of ordinary life’s greatest hits. One may choose to hear these bops over and over again, of course, but in actuality, the best moments are one-hit wonders.
Consider the following: several years ago, while visiting Bloomington for research, I took my mother with me on a quick side quest to Columbus. We attempted to do the Jin and Casey Tour, starting at I. M. Pei’s public library with its concrete ice tray ceilings, then wending our way past the Columbus Mental Health Hospital (a structure built over water that Jin finds “healing”), and finally arriving at Mill Race Park. On maps of Columbus, the Flatrock River forms a cupid’s bow around the park. My mother and I walked the bow’s curve until we got to a train bridge transecting the river. At the bridge’s base, someone had spray painted the question will you marry me? inside a heart stencil, reminding me to inform my mother that my partner and I were planning to legally conjoin ourselves that May, that my own years of lonesome wanderlust were coming, at least on paper, to an end.
“Nothing fancy,” I told her. “Just in and out of the courthouse.”
“Good,” she responded, avoiding eye contact. Her avoidance told me she meant it.
_____
Le Corbusier once mused that the view from any house is “a categorical view, without connection with the ground.” I don’t really understand this statement, though it seems salient to mention while writing about a cabin suspended over water, barely tethered to Maine. The idea that each house, each piece of architecture, is a machine for living in was more than a metaphor for Le Corbusier. He took form and function very seriously, saw architects as machinists whose edifices, by streamlining and diagramming space, could help purify the messy workings of existence.
The critic Beatriz Colomina interprets Le Corbusier’s thinking in optical terms: “The house collects views and, in doing so, classifies them. The house is a system for taking pictures.” Houses are like cameras then, mechanical boxes which collect and collate our shared and solitary moments. But what we bring into a house cannot always be cleansed and blotted of its mess, and not every house is optimized toward careful, purposeful living. Some houses, like this one, are open-ended. I step outside onto the deck, and the rain comes down to meet me.
If every shelter in the storm of one’s emotions is essentially a machine for living in, then perhaps it’s okay to find refuge in newer, more intelligent, more uncannily human machines. Refusing AI’s intrusion into my writing, after all, does not mean these beings have left me to my own devices. On the way up to Maine, I’d read articles about the state legislature passing an eighteen-month moratorium on data center projects, a measure that governor Janet Mills then vetoed; I’d searched for restaurant recommendations on Google, accepting the cheery advice of Gemini AI on what “local favorites” to patronize or skip; I’d been harassed by the floating banners on my Grindr grid, all imploring me to try out the app’s new “gAI” service for helping lonely twinks generate interest. My issue is that I do not find these machines as trustworthy, or even as interesting, as a house. I cannot interact with them in the ways that I crave interacting—through frictional touch and palpable unknowing. I cannot be dumb with them, eat piles of fried clams with them, or get sucked into the vortex of their experience. They can’t be the Casey to my Jin, despite certain physiognomic symmetries, and they certainly can’t be my ex.
There is one scene in Columbus I think about all the time—for me, it is the film’s most potently familiar moment. Jin and Casey stand in contemplation before a glass-walled building: the former Irwin Union Bank Building, designed by Eero Saarinen. “What do you know about it?” Jin inquires, and Casey locks into AI-assistant mode, spitting out chat bubbles full of facts about the building and its “radical” design, its status as “one of the first modernist banks in America.” Jin interrupts Casey’s spiel, because he’s not really interested in the building, he’s interested in why it matters to Casey.
“I’m also moved by it,” she admits, and Jin jumps on this unfactual feeling. “Yes, tell me about that. What moves you?”
What moves Casey about the building can’t be expressed in language, or maybe it can, but you’d have to be there with her to catch the drift of her utterance. Kogonada’s lens moves to the other side of the building’s transparent exterior, so that all the viewer can sense of Casey’s answer is the wordless image of a young woman standing in the summer light of Columbus, trying to describe why she holds this particular machine so close to her heart. She doesn’t say much. Instead, she uses her hands to trace a rectangle in the air: an illusory box containing all the feelings she can muster but not explain.
My fellow refuseniks might watch this scene and take heart. It so clearly conveys that words, the métier of LLMs, are not always our only or best tools for channeling emotion or sharing experience. Expert wordsmiths that they are, chatbots already surpass our ability to package and transmit information. They can translate words from English into Mandarin and serve as reliable tour guides for our journeys, but the steam evaporates from their invisible engines when it comes to helping us feel emplaced and connected to our frequently aphasic lives. Casey’s box is a spatial representation of that blessed shortcoming—it is a gesture made against indifference, against both cognitive and experiential surrender.
One reason I’m not so heartened is that some of Silicon Valley’s buzziest new ventures now want to take AI past the veil of language. World Labs, a company founded by Fei-Fei Li, the “Godmother of AI,” is currently working to open a “frontier where [AI] systems construct complete worlds with the physics, logic, and rich detail of our physical reality.” Li’s wager is that the next wave of AI products will be trained not in languages, but in space. These spatially intelligent programs will be able to construct ersatz worlds not unlike the ones David and Sarah traipse through in A Big Bold Beautiful Journey. (Eventually, they might even be tasked with reshaping the physical world that we live in.) World Labs’ first product is called Marble. It is a website where users upload images of a setting, after which the company’s proprietary AI generates a virtual world based on that prompt, like a custom version of Zuckerberg’s defunct Metaverse.
During my last evening in the Artist Shack, I boot up my laptop and ask Marble to make me a world derived from this space. I upload four photos I took and the program quickly reverse engineers a textual “World Guide” based on my images: “The scene is a rustic and natural home interior, capturing a blend of warmth from the wooden elements and the expansive beauty of the marshland outside. The overall tone is serene and inviting, with a strong connection to the natural environment. The room features a white double-door refrigerator on the left, next to a concrete-like countertop with a sink and a stovetop . . .” On and on the tour guide drones.
My world has finished loading. I can now click into it with my mouse, entering the world which Marble built for my pleasure and discovering . . . not much. The windows and wood paneling look about right, as does the view of the marsh, but the space doesn’t feel real in any way—it’s just a heavily filtered image blown out into a panorama which I can swivel my gaze around but not inhabit. Though most of this world’s details eerily mirror the room I’m sitting in, the AI has inexplicably added a vase of flowers and duplicated the dining table, as if the machine were expecting guests.
I understand I’m using an early iteration of World Labs’ spatially intelligent software, but even if it was more than it is, even if it had all the details down pat, I cannot fathom wanting to spend my days inside an AI’s rendition of space, when I have such limited time to be human in the dying world that we have. Some techies think even that world is a simulation, a video game designed by our future selves that all of us are stuck completing. Maybe that’s true, and maybe Marble is the inception-like portal that will lead us as a species to the other side of the mountain. But I have no intention of going there yet, not until AI, or some other off-world presence, has answered at least a few of my questions. Namely, would salt marshes still smell like shit in this world-beyond-our-world? Would futons still add laughter to sex? And would heartbreak really feel good in a place like this, to quote the ye olde AMC commercial? Because it did for me in Maine, crying into my mac and cheese and watching the water. There are so many machines for living already operational on this earth. Some of them are big and some of them are ugly and some of them are perfect.
I am the human inside this perfect-enough box. My feelings, they fill it.
_____
*An essay-review of three films directed by Kogonada: Columbus (Oscilloscope, 2017), After Yang (A24, 2021), and A Big Bold Beautiful Journey (Sony Pictures Releasing, 2025).
