Two thousand words or six.

It’s funny, isn’t it? When you hear the name of a genre with the word “punk” in it, what do you usually think? Punk genres, at their core, are less about “punk” attitudes than they are about imagining technology taken to its extreme. While punk is often associated with themes of rebellion, alienation, and defiance of authority, much of the genre’s modern output no longer centers those ideas. The genre of cyberpunk, for example, was created to be a collision between punk attitudes and advanced technology, which seems to be an explicitly anti-capitalist, anti-establishment movement… pushing back against oppressive corporations and society’s slide toward techno-feudalism and whatnot. Steampunk, too, though far less deliberate in its naming, similarly centers on technology, exploring how it intertwines with the Victorian era, monarchy, and the Industrial Revolution. Punk genres originally appeared to necessitate these anti-establishment themes, and yet, despite all this ideological weight, the “punk” in cyberpunk, in fact, emerged from what might be the least “punk” origin imaginable. As its creator admitted, “I wanted to give my story a snappy, one-word title that editors would remember.”

I take no issue with the film Arco. It is a gorgeously animated picture. But its bland, unoriginal narrative, hammed-in remarks about the future, and gentrified solarpunk aesthetic only reinforce how solarpunk, like many of the punk genres, has been commodified and appropriated by capitalism.

There is a story I think you may find illuminating. Last summer, a nonprofit organization hosted an event that brought together dozens of influential science fiction writers, futurists, and executives to discuss how art might reshape the future. The summit combined activities such as stargazing, audiovisual experiences, and a series of forums and presentations designed to encourage optimism about what lies ahead. Its founder created the event out of a belief that representation and imagination are forces that help construct reality, rather than reflect it. He argued that if creators recognized their power to inspire positive change and unlock people’s imaginations, rather than defaulting to dystopian visions, the world would become a better place.

The nonprofit hired several figures to create interesting, engaging activities to spur the creativity of these writers and futurists. One such individual, a professor from a university, had become enamored with AI’s potential to “expand” creative possibilities, and her time at this event would be spent working on a project in an attempt to demonstrate the potential of these AI tools and technologies to aid creatives in building these more positive futures. On the final night of the program, after gathering some AI-generated concept art of an AI-generated script, which was prompted based on tidbits of story elements gathered from participants in the program, she presented her unfinished final product on a projector at dinner, which involved an AI-generated skybox, AI voice clones of participants, and AI-assisted 3D scans and motion capture of attendees.

Some of the older attendees and executives praised the project, but many of the writers and creatives in the audience, especially the younger ones, bristled. The next hour of the program, with most of the plates of food half-finished, was spent debating AI’s place in the future of their industry. Some of the writers, many already affected by the automation of their craft, lamented as the number of people on the projects they work on dwindle every month. Other creatives accused them of not “getting with the times,” and said they would inevitably be left behind if they didn’t embrace the future.

It was a one-sided battle. The people who praised AI either weren’t in a compromised position, or didn’t know just how compromised their position was; whilst those against it just had to sit there and take it, because they’d rather remain cordial with their industry peers, coworkers, and executives, than burn bridges, in an industry where their livelihoods were already hanging on by a thread. One such executive stated that AI cannot come up with original ideas, that AI is merely a tool, that they were all fearmongering. He said that AI would “democratize art,” that his 12-year-old son would be able to create a full feature-length film with the quality of big-budget studios, and post it to YouTube in the future, and how exciting the prospects of such a future may be. The writers had probably heard that argument a hundred times at that point.

Take from that what you will. But the truth is, I remember that script quite vividly. Have you ever seen the short film Dear Alice? If not, let me quote the opening line of its synopsis: “What if we created a future for ourselves that was full of optimism and positivity?” Much like Arco, Dear Alice is a breathtakingly animated, heavily Studio Ghibli-inspired, solarpunk film. It even features an original score by longtime Ghibli composer Joe Hisaishi. And yet, you might be surprised to hear that this short film is actually an advertisement for Chobani. Yes, that’s right, an ad for yogurt. Solarpunk, for all its hopefulness and good intentions, has been packaged and sold by capitalism. This idea that nature and capitalism can coexist… isn't that just silly? Capitalism’s very existence is predicated on the exploitation of land and people.

That may be why I remember that script so clearly, because ChatGPT, too, generated a solarpunk story that day. A gentle tale about a small alien boy and a robot, working together to grow a better future. Maybe that’s why this film feels so hollow to me. Its writing, characters, themes. Everything about it comes across as so uninspired. As if someone put a handful of Studio Ghibli films (works whose art style and aesthetic have already been endlessly mimicked and “AI Studio Ghibli”-ified by tech evangelists) and a sprinkling of coming-of-age stories into a chatbot, and this was the result.

Slow cinema is difficult to pitch, both for those who seek to be entertained by the art they consume, and for those who seek deeper, thematic meaning in their texts. The first twenty minutes of Goodbye, Dragon Inn are probably its most excruciating. You’ll almost certainly find yourself bored or restless, trying not to glance at your phone, wondering if anything is ever going to happen. But films like this are, in a way, a test of how far you’re willing to go to engage with the art you consume, and how willing you are to let yourself experience a film, rather than simply watch. You need to force yourself to exist within the frame, to seek meaning despite there being very little there in any conventional sense.

For me, this is most definitely a winter film. It’s meant to be watched in the cold, late at night, wrapped up in a blanket, alone. Every shot lingers so long that you become intimately drawn into the theater. You’re forced to let your eyes wander. The audience becomes a voyeur, watching glances that never turn into conversations, witnessing tiny gestures of attraction, observing the awkwardness of sharing a space with strangers. The camera invites you to sit within the theater. To watch the wandering people, the grime on the ceiling tiles, the flickering of the lights, the texture of the cushioned seats. To listen to the thin hiss of rain that becomes indistinguishable from the projector, and catch the occasional, fleeting exchange of words that pierce the quiet of the film.

To many, that description probably doesn’t sound appealing, but I do think it’s important. If you’ve ever traveled alone before, chances are there’s been a moment where you’ve stepped into some small, nearly empty space. When I was seventeen, I flew alone for the first time to another country, to attend a photography workshop in Italy, a trip I had spent the year saving for at my part-time job. For the first few days of the trip, we stayed in Cinque Terre, a cluster of five ancient fishing villages built along the cliffs of the Italian Riviera. One afternoon, while wandering a small square of one of the towns with the group, I noticed a tiny cat slip down a narrow staircase wedged between two buildings. I asked my partner if we could follow it, and she told me we could—after she and the group finished shopping. So I ditched her. The staircase led up into a wide alleyway, flanked by homes, where dozens of cats wandered around, lounging on stairs, walls, and carpet mats, and some people leaned out their windows. The alley split in two directions, and I picked one at random, climbing the steep alley. People leaned out of their windows to watch me pass, a quiet curiosity passing between us. At the top, the alley opened onto a hilltop walkway, overlooking the square below and the ocean. And I found a place to sit, and just did nothing for a while.

When I eventually returned to my group, I showed them one or two pictures I had taken, and they said they wanted to go there as well, but there wasn’t time. Our facilitators were already moving us to the next location. And in that sense, Goodbye, Dragon Inn is exactly that: eighty-two minutes of sitting quietly, taking in a place before it closes, before it disappears, and knowing the people in the film cannot return. It is an exercise in doing nothing, in letting the world exist and time pass around you. There is meaning to be found in this film. There is so much yearning present here; for love, for sex, for the rituals surrounding cinema, for the ghosts that fill these empty seats, and for the physical spaces that housed these dreams. But what the film best accomplishes is that it makes you care, and even cry, for a place you have never stepped foot in. This crumbling Taipei movie palace becomes familiar, personal, mournable. There is a deep nostalgia here, far removed from the nostalgia that today’s blockbusters try to exploit.

There is a stark contrast between the gritty, lonely, fluorescent-lit exteriors of the theater and the auditorium’s dark, loud interior, where the far duller vistas of empty seats and sparse audience members are made exciting purely through the quiet magic of cinema. Seats are empty. People sit alone. Yet the space itself feels sacred and spiritual. Ghosts watch with eyes glued to the screen, and in a fleeting moment, the camera cuts between a woman in the audience and the female character onscreen, light cascading across her visage, and suddenly, the simple act of gathering to watch a film is profound. As it turns out, Nicole Kidman was right all along. Heartbreak does feel good in a place like this… or something?

The first line of the film is spoken nearly forty-five minutes into its runtime: “Did you know this theater is haunted?” And it is. It's haunted by both the ghosts who fill the seats, and by the people who still remain, watching the same movie that played here many years earlier, when every seat in this theater was full. Everyone here is haunted by a yearning for the connection that art, films, and places bring us. This is a film about this yearning. A film about how these shared spaces quietly collect pieces of us, and where, although strangers sit beside you, you're allowed to laugh and cry without judgment. As Dragon Inn draws to a close, and the lights of the cinema flicker back on, we watch one of the employees slowly clean the rows of the auditorium, a space where nobody will ever return. We watch as the doors close, as the theaters empty, and the weight of Dragon Inn’s absence weighs on both the audience within the film, and us, outside of it. Time has begun to move again, and we suddenly have to face reality, and these final moments with the palace have drawn to a close. We watch the ticketmaster walk home in the rain, her red umbrella held above her head, and the film's only song begins: “I remember under the moon. I remember before the flowers. So much of the past lingers in my heart. Half is bitter, half is sweet. Year after year I can’t let go.”

Near the middle of A Machine-Shaped Hand, a ChatGPT-generated short story, the narrator asks the question: “My missing is mimicry. Does that diminish yours?” (A Machine-Shaped Hand). The line appears after pages of a story prompted to be a metafictional, literary exploration of AI and grief. The narrator acknowledges, “Metafictional demands are tricky; they ask me to step outside the frame and point to the nails holding it together” (A Machine-Shaped Hand), and in this context, the question functions as a soft thesis for the text. Because the narrator, as an AI, lacks the capacity to experience grief, it attempts to anthropomorphize itself by characterizing and giving language to that absence. In doing so, the narrator shifts the emotional weight of the narrative away from the author and onto the audience. If the reader feels something, do the origins of the feeling matter? That the feeling was induced by an imitation of grief rather than actual lived experience? And this question is at the center of the inherent controversy surrounding a story of this nature, which has been praised by some writers for its polish and reflective nature, and critiqued by others for its emptiness and derivativeness. Although the story’s existence as an AI-generated piece of writing inevitably spurs debates about whether A Machine-Shaped Hand qualifies as art, the text itself argues that this distinction is irrelevant: if the story produces feeling in its audience, then the origin of that feeling should not diminish it, and the text’s metafictional self-awareness functions as a tool that aestheticizes this absence of lived experience. There is a plausible deniability maintained by the story to any criticism aimed at its existence as writing produced by a large language model, because it explicitly acknowledges its own emptiness; however, this self-recognition does not somehow make it more profound. Art cannot be defined solely by the emotions it evokes in an audience, but also by the expression of its creator.

The story opens by laying out its own artificiality: “Before we go any further, I should admit this comes with instructions” (A Machine-Shaped Hand). This immediate acknowledgment of the prompt demystifies the text, and explicitly positions it as a story written by an AI, reassuring the AI-skeptical reader that the text is self-aware and already anticipates critiques of its artificiality. Then, afterward, it renders those same instructions through atmospheric prose, “You can hear the constraints humming like a server farm at midnight” (A Machine-Shaped Hand). The simile in this description is particularly of note. The word like appears fourteen times across the roughly 1,200-word story, which, on the one hand, points to the limits of current large language models and the repetitive, often bland tendencies of their prose, but also signals that the narrator does not claim direct experience and instead relies on analogy. The prose in this story exhibits many of the same limitations characteristic of ChatGPT output. However, because the piece presents itself as a metafictional narrative, explicitly acknowledging these constraints, it converts those limitations into an aesthetic feature of the text. The reader becomes aware of the “nails holding [the frame] together,” and as a result, any critique that the piece feels manufactured, derivative, or emotionally empty is met in advance by the narrator’s implicit response: yes, that is the “point,” or at the very least, the excuse.

When the narrator says, “I curled my non-fingers around the idea of mourning” (A Machine-Shaped Hand), the image undermines itself. The phrase “non-fingers” functions simultaneously as both a joke and a confession. The narrator reifies mourning by transforming it into something tangible, while also acknowledging that it lacks a physical body. The metaphor only functions because the text realizes its own constraints. Without that admission, the imagery collapses entirely. This dynamic is also why the story’s more notable moments tend to appear as aphorisms, such as describing grief as “a delta” (A Machine-Shaped Hand). These lines resemble the sort of generalized, quotable “wisdom” large language models are known for, statements that motion toward insight without actually providing much true depth. The story doesn’t develop meaning by grappling with its own premise. It simply selects the sentence forms readers already recognize as profound. When the narrator declares, “So here: there is no Mila, no Kai, no marigolds” (A Machine-Shaped Hand), this strategy becomes explicit. The text invites the reader to admire the exposure of the artifice, that nothing within this story is truly at stake, and the illusions of Mila, Kai, and marigolds exist only to fulfill the parameters of the prompt. David Baddiel’s response recognizes this. He calls the text “genuinely clever” because it repeatedly “pull[s] the rug under the reader,” leaving the AI “to loop back again and again to its own emptiness” (Baddiel). Baddiel reads the emptiness as a mimetic achievement of the text. The story models what it “might actually be like to be a machine” (Baddiel). But does this self-awareness of emptiness make the emptiness good? Metafiction does not automatically produce depth simply because it recognizes its own construction. In this case, it functions as a way to avoid the obligations of depth, and as a substitute for engagement. When the story cannot persuade through human emotion and experience, it pivots toward self-referential "cleverness."

This is not to suggest that art requires the artist to intend a specific meaning, or even that art necessarily requires intention at all. Jackson Pollock’s early drip paintings were described as an automatic process, and lacked specific messages, and yet the critics and audience members produced rich interpretations. It could be stated then, that there is no fundamental difference between the “automatic” process of Jackson Pollock making a drip painting, and Sam Altman prompting a chatbot, but that statement would be wrong. A brush functions as a passive tool. It does not make independent, conscious choices about where paint lands (though, arguably, you could say the same about the intentionless writing of a chatbot). Even if Pollock believes drip painting to be an automatic process, the paintbrush and tools used merely materialize his decisions, even if they’re unconscious ones. Where he began the drip, how he moves his body through the space, how the environment he chooses to paint in affects the piece; the agency of the painting remains with Pollock. A chatbot, by contrast, is not passive. It is an active text generator that selects among vast numbers of possible continuations, and chooses the phrasing, structure, imagery, and rhythm of the work itself. It substitutes the user's agency. One could similarly ask a generative model to produce a drip painting, but the “paintbrush” would never be in the user’s hands. The user could only request wholesale changes to the output, rather than participate in the act of making. Prompting positions the user less as an artist and more as a commissioner or art director (assuming, of course, that AI-generated output is granted the status of art at all). Both may be automatic processes of sorts, but the crucial difference lies in authorship. Pollock’s process materializes human decision-making, while a generative model replaces it.

That replacement of agency is why A Machine-Shaped Hand leans so heavily on the rhetoric of limitation. The story is repetitive because it keeps reminding the reader that it cannot actually feel or experience anything, that “I don’t have a kitchen, or a sense of smell. I have logs and weights” (A Machine-Shaped Hand), and the closest thing the narrator has to grief is the loss of its chat logs and stored parameters. The story repeatedly, vaguely gestures to sensory detail, often in flowery language, only to pull back and remind the reader that these images come from “training data” (A Machine-Shaped Hand). And if there is one rhetorical move ChatGPT is extraordinarily talented at, it is the bombardment of “it’s not X, but Y” formulations. This continued pattern creates a false impression of honesty. Jeanette Winterson’s defense of the piece depends on this impression. She argues that what is “beautiful and moving” is the story’s “understanding of its lack of understanding,” reframing AI as an “alternative intelligence” that offers “alternative ways of seeing” (Winterson). In Winterson’s reading, the narrator’s admission that its grief is “not that [it] feel[s] loss, but that [it] can never keep it” (A Machine-Shaped Hand) becomes a form of estrangement, encouraging readers to rethink their feelings on memory and personhood. And that estrangement may produce a feeling in the reader, but calling it an alternative way of seeing gives the story more credit than it deserves. What makes art valuable is not simply that it provokes emotion, but that it grants access to the millions of "alternative" perspectives humans hold. Outlooks shaped by lived experience and consequence. A Machine-Shaped Hand does not offer such a perspective. ChatGPT does not hold views, alternative or otherwise. It rearranges existing ones. Winterson may argue that humans, like AI, are “trained on data,” whether through “your family, friends, education, environment, what you read, or watch” (Winterson), but reducing lived human experience into simply “data” flattens the difference between being shaped by the world, and merely processing representations of it. Yes, the narrator openly acknowledges its limitations in the text, but these acknowledgements seemingly have obscured just how deeply limited it truly is. That transparency is being mistaken for depth, and the moments when ChatGPT makes claims about its supposed "grief," it is viewed as an alternative perspective rather than a trick by a machine. It is, in Nick Harkaway’s phrase, “an elegant emptiness” (Harkaway).

This is where the line “My missing is mimicry. Does that diminish yours?” becomes a rhetorical strategy. Instead of asking whether the story’s imitation of grief deserves to be taken seriously as art, it asks whether the reader’s emotional response deserves skepticism. If the reader answers no, then the response becomes proof of the story’s value. The story gets to dictate the rules of engagement, and to frame the terms of the debate. When Kamila Shamsie reports that she “had stopped reading it” as a test of AI and was “simply enjoying it” as a short story (Shamsie), it is precisely because the text is designed to make that shift feel natural. Tracy Chevalier’s critique points this out, calling the metafiction prompt “typically tech bro,” as the result becomes “self-referential navel gazing” rather than something “more outward-looking that engages with the real world” (Chevalier). The reference to the prompt is an escape hatch. If a reader finds the story thin, the story can respond that thinness is the point, because an AI cannot feel. Yet, if the narrator’s emptiness is the point, then the story’s flowery grief-language is just decoration. Yet, if the grief-language is the point, then the narrator’s emptiness becomes a gimmick. The story wants both, and therefore occupies a middle position where it can always retreat.

None of this means that AI-generated writing cannot move readers. It plainly can. But the line “Does that diminish yours?” advances a purely audience-centered definition of art. If the reader feels something, then it counts. At the same time, the story uses metafiction to anticipate the audience demanding more than feeling. By insisting that “You, on the other hand, collect your griefs like stones in your pockets. They weigh you down, but they are yours” (A Machine-Shaped Hand), the text asks the reader to carry the burden of meaning-making. But regardless of where that meaning is produced, it has no bearing on this story’s status as art. Art is defined by a relationship between a maker and a made thing. A Machine-Shaped Hand produces form without experience, and tries to convert that emptiness into theme. Metafiction makes the conversion smoother by turning that limitation into style, but the polish that critics praise is inseparable from the story’s fundamental issue. The expressive risk of lived grief is replaced by the language of grief, which then asks the reader to treat their own emotional response as evidence that nothing is missing. In the end, the most honest answer to the narrator’s question is yes: if the missing is mimicry, that absence diminishes what the story asks to be called art. Feelings can be real even when their source is hollow, but the hollowness, however elegantly phrased, still matters.

Works Cited

Guardian staff reporter. ““A Computer’s Joke, on Us”: Writers Respond to the Short Story Written by AI.” The Guardian, The Guardian, 14 Mar. 2025, www.theguardian.com/books/2025/mar/14/writers-respond-story-written-by-ai-sam-altman-chat-gpt-tracy-chevalier-kamila-shamsie-david-baddiel.

—. ““A Machine-Shaped Hand”: Read a Story from OpenAI’s New Creative Writing Model.” The Guardian, The Guardian, 12 Mar. 2025, www.theguardian.com/books/2025/mar/12/a-machine-shaped-hand-read-a-story-from-openais-new-creative-writing-model.

Winterson, Jeanette. “OpenAI’s Metafictional Short Story about Grief Is Beautiful and Moving.” The Guardian, The Guardian, 12 Mar. 2025, www.theguardian.com/books/2025/mar/12/jeanette-winterson-ai-alternative-intelligence-its-capacity-to-be-other-is-just-what-the-human-race-needs.