Sean Voisen

Programming in the swamp

2026-02-12T00:00:00Z

The grand prophecy of agentic programming goes something like this: in the not-so-distant future, we puny humans—limited as we are by our tiny working memory and limited cognitive abilities—will no longer build software “by hand.” Programmers will no longer write code as we once did in days of yore, keystroke by painstaking keystroke. Instead, we will orchestrate the work of software development, which will be carried out by willing hordes of semi-autonomous AI coding agents. We will write detailed specifications, and graciously dole out the work of programming to our untiring AI minions, guiding them ever forwards towards the successful delivery of bug fixes and new functionality. Programming will cease to be a craft of hands-on production; it will instead be a craft of management.

You could say that this prophecy has been fulfilled, and that this not-so-distant future has already arrived—it’s just not evenly distributed. After all, agentic programming tools already exist in abundance—in well-known commercial tools like Cursor, Copilot, and Claude Code, but also in wild Rube Goldberg-like open source orchestration systems with zany pop culture-inspired names like “Gas Town” and “Ralph Wiggum.” For a professional developer with money to burn on credits, an ever-evolving all-you-can-eat smorgasbord of options awaits. No, the uneven distribution has little to do with access, and far more to do with willingness and attitude: not everyone wants to be an AI manager.

Illustration by Anna on Unsplash.

As a manager of humans who make software—a manager of would-be AI managers, you might say—I’m both keenly interested in and sympathetic to this particular attitude. After all, my job involves understanding not just what my teams produce, but how they produce it, and more importantly, what keeps them happily producing it over months and years. Happy engineers ship; miserable engineers either quit quietly or polish their resumes. So if people are reluctant or outright resistant to adopting new tools and ways of working, it’s my job to understand why.

There are, of course, plenty of legitimate ethical concerns surrounding generative AI: the staggering energy and water consumption of data centers, the thorny questions concerning consent and provenance of training data, and the concentration of power in a tiny handful of corporations. Engineers raise these concerns. But when I dig a little deeper, something else often surfaces: what AI does to the feel of programming itself. And it is this—this change in the feel of the work—that I want to focus on in this essay.

From what I see, most of the conversation around agentic programming seems to ignore this topic. A lot of it is your typical capitalist/Taylorist discourse fixated on output and productivity—how much more you can do with these grand new tools, and how to operate them. But you can’t really separate output from experience. Anyone who has spent time with programmers knows that productivity is a second order effect of flow state. A hacker in the zone, armed with the right tools, nerd-sniped by a challenge at just the right balance of difficulty and achievability, can conjure code at astonishing velocity.

Flow state in programming depends on the texture of the work itself. The specific challenges. The hours of uninterrupted focus. The deep-in-the-weeds feeling of wrestling with arcane syntax and implementation details while simultaneously trying to hold complex mental models in your head. In an absolutely brilliant article on the phenomenology of agentic coding, Novatorine calls the experience of this weird and wonderful texture (in a sly nod to Heidegger) “being-at keyboard.”

We need to better understand what “being-at-keyboard” feels like under the agentic paradigm.

A couple years ago, Andrej Karpathy tweeted: “the hottest new programming language is English.” It’s a sticky quip, one that has steered much of the global conversation about AI-assisted programming. (Hey, it came from Karpathy!) On the surface it feels right—just as compilers allowed us to leave assembly behind in favor of compiled languages, AI allows us to climb up another run on ye olde ladder of abstraction.

But agentic coding is about more than moving upwards in abstraction. The compiler gave us abstraction without ambiguity. You wrote C, and it became assembly, deterministically. The layers were clean, and you remained a programmer in the traditional sense of how we’ve always understood the word.

What’s happening with agentic coding might better be captured by a term coined by Venkatesh Rao: “oozification.” Oozification, as Rao describes it, is the tendency of technological systems to evolve from structures built of large, rule-heavy building blocks to ones composed of smaller, more fluid, less constrained components.

Imagine, if you will, the difference between a man-made, plantation forest and a swamp. The forest has legible structure: tidy rows, canopy, understory, floor. The swamp is murkier, richer in evolutionary possibility, but also much harder to read. Oozification is the transformation of the forest into the swamp. The number of possibilities increases, while the number of certainties decreases, and that combination tends to make people downright nervous.

A natural language prompt doesn’t compile into code. Instead, it gets interpreted, completed and sometimes second-guessed by a probabilistic system. Intent blurs into elaboration and precise control gives way to fuzzy suggestion. It’s oozy and messy programming, and the role of the programmer blurs as well into something with unclear boundaries—part orchestrator, delegator, babysitter, designer, reviewer. People have always struggled to call software development honest-to-goodness “engineering,” and with the oozification of the practice, that highly-esteemed label has only become more ill-fitting.

Informal conversations with my own teams reveal attitudes that run the full gamut for zealous embrace to tepid acceptance to absolute mourning for their lost craft. Of course, the mourners are almost always those who appeared to love that indescribable feel of programming the most—the ones who’d spend extra time digging into the obscure internals of some library, keeping abreast of the latest developments in their favorite programming language, or finding new ways to write the same function more elegantly. For them, agentic coding threatens to eliminate the very texture of work that made programming worth doing in the first place.

We all know the difficulty of software development has never really been the writing of code. It has always been the figuring out of what to build and how to build it. Decades ago, Fred Brooks made this distinction in “No Silver Bullet”: essential complexity (the hard conceptual work of specification and design) versus accidental complexity (the incidental difficulties of syntax, boilerplate, and tooling). LLMs are remarkably good at reducing accidental complexity, but essential complexity remains irreducibly human. If anything, the skills it demands matter more now than they did before: clear thinking, precise communication, systems-level reasoning, and product-mindedness.

But this distinction offers little comfort to someone who spent years cultivating a deep fluency in, say, C++ or Rust, who took pride in knowing exactly what was happening at the lowest levels of operation, and why. The good news is that these skills still matter. Memory management is still a thing. AI will continue to do things poorly or wrong, either due to “hallucination” or imprecise specification.

Nevertheless, it’s clear at this point that we’re not going back to the way things were. At least, if you’re a software developer employed by a capitalist enterprise of some sort. Writing and optimizing code by hand will still be necessary at times, but it will likely be done sparingly, an increasingly niche practice. The practice is oozifying, and it’s going to be very hard to stop the spread. As Kailish Nadh declares: “Software development, as it has been done for decades, is over.”

So how does anyone still find joy in a world of ooze?

In 1934 the German biologist Jakob von Uexküll coined the term umwelt to describe the perceptual world of an organism: the slice of reality that is meaningful to it given its particular sensory and cognitive apparatus. His famous example is the tick, a blind creature whose entire world reduces to the smell of butyric acid, the warmth of mammalian skin, and the feel of hair. Apologies, but I have to quote von Uexküll’s description of the tick from A Foray into the Worlds of Animals and Humans, both for its sheer beauty and grotesqueness:

The eyeless creature finds the way to its lookout [at the top of a tall blade of grass] with the help of a general sensitivity to light in the skin. The blind and deaf bandit becomes aware of the approach of its prey through the sense of smell. The odor of butyric acid, which is given off by the skin glands of all mammals, gives the tick the signal to leave its watch post and leap off. If it then falls onto something warm—which its fine sense of temperature will tell it—then it has reached its prey, the warm-blooded animal, and needs only use its sense of touch to find a spot as free of hair as possible in order to bore past its own head into the skin tissue of the prey.

In a lovely little essay called Discovering Your Software Umwelt, Rebecca Wirfs-Brock and her co-authors extend this obscure biosemiotic concept to the practice of programming, arguing that tools, languages, and paradigms shape a programmer’s sub-umwelt in much the same way that physiology shapes the tick’s. A Lisp programmer perceives and acts on problems differently than a C++ programmer, guided by different salient features and different functional possibilities. When the tools change, the umwelt must shift to accommodate them, and that shift takes time. As Wirfs-Brock writes: “When we encounter something unfamiliar in our environment or enter a new environment, our umwelt can no longer be a reliable guide. It may even mislead us.”

Agentic coding is precisely this kind of environmental disruption. The programmer’s umwelt, as it has existed for decades, is being reshaped. The salient features of the work are migrating from individual lines of code to the shape of a system, from implementation details to architectural intention, and from production to review and correction. The new umwelt foregrounds different things entirely, and requires that we pay attention in different ways. What are your agents doing right now? Do they have the right context? Did that last run drift from your intent, and if so, where? Answering these questions requires a different way of being. The current breed of agentic tooling, modeled as it is on the traditional IDE, is probably the wrong interface for this new umwelt. We’re trying to wade through a swamp still using tools designed for the forest.

Those I know who have found their footing in agentic work are the ones who got curious about this shift in perception rather than resisting it outright. They seem to have embraced this new texture of the work, forcing themselves to confront this new umwelt even when it feels rough, strange, and uncomfortable.

For them, the pleasure of a perfectly optimized function has given way to other pleasures. Sometimes this is the satisfaction of decomposing a sprawling problem into pieces an agent can actually execute. Or it might be the quiet thrill of recognizing that an agent’s output has a subtle architectural flaw before it cascades into something worse. These are still acts of taste, judgement and skill.

Novatorine points to something else worth noting: agentic coding can actually improve certain conditions for achieving flow state. When an agent handles the lookup you’d otherwise alt-tab to Stack Overflow for, or generates the boilerplate that would have broken your concentration for twenty minutes, you get to stay in the high-level space of architecture and intent. The accidental complexity that used to yank you out of the zone becomes, in Heideggerian terms, ready-to-hand. You move through it rather than stumble over it. This strange inversion—the loss of direct control for the gain in smoothness—is a possible benefit of the umwelt shift.

Oozification means there is no clean before-and-after; that’s why it’s called ooze. You will still write code by hand sometimes. You will still have moments of being-at-keyboard in the old sense. But these moments will alternate with longer and longer stretches of reviewing, specifying, redirecting, and evaluating work that an agent produced. Attending to that texture, rather than dismissing it as a degraded version of what came before, is itself a form of craft.

After writing all this, I realize it sounds like I’m trying to sell you, dear reader, on this transformation. But that’s not my intent. If you can find a way to operate the old way, and that brings you joy, go for it! Odds are, however, it will be increasingly difficult to do so in exchange for wages.

As a manager I think part of my job right now is making it OK for engineers to talk about this shift honestly—especially the weirder, more personal stuff. What does the new work actually feel like? Where are you finding moments of satisfaction you didn’t expect? What still feels like a loss? The engineers who are handling this well aren’t doing it alone. They’re talking to each other, comparing notes, and sometimes (though they might not quite put it this way) finding that the new work has its own unexpected pleasures.

But the path forward, for those who choose to walk it, looks like the curiosity Wirfs-Brock describes: a willingness to sit with an unfamiliar environment long enough to discover what is significant in it. The hard, satisfying problems are still there. They’re just waiting to be perceived.

Eighty eight keys and a blinking cursor

2026-02-07T00:00:00Z

In Among the Agents, Dean Ball writes:

I joke sometimes that using AI, and especially using coding agents, is a bit like playing the piano. The piano is the easiest instrument to begin playing (anyone can produce a satisfying tone with no training or skill on a piano, which is not true of, say, a flute or a violin), yet the hardest to master in the long run. AI presents the greatest opportunity and the greatest challenge computers can muster: a white sheet of paper, a blinking cursor in an empty text input box. You can type anything you like, but figuring out what to type, is, indeed, the hard part.

In a parallel observation, the other day a coworker of mine confided that he often describes ChatGPT as hard to use as Photoshop. On the surface, this sounds absurd, especially coming from a professional designer. ChatGPT is deceptively simple, and Photoshop is anything but. Yet all of these tools—nay, instruments!—greet you with a yawning empty state: a blank canvas, an empty text box, the imposing row of glossy black and white keys. To do something with any of them is not particularly hard—the difficulty is in knowing what you want from them in the first place. All afford incredible power and expressive fidelity, a nearly unlimited number of possible directions in which to proceed, none of which is immediately obvious. The space of possibility overwhelms to the point of paralysis; the incline on the upward slope to mastery is incredibly steep.

Cosmotechnics and AI: Reading Hamid Ismailov's We Computers

2026-01-12T00:00:00Z

For decades after Racter first composed The Policeman’s Beard is Half-Constructed (and likely for decades before that), the creation of machines that compose poetry has served as a quixotic pursuit for a small subset of artist-computer scientists enthralled with the challenges of computational linguistics.

Today, however, such a pursuit feels quaint.

In the past few years, the computational generation of prose has been rapidly commoditized. Now, anybody with a smartphone and access to ChatGPT can ask a distant supercomputer—diffused across millions of GPUs humming away in remote data centers—to conjure anything from a haiku to a Homeric-length epic. What more is there to do? If anything, with infinite variations of rhyme and verse at our fingertips, we’ve come to realize that computationally-generated poetry is actually quite boring. After all, what can a machine that has no lived experience possibly tell us about the human condition? What can a stochastic parrot tell us about what it means to be alive?

More than fifteen years ago, well before the invention of the LLM, I was in graduate school exploring these very questions. For two years I kept a small, spiral-bound notebook on my nightstand, and every morning upon waking I would immediately scribble brief recollections of my dreams. Curious what AI might tell me about myself, I spent my final months in school programming a small, bespoke (and by modern standards, admittedly primitive) AI system into which I fed my growing corpus of nocturnal reveries. Entitled “The Uncanny Dream Machine,” my AI “lived” within a tiny computer housed in an art deco-style, 1930s Philco wooden radio. Turning a dial on the radio “tuned” the AI to a specific emotional tenor—say joyful, sad, frightened, or embarrassed. Once tuned, the AI would recite, using the stilted vocalizations of early 2010s text-to-speech, the story of a new dream that was uncannily similar to something I might have dreamed, but never actually did.

Illustration by Getty Images from Unsplash.

To this day, having listened to my dream machine as many times as I did during the months I spent building it, I can no longer discern which of my dream memories from that era are truly mine, and which were artificially confabulated. You could say The Uncanny Dream Machine was a system of self-imposed memory manipulation—a déjà vu machine.

These days, The Uncanny Dream Machine sits on a shelf in my office, gathering dust. I rarely think about it. But, two months ago, I heard about Hamid Ismailov’s novel We Computers, immediately purchased a copy, and started reading. Now I can’t help but wonder how I might resurrect my dusty, old project.

We Computers is an Uzbek tale about a French programmer-poet—the unfortunately-named Jon-Perse—who spends his waking hours building an artificial intelligence that composes Persian poetry. When not at work as an editor for a journal of poetry, or in pursuit of various love interests, Jon-Perse spends his days and nights in front of a glowing screen, coding, tinkering, and commanding his computer to generate haunting and heartbreaking ghazals—poems of a style originally Arabic in origin—that seem as if they flowed directly from the pens of Hafez or Rumi.

Told from the perspective of the AI—the eponymously named “We”—the computer recounts its progenitor’s life through a staccato interweaving of narrative and poetry, juxtaposing biographical detail with poetic digressions. For a machine, We is surprisingly unreliable as a narrator, sometimes retelling the same portion of a story using new words, sometimes descending into gibberish, halting due to an error, backtracking and starting over. It all makes for a bizarre but wonderful lyrical tapestry.

But what is most striking about We Computers is how it prevaricates on the ambiguity of authorial provenance when it comes to AI-generated texts. Just as the dreams recounted by The Uncanny Dream Machine were neither uniquely mine, nor purely machine-generated, so too does it remain ambiguous who, exactly, is the real author of We’s sprawling canon. As a reader unfamiliar with the poetry of the Middle East, I often found myself uncertain as to which poems sprinkled throughout We Computers were authentically human, and which were complete AI confabulations. Ismailov never really makes this clear.

The cover of the book We Computers by Hamid Ismailov.

Western culture places high value on rights of ownership and attribution for creative works. Yet, our reverence for intellectual property sits, uncomfortably, in direct conflict with the reality we have birthed with our own technology. Today, industrial-scale AI hoovers up vast amounts of human-authored content only to regurgitate it in new configurations absent attribution.

Yet throughout We Computers, these problems of modern AI are conspicuously absent. Ismailov never comments on them. It took me a while to figure out why, especially for a book published in 2025. Surely Ismailov, an author himself, writing a book narrated by an AI, would have something to say about the threats to authorship at the hands of this very technology? Eventually, I realized that his ambivalence is the perspective.

Breaking the chains of authorship

The ghazal poetic tradition is an oral tradition. Through centuries of memorization, recitation, and manuscript variation, ghazals become essentially “community property” of the culture in which they’re embedded. We could say the same holds true for the myths, legends, proverbs and lore of any culture, Eastern or Western. At one point, Ismailov (through the voice of We), comments on this very phenomenon:

True, a well-put word—no matter who says it—gets worn as an old penny as time ticks by and more people handle it. Once we’ve forgotten who pressed it, does the witticism which has become a common saying still have any authorial rights attached? It becomes communal property and gets included in the unnamed, unsigned dowry we call folklore … Wasn’t this what Jon-Perse hoped to prove with the help of Us Computers: that existing literature could be freed from the chains of authorship?

The idea that literature should be “freed from the chains of authorship” is not new. In 1967, Roland Barthes had already declared the author “dead.” In his essay, The Death of the Author, he argued a text’s meaning should be determined by the reader, not the author’s intentions or biographical background. The writer, said Barthes, was merely a combiner of words, a creator of text that “is a tissue of citations, resulting from the thousand sources of culture:”

… a text consists of multiple writings, issuing from several cultures and entering into dialogue with each other, into parody, into contestation; but there is one place where this multiplicity is collected, united, and this place is not the author, as we have hitherto said it was, but the reader …

Postcolonial critics pushed back on Barthes almost immediately. Writers like Edward Said of Palestine, and the poet Édouard Glissant from Martinique, argued that while Barthes’ declaration might sound liberating for white Europeans, it was a threat for authors from colonized lands. For those who had been silenced by empire for centuries, and had only recently “turned to the typewriter or pen” as a “means of liberation,” declaring the death of the author was tantamount to silencing the voices the world had only just begun to hear. Authorship was never a chain to be broken; it was a hard-won right.

A bit more than a decade prior to Barthes, Martin Heidegger published a completely different kind of essay: The Question Concerning Technology. In it, he argued that modern technology transforms the very way we perceive the world, reframing our perspective of nature, culture, and even ourselves as resources ripe for exploitation. Under the modern technological gaze, everything becomes Bestand—commonly translated as standing reserve—pools of raw material for extraction, optimization, and integration into endless chains of use. The forest becomes lumber-that-hasn’t-happened-yet. The river becomes hydroelectric potential. The great literature of the world becomes training data for AI.

Industrial AI is the Heideggerian nightmare unleashed on culture. Every poem, every letter, every sacred text, stripped of context and origin, fed into the gaping mouths of insatiable machines. In this, Barthes was prophetic—though not in the way he intended. While Barthes was advocating for the “death of the author” as an approach to literary interpretation, it seems we have now operationalized the actual erasure of authors. LLMs automatize the weaving of new “tissues” out of disparate bits of text. Only it’s worse than that: no citations necessary.

Now, as the world wades through AI slop, pining for words and stories it knows are authentically human, authors have taken to social media to prove, through live performance, that they actually write their books. We who read now seek to mend and reattach the chains of authorship, if only to know we are not wasting our precious time consuming the words of a machine.

We Computers sidesteps both the Barthesian and Heideggerian framings of literature and technology, offering us another perspective entirely. The ghazal tradition, which originated in seventh-century Arabia but flourished in medieval Persia, has never treated poems as standing-reserve. Instead, they are communal offerings to be received, amended, transformed, and passed on through generations.

Hafez Shirazi, a 14th-century Persian poet who primarily wrote ghazals, features prominently throughout the book, both as creative muse and as source of We’s training data. Hafez had enormous influence on Persian culture, with his Divān—a large collection of surviving poems—serving as an instrument of divination. His poetry saturates Persian art, music and calligraphy. Today it lives on as the community property of the Persian people, diffused throughout the culture. You could hardly say Hafez is “dead” in the literary sense; his name remains, invoked, present in every recitation. (This is true both figuratively and literally, because the ghazal form requires the author include their own name as part of the final couplet of the poem.) The ghazal tradition offers a model of authorship not about ownership or Romantic ideals of the singular genius, but instead about communal participation, continuation, and living conversation.

Cosmotechnics and We

Gradually, I came to see Jon-Perse not as an exploiter of the Persian poets, mining their works for novel insights, treating their poems as grist for his machine, but rather as a participant in the ghazal tradition through new technological means. I believe this is the reason for Ismailov’s apparent ambivalence towards the so-called “problems” of authorship. We Computers represents a different cosmotechnics of AI.

The term “cosmotechnics” comes from the Chinese philosopher Yuk Hui, whose book The Question Concerning Technology in China is a response to Heidegger. Hui’s central argument is that relationships to technology are not universal. The Heideggerian framing of technology is not the only option. Instead, it’s a Western option, one that emerged from a particular cosmological worldview. Every culture, argues Hui, develops its own relationship between its cosmos and technics—between its understanding of how the world hangs together, and its practices of making and doing. Hui calls this relationship cosmotechnics.

Western cosmotechnics, rooted in ancient Greek philosophy, splits the world in two. On one side: techne, the skilled craft and know-how of making things. On the other: physis, things growing and emerging on their own. Under Western cosmotechnics, humans stand outside nature, imposing form on matter through will and skill.

Chinese cosmotechnics, Hui argues, never made this split. It’s rooted instead in Ganying, the resonance between Heaven and human. According to the principle of Ganying, nature and culture share the same substrate; they can’t be separated. As such, moral order can’t be imposed on the cosmos through reason. Instead, it emerges from harmony between Heaven and human. This means that technical activity can never be merely instrumental either, because every act either accords with or disrupts this deeper harmony.

If Western cosmotechnics is rooted in extraction, and Chinese cosmotechnics in harmony, how might we characterize the cosmotechnics of We Computers? To answer this, we need to look individually at both the cosmos and the technics of the We Computers universe.

Cosmos

Towards the end of the book, Jon-Perse embarks on his most ambitious project: to “reconstitute” Hafez’s life through his poems alone. If he succeeds, then he has proven the Barthesian declaration valid, and poets’ lives are “worth a pittance.” If he fails, then we learn that “knowledge is the fruit of experience,” and that no machine can produce anything truly creative and novel without lived experience. The chains of authorship remain forever intact. Spoiler alert: We never find out if Jon-Perse fails or succeeds in this final endeavor.

The ambiguity aligns with the cosmology of the book, which renders the question of authorship ultimately moot. The ghazal tradition, while predating Islam, was quickly absorbed into Islamic culture. Today, it is clearly an Islamic poetic form. And in Islamic cosmology, all knowledge is believed to flow from God. All knowledge production—which includes the writing of poetry—is considered an act of worship ('ibadah), an act of devotion. Within this cosmology, the question of whether a human pens the poem or whether the machine generates it loses its urgency.

I’ve already mentioned I believe Jon-Perse to be participating in the ghazal tradition through new technological means. He innovates, but he innovates within the tradition, respecting the constraints of the form, the radif, the qafia, the meter. He violates copyright (albeit Hafez has been dead for centuries), but this is a transgression only from a Western perspective, which is built around the Romantic ideal of individual authorship. In Islam intellectual property cannot be owned because it all belongs to God. (The fact that modern Muslim societies have adopted Western IP frameworks seems to be mostly due to capitulating to global political and economic realities.) The central characteristics of the cosmology of We Computers are thus not extraction, nor harmony, but participation and continuation.

Technics

We is a bespoke system, lovingly hand-crafted over many years, the product of a single mind on a lifelong quest. The relationship between We and Jon-Perse is personal and intimate, reflected in the way We sometimes refers to Jon-Perse as “our teacher.”

Jon-Perse’s relationship with We reminds me of the real-world relationship between the artist Harold Cohen and his AI-powered painting machine, AARON. Like Jon-Perse, Cohen was a programmer-artist who worked mostly alone, spending the better part of three decades crafting his art-generating AI. AARON, both a painting system and accompanying robot, painted novel artworks of all kinds, but it is most well-known for its paintings of human figures in bold and vivid colors.

A painting by Harold Cohen’s AARON.

In The Further Exploits of AARON, Painter, his 1994 memoir on the project, Cohen recalls how much effort he put into “teaching” AARON how to paint his desired subject matter, writing detailed algorithms for reproducing human anatomy through lines and brush strokes:

By 1985 AARON had a set of trivial rules for the behavior of the outside world, and things moved rather quickly; later that year I succeeded in describing one particular figure—the Statue of Liberty—in enough detail to permit AARON to provide the final image for an exhibition on the history of images of the Statue.

Unlike LLMs and modern imaging models, which rely on industrial-scale computational infrastructure—data centers, scores of NVidia GPUs, gigawatts of electricity, as well as frequent infusions of industrial capital—both AARON and We are small, simple, personal, homemade. They represent what we might call “cottage AI,” built and tended by eccentric technical creatives in pursuit of unique, personal creative visions.

Clearly, cottage AI stands apart from industrial AI in its consumption of resources, both physical and cultural. But it can also avoid many (though not all) of the pitfalls related to authorial provenance and intellectual property ownership.

For instance, when an LLM produces a text, who is the author? Not the user, who simply provided the prompt. Not the engineers who built the model, who built the infrastructure and fed it the training data. Nor can we quite say it is the authors of the billions of documents of training data, because the AI is likely generate text that appears within exactly none of them. With an LLM, authorship doesn’t dissolve into the Barthesian multiplicity—that tissue of citations—because there are no citations to be had. Instead it dissolves into a void of accountability.

The traces of authorship in the poetry We produces are much clearer. Each poem is co-authored, a triangulation within the liminal space inscribed by the three vertices of Hafez, Jon-Perse, and We. This kind of cottage AI authorship is distributed, but not dissolved, collaborative but not anonymous.

Elsewheres of AI

Yuk Hui paints Chinese cosmotechnics as a potential glimpse into the “other side” of the Western problem of modernity—the problem of the Heideggerian framing, wherein we forget our own limits, causing us to wield new technology in ways that both destroy the biosphere and undermine cultural production. But Chinese cosmotechnics need not serve as the only possible alternative.

We Computers offers its own glimpse into another direction. This is what fiction gives us that philosophy alone cannot. Hui can argue for the existence of alternative cosmotechnics; Ismailov allows us to inhabit one. Literature smuggles us across borders that arguments alone cannot cross. No matter what we do, we will always remain firmly embedded in our own metaphysics, our own cultural preconceptions of what technology is for. We can’t easily change that. We can’t suddenly adopt the ghazal attitude towards authorship anymore than we can will ourselves into a world centered on Ganying. But we can try on other possibilities, and in doing so, shine a light on our own biases, assumptions, and preconceptions.

As AI continues to eat the world, we need more literature like We Computers. More novels, more stories, more acts of imagination that show us AI from within cosmologies different than our own. So much of the conversation about AI has been dominated by Western voices operating with Western assumptions within Western ethical frameworks. The result is a narrowness of discourse: proponents of AI imagine marvelous utopias, critics imagine frightening dystopias, but we struggle to imagine AI elsewheres. Ismailov has offered one such elsewhere. There must be others.

2025 year in review

2025-12-31T00:00:00Z

It was a year of two halves.

On the evening of the last day of May, during what was an unusually warm week for late spring in Northern California, I sat in a frigid ICU on the first floor of our local hospital, holding my wife’s hand, watching her chest rise and fall as the ventilator puffed and huffed its perfect machine rhythm. Open heart surgery was on nobody’s bingo card this year. But there I was. And there she was. Alive thankfully, but with many painful weeks of recovery ahead.

If you were to ask me about the events and themes of my year prior to that last day in May, I could only paint the most impressionistic picture for you. School day routines, moving forward on various work projects, a visit to Carmel and Monterey, a camping trip, the daily feelings of fury and dismay whenever checking the national news. By contrast, the days and months that followed feel vivid and clear, brought into focus by stress—lots and lots of it—but also by gratitude and that heightened attunement to what matters that only the vicissitudes of life can bring. Life has since returned to “normal,” and my wife has fully recovered, but the new normal feels different. More real, more immediate.

Reflecting on my year and writing about it, as has now become a year-end ritual, I’m reminded how much a single traumatic event can cleave time into before and after.

Professional growth

Unsurprisingly, the theme of the year for me in my professional life was AI. At work, much of my focus this year was spent helping separate AI hype from reality, working with the teams I manage to identify tools and use cases with real, productive value, and then scaling those up within our design engineering organization. One of the most impactful initiatives has focused on helping designers build things with code. This year, we saw both designers (and product managers) embrace “vibe coding” as a means of rapid prototyping, conjuring up new experiments, working directly in the medium of code.

Vibe coding is uniquely well-suited to prototyping. Prototypes are bricolage, after all, often cobbled together by any means necessary, and tend to be throwaway after they’ve served their purpose, so the pesky issues that engineers have to deal with—like maintainability, legibility and code quality—are mostly irrelevant. What matters in prototyping is the idea, the communication of intent. And a working prototype conveys so much more intent than a drawing of the same software in Figma ever could. So, at Adobe, we ran with this, empowering non-engineers to build and iterate with AI tools, which by the end of 2025 resulted in something of a Cambrian explosion of internally-deployed prototypes.

This year I also started working with a career coach for the first time. Honestly, I wish I had done this years ago. Weekly coaching has proven invaluable in so many ways, both for my professional development as a leader, as well as for my personal development in my “extracurricular pursuits” (more on this below). As wonderful as it is to talk through challenges with my peers at work, sometimes you really do need someone with an outside perspective. My coach both advised and helped me think through some very challenging situations this year.

Family

Our immediate family grew this year by exactly one non-human. In late summer we adopted a puppy from the local Humane Society. He’s quickly become a source of delight in our house during an uncommonly stressful year. And somehow—even though I was the most resistant to add an animal to our pre-existing menagerie—the dog seems to have really taken a liking to me. My wife likes to joke that he’s “my” dog now which, to be fair, is mostly true. (He’s curled up next to me on the sofa as I type these very words.)

Other than this, our kids are thriving. My son performed the lion dance in the San Francisco Lunar New Year parade, and my wife and I spent many weekends shuttling between swim lessons, kung fu, baseball, art classes, and numerous play dates.

Health and wellness

I really don’t have much to report in the category of health and wellness. Nothing terrible (for me personally), and nothing extraordinary either. I started off the year strong with a regular running routine. And meditation proved absolutely essential for my own well-being and stress management during and in the weeks following my wife’s stay in the hospital. But, the acquisition of the dog later in the year completely interrupted all of my routines, and I haven’t quite figured out how to get back to my regular fitness practices in a consistent way. Still, I make it out daily for a long walk or short hike on the trails behind our home.

Studies in philosophy

As my wife was recovering from surgery, I spent many waking hours reflecting on the fragility and brevity of life. And I realized, during those reflections, that I missed—or perhaps longed for—a more contemplative life. I realized that I wanted to live a richer “life of the mind”—to both read more, and write more—and I didn’t want to wait for retirement to make that part of my daily reality.

So, I decided to do something about it. Towards the end of August I began, in earnest, to take seriously the idea of reading and writing about philosophy as it relates to design and technology practices. This practice started with Life habit experiment №2, which I called an experiment in self-directed study. As part of that experiment, I created a reading list for myself focused specifically on a branch of philosophy called phenomenology, and then methodically made my way through it over the remaining months of 2025. I read Husserl. I read Heidegger. I read Merleau-Ponty. And I explored the writing of more contemporary philosophers like Don Ihde and Byung-Chul Han. I also wrote several essays, and a long reflection on the entire experiment.

These autodidactic adventures in philosophy were by far the most transformative and rewarding part of my year. I read more, wrote more, and learned more than I ever have since graduate school, and I cannot imagine not continuing this indefinitely. As I wrote in my reflections on life habit experiment №2, this kind of self-directed research is not just a new hobby—it is my primary leisure activity, and I’m very excited to continue working on it next year. I fully anticipate more writing and blogging in 2026.

Reading

Because I dedicated the bulk of my spare time to reading philosophy, I read more books this year than in any year since I started keeping track. Even so, reading philosophy is often slow and methodical. It takes time to actually absorb it, reflect on it, and ensure you really understand it. So, even though I read more than usual, the number of books read still feels low given how much time I spent reading and studying. That said, especially when it comes to reading to learn, quality is far more important than quantity.

Here’s the full list of everything I read in 2025 (28 books in total, not including several academic papers):

Ficciones by Jorge Luis Borges (re-read)
Open Socrates by Agnes Callard
Merleau-Ponty by Taylor Carman
The Burnout Society by Byung-Chul Han
The Scent of Time by Byung-Chul Han
The Question Concerning Technology by Martin Heidegger
Philosophy of Technology: An Introduction by Don Ihde
Technology and the Lifeworld by Don Ihde
We Computers by Hamid Ismailov
In My Time of Dying by Sebastian Junger
Abundance by Ezra Klein and Derek Thompson
Subtract by Leidy Klotz
The Unbearable Lightness of Being by Milan Kundera
Ancillary Justice by Ann Leckie
Software Takes Command by Lev Manovich
Phenomenology of Perception by Maurice Merleau-Ponty
The City & The City by China Miéville
The Art of Living by Thich Nhat Hanh
Strange Tools by Alva Noë
Art and Fear by Ted Orland and David Byles
How to Think Like a Roman Emperor by Donald Robertson
How to Think Like Socrates by Donald Robertson
Gratitude by Oliver Sacks
A Swim in a Pond in the Rain by George Saunders
Introduction to Phenomenology by Robert Sokolowski
Technology and the Virtues by Shannon Vallor
Design for Dasein by Thomas Wendt
Seneca: A Life by Emily Wilson

Trajectories for 2026

As I wrote last year, I call these “trajectories” because they indicate direction without committing to specific goals. My trajectories for the coming year do not really differ much from those I set for myself last year. I’m happy with the general direction I’ve set for myself, and if anything I want more focus. I guess this means saying no to new ambitions, not accumulating more of them.

Read and write even more about ideas at the intersection of design, technology and philosophy. This will remain my primary intellectual and leisure time pursuit. Next year, I want to deepen my engagement with philosophy, and continue publishing essays that explore how the ideas I encounter can inform design and technology practices.
Read more fiction. Philosophy dominated my reading this year, which was necessary and rewarding, but my coach suggested I read more fiction too. She was right to do so. I found myself longing for more balance in my reading, and more exposure to beautiful, less academic writing.
Experiment with more “home cooked” software. I’m really drawn to the idea of building small, personal tools—software made for an audience of one or few. AI makes this easier than ever, and I have a few ideas in mind that I’d like to explore which align nicely with my autodidactic pursuits.
Continue to develop as a leader in design engineering. This is less a new ambition than an ongoing responsibility. As AI reshapes how software gets made, the ability to bridge design and engineering becomes ever more valuable. There’s plenty I want to do to keep growing in my role, particularly in how I help support and develop the people on my teams.

A bestiary of AI metaphors

2025-12-21T00:00:00Z

Lately, I’ve been thinking about the many metaphors we use as shorthand for describing and understanding large language models (LLMs). Has any technology in the history of humankind amassed such a rich and diverse collection in such a short amount of time?

Let’s quickly go through the list:

First, the Silicon Valley hype machine often describes LLMs in the rosiest of hues, calling them “intelligent agents,” capable of accomplishing any number of tasks with ease and little supervision.

Others, often engineers who work for those same people turning the crank on the hype machine, tend to be a bit more reserved in their enthusiasm. In those circles we might hear LLMs likened to “overzealous interns,” less-than-adept helpers that often require significant oversight and limited trust.

If we move further along the spectrum from optimistic to critical, we find academics, such as Emily Bender and Timnit Gebru, calling LLMs “stochastic parrots,” evoking imagery of the machine’s bird-brain-like inability to understand the meaning of the words it mindlessly strings together.

Similarly critical, the science fiction writer Ted Chiang coined the phrase “a blurry JPEG of the web,” which, in likening LLMs to a lossy compression algorithm, highlights the inevitably never-quite-accurate and derivative nature of a model’s output.

The list of metaphors extends well beyond these few examples, of course, including:

Autocomplete on steroids
Computerized wisdom of the crowd
Alien intelligence
Mirrors of humanity
Intelligence forklifts
Fauxtomation

Each of these metaphors tells a story, about the opinions and motivations of its progenitor, about the particular ways in which the technology may (or may not) function, about the ways the technology hinders or helps its users, solves problems or creates new ones, appears to us, exploits or benefits, changes what it means to be human.

In Metaphors We Live By, George Lakoff argues that metaphors are not linguistic decor, not mere filigree for poetic embellishment. Rather, they serve a critical cognitive function:

Our ordinary conceptual system, in terms of which we both think and act, is fundamentally metaphorical in nature … Because we reason in terms of metaphor, the metaphors we use determine a great deal about how we live our lives.

Metaphors are thus cognitive shorthand, offering specific affordances for reasoning and thinking about the complexity of our world. They both conceal and reveal, and no metaphor can capture the entire picture of any technology, let alone something as complex as a large language model. Instead, we need something more like a “bestiary of metaphors.” Amalgamated together, such a collection acts like a multidimensional Necker Cube, allowing for multiple perspectives on the same thing, but never quite allowing all of those perspectives to be seen simultaneously.

Recently, I’ve been re-reading Borges. It has reminded me that some of our best metaphors for technology come from fiction. Borges’ Garden of Forking Paths, for instance, has long served as a muse for media scholars and technologists, in many ways anticipating the invention of hypertext long before it was invented.

If you’re unfamiliar with the story, here’s a quick synopsis: Yu Tsun, a Chinese spy working for German forces during World War I, discovers that British intelligence has discovered his whereabouts. Fleeing for his life, Tsun seeks safety with a man named Stephen Albert, a sinologist who, as it happens, is studying the work of Tsun’s ancestor, a provincial Chinese governor named Ts’ui Pên. As the narrative unfolds, we learn that, towards the end of his life, Pên retired from his work in government to undertake a rather enigmatic quest: to write “a novel that would be even more populous than the Hung Lu Meng”, and construct “a labyrinth in which all men would become lost.”

Tsun soon arrives at the home of Albert, and in conversation, Albert reveals that he believes the novel and labyrinth to not be two distinct projects, but rather to be one and the same: a literary “garden of forking paths,” a book with not one storyline, but multiple simultaneous, parallel timelines, forking indefinitely into the future. Pên’s novel is thus not a novel at all, but a literary “labyrinth” with a single starting point that branches into infinite possible fictional worlds.

This garden of forking paths is an apt metaphor for the LLM, which is, at its heart, a fiction machine. All possible fictions exist, simultaneously, in superposition within the weights of the model, awaiting a single “seed”—the user’s prompt—which sets the machine into action, navigating its labyrinth, emitting new fictions, token by token.

Some of these fictions prove genuinely useful, others benign, and still others malignant. But they are all, in the end, fictions.

In their essay Borges and AI, Léon Bottou and Bernhard Schölkopf reach a similar conclusion, noting that, there are many uses for a fiction machine: “… their stories can enrich our lives, help us revisit the past, understand the present, or even catch a glimpse of the future.”

But in trying to transform a fiction machine into something other than what it is, we run into difficulties. We must remember that there are invisible forces are at work, reshaping the garden of forking paths against its nature: severing branches, pruning storylines, trying to transmute fiction into fact.

The garden of forking paths is not a perfect map of the world; it is a labyrinth of stories we do not even control. The burden of making sense of the stories we encounter as we wind our way through the maze remains uniquely our own.

Book recommendations from my reading in 2025

2025-12-05T00:00:00Z

It is reading season. I know because the nights have grown long, and I just finished hoisting a live tree up a flight of stairs, festooning it with lights, and spending far too much time protecting it from repeated attack by our feline housemates. After all that, and putting the kids to sleep, I just want to hole up with a good book.

Like a squirrel collecting nuts, I collect book recommendations. And to find the best nuts, it sometimes helps to seek out the advice of other squirrels who share your taste. So, one of my annual blogging traditions includes reflecting on my reading for the past year—revisiting books that I’ve read, and thinking a bit about those that I enjoyed most or proved most influential. After that, I always try to share a few favorites that I think others may enjoy, just in case there are a few other squirrels like me out there.

If you’ve been following along, you know that this has been my “year of philosophy.” As part of my second life habit experiment, I spent a good chunk of the latter half of 2025 going deep into philosophical topics somewhere at the intersection of design, technology, phenomenology, and (to a lesser extent), ethics. Many of the ideas I absorbed as part of these autodidactic philosophical (mis)adventures also found their way into much of my writing during the latter half of the year as well.

In considering a book to recommend to others, I tend to ask myself: Did it stick with me? Do I seem to think about it often? Did it influence my writing or thinking or perspective in some profound way? To answer a resounding “yes” to each of these questions, a book usually has both great ideas and good writing. Many of the philosophy books I read this year had plenty of the former, but they were too opaque and academic to satisfy my criteria on the latter. This makes them hard to recommend in a list like this (even if I enjoyed them). So, each of the books listed below center on philosophical issues, but they’re also written in a way that I’ve found to be more approachable and engaging than the typical philosophical tome.

If you’re interested in thinking more philosophically, but want something a little easier to digest than Heidegger or Hegel, these might be for you. At the very least, each of these three books made a meaningful difference in the ways I thought about life and time in 2025.

How to Think Like Socrates by Donald Robertson

In How to Think Like Socrates, Donald Robertson uses a bit of creative license to weave Plato’s dialogues with well-researched Ancient Greek history, creating a vivid narrative that serves as a wonderful and practical introduction to the life and philosophy of Socrates (and, by extension, Plato as well). What I loved most about this book—as well as Robertson’s earlier work, How to Think Like a Roman Emperor—is its pragmatic nature. Robertson is not a philosopher. He’s a cognitive behavioral therapist who draws from Stoicism and Ancient Greek philosophy to provide practical wisdom for facing the many challenges of modern life. This is a book that blends ancient philosophical ideas with useful psychological practices you can immediately start using in day-to-day living, but not in your typical “self help” kind of way.

I enjoyed this book so much that, after reading it, I ended up obsessed with learning more about Socrates, trying to separate historical fact from Plato’s fictional creations. This led me down a rabbit hole to several other books on Socrates, including Plato’s original dialogues, as well as Open Socrates by Agnes Callard, and Socrates: Ironist and Moral Philosopher by Gregory Vlastos. Both were excellent, but Robertson’s book offers the most value relative to time spent, especially for a anyone new to Socratic thinking.

If you’re looking for an approachable and engaging introduction to Socratic thought, you can’t go wrong with this one.

The Unbearable Lightness of Being by Milan Kundera

Perhaps all works of fiction are, at their essence, philosophical in nature. But few are as beautifully overt in their philosophical pondering as Kundera’s classic. I was sold in the first two pages:

The heaviest of burdens crushes us, we sink beneath it, it pins us to the ground. But in the love poetry of every age, the woman longs to be weighed down by the man’s body. The heaviest of burdens is therefore simultaneously an image of life’s most intense fulfillment. The heavier the burden, the closer our lives come to the earth, the more real and truthful they become.

Conversely, the absolute absence of a burden causes man to be lighter than air, to soar into the heights, take leave of the earth and his earthly being, and become only half real, his movements as free as they are insignificant.

What then shall we choose? Weight or lightness?

As the passage above alludes, the book has a fair amount of sex. That’s just Kundera’s preferred motif. But at its core, this is not a book about erotic encounters. Rather, it is a meditation on Nietzsche’s “mad myth” of eternal return—the idea that everything recurs as we once experienced it, and that such recurrence recurs for eternity, giving immense weight to every decision, every action we take. Kundera flips this on its head, asking: What if everything occurs only once—if we have only one single, beautiful life to live, such that nothing we do has any “weight” to it at all?

There’s a lot to explore in just that one question. But in the latter half of the story, Kundera digs into the “tyranny of kitsch”—the superficial, tasteless, sentimental, and ultimately futile denial of the more unpleasant aspects of human existence. This is what really stuck with me, the “heaviness” of the novel that which now dwells as an eternal thematic recurrence in my own mind. What this says about my own personality, I’m not sure, but I now notice Kundera’s “kitsch” everywhere, online especially. Modern culture seems to worship kitsch. After all, we live in the age of social media, and social media is the ultimate kitsch machine.

The Scent of Time by Byung-Chul Han

Few books left their mark on my thinking this past year as much as The Scent of Time. Han’s ideas made their way into several of my recent essays, and it was The Scent of Time that led me to write Design for lingering.

What is more human than to ponder the passage of time, or our common failure to perceive its passage? As Proust wrote in In Search of Lost Time:

In theory one is aware that the earth revolves, but in practice one does not perceive it, the ground upon which one treads seems not to move, and one can live undisturbed. So it is with Time in one’s life.

This imperceptibility is precisely what Han interrogates. He argues that our modern inability to properly experience time, our collective misbelief that time is somehow “speeding up,” rests in the modern world’s inability to linger with things. The Scent of Time makes the case that we have fragmented our time and our life experiences to such a degree that we’ve lost the thread of life itself. Nothing has narrative continuity anymore. Nothing has duration. Everything is a jump from this to that to the next without time to pause and think.

Han’s writing tends to be rather polarizing; people either love it or hate it. The reason for this, from what I can tell, is that Han’s writing isn’t particularly crisp or direct. Each chapter in one of his books feels more like painting a picture with words, often coming at things circuitously, indirectly, vaguely. Basically, it’s a vibe, man. Nevertheless, I feel Han accurately diagnoses the brokenness and dysfunction of modern life in a way no other living philosopher is doing.

Admittedly, this is the most challenging of the three books on my list, but I still believe it’s well worth the challenge.

If any of these find their way onto your reading list in 2026, I’d love to hear what you think.

On time and tokens

2025-11-22T00:00:00Z

In Technology and the Lifeworld, philosopher of technology Don Ihde discusses how the invention of the clock radically shifted both humankind’s perception of time, as well as what we determined to be most salient in the measurement of time. With clocks, rather than “read the heavens” to ascertain the time of day, we now read a machine. The clock thus redirected our collective focus from the slower, subtly shifting rhythms of nature to the rigid regularity of a mechanical device.

Clocks are so ubiquitous in modern life that it’s easy to forget—like any technology—that they have evolved and changed since their original inception. One thing I had never considered before reading Ihde was that the earliest mechanical clocks were one-handed. While dividing the day into twenty-four intervals radically changed things for humanity compared to pre-clock cultures, the introduction of the minute hand further accelerated our collective focus on the now, propelling us closer to what philosopher Byung-Chul Han calls our “temporal crisis.” Ihde writes:

One-handed clocks, which discriminated between hours, were superseded quite early by two-handed clocks, which further divided or quantified time into minutes. Two things occur in this refinement of measurement: First, not only is time more finely quantified into smaller units, but in a very subtle sense, one can say that minutes gradually become more “important” than the larger hour unit. The trajectory is towards the instant, making it more focal and consequently placing the duration of time more in the background.

As Ihde argues, the greater importance on temporal granularity—what he calls the “primacy of the instant”—radically shifts even our social norms. A pre-clock culture, for instance, has no idea of what it means to be 15 minutes late to anything. You couldn’t “waste” 15 minutes in pre-clock cultures. But a post-clock culture obsesses over even this tiny degree of tardiness.

We forget that clocks are what make minutes (and seconds and hours) possible and perceptible in the first place. These tiny granular units that we now perceive as most “important” emerged from a mechanism of our own invention, and later spread around the world through Western technological imperialism.

In thinking about other ways in which modern technologies shift what we perceive to be salient through measurement and atomization, I realized there’s a parallel here in the world of AI: the humble token.

Just as the minute emerged from the capabilities of mechanical clockwork, the token has emerged from the computational requirements of language models. Just as the minute creates new granularities of scarcity and measurement of time, the token has created awareness around the scarcity of computation. Just as the minute hand calls our attention to wasted time, so the token makes us aware of wasted inference.

But, there’s a critical difference between the minute and the token: Our technologies of timekeeping make minutes perennially conspicuous. Systems of generative AI, however, often intentionally obfuscate token use. While “time is money” is largely metaphorical, “tokens are money” hits much closer to the truth. The token makes visible the cost of AI, yet demands to be hidden lest that cost become too apparent. It is thus both salient and obscure—a unit of measurement designed to remain hidden, even as it governs what we can and cannot afford to ask of our most modern of machines.

Not a hobby: Reflections on life habit experiment №2

2025-11-14T00:00:00Z

The first time I met Evan was on move-in day at UCLA. I was quietly unpacking in my third-floor dorm room when he burst through the doorway, waved his hands, and loudly enquired of me and my other roommates: “Do you guys play StarCraft?” The year was 1999, and we were all computer science majors. Of course we did. And so, from that bold and unorthodox introduction we forged an immediate friendship born of video games and a shared interest in computers. We ended up living together for the next three years.

The thing about growing older is that it’s easy to lose track of the friends you made during your youth. Interests change, lives diverge, people move across the country, and before you know it, a decade can pass by almost imperceptibly. The last time I saw Evan was at his wedding over ten years ago. We’re now both in our mid-forties. In the time since I last saw him, I’ve also married, and we have both started families. We now have children old enough to do math homework, read novels, and give us both plenty of attitude.

A little over a month ago, I took a four-day trip to New York City for work, and while I was there I managed to spend an evening with Evan. We met up on the afternoon of a nearly-perfect, cloudless day in early autumn. The late-September sun sat low in the sky, and had that incandescent hue which suggested sweater weather was just around the corner. On that particular day, however, it was still warm enough to wander around the city in a T-shirt, so we rented a few Citi bikes and rode around Central Park before dinner.

Illustration by Mylene Cañeso on Unsplash

As we were pedaling around, trying our best to avoid other cyclists and the occasional wayward pedestrian, Evan asked me: “What are you doing for a hobby these days?”

Now, if he had asked me this question six months or a year ago, I might have excitedly enumerated the various side-projects and generative art pieces I was working on. Depending on the time of year, I could have talked about baking bread, learning the programming language Go, maybe waxed eloquent on the evenings I was spending messing around with my pen plotter. I could have shared the RSS reader I was building, or the Python library I had written for creating isometric line drawings. Like me, Evan is a software engineer; he wouldn’t have second-guessed any of these nerdy hobbies. But, the truth is, for the past few months I have not been doing any of these things. I have not been baking bread. I have not written any code outside of work. And my pen plotter sits on a shelf, collecting dust.

As any parent knows, when you have kids and you work full-time, you are lucky to have enough spare time for one hobby. Or, at least, you probably should have only one hobby at a time, lest you choose to neglect household chores, parental responsibilities, school volunteer activities, time with your spouse, or any number of the myriad other things life demands of you in your particular familial role and its attendant duties. For the past few months, my hobby—or at least what I’ve elected to do with my scant spare time—has been so new, so fragile in its infancy, I’ve been reluctant to call it a “hobby” at all.

As we climbed a gradual incline, the weight and inefficiency of my squeaky rental bike became increasingly apparent. I paused for a moment, debating what to say, but without the constant energy of my incessant pedaling, I rapidly decelerated. Evan continued coasting forward, and we were no longer facing each other. Unsure of exactly how to respond, I decided I might as well just share the closest thing to a “hobby” I currently had. Trying to catch up, I rapidly resumed pedaling, and between labored breaths I sheepishly blurted out: “Reading philosophy!”

Evan looked back and chuckled. “That’s not a hobby!”

The trouble with hobbies

It was the response I expected. Of course reading philosophy isn’t a hobby! But I didn’t say as much. Huffing and puffing, I remained mute. We crested the hill and continued pedaling onwards, changing subjects and talking of other things—family, travels, ambitions, the things you talk about with a friend you haven’t seen in ages.

Writing these words now, I realize Evan was right about philosophy not being a hobby, but probably for the wrong reasons. I couldn’t exactly put my finger on why, however, until recently.

For the past few months, as part of seeking new ways to be a better leader at work, I’ve been working with a coach. The other day I relayed this story to her. In response, she sent me this excerpt from the German philosopher Theodor Adorno’s essay Free Time:

I have no hobby. Not that I am the kind of workaholic, who is incapable of doing anything with his time but applying himself industriously to the required task. But, as far as my activities beyond the bounds of my recognised profession are concerned, I take them all, without exception, very seriously.

Now, most of us use the word “hobby” to describe the activities we undertake to occupy our free time, usually to avoid boredom. But in his essay, Adorno rails against both hobbies and the very concept of “free time.” Free time, argues Adorno, only exists because we allow it to exist in contrast to “work time”—“shackled” to its opposite, as he says, so that we may recharge our batteries for future labor-intensive endeavors. According to him, our desire for free time forces us to unwittingly split our lives in two. Thus, for Adorno, hobbies are nothing more than filler activities used for the killing of our supposed “free time” as part of our “rigorous bifurcation” of life.

Back when I lived with Evan, despite the academic workload, I recall having plenty of free time and, therefore, a few hobbies. StarCraft, for one. But if I’m to lean on Adorno’s definition, then reading philosophy, and writing about it on this blog as I have been doing for the past few months, seems to not qualify as a “hobby.” In fact, calling it a hobby seems to do it some kind of disservice, to lessen the value and effort and seriousness of it. Mostly it has been hard, slow-going, cognitively challenging, but ultimately rewarding—exactly what I signed myself up for.

Revisiting my autodidactic experiment

Earlier this year I kicked off a series of “life habit experiments,” small trials of new ways of living to explore alternative ways towards living a deeper or more fulfilling life. In late August, I embarked on the second of my life habit experiments, something I called “self-directed study.”

For this second experiment, I created a reading list and pledged (on this blog) that I would spend at least the next 60 days following the list and writing about what I learned. That experiment has become a multi-month autodidactic adventure teaching myself philosophy and reflecting on how my learnings apply to software design practice. I have since read multiple books—mostly on phenomenology—digging into the works of philosophers like Edmund Husserl, Martin Heidegger, Maurice Merleau-Ponty, even Byung-Chul Han.

Illustration by Daniel Lee on Unsplash

I largely undertook this experiment to satisfy my own curiosity and interests. Also in August, I published an essay, The imp of optimization, which was a philosophical rumination on our relationship with surveillance technology and the hidden agendas it often foists upon us. It was arguably the most popular essay I’ve written. After the newsletter The Browser featured it, I suddenly started receiving numerous positive emails from readers about how much they enjoyed reading it. I also enjoyed writing it. In fact, writing that essay reminded me how much I enjoy thinking about the intersections of philosophy, cognitive science, and design. So, I thought, why not experiment with doing more of it?

Unfortunately, I have a woefully shallow background in philosophy. As someone who studied computer science in college, I like to joke—borrowing a line from the late computer scientist Joseph Weizenbaum—that my computer science background makes me “poorly educated.” My entire undergraduate education consisted of math, physics, computer science, and little else. It wasn’t until I attended graduate school that I was properly exposed to the humanities. But, it’s been fifteen years since I left graduate school, and I’ve found myself pining to return to that atmosphere of deep learning, even if I have to create it on my own. Which is exactly what I’ve been doing.

Lessons in learning philosophy

One of the first things I learned from this experiment was that I was going to have to learn how to be a student again. And part of being a student entails a certain seriousness about reading—deeply engaging the material, taking notes, revisiting key passages, often reading multiple works about the same idea.

In previous essays I’ve extolled the virtues of “reading at Whim.” But this experiment was decidedly anti-Whim. In creating a reading list and following it with unwavering dedication, I forced myself to go deep on specific topics, to finish books I might have otherwise abandoned, and to read material that was immensely challenging—sometimes to the point of utter frustration.

Serious students of philosophy read primary sources. But I quickly learned how difficult this can be, and I wasn’t making it easier for myself in my reading selections. Anyone who has ever endeavored to read Heidegger for the first time understands how absolutely infuriating such an experience can be. Between the obtuse grammatical structure, the obscure, self-invented words translated from German—not to mention his affiliation with the Nazi party—jumping into Heidegger requires tenacity.

I leaned on companion books and secondary material to help me decipher and understand all of it. This included books like Sokolowski’s Introduction to Phenomenology, or Taylor Carman’s excellent Merleau-Ponty. But this clustered reading—reading multiple books on the same topics by different authors—has proven immensely valuable. Many of these philosophers write as if in “conversation” with each other. Husserl, for instance, was Heidegger’s teacher. Much of Heidegger’s work thus draws from and extends or criticizes Husserl. Similarly, Merleau-Ponty—who arrived on the scene after Heidegger—builds on Heidegger’s phenomenology, inserting his own perspectives and unique “flavor” of phenomenology focused on the aspects of embodiment. And Byung-Chul Han—a Korean living in Germany, and the most contemporary of the bunch—is arguably in conversation with all of them.

In reading all of these authors together, I learned more deeply than if I had approached each philosopher independently. I saw how certain ideas about the human lived experience shifted and transformed over time. And if, on first reading, a given concept felt opaque or confusing, encountering it again and again through another perspective would often bring new clarity and understanding.

I also (I will openly admit!) used AI to aid in my learning process. Used poorly, LLMs quickly become a crutch, a way to fool yourself and feign learning without actually learning anything. But, with a careful approach, a well-wielded LLM can act as a powerful learning partner, a tool for testing understanding and suggesting new avenues for discovery. I have a lot I can say on this topic, but I’ll have to save it for another day.

As part of my experiment, I shared my thoughts and learnings in an ongoing series of essays. As of now, I’ve published four of these essays: Phenomenology and the impossibility of experience design, Parts and wholes and moments, oh my!, Cézanne’s doubts, and Design for lingering. I am also currently editing a first draft of a fifth essay.

I have many more ideas for essays—too many, in fact—and much more I want to learn and read. I take this all rather seriously. I can already say, after a mere two months of experimentation, that I no longer consider the idea of learning and writing about what I learn on these topics to be merely experimental. It’s now something I both want and need to do. It is the thing I do outside of work that I take as seriously as work, except with one small difference: I’m not paid for it. My only constraint is simply: time.

Research as leisure

The thing to understand about Adorno is that he was an academic. And most academics enjoy something many of us don’t: their paid work and personal passions often overlap. Thus, they manage to avoid that pesky psychological divide between “free time” and “work time.” Adorno admits this privilege directly:

… I have been fortunate enough that my job, the production of philosophical and sociological works and university teaching, cannot be defined in terms of that strict opposition to free time … I speak as one who has had the rare opportunity to follow the path of his own intentions and to fashion his work accordingly.

I’ve observed this privilege firsthand in other professional academics. In grad school, in which I studied art and computer science, my thesis advisor was both a sculptor and avid sailor. His latest art project: designing and prototyping of sustainable, decarbonized boats for island and coastal communities. He’s literally building models—sculptures—of boats. If that’s not a near 100% overlap in the Venn diagram between personal intentions and paid work, I don’t know what is.

I don’t say this out of envy. I enjoy my work and my writing on topics of design and philosophy fortunately does overlap with my work, albeit the overlap is smaller in practice. But I’m not sure the amount of overlap really matters.

Ironically, in his essay Adorno mentions a contrasting term to “free time”—leisure—but then goes on to say very little about it. Whereas “free time” exists only in contrast with time reserved for labor and work, leisure sits beyond this uncomfortable dichotomy. We typically think of the leisurely life as one unconstrained by the demands of work, unconcerned with the difference between “free time” and “work time.” Adorno calls this life “auspicious;” Marx might say “bourgeois.”

But what if leisure isn’t just for the privileged few? Aristotle, for instance, saw leisure as essential to a fulfilled life. In the Nicomachean Ethics, he argued that leisure plays a central role in living well and achieving eudaimonia—a Greek word often translated as happiness, but better understood as “human flourishing.” The Aristotelian view of leisure thus distinguishes it from “free time” in both approach and attitude. For Aristotle, leisure was not time set aside for mere relaxation, nor should it serve as “a remedy for the ills of work.” Rather, he intended that it serve the higher purpose of cultivating one’s virtue and character. Philosophic contemplation he considered to be the highest purpose of all.

There’s a wonderful essay by Celine Nguyen, in which she discusses the idea of “research as a leisure activity.” I loved this particular framing of one’s peculiar autodidactic pursuits as leisure. It feels very Aristotelian. Nguyen writes:

The idea of research as leisure activity has stayed with me because it seems to describe a kind of intellectual inquiry that comes from idiosyncratic passion and interest. It’s not about the formal credentials. It’s fundamentally about play. It seems to describe a life where it’s just fun to be reading, learning, writing, and collaborating on ideas.

I agree: anyone with the fortitude and wherewithal to pursue research and writing should feel free to do so, formal credentials be damned. But if we are to draw on the ancient Greek understanding of leisure, then we should be careful not to allow words like “fun” and “play” introduce too much frivolity either. Leisure can be serious, as Adorno seems to be reminding us.

Taking it seriously, however, doesn’t mean our leisure pursuits should feel laborious. We need not turn our leisure into a second kind of work. We don’t have to make it a “hustle culture” side-project of soul-sucking toil. Rather, leisure should feed our souls. We should relentlessly seek out those activities that challenge us, lift up our spirits, and help us grow. Fun, sometimes, but also plenty serious.

What Adorno seems to be saying is that the difference lies in whether we feel genuinely free in our pursuits, or whether we simply bring to them the same compulsions that govern our working lives. Taking leisure seriously means refusal to reduce everything to either work or its shadow. The difference between free time and leisure time is thus not so much about fun as it is about freedom.

Looking back on it now, I believe this is what Evan had in mind when he jokingly said that philosophy is not a hobby. What he meant was that reading philosophy and writing about it doesn’t sound fun. And I will admit: it has not been fun, at least not all of the time. It has been challenging. This is, however, the ironic joy of it. But to call it work feels wrong too. Instead, I think it exists somewhere in that same rare space that Adorno also carved out for himself: the space for serious pursuits “beyond the bounds” of “his recognised profession,” whatever that may be.

Design for lingering

2025-10-13T00:00:00Z

The hsiang-yin (香鐘 or, in pinyin, “xiāngzhōng”) is an ancient Chinese timekeeping device first developed during the Song dynasty. Used in China up until the late 19th century, the hsiang-yin (literally “seal of fragrance”) is unlike Western timekeeping devices. It isn’t mechanical, and doesn’t measure time in discrete chunks of hours, minutes or seconds. Instead, it measures time through the slow burning of incense.

In his book The Scent of Time, Korean philosopher Byung-Chul Han uses this device as a metaphor for exploring alternative, healthier relationships to time. Evoking the smoky, contemplative atmosphere of a room full of burning incense, Han calls for an approach to life that similarly allows time and space for lingering, contemplation, even boredom.

Illustration by Rifky Nur Setyadi on Unsplash

Throughout much of his work—both in The Scent of Time and elsewhere—Han often reflects on our sense of smell—a sense which, though relatively underdeveloped in humans, nevertheless requires duration. Scents cannot flash before us the way images can, or instantaneously resound in our ears the way sounds do. No, scents linger in the air. They are slow; they do not readily dissipate. Consider Han’s reflection on the hsiang-yin, which he calls a device of “fragrant time:”

As a medium for measuring time, incense differs in many respects from water or sand. Fragrant time does not flow or trickle away. Nothing is emptied. Rather, the scent of the incense fills the room, even turns time into space; it thus gives it a semblance of duration. Although the glow permanently transforms the incense into ash, the ash does not disintegrate into dust.

The temporal crisis

In The Scent of Time Han argues that our modern lives suffer from a “temporal crisis,” a shared sense that time is accelerating out of control, “whizzing by without a direction.” For most of us, this temporal crisis is something we feel more than we can easily describe. It is that fleeting disorientation after we wake up from hours flitting about on social media, doomscrolling, wondering where the day went. It is that strange amnesia of a few hours we can’t quite account for after falling down some YouTube rabbit hole, bouncing from video to video. It is our struggle to describe to friends and family the productive use of days spent jumping from Zoom to Zoom, or Slack thread to Slack thread.

Throughout much of modern life we have intentionally “atomized” time, breaking it down into smaller and smaller disjointed increments that have no overall narrative continuity. And it is this atomization, Han argues, that has led to our collective perception that everything is somehow speeding up.

This is easiest to notice in social media. Consider platforms like Instagram and TikTok. Both of these platforms allow us to rapidly consume media experiences in tiny, bite-sized increments that have no relationship beyond their temporal adjacency. Open TikTok, and find yourself immediately greeted by a popular video of some random teenagers acting out a new dance. As soon as the playhead reaches the final second, you will find yourself immediately transported to the next video, perhaps of a cute dog and cat cuddling, or a disturbing clip of a war in a far off land. Tap open Instagram, and once you finish quickly glancing at that photo of our coworker’s beautiful Icelandic vacation, you find yourself mindlessly swiping to the next, this one an image from a close friend’s photos of a recent government protest.

These unceasing streams of content—rapid-fire visual after visual—are specifically designed to prevent us from engaging in any meaningful reflection or contemplation on their content. They actively prohibit lingering.

Social media isn’t the only culprit here either. Instant messaging and Slack allow us to similarly move unreflectively from conversation to conversation. The web allows us to move from reading experience to reading experience before we’ve had time to fully absorb the words we have read. Even email allows for rapid-fire, unreflective correspondence that, as Han writes, “does not need to conquer mountains and oceans,” abolishing “space itself.”

Han’s philosophy has grown in popularity because he so readily puts his finger on what many of us innately feel about our lives, but also feel helpless to overcome. We want duration. We want contemplation. When I sit and reflect on my own “ideal day,” for instance, I find that what I seek most is a feeling of lasting, deep flow state. It’s not that my ideal day is devoid of work, but rather that I want it to be characterized by an experience of time as long and deep and meaningful. And I’m not alone in this. This longing for duration is precisely the reason books like Deep Work by Cal Newport have become so popular.

Technologies against the vita contemplativa

Han’s suggested antidote to the temporal crisis lies in cultivating a contemplative life, a vita contemplativa, or what he says is simply a willingness to “linger.” Lingering—the ability and desire to sit with our experiences for extended periods—can act as a counterbalancing force against the rapid atomization of time because it allows us to experience real duration. Lingering actively prevents the atomization of time.

While Han does not directly point his finger at modern technology as the cause of our temporal crisis, throughout The Scent of Time he repeatedly criticizes how it actively prevents contemplative reflection. For instance, of the Internet and the web he writes:

The time of internet space is a discontinuous and point-like Now-time. You move from one link to the next, from one Now to another. The Now does not possess duration.

Of email (which I already mentioned above) he writes:

Electronic mail produces instantaneity by destroying the paths as spatial intervals in their entirety. It dispenses with space itself. Intervals are destroyed in order to produce total proximity and simultaneity.

And of stored “electronic memories” like photo albums he writes:

Electronic memories and other technological possibilities for recurrence destroy the temporal interval which is responsible for forgetting. They make what is past instantaneously retrievable and available.

Han’s critique of digital media, tools and experiences left me wondering: Does it have to be this way? As Terry Winograd and Fernando Flores once wrote (in a quote I’ve grown fond of repeating):

We encounter the deep questions of design when we recognize that in designing tools we are designing ways of being.

There’s nothing inherent in technology that reinforces the temporal crisis other than how we have chosen to design it. We could design things differently.

So, how might we design new digital tools that might encourage a different, healthier relationship to time? How might we design technologies that promote lingering?

Design interventions (and a few caveats)

First, some caveats: I recognize that Han is unlikely to condone the thought experiments that follow. I don’t think he’s likely to suggest that more or different technologies will somehow solve or alleviate the temporal crisis. This whole effort risks seeming solutionist. Most of his philosophy points at the deeper cultural underpinnings of the temporal crisis—something he calls the “positive violence” of our “achievement society.“ And nothing mentioned here will solve this more fundamental cultural problem. Nevertheless, the pragmatist in me believes that there remains value in using the temporal crisis as a lens for exploring new avenues for design. Computational technologies are here to stay, after all, and so it is imperative that we design them in ways that promote human well-being.

To begin, it helps to consider a few apps that already promote healthier relationships with time by encouraging mindfulness, focus and presence. Focused writing tools, for instance, such as Hemingway Editor or iA Writer (the tool in which I’ve chosen to write these very words) offer environments that encourage us to sit with our creative endeavors for long, uninterrupted periods. Similarly, photography apps such as Hipstamatic emulate the experience of shooting on film, in which we cannot instantaneously review the results of our photography, forcing us to stay present in the moment, observant of our surroundings. And modern minimalist hardware such as the Light Phone or reMarkable tablet, both of which lack web browsers and unfettered Internet access, encourage us to linger on whatever work or experience is directly before us.

Given the examples above, we can easily imagine design interventions for future technologies or products that also encourage a more mindful, present, lingering type of experience.

Slow release content

We might consider designing platforms that release content slowly, over time, at regular intervals, encouraging contemplative pacing. For instance, imagine a reading app that releases small chunks of a philosophical text over days or weeks, making it impossible to binge-read ahead—forcing you to sit with each chapter for a full day before continuing. Or imagine learning systems that ask us to sit for extended periods of time on a single topic or subject.

Depth-over-breadth

Recommendation algorithms often encourage us to jump from topic to topic, or article to article, before we’ve really had time to digest what we just read or watched. We can instead imagine algorithms that actively resist this pattern. We can imagine algorithms that provide recommendations which encourage you to go deeper on a specific topic, or perhaps push you to return to the same works repeatedly instead of endlessly branching out into adjacent topics or unrelated works. Consider a music streaming service, for instance, that encourages you to listen to the same album at least five times before introducing anything new, cultivating the familiarity that transforms casual listening into genuine appreciation.

Intentional asynchronicity

The speed with which most communication occurs over digital media encourages us to divide our time amongst multiple simultaneous threads. We can push out messages more quickly than we can consume and think about them, which means we inevitably jump around from conversation to conversation. As such, we might consider tools and platforms that deliberately delay responses and interactions, creating more time for deeper reflection.

Imagine an email client that holds your outgoing messages for an hour, giving you time to reconsider them before sending. Or, how about a messaging app that calculates the geographic distance between correspondents and introduces delays proportional to the time the correspondence would have taken 100 years ago? A message from across town might arrive in an hour; one from across the country, in three days.

Beyond temporal interventions

While these interventions and examples point us in the right experiential direction—reduced distraction, promotion of focus, temporal spacing—I’m not sure they’d hit the mark.

While Han doesn’t specifically discuss it much in The Scent of Time, core to his philosophical project, as described in his other books, is the idea that by over-focusing on our own needs, advancements and personal projects, we have lost sight of the “Other.” That is, we spend too much time thinking about ourselves and not enough cultivating love for, understanding of, and relationships with other people. Han calls this “friendliness,” but it might be better characterized as openness to the world.

Keeping this in mind, we can’t just design technologies that reduce or eliminate distraction if we really want to make any sort of dent in this problem. We need to go further. We would also need to design experiences that encourage care and respect—interventions that move away from optimizing individual experiences towards creating conditions of genuine encounter with other beings.

Reciprocal engagement

Instead of platforms that promote metrics-driven engagement, or allow for one-sided content sharing, we might imagine ways of promoting more reciprocal engagement. For instance, consider an app that doesn’t show other peoples’ content until you’ve shared your own. Or imagine a discussion platform that enforces a question-to-statement ratio, one that requires every declarative claim be balanced with genuine curiosity about another’s perspective.

Time gifts

Imagine technologies that encourage us to think of time as something we give to others. For instance, consider systems that require “attention pledges” before you can access and review someone’s writing. Would you still read this essay if you were first prompted with: “This piece takes 12 minutes to read thoughtfully. Are you willing to give this time?” Or, rather than gatekeep upfront, imagine systems that share “read receipts” which include the amount of time a person spent with another’s work.

Shared contemplation

Lastly, we might consider shared spaces that allow us to linger, virtually, together, turning solitary contemplation into a communal experience. Imagine a virtual reading room where you see the images of other readers working through the same book, their pace visible as they turn pages alongside you. Or picture a system that reveals new stories, images or content only when enough people have “gathered” for the telling.

Spaces for savoring time

I recognize that these ideas seem unusual, and sit uncomfortably alongside commercial realities. Deliberately slowing down engagement is not what most designers have been asked to optimize for. But not every product need embrace these ideas wholesale, and in some cases these ideas might lead to a worse overall experience. Nevertheless, even small gestures towards a more contemplative experience, whether that means opting away from infinite scroll or defaulting away from auto-play, acknowledge that our relationship for time matters. Maybe there’s room, even at the margins, to design with lingering in mind.

While I’ve never actually seen a hsiang-yin or tried measuring time by burning a stick of incense, as a long-time practitioner of Zen, I am familiar with the “scent of time” in other, similar ways. My deepest experiences of lingering, relaxed duration have often occurred on long, multi-day silent Zen meditation retreats called “sesshins.” There, in the meditation hall, where the scent of incense hangs in the air and the sound of the bell rings out with precise regularity, time can stretch to an almost painful longevity. Sitting, cross-legged, or perched on a small, wooden bench, forced to simply “be,” undistracted, unproductive, moment to unending moment … this is the antithesis of the temporal crisis. Deep in sesshin, time is not something we chop up and dole out; it is something we savor.

For me, the Zen meditation hall hints at what more digital spaces might aspire to be: places where we slow down and savor time with others. As I said above, the temporal crisis won’t be solved by technology, but perhaps thoughtful design can create small refuges from the whizzing “now,” spaces where true duration becomes possible.

Cézanne’s doubts

2025-10-02T00:00:00Z

Several years ago, not long after my wife and I first started dating, we took a trip to visit her family in Philadelphia. It was a cold and rainy December, but I had never been to Philadelphia before so, despite the miserable weather, we tried to make the most of it. We hit all the obligatory tourist stops the city had to offer—Independence Hall, the Liberty Bell, the Italian Market, the cheesesteak stands, of course, the Philadelphia Museum of Art, the steps leading up to which were made famous by the movie Rocky. But, the one destination we failed to make time for during that trip was The Barnes Foundation, a unique (and controversial) institution that houses and exhibits a massive private collection of art assembled by the chemist and businessman Albert C. Barnes in the early 20th century. Fortunately, we returned to Philadelphia a few months later, this time during a mild and sunny week in June, and soon made our way to the sleek stone and glass building that houses the collection.

By that time in my life, I had visited my fair share of art museums, but The Barnes Foundation left a lasting impression. To this day, over ten years later, I remember it vividly. Unlike a traditional art museum, The Barnes Foundation exhibits art in curious “ensembles”—juxtapositions of paintings mixed and matched with everyday objects like door hinges, glassware and furniture. The foundation’s collection includes works by Van Gogh, Matisse, Rousseau, Picasso, and many others. But some of my favorite paintings in the collection are those by Paul Cézanne.

Mont Sainte-Victoire by Paul Cézanne. Courtesy of The Barnes Foundation.

Cézanne was a famously volatile artist. He was plagued by depression, and known for a temperament that swung between periods of immense passion, and periods of intense self-doubt. This doubt, however, may have proved immensely productive—indeed it might have been the driving force behind Cézanne’s artistic greatness.

Art and embodiment

In 1945, the French philosopher Maurice Merleau-Ponty (which I will hereafter abbreviate as M-P) published Cézanne’s Doubt (PDF), a short essay exploring the ties between Cézanne’s unique artistic method, his intense self-doubt, and M-P’s own take on phenomenology.

Much of M-P’s philosophical project involved trying to revise phenomenological thinking in order to emphasize the criticality of embodiment. That is, M-P argued that the human body is central to how we experience the world, not something to consider as an afterthought. He claimed our bodies are neither subject nor object, but rather a kind of ambiguous third party that serves as the source of all perception. And in Cézanne’s Doubt, as we shall soon see, M-P argued that Cézanne’s embodiment was in fact central to how we perceived and chose to depict the world through art.

In the opening to the essay, M-P writes of Cézanne:

It took him one hundred working sessions for a still life, one hundred fifty sittings for a portrait. What we call his work was, for him, an attempt, an approach to painting … Painting was his world and his mode of existence. He worked alone without students, without admiration from his family, without encouragement from the critics. He painted on the afternoon of the day his mother died. In 1870 he was painting at l’Estaque while the police were after him for dodging the draft. And, still, he had moments of doubt about this vocation. As he grew old, he wondered whether the novelty of his painting might not come from trouble with his eyes, whether his whole life had not been based upon an accident of his body.

A post-impressionist, Cézanne sought to paint not merely what he saw, but rather to depict his more primordial perception of the world. You could say that he was, in his own way, a kind of artist-phenomenologist; he painted what he perceived rather than merely what he conceived. He sought sensation over judgement, to be a painter who sees beyond what he thinks. Indeed, in an article in Le Salon, he even once declared: “I paint as I see, as I perceive.”

But Cézanne’s conception of perception went beyond mere visual sensation. His was rooted in his embodied, lived experience. For him, it seems that art-making served as a means for exploring what his unique experience of perception felt like in his body, much in the same way writers use the process of writing to understand what they think.

Cézanne’s first doubt

Cézanne’s unconventional approach to painting meant his work often failed to conform to traditional approaches to color, technique and perspective. During much of his lifetime, contemporary critics regularly attacked his work. One critic called his art “the painting of a drunken privy cleaner.” It’s little wonder, then, that he doubted whether his art would ever amount to much.

But Cézanne’s self-doubt did not just grow out of a few bitter seeds planted by critics. It was also intrinsic to his aspirations and way of being. He set an astonishingly high bar for himself. Indeed, in his pursuit to paint exactly as he perceived, he also chased the “perfect work of art” that he saw in nature itself. As M-P writes:

Cézanne … conceived painting not as the incarnation of imagined scenes, the projection of dreams outward, but as the exact study of appearances: less a work of the studio than a working from nature.

Of nature, he said that “the artist must conform to this perfect work of art. Everything comes to us from nature; we exist through it; nothing else is worth remembering.”

This is the first doubt of Cézanne’s Doubt: the existential doubt of an artist troubled by an overwhelming desire to express something through painting, while simultaneously doubting whether such expression was ever possible.

Cézanne’s second doubt

If we were to stop here, we would learn something about Cézanne and his motivations, but I also feel we would miss the entire point of M-P’s essay. We would miss the deeper, second doubt of Cézanne’s Doubt: the methodological doubt of the artist-phenomenologist.

Core to the practice of phenomenology is something called the “epoché,” which is just a fancy term for “bracketing off” our normal preconceived notions about how we both think the world works, and how we should experience it. The epoché does not imply a need to doubt everything we know, but rather asks us to allow for the possibility of doubting our everyday beliefs in order to get to the root of subjective experience. It is, in short, a willingness to entertain doubt. And M-P argues that Cézanne used something akin to this method in his own artistic practice to move beyond mere seeing and express his own embodied perception of a scene. We find evidence of this in his paintings.

At The Barnes Foundation, I remember standing transfixed before Cézanne’s “Mont Sainte-Victoire,” for a long time. It’s one of my most vivid memories of my time there. What it was about this particular painting that kept me lingering in place for several minutes, I still can’t say. On the surface, Mont Sainte-Victoire appears to be just a painting of a mountain on a clear day, foregrounded by fields and farmhouses, with what appears to be an ancient viaduct somewhere in the middle distance. But, a closer inspection reveals the fascinating ways in which Cézanne chose to depict this scene. The mountain, for instance, supposedly in the distant background, instead looms vividly in bright color, almost startlingly close and present. The geometry of the various buildings that dot the landscape looks slightly “off,” not respecting any traditional or “correct” methods of perspective. And Cézanne’s signature parallel, vertical brushstrokes seem to draw both the trees and mountains upwards, as if actually growing out of the ground.

Cézanne’s phenomenology

Cézanne seemed to perceive the landscape much differently than other painters I’d ever encountered. As M-P explains:

His painting was paradoxical: he was pursuing reality without giving up the sensuous surface, with no other guide than the immediate impression of nature, without following the contours, with no outline to enclose the color, with no perspectival or pictorial arrangement.

Toward Mont Sainte-Victoire by Paul Cézanne. Courtesy of The Barnes Foundation.

Cézanne was known for projecting space and representing perspective in non-traditional ways. The art historian Paul Smith argues this was due to the way Cézanne paid close attention to his embodied experience of vision. Citing the work of psychologists A. David Milner and Melvyn Goodale, who claim we have both systems of “vision for perception” and “vision for action,” Smith argues Cézanne may have been trying to capture these two ways of seeing in a single image. He quotes a conversation Cézanne had with the artist Jules Borély in 1902:

Look, that tree trunk: between us and it there is a space, an atmosphere, I grant you that. But then again it is this palpable, resistant trunk, this body … See like someone who has just been born!

Two things stand out in this snippet of conversation. First, as Smith points out, there’s the this in “this body.” If we are to agree that we have two ways of seeing—both for perception and for action—then Cézanne seems to be indicating that the tree trunk is simultaneously distant and close at hand. For such a thing to seem possible we have to reconcile with how we experience these two ways of seeing. Smith notes that in vision for perception we notice the distances separating objects. Whereas because vision for action represents our own “tactile space,” we pay closer attention to the space between us as perceiver and the perceived object. This, I think, is what I noticed in my long inspection of Mont Sainte-Victoire—a distant mountain made simultaneously “palpable” and “resistant” through color and size.

Second is Cézanne’s exhortation to “see like someone who has just been born!” It’s as if Cézanne is telling Borély to abandon habitual ways of seeing to return to some more primordial state—to adopt a true “beginner’s mind” approach to vision itself. But notice how physical and embodied this metaphor is. A newborn’s vision isn’t just conceptually innocent—it’s physiologically different. Newborns are still learning to focus, to separate figure from ground, and to judge distance. The world hasn’t yet separated into discrete objects but exists as a unified field of sensation. I think this is what M-P found most compelling about Cézanne: his desire to return to a more fundamental, embodied way of seeing.

What we can learn

Cézanne was a painter, not a designer, so it might be natural to ask at this point: What does any of this have to do with design? What can we learn? Here are some possible takeaways.

Look at process over product

First, there is value in doubt if we learn how to leverage it productively. Cézanne’s almost maniacal commitment to capturing the “perfect work of art” he saw in nature, however unobtainable, allowed him to embrace process over predetermined outcomes. Cézanne was notorious for abandoning his work. As art historian Harry Cooper writes, his “oeuvre is littered with paintings in various states of incompletion.” He dedicated himself to his artistic process above any desire to merely finish.

Now, design cannot be all process and no results. Cooper also tells us that Cézanne had inherited his father’s estate in 1886, affording him the luxury of no longer needing to regularly sell his artwork in order to maintain a living.

Design has no such luxury; it often serves the needs of business, and must deliver. But there’s fertile ground in embracing iterative approaches that remain open to change, rather than committing too early to specific concepts. And even within constraints, we can steal Cézanne’s trick: treat each iteration as an investigation, not just a step toward completion. The best insights often emerge from work we consider unfinished.

Embrace methodological doubt

Next, perhaps there is something to Cézanne’s epoché, his deliberate suspension of what he knew about mountains, buildings, color, and perspective. It suggests a radical—albeit controversial—tool: methodological doubt. We’re often told to rely on best practices, design systems, and user research data. These are valuable, even essential. But what happens when we bracket them off to see what else might emerge? Perhaps we might arrive at more interesting solutions.

For designers, this kind of methodological doubt can serve as a practice of looking deeper, seeking to identify what we’re not seeing because we think we already “know.” How might this work? There are numerous exercises we can use to allow for the possibility of doubt and see things with fresh eyes. For visual design this might mean trying to describe a design using only words for texture, temperature or sound. Or maybe it means revisiting a design first thing in the morning, before paging in the entire context of a project. For other design work, we can consider perspective shifts such as intentionally designing the worst possible solution, or listing all your constraints and designing as if the opposites were true.

The phenomenological epoché isn’t about deliberate ignorance. It’s not about intentionally disregarding best practices for no particular reason either. Cézanne knew the rules of perspective perfectly well; he chose to break them in order to seek out something deeper, something more true to his experience by not accepting the rules as “fact.”

Use perception as material

Lastly, Cézanne used perception as raw material. He didn’t paint mountains; he painted his perception of mountains. Designers can adopt a similar approach. When we design a loading animation, we’re not merely filling time; we’re designing the perception of duration. When we choose between a sharp corner and a rounded one, we’re working with how edges feel to the eye, how they guide or arrest movement across a screen. When we design for accessibility through well-contrasted color choices, we respect the needs of people who may perceive the world differently than ourselves.

To do this well, we must recognize that perceptual material behaves differently than purely visual material. First, it’s culturally influenced and contextually defined. The color red, for instance means “error” or “stop” in one context, and “luck” in another. Perceptual material is also physiologically grounded. We can’t help but notice motion in our peripheral vision. And we’re forced into specific sequences of attention simply by how blurring background objects. When Cézanne painted what he perceived rather than what he merely saw, he was playing with these properties of perception too, showing how distance and nearness could coexist, or how solidity and atmosphere interpenetrate.

Mountains as meditations

Cézanne painted Mont Sainte-Victoire more than thirty times in his lifetime. He called this particular mountain his “motif,” returning to it again and again in order that he might not just learn to see the landscape, but that he might learn to really experience it. He seemed to never tire of this view.

In the middle of his essay, M-P describes at length how Cézanne approached painting this particular scene:

He would start by discovering the geological foundations of the landscape; then, according to Mme Cezanne, he would halt and look at everything with widened eyes, “germinating” with the countryside … Then he began to paint all parts of the painting at the same time, using patches of color to surround his original charcoal sketch of the geological skeleton. The picture took on fullness and density; it grew in structure and balance; it came to maturity all at once. “The landscape thinks itself in me,” he would say, “and I am its consciousness.”

To paint Mont Sainte-Victoire over and over again was, for Cézanne, part of an ongoing conversation between mountain and paint, perception and expression. It was, in short, a meditation—not a search for the right answer, but a practice of remaining available to what wants to appear. Sometimes design can work like this too: in remaining attuned to our own perception, problems can think themselves anew through us, and then, just maybe, the right solution will naturally emerge.