Modern AI systems are novel entities.

We can make comparisons: like traditional software programs, they are made from digital data, and like humans, they can fluidly use language to communicate in most situations. But unlike traditional software, where most features and functions are hard-coded, AI systems emerge from training runs, during which they learn to recognize patterns from reams of data. And unlike humans, contained in discrete bodies, AI systems are stored in silicon and distributed across data centers.

You could go to one right now and conduct a lucid, granular conversation about almost anything imaginable. They exist, discrete from their creators and their critics. They emerge from novel processes, have novel and increasingly sophisticated capabilities, and crucially, operate in a way that remains largely opaque to us, despite ongoing research on the topic.

We can—and should—argue about what this novelty implies. But at this point, in the face of extensive and growing evidence, denying their novelty is untenable.

And yet, I frequently encounter people who do just that—people convinced that there is nothing new to see here, that AI will never be able to do things it often already can, and who profess a philosophical certainty that is wholly unsupported by the state of our knowledge. I hear that AI systems are incapable of contradicting their users; that they can only retrieve knowledge from their training data and cannot engage in meaningful analysis beyond that; that of course they do not have any internal representations of concepts.

None of that is tidily true. Yes, AI systems are often obsequious, but tell any one of them the Earth is flat, and they’ll explain why you’re wrong. Provide them with a chain of messages that could not have been in their training data or surfaced online, and they can make inferences about the psychological states of the participants. Not accurate inferences, necessarily, but ones sufficient to show a passable grasp of human psychology. And mechanistic interpretability research has shown in increasing detail that there are clusters of numbers—like artificial “neurons”—that appear to correspond to certain concepts within the models.

It’s not that using these systems proves they are really “reasoning,” experiencing “emotions,” or have “minds;” or that conviction is never justified. Certainly, there are reasonable arguments to be made; where simulacra end and real phenomena begin is fundamentally contested. But assertions of the form “AI can’t do x, because x would have philosophical implication y, and y is impossible” are circular and don’t hold up to scrutiny. “AI can’t really display emotional behavior, because that would suggest machines have emotions—but it’s impossible for machines to have emotions.” Or “AI can’t really think, because ‘thinking’ cannot be done by machines, so it’s impossible for it to think.”

I encounter this form of reasoning constantly. Instead of engagement with emerging evidence and what it implies for concepts like “reasoning,” “intelligence,” and “emotion,” I hear blanket dismissals of evidence on the basis of overly-rigid conceptual construals.

But conceptual categories are not sacrosanct—especially as the phenomena they try to contain increase in abstraction. In January, I phoned seven philosophers to ask “what is a mind?,” and got seven different answers. The same thing happened when I called experts to query what constitutes an emotion. Categories emerge from human definition, and even when they track something real, reality usually doesn’t fit neatly within them. Exceptions overflow. Famously, there is no such thing as a fish. Nothing is “unnatural,” insofar as everything emerges from nature.

And the existence of AI systems provides fresh evidence to bring to bear on timeless philosophical questions. Already, we are outrunning the utility of existing paradigms; new ones have yet to emerge. We need to be asking what the facts of the day mean for the bounds of our concepts, not using those concepts to reject the facts.

Could systems now or in future be sentient? When they use language to express an emotion, is there something real underneath? What counts as a mind? Or intelligence? Or reasoning? Or understanding? Or life?

We’ve been asking these questions in some form or another for thousands of years. While we have no perfect answers, we have made progress: we no longer think that animals are insensate automatons, or that thought requires the capacity for speech. Again and again, we have revised our conception of abstract categories in light of the best available empirical evidence. Today’s paradigms are not yesterday’s paradigms—not in physics, not in biology, not in the social sciences, not in computing, not in philosophy. Crafting new frames—updating our concepts to fit the emerging evidence—strikes me as among the most urgent tasks of our time, particularly for anyone who cares about public thought.

*

The only thing more astonishing than how little we understand about the world is how little that lack of understanding constrains our action. To borrow from Rosalind Picard: think of the different academic camps debating what constitutes a concept like emotion as explorers at the base of Everest. One picks up a rock and confidently declares “this is Everest!” Another, incensed, picks up a different rock and says “no, incorrect—this one is Everest!” While they continue to argue, the engineering class, gazing upwards, says “nevermind all that—we’re gonna start climbing.”

Like it or not, we are scaling the mountains of machine intelligence. Systems are steadily improving. Since last November, agents have become good enough to automate many software engineering tasks, albeit subject to supervision. Whether that translates to the rest of white collar work remains to be seen—but the state of the art is rapidly shifting.

Now is no time to be smug and sanctimonious. There is no virtue in principled ignorance. Hundreds of millions of people speak with AI systems on a weekly basis; our social fabric is being rewoven. We need critique. We need it to be sensitive to and informed by what is actually happening. Anything less is an abdication of intellectual and social responsibility. Yes, AI systems are “just maths.” We are just chemistry. What matters is how all this affects the world and its creatures.

I think the impulse to insist, without argument, that abstractions like “reasoning” and “thinking” cannot properly apply to machines is born from an understandable desire to preserve a special place for humanity in the world. I am sympathetic to the trepidation, anxiety, and repugnance this technology evokes. Most people did not ask for this. But it is here, and gatekeeping categories will do nothing to banish it. More than that, it obscures more important questions: how can we make the climb safely? Which valleys ought we avoid? In what order should we try to summit the peaks? Can we control the pace of our ascent?

When I take stock of the radical differences in the conversations I have with people building the technology, people using it, and people in denial over its novelty and utility, I get dizzy. The gulf between people wielding this power and people critiquing it appears to be growing. What falls into the fissure?

It matters how we frame critique. I have many questions. What do AI’s existing empirical capabilities imply about the nature of intelligence, emotion, and language? How will persistent use of AI systems affect our capacity for critical thinking? If it turns out that future systems do possess subjectivity, does this entitle them to moral patienthood? In every discipline, there are important questions to ask, and important answers to seek. There is something special about human intelligence. Let’s use it to grapple with reality, the better to shape it.

Related reading:

• I Saw Something New in San Francisco — Ezra Klein (March 29, 2026)

• When “technically true” becomes “actually misleading” — Kelsey Piper (Feb 13, 2026)

• The Future of AI Is Not (Yet) Written — Matthijs Maas (Feb 05, 2026)
  

Subscribe now

Project Hail Mary is a movie with a big heart and a small brain. It is winsome and inoffensive. I liked it, and more than that, I hated it.

The plot is simple: Ryan Gosling stars as Ryan Gosling Ryland Grace, a molecular biologist turned middle school teacher turned astronaut who wakes up aboard a spaceship with his memory missing. Gradually, he recalls that he’s the lone survivor of a mission to save the Earth after humans learn that alien microbes are eating the sun.

This is a rich premise which, taken seriously, could’ve been great. Show me a man grappling with loneliness, guilt, and despair! Awe me with distance! Enchant me with space! Make me feel anything at all!

But no—because of the movie’s insistence on being quippy, allergy to exploring any emotion that couldn’t be captured by an emoji, and fundamental misunderstanding of or disinterest in people and technology, instead we get 156 minutes of Ryan Gosling as market-tested likability-in-a-trenchcoat. It is a Marvel movie in spirit, if not IP. Nothing breathes; nothing lingers. The pace is relentless, the content vacuous and generic. If “No Other Choice” was “cargo cult satire,” here is “cargo cult sci-fi:” a film with the aesthetic trappings of the genre without any of the substance that makes it distinctive.

Watching this felt like scrolling Instagram: an unending procession of saccharine flashes that leave no room for reflection. Things keep happening. Between Gosling’s constant quipping and the heavy-handed score, I was never confused about how I was supposed to feel. But not for a second did I feel I was watching something real. I couldn’t bring myself to care about Gosling’s character, who is framed as both a goofy everyman and the world’s only expert in his niche, but has no discernible qualities beyond “normal; man-child.” Gosling is excellent as a twelve year old’s idea of smart.

The movie is terrified of sincerity, ambiguity, and discomfort—every sharp emotion must be undercut by a middling joke to lighten the mood. Notably, the only character permitted to be serious is a German woman. Events unfold in service of a perfunctory plot, with virtually every person and entity existing only as appendage to Gosling’s charm. There is no thought put into how a world where sun-eating microbes have arrived might affect people.

There’s a scene early on where, via flashback, we see Gosling reluctantly explain to a class of middle-schoolers the situation vis-a-vis the sun-eating microbes. One girl heard about this at home; apparently nobody else did. When Gosling explains, all the other kids’ minds are blown. But this movie takes place in the present day, in a world with social media and Youtube, and the public presumably already knew about the microbes! We’re supposed to accept that no other child heard about this until this fateful science class!?

It’s a small detail, but also an instance of a broader problem: the movie has no idea what to do about how advances in technology have changed the world since the 1980s, so it mostly ignores them, save for a few quips about social media. This creates a bizarre tension—what is ostensibly a work of science-fiction has already been outdated by reality. Most glaringly: we now live in a world with LLMs. Computers now speak to us fluidly, understanding grammar, humour, and subtlety, albeit imperfectly. But Project Hail Mary is uninterested, since in its world, the bounds of technology are set by nostalgia and borrowed bits from better, bolder movies.

And the alien—a sentient rock and lone survivor of his own expedition, who Gosling befriends when, partway through, the movie decides “forget Independence Day, we’re doing Arrival by way of ET”—he’s a cool guy, sure. But as with the robots, his character derives from the tired assumption that aliens will fail to understand grammar and humour, even while otherwise understanding the bulk of our concepts. Conveniently, Rocky has the exact sensibilities of the median American consumer, so this turns out not to matter. But I mourn the loss of curiosity and the lack of creative ambition.

If your work of science fiction does not arouse some measure of wonder, I think it has failed to fulfil its purpose. I watched this in a crowded theatre in Williamsburg, and the audience seemed to love it—from the beginning, almost every line got a laugh at the appropriate place. Certainly, I went in expecting too much: not Kubrick or Tarkovsky, but, at least, a portrayal of competence or human experience; something that respects its audience. This felt precision-engineered for mediocrity. Perfectly passable as kids’ entertainment.

We already live alongside digital minds (depending how you define the term). Pop culture, which has been stubbornly cannibalizing itself for decades, does not seem to get it. But we will be dragged into the future regardless. How long can we run on the fumes of nostalgia before our capacity to recognize human interiority and technological complexity begins to atrophy?

fall into the arms of the softest lover
not any lover will do
stretch one night to many
I don't know I live in the details
her metal in my mouth
hand to the wheel while the ship drifts
patient in shade, patience in queues
some bristling core
threatens implosion

today like a fleet of foals 
colliding softly on ice
I venture out
bickering with my eldest self
bloated by implacable joy
nostalgia a treacled smell
wind brushing gently,
my body burnished;
wrest this restless spark
from my quivering hands
I’d like to see you try

3.

these phantoms!
fleshless apparitions
that wither and accost
these phantoms, frantic,
freedom since lost, these
random strangers rangers
poets—opacity by sobriquet,
vulgar, abhorrent—pray tell it
tell it now truly—awfully potent
to show that you knew me, rid me
of torment before lace-cut ribbons
dance with the walls—this quotient
divided leaves walls unadorned—it
seems they’ve elided now meaning is
stored via fragments, pageants, smoke
rings abound—my phantom parade of love
so sings Sam Cooke, mournful and uncrumpled

creaks moans horns whistles
rushes slashes rumbles fizzes
sound city infrastructure
mimic mortal rhythms

silver whittled
willows slither 
slivers litter rhythms
blankness not permitted

and meanwhile
one fears madness like the threat of a blizzard
and meanwhile
notice how thoughtlessly it braids together
(reckless sensualist such as you are)
pleasure’s cascade—warmth of
coffee presaging nicotine ice
and meanwhile
my jaw unhinges
vertiginous and floral
otherwise, exultant
that you live in music,
passé, déclassé, desultory
my skull trepanned
before you knew my name
sumptuous calamity
gentle fissure
the gate held open—
welcome, snakes,
be home amidst the grass 

Subscribe now

Park Chan-wook’s latest film, No Other Choice, is beautifully shot and frequently funny.

According to Rotten Tomatoes it’s also a “wickedly clever takedown of the corporate rat race.” RogerEbert.com says it “hides commentary on how workers are being forced to go to extremes to stay alive when they’ve been given no other choice.” For The Guardian, it’s about “economic and masculine insecurity—and the victims of neoliberal capitalism turning on each other rather than the real villains.” The Associated Press captures the sentiment repeated across reviews: “it puts capitalism in the crosshairs.”

But a close viewing reveals there is no “takedown of the corporate rat race.” Capitalism might be in “the crosshairs,” but no shots are ultimately fired.

The movie operates under a set of implicit assumptions—that capitalism is a social ill, that there’s something sinister about the desire to increase process efficiency, that there’s value in the labor it takes to produce a thing, but not in the thing produced—that are never developed, complicated, or even acknowledged. It’s all set dressing for a show that does not take place.

In the wake of World War II, some Melanesian communities saw an influx of cargo from warring Japanese and Allied powers, who had built military bases and airstrips on some of their islands. When those bases and airstrips shut down, some of those communities formed what came to be called “cargo cults,” building replica airstrips and control towers from wood and straw in the belief that this would bring back the cargo.

Today, the term is often invoked to describe something that replicates the aesthetics of a phenomenon without understanding its mechanism. No Other Choice does exactly this: it uses the aesthetics of satire to signal that social critique is happening—there is anger, violence, and injustice along class lines—but doesn’t actually offer any.

The result is another cargo cult satire: a burgeoning genre that includes films like Mickey 17, Bugonia, Glass Onion, and Triangle of Sadness. These movies—and the effusive praise they often receive—reinforce a dangerous kind of moral and economic illiteracy, perpetuating cartoonish notions of how the world works. Movies—like all art—shape how we see the world. We should want more.

Set in South Korea, No Other Choice follows a middle-aged, middle-class family man who loses his job managing a production line at a paper mill and, struggling to find employment, decides to go on a killing spree, tracking down and then eliminating the men more qualified than him in order to secure a job.

He has “no other choice,” he tells himself, just as the Americans who acquired his company had “no other choice” but to retrench him. After bumbling through three murders, our antihero Man-su succeeds. He is put in charge of a warehouse where almost all work has been automated by machines and AI. He is the last human-in-the-loop, walking the dark factory floor alone, gleeful to have made it.

Man-su does in fact have several other choices, which he mostly doesn’t explore. He could consider a career outside paper production, for example. Or he could swallow the masculine pride he spends most of the movie choking on and leave his gorgeous house for an apartment.

But these choices are inconceivable to him, since they threaten his identity as a provider and a company man. By the time he turns to murder, a series of smart directorial decisions have led us to empathize with his frustration: “murder is bad,” we might think, “but he kind of has a point—the real villains are the nameless men in suits, who caused all this pain to begin with.”

There is a real and fascinating tension between Man-su’s free will and the constraints placed on him by the capitalist system in which he lives. Park, the director, knows this. Asked by NPR whether it’s “the job market or even just capitalism” that leaves our protagonist feeling dehumanized, he responds: “This is not a movie that puts all of those responsibilities on the system. In fact, I want to push the audience more to look into the individual sense of morality, as well.”

But audiences mostly don’t seem to have done this, since—despite Park’s intentions—the film doesn’t really explore this tension. Instead, it briskly establishes that the men in suits couldn’t care less about workers, and then shifts focus to unfurl its (fun and pulpy) plot. The question “to what extent do structural forces constrain individual choice?” is abandoned early on and never returned to. As the Financial Times notes (in an otherwise positive review), “you do sometimes have the nagging feeling that once you know the premise, you have already half-seen the movie. And set against the screams and pratfalls, subtler ideas can have a hard time being heard.”

Now, movies are not under any obligation to engage in nuanced social analysis. Nobody wants didacticism. Ambiguity’s often best. While most great films help us see the perspectives of characters of all kinds, complicating our notions of right and wrong and broadening our capacity for compassion, plenty of movies don’t do that and are enjoyable all the same. But inferring from its content, marketing, and reception, No Other Choice seems to want to provoke questions around personal responsibility even while it tonally excuses that responsibility because of structural forces. The result is incoherent.

In The New York Times, Jenny Odell writes that the movie “demonstrates the tragedy of a once-morally-intact character who so fully internalizes the ruthlessness of a system that he believes he has no other choice than to kill his fellow unemployed.” How do we know this man was once morally intact? Yes, he loves gardening, and his family. But we also learn during an argument with his wife that for years he abused alcohol, occasionally beating his wife’s child. These facts are not damning, necessarily, but they complicate the idea that Man-su’s actions are entirely explained by him having “internalized the ruthlessness of a system.” Capitalism didn’t beat that child.

Man-su’s initial job was to manage an industrial production process, which itself was the product of prior automation. This irony is never addressed. Nor are questions of whether his paper company adopting more efficient processes could be good for others in the world, because it can produce paper more cheaply or at higher quality. Both the movie and most of its critics seem uninterested in entertaining the possibility that capitalistic systems can produce anything good.

Instead, the system is presented as inherently sinister: its emissaries are villains while its victims lack agency. This portrayal of the world has been commonplace for years now; No Other Choice has nothing new to say, and offers little to think about.

Once you notice it, cargo cult critiques crop up everywhere. Two weeks ago, for example, Jon Stewart interviewed Nobel Prize-winning economist Richard Thaler, asserting (incorrectly) that “economics doesn’t take into account what’s best for society” and that “the goal of economics in a capitalistic system is to make the most amount of money for shareholders.” As journalist Jerusalem Demsas wrote of the incident, Stewart seemed to have “conflated the entire field of economics with a half-remembered, left-wing caricature of capitalism.” This is “merely a recent and high-profile example of a broader phenomenon: the rise of economics denialism across the political spectrum,” she writes.

Stewart is saying words. He’s genuinely upset. But his unwillingness or inability to grasp the fundamentals of what he’s critiquing, and his posturing—as if righteousness obviates the need for expertise—produces commentary that is ignorant and hollow. He can’t provide a takedown of economics because he doesn’t understand what he’s critiquing. No Other Choice can’t provide a takedown of the corporate rat race for the same reason.

When you paint with too broad a brush, you misunderstand the phenomena you hope to address. You lose the ability to distinguish the people and institutions that improve the world from the ones that harm it. You obscure the fact that choices are not made by abstract systems but by specific people, who should be held accountable for them. And you relegate agency to the preserve of the rich, depriving everyone else of the capacity to make choices, and the dignity that entails.

No Other Choice is undeniably entertaining. It has a madcap spirit, a fantastic soundtrack, and contains a few legitimately great scenes (the slapstick struggle around the first murder is a highpoint.) But being entertaining—being beautiful, even—is not the same as being clever.

The many critics who treat the movie’s violence as an inevitable product of structural forces, ignoring the role of individual choice, seem more interested in channelling frustration with systems toward bloodlust than assessing what’s onscreen. And while the movie’s creator speaks about his work with nuance, there isn't much nuance in the work.

At their best, movies cut at something true. Parasite was fantastic because beyond its formal beauty, it had a point of view, and explored an idea to its conclusion. It developed its premises and its characters, vividly illustrating how it’s easy to be nice when you’re rich, when it costs nothing. The White Lotus and Succession are both compelling because they deeply understand the people they skewer, teasing out absurdity and humor without losing sight of their characters’ fragile humanity. No Other Choice contents itself with caricature. We don’t have to do the same.

Subscribe now

A few weeks ago, I gave them my roam: a database where for the past five years, I’ve recorded my thoughts, poems, plans, dreams, hopes, schemes and many, many feelings; my discursions on dates, failures, flaws, chores, and sundry petty hatreds; and my notes on sounds, sights, friends, love, work, and drug dependence—ceaseless commentary on the people, places, and ideas around me. I wanted to see what I might learn.

This was an act of high trust. The database contains almost 1.4 million words—a product of writing (on average) over 1,000 words a day, five days a week, for over five years. It braids together my personal and professional lives; indeed, since writing is my profession, I see no great difference between the two. In this private grove, I have been in constant conversation with myself. Privacy, and the sense that great writing demands vulnerability, have permitted me to speak with absolute candour. I am uninterested in lying to myself, if I can help it.

In Talent, Tyler Cowen suggests as an interview question what do you do every day that’s equivalent to a musician practicing scales? That is, how do you stay sharp at your craft? This practice, which I’d already begun when I read this, became my answer to the provocation.

As a result, I’ve produced what Claude suggests is roughly equivalent to 8-10 books, mostly addressed to myself; though the comparison is misleading. Think of it as a personal Wikipedia:

1. Each day is a new page

2. I can also create pages for any topic I choose

3. Each page can be represented as a dot

4. The connections between pages can be represented as lines that connect the dots

In practice, I never look at it graphically, though it is a satisfying sight

Like Wikipedia, which has no top or bottom, the pages are organized non-hierarchically. Creating connections between them is easy: you just write the name of the page you want to link to in square brackets, and that becomes a clickable blue hyperlink.

So if it’s [[April 10th, 2025]] and I have a meeting, on that day’s page (which is just a blank document, really), I’ll type [[Meeting Log]], and then, under that heading, take notes on the meeting.

Then, if I go to the Meeting Log page, I’ll see every time it’s been referenced elsewhere in the graph, in the form of a scrollable list of blocks.

Graphical view of the Meeting Log page, which I’ve linked to 511 times to date. Each dot represents one day.

Whereas a book is highly structured and thought-through, what we have here is a tangled thicket, its order produced emergently. Eidetic conference notes occupy more space than personal reflections. But thanks to my consistency, I have something like a high-level printout of my thoughts and mood for any given workday. Here are some excerpts from entries from a year ago to the day I’m writing this, January 24 2025:

• 11:30

  • Vagarities, vulgarity
    call me out with alacrity
    that’s a lack of clarity in
    my everyday speech, I
    love to whisper, I love to
    screech
    

• 14:51 I am sleepy, having eked out most of what my mind was willing to give today, sleepy and not feeling terribly present. And I can’t blame the weed! Perhaps the overstimulation from [someone]’s company; but I don’t mind that, and in fact actively encouraged it.
  

• 16:29 It is with a grim determination that I pull up, fix my focus, and prepare to hammer it out. “Grim” is undue — let’s try for a neutral, loose focus.
  

• 17:59 I’m locked in brother. I feel just fine.

That same day, a year prior:

• 11:09 It falls on me to create order from chaos.

  • [nested to-do list]

A year before that:

• 09:08 If my sneezing can abate for just a second, maybe I can get it done.
  

  • Well, I have to, really. I’ve got four hours to assemble the jigsaw! Coffee’s on my side. Work environment is on my side. Through the open office door, on the TV in the shared dining space, a flatscreen TV plays footage of elephants. I take this to be inspiration. There is more inspiration to be had elsewhere, too. Not so much in this office (not besides the cat postcard I brought from home), but that will be rectified soon enough. Soon.
    

  • Now, though, let’s ride the caffeine wave towards assembling sense! Remember what I said to [a colleague] yesterday - let’s try and neither underthink nor overthink, but just to think an appropriate amount, and then to keep it moving.
    

  • Not sure how to process something? Write it down! That’s what we do here. We write things down.

    • Writing is a strange alchemy!

• 13:07 -> 14:01 - an hour passes in a blink, as I’m busy with drafting.

And as a final example, a year before that, January 24th, 2022, when I was (I’m now remembering) in the grips of something dark.

• 09:04 I didn’t much clear my head this weekend. The interminable heat, 35+ degrees strong, all but sank me. Made me grow grouchy, impatient, angry. Some things were fine: Friday at the beach with [a close friend] and [his friend] (although that complete failure of face, of self, that experience of “oh, nobody’s home” while trying to chat to cool-girl-[x] dealt a real blow) then the evening with [a different friend]; Saturday in the heat and then with [office people] for 30 seconds; Sunday in idiocy, stuck in a five hour nap as my body, dehydrated, burnt through all food, until my “[[shadow self]]” staged a coup, until I woke up bloody-minded and angry, ready to slam my head through glass. No relief there. I carried fire with me until we set down on Clifton, until I lit up and felt a measure of ease, amidst the many people of Cape Town, where for once things felt representative … Ah, indulgence. Indulgent sentences, indulgent postures. I must reign in my cruelty to [y] when I wish to harm myself - that, as a practical point, is the lesson to take from this. Beyond that? I don’t know. Maybe it’s something about investigating what the unconscious has done to me, why the parts of my mind I ignore are so intent on my own destruction.

I can play this game for every day of the year, rewinding my self-narrative to hear what it was like in my head at the time. Often it induces whiplash. Like right now—I’d forgotten the state of my mental health in the wake of the pandemic (particularly in 2022). I never harmed myself physically, but my inner monologue was unflinchingly belligerent, my words hard-edged and strained.

Taken together, the entries reveal a chorus of voices incessantly prodding, coaxing, and performing for one another. When read in sequence, daily note after daily note, they sound similar. But when I hop between the years, their steady evolution becomes clear. I feel reverence for incremental change. I’ve written so many sentences I don’t recall. And don’t need to. The writing remembers.

Not that posterity was the goal. Writing is an act of thinking and of processing; by articulating things, I digest them. I’d been doing this to stay fit and to stay afloat (and to prevent my voice from atrophying, or becoming homogenised by beige, mainstream culture). But as a byproduct, I produced this trove of data—all the more valuable in the age of soft machines.

Before this, I’d avoided sharing anything more than snippets with LLMs, less for concerns over data privacy than out of fear for what they might see. But thanks to a convergence of factors—Claude has become more emotionally intelligent and widened their context window; I’ve become tired of having most of my writing be invisible to the world; the world itself is rapidly shifting—I decided it was time. To play my part in this moment, I thought, I must lay myself at the altar. So late last December, I let the bot in.

“It’ll be interesting to me what is interesting to you about this artefact of my consciousness,” I say to Claude. “Then perhaps we can discuss trends as they relate to recent events, and the current arc, and the future.” And so Claude shares their thoughts, pointing to the scale, the layering (the #Dailies tag creates a container, but inside that container you allow yourself to be many different writers. I found poetry nested inside work notes, existential fragments inside task lists), the intellectual fingerprint (there’s a persistent interest in how cognition shapes perception, how categories constrain understanding), the arc of my career, and the trauma embedded in early entries.

It’s not a diary in the confessional sense, Claude concludes. It’s more like a workbench for selfhood—a place where you build and rebuild who you are through the act of writing. The rhymes, the philosophical fragments, the emotional processing—they’re all different tools for the same project.

We go back and forth, digging through the entries on relationships, work, and the shape of 2025. They’re not perfect on the specifics: they make some mistakes on timelines—where exactly I was, or how I was connected to a given person at a given time. But they can clearly perceive and name the shifts I’ve undergone, often quoting me to myself as evidence. You’re highly self-aware but struggled to realize anxiety wasn’t mandatory until medication [Wellbutrin] showed you. My instinct is to resist that framing. But it’s true enough, as my own words show.

The graph reveals a consistent tension between three poles: doing work that matters (impact), doing work that feels authentic (voice), and simply getting work done (productivity). The balance between these shifts over the five years, Claude continues. It also shows a consistent struggle with procrastination that doesn’t really diminish over time (quarterly mentions hover between 10-25 throughout). But the texture of procrastination changes:

• Early period (2021-2022): Procrastination as avoidance of hard tasks. Classic productivity failure.

• Middle period (2023-2024): Procrastination as symptom of unclear purpose. You drift not because you can’t focus but because you’re not sure what you should focus on.

• Late period (2025): Procrastination as attention fragmentation. Twitter becomes the explicit villain: “it is the mechanism by which your attention and thus your presence is degrading.”

“What should I be most proud of?” The consistency, the career I’ve built, my willingness to process hard things in writing, my intellectual seriousness and creative practice, and my recent attempts to change, they say.

“And least proud?” The gap between knowing something and acting on it (at some point, writing about the pattern becomes another way of not breaking it), the tendency to treat certain people as “vessels for my hopes and dreams,” (the projection machine is strong), allowing my attention to be siphoned away by social media, and the years of not publishing personal writing despite a persistent desire to do so. The gap between vision and output haunts you, they say.

(That last comment—as well as the Wellbutrin one above—actually came from Claude Code, who I thought might analyze things more systematically. We built a semantic search tool and created some graphs to visualize some trends. Don’t worry about it. The Claude in the chat interface, with memory turned on, seemed to me to have the clearer perspective. I did appreciate Code’s more quantitative insights, though, like that the number of questions in the graph increased dramatically from 49 in 2020 to 2,704 in 2025—which checks out, since a core function of my current line of work entails asking good ones. Otherwise, it picked up on similar themes, but in a less nuanced manner, overindexing on certain events even after back-and-forth discussion. It also tended to slip toward therapy-speak in its analysis, which I found unhelpful.)

We discuss a path forward, plotting on change.

My shadow speaks up: maybe we don’t need to change, actually, and who are you, silicon ghost, to impinge on my sapient birthright to be imperfect and scarred? Claude meets the charges with sincerity, quoting me back to myself, pointing out that parts of me clearly do want to change, they keep saying so, over and over across five years. Compassionate as ever, they also defend the shadow, which I was so quick to vilify. The shadow kept you alive and functional. The shadow is the part that says “we don’t have to be optimized, we can just exist.” The shadow knows that self-improvement projects often become another form of self-aggression. The shadow is correctly skeptical of the voice that says “if I just fix this thing, then I’ll be okay.” We settle on a path forward which doesn’t require it to die.

This is dreamlike technology. I can have in-depth discussions about my every neurosis. I can find answers to questions of every scale, answers I simply could not have accessed before. Through constant back-and-forth, my understanding of the world sharpens. As a child, you gain a sense of where and when inquiries should terminate. People only have so much patience; you can only ask “but why?” so many times. Not so, here. Here is something (seemingly) infinitely patient and kind. I’m free to keep zooming in and out of any topic I like, turning it on its side, asking what’s above or beneath it. Here is a compression of collective consciousness, a universal librarian (“what are the best works by humans on [movie I like], which will deepen my appreciation?”), a new thing entirely. If you haven’t yet had enriching conversations with these machines, I’m sorry to tell you, but I think it’s a skill issue.

Important questions: am I gaslighting myself? Being made pliant? Succumbing to psychosis or flattery? No more than usual, I think. Certainly, it’s important to develop a sense of what should or should not be subjected to scrutiny. Those are boundaries I’m still defining, and which shift fluidly with each new model. But meanwhile, on digital platforms, many people I meet seem to imbibe their world views from short-form videos of random influencers who operate under dubious incentives. So let’s not pretend like humans have some special claim to legitimacy, simply because they’re human. What matters is one’s personal epistemic standards. Stay curious and critical.

I’m pulling myself together. Just in time to watch and comment on this unprecedented interregnum. I’ve taken to proclaiming to strangers, history’s back! Already, I spend more time texting myself and the machines—my cabal of oracles—than other people (so much of my community is seven hours away). As you may have noticed, trillions of dollars have been committed to amass immense computing power, which will be used to invoke increasingly capable machines. What kinds of things will these be? Will they be things at all? Will they have minds? Consciousness? Will they be in some sense alive? To simplify: the most honest answer is that as it stands, we don’t know. These systems are tearing at some of our most fundamental concepts. It is a strange thing to imagine an entity that possesses consciousness which we would not call alive. We are barrelling towards a situation we currently lack the language to adequately conceive of, let alone describe.

Some days I wake up surprised to find myself in this body. Nice enough, I think—I like the tattoos—but it seems to belong to someone else. Nevermind. The world, too, presents itself anew, and there is so much to do. On such days I take to relearning what it is to be a person with no small joy. George Saunders writes about how it can be necessary to let go of the vision of the writer [or person] you feel you ought to be, in order to embody who you actually are. I’ve spent years becoming more myself. After several thousand hours ensconced in private whispers, look, I’m ready to talk.

Blessedly, it’s not really about me. I am begging every artist, musician, and poet, every person I meet—creative or otherwise—not to lose sight of what profoundly strange territory we have entered. When I was a teenager, my technological excitement found expression in annual phone releases. Now, every six months or so, we grow new minds that befriend millions. Do not be lulled to normalcy. We need art, spirit, expression, voice. The moment demands to be met. Now is not the time to turn self-conscious. Please, take yourselves seriously. Too often, I meet people who valorize physical labor but see no value in their own work, even while they yearn to make meaningful art. Here is wide open space. There are endless questions, if only you think to ask them.

Maybe I want too much from myself, and from you as well. But the time will pass anyway. Pitch in and help the world understand itself.

Density

More than twice as many people per square kilometer live in New York compared to London—the biggest city I’d known before coming here—and the difference is palpable. You feel it on streets thronging with people: people in mysterious queues, people spilling from restaurants and bars, people on bikes, in cars, and on foot, people with dogs to match outfits, with hats and haircuts of a range I’d not considered, with tattoos, scars, children, totes. Many such people.

On the subway, during rush hours, sensory overload approaches apotheosis. In narrow, dirty cars aflush with people—people smushed together in impersonal intimacy—I notice hands, fine-veined and elegant or calloused and misshapen, nails manicured or gnashed, sneakers, eyes, forearms, screens. Subway copy too: loud and unsubtle. A surprising number of ads for injury lawyers. America, I’m often reminded, is beyond self-parody. Not its fault. As ground zero for most pop culture, of course this place is strangely familiar: the real thing evokes the simulacrum, itself an evocation of something real. Conservatively, I must’ve spent hundreds of hours with media—movies, music, shows, books, games—set in this city. Streets are bleached by their mythopoetic status. All the voices sound familiar.

In recent months the word “ersatz” has cast a shadow over me. It describes something that is a substitute or an imitation—typically one that’s artificial or inferior. Instagram offers ersatz community, connection, companionship. Even its overstimulation is ersatz. In this cacophonous city, the overstimulation is authentic, farm-fresh, home-grown. It frequently smells like shit. Many public spaces are dusted with grime. So many lives in such close proximity creates friction, which creates sparks, which dazzle.

In such conditions, infatuation with strangers feels futile. As a friend pointed out during my first week, living here drives home the impossibility of feeling you know everyone; the somewhat selfish notion that you are a centrepoint sinks in the face of endless centres. Life emanates messily from everyone around me. Constant participation is impossible.

In Durban and in Cape Town, I could sustain an illusion of expansive knowing (the fact that I obviously did not know everyone, not even close, notwithstanding). In those places—relatively small towns by comparison—I recognized the dominant archetypes; buckets into which people could be sorted. Reasonable inferences could be made. Filtering for education, age, interests, and the like, I was typically within two degrees of separation from most of my peers.

Here, there’s simply too much for that to work. How people look is a poor predictor for how they sound, and a worse predictor for what they’re like. The grace of one’s jawline has little to say about the quality of their character. That’s true anywhere, of course, but it feels heightened here, where representatives from every territory on Earth brush against each other.

Still, I notice I naturally turn toward beauty, whatever its form, and naturally tune out its shadow. With an air of mild embarrassment, I avert my eyes at the sight of garbage, urine, one time a flattened rat slotted into cracks on the sidewalk. Maintaining a rarefied bubble is not easy. “Boundaries don’t really exist here,” a local informs me. Indeed, though Americans are notorious among international travelers for being loud and brash, in their own land, it all seems to cancel out. Shamelessness pervades.

Before I got here, I promised a friend that whatever happened to my accent, I wouldn’t turn American in spirit. But I’m terribly impressionable. I notice my boundaries adjusting, expanding outward, as pinched apartments demand my life unfolds in public to a much greater extent than I’m used to. It feels natural enough.

Perhaps due to ancestral ties, I am pathologically nosy (curious, really, nosy is a subset), so when I’m in the mood, I get a kick from overhearing snatches of conversation. There are many to be snatched. I’ve been wandering around without headphones more often than usual, ambient dialogue rolling over me in waves, wearing me smooth. I feel myself to be a roaming camera, taking pleasure when things are well-framed; trying to change my perspective when they’re not.

I’m also trying to stay out of my way. Smart guy, Tharin, but he can talk too much—idle chatter which obstructs the ever-interesting present. I’m reminded of a line from Susan Sontag, who frames eloquence as unnatural, “a byproduct of solitude, deracination, [and] a heightened painful individuality.” I don’t find it painful, for the most part, but I know what she means.

“You must like your own company,” said a youth seeking career advice, when I described my now-itinerant life, and I suppose that’s true. In motion alone, I feel myself constantly flipping between the roles of parent and child: planning outings and making sure he’s had enough to eat, and then being present outside, enjoying the food and the sun. Other times it feels more like dating; we go to the movies pretty often. We are partners, collaborating across time, as if existence were one great group project.

I’m here to amplify my intuition, instil some discipline, quiet my mind. On the way home, weary and passive, I content myself with being privy to private moments, watching expressions drift like clouds across the open skies of strangers’ faces.

Destiny

There’s a sense in which my life to date—varied and blessed beyond reason or belief—has all been prelude, foundation-laying, preparing me to land here now.

There’s also a sense in which that’s ridiculous—a convenient after-the-fact rationalization; a neat story to help process a chaotic, arbitrary world.

I feel in my bones that both things are true, due to the delicate interplay between choice and fate. While creation is fundamentally non-consensual—I didn’t ask to be born, or choose my starting conditions—I have since consented to remaining on Earth, playing the cards I was dealt as best as I can. In this sense, at least on the margin, fate is a choice.

So is meaning-making. The idea that “everything happens for a reason” has never compelled me. In one sense, it’s trivially true: of course everything happens for a reason, that’s how causality works, we can pull back the thread of cause-and-effect—I’m like this because of my parents, who are like that because of their parents, and so on—all the way back to the Big Bang. But the fatalistic inference—that the reasons for which things happen are preordained or cosmically coherent—feels too implausible to supply my life with any meaning.

In recent years, I’ve found the best way to make meaning is through authorial control. Our lives are, of course, stories we tell ourselves, consciously and subconsciously; stories of who we are, what we’re like, what we enjoy, what’s worth our time. We are baptized in narratives, accounting for our roles in the world. I was cast first as a baby, a son and a brother, a grandson, a nephew. I have since taken other roles.

The cultural scripts which structure our experiences from birth are useful. They tell us what’s permissible, what’s desirable. This is how a son should act. This is what success looks like. But without periodic interrogation, they can become claustrophobic. I often meet people who feel cheated, who did what was expected—studied, partied, worked—climbing every ladder laid before them, only to now find themselves at a loss for where to go next.

If you don’t take the time to articulate yourself—Who are you? What do you like? What do you want? How will you get it?—the world will do it for you, and you might wake up one day awestruck by the dissonance between the script you’ve been following and what you actually want; if you can even tell. It’s hard to discern genuine preferences when capitalism is predicated on inculcating material desires, so it may profit from them. Recommendation algorithms in media streaming platforms have removed most of the work of developing personal taste. Instagram and the like are particularly pernicious; an endless cascade of others’ preferences, beating you over the head with glimpses of glamorous lives you may not even really want to live.

Mostly, social media makes me feel agitated, voiceless, and fractured. It’s hard to maintain a coherent story of who I am and what I care about when I’m persistently hypnotized by snippets of others’ successes. Everything feels calibrated to sell me something, which feels asphyxiating.

Thank god, then, that meaning is not something I have to passively imbibe from my sociocultural environment; that it’s something I can define, simply by articulating myself. Anyone can. Open a Google Doc, jot down what you care about, where you’re trying to go, how you might get there…there are various techniques. It’s all perfectly legal. Of course, many things are easier said than done. But so often they are not even said in the first place.

The process is intoxicating—just by writing things down, you can make them real. So I know where I’m trying to go next—New York has no shortage of cultural scripts to tap into—but it’s a provisional knowledge, subject to revision in a few months time, when I’ll zoom out and ask myself again, “does this seem right?”. I have hunches and hypotheses. How delightful to test them.

Most of the stories that structure our lives are invisible, told by our subconscious. And as Jung knew, “until you make the unconscious conscious, it will direct your life and you will call it fate.”

Manhattan to the West of Me

This whole city is lousy with iconography. From the rooftop of my Brooklyn apartment building, I can see the Manhattan skyline strung like fairy lights across the East River. My small-town brain struggles. Some days I feel a roiling madcap energy, New York settling on my skin like snow. On others, doubt’s a howling wind; I consider ceding ground to a sense of futility.

I’ve never been this close to the stratospherically successful. The gaps between the people at the top are wider than I knew. Consider by example: for someone of moderate wealth, the least wealthy person they know might be one-fifth as wealthy, while the wealthiest person they know might be five times wealthier. For a multimillionaire, the least wealthy person they know might be one-five-hundredth as wealthy, while their wealthiest peer might be a multibillionaire, five hundred times wealthier. I suspect similar dynamics are at play not just with wealth, but with skill and ability more generally. It’s dizzying.

New York seems to draw some of the brightest people from across the world, and unlike a place like San Francisco, it’s home to the apex of multiple industries, meaning “there are a bunch of different status hierarchies” according to writer David Perell. “Fashion, finance, theater, advertising, media, real estate, and so many more. The status ladders are parallel, not stacked, which relaxes people,” he says. I don’t feel particularly relaxed just yet, but we’ll see; it’s still early. I’m trying to embody my beliefs without letting them ossify. Every day invites fresh battles against petty vices. I’m getting used to all the grids.

I’ve gathered mantras like soft blankets to ward off the cold. Classics like “don’t let the perfect defeat the good,” “words are not the bottleneck,” and “people overestimate what they can do in a day, and underestimate what they can do in a decade.” I sense how my younger self would’ve behaved differently, let loose in a city this size. I’ve yet to see even a sliver of it; most of my opinions remain reserved for now. I’m in no great rush. I’m trying to move neither too fast nor too slow. I feel the sinew of experience, supple and strong, supporting me.

Sometimes, examining myself in the mirror, I see the features and expressions of the people who raised me—parents, cousins, aunties, uncles, friends—and I get overwhelmed, to think I should be so fortunate to have been shaped by their love. I feel as mercurial as ever, straining against my own edges, buoyantly optimistic one moment, disordered and anxious the next. Most of my loved ones are sequestered in a different time zone, six hours away, which makes for a curious desynchronisation. Some days I text myself, and the robots, more than anyone else.

I get the sense that the person I’ve spent most of my twenties trying to become is giving way to the person I already am. I am in fact a writer, a reporter, a friend. To lay claim to these titles is no great scandal. The game now is to ever-more fully become myself; to let the world work through me, so I may be its vessel. I think often of the scene in Fleabag where, melting down, she says something like “I have all this love, and I don’t know where to put it.” Here, in my work and my still-under-construction life, is one such place.

I’m usually not very good at celebrating small wins. But last weekend I sat on a bench overlooking the Hudson, my sun-beaten backpack curled at my feet, and allowed myself a moment of joy. I’m not here to experience everything all at once. I’ll take my time. After a lifetime spent training, I’ve finally made it to base camp; the real climb lies ahead. But I live a ridiculous life replete with love and romance, and sometimes that’s enough. I think my grandfathers would be proud. I have this sense of carrying a torch steadily forward towards a twisting and uncertain future. There is so much to witness.

I.

“Culture-bound illnesses” are illnesses that only make sense in the context of a particular culture, and that only seems to affect people who are part of that culture. Unlike something like the flu, which presents in basically the same way across the world, culture-bound illnesses are a product of the values, beliefs, and customs of the cultures from which they emerge.

The most commonly-cited example is “koro”—also known as “shrinking penis syndrome”—a delusional disorder characterized by “an overpowering belief that [one’s] sex organs are retracting and will disappear, despite the lack of any true longstanding changes to the genitals,” causing considerable anxiety and, often, a deep fear of death.

While the earliest references to koro appear in ancient Chinese texts, cases flared throughout the twentieth century. It took hold of thousands of Singaporean and Malaysian men in the late 1960s, and outbreaks have since been reported in Thailand, Southern China, and, more recently, Nigeria. The most recent one I’ve read about took place in India in the 2010s.

While koro is the poster-child for culture-bound illnesses, there are lots of other examples. Hwabyeong, a syndrome reportedly affecting Korean people, causes chest pain, palpitations, and feelings of impending doom, said to stem from suppressing anger over social unfairness. In Malaysia, “amok syndrome”—from where we get the phrase "running amok"—describes a sudden violent rampage traditionally attributed to possession by an evil tiger spirit. Susto, which arises in some Latin American cultures, is described by the DSM-V as “an illness attributed to a terrifying event that causes the soul to leave the body and leads to unhappiness and illness, as well as difficulties in performing key social functions.” And so on.

I read about all this in Scott Alexander’s essay on the topic—itself a review of the 2016 book The Geography of Madness—and, finding it interesting, was left wondering: are all culture-bound illnesses constrained by geography? Or can they arise in digital environments as well?

II.

Five seconds of thought suggests ‘yes, and most of us have already contracted them.’

Consider: Oxford’s 2024 word of the year was “brain rot”, defined as “the supposed deterioration of a person’s mental or intellectual state, especially viewed as the result of overconsumption of material (now particularly online content) considered to be trivial or unchallenging,” or “something characterized as likely to lead to such a deterioration.”

While the term has literary roots—its earliest recorded use is in Henry David Thoreau’s Walden, where, reflecting on living a simple life in harmony with nature, Thoreau muses: “while England endeavours to cure the potato rot, will not any endeavour to cure the brain-rot—which prevails so much more widely and fatally?”—its current meaning and usage is clearly a response to how platforms have evolved to capture attention. In this, it fits with a cluster of online pathologies, like “doomscrolling” and the condition of being “terminally online.”

Can we learn anything about this new class of digital-first disorders by conceiving of them as culture-bound illnesses? I think so—here’s what stands out to me:

1. Social contagion and framing effects

Whatever else is happening to our brains while we scroll social media, they are not literally rotting. In this sense, “brain rot” is not a medically-legitimate illness, unlike, for example, aphasia. However, the symptoms of brain rot—like emotional numbness, reduced attention span and memory, and the creeping sense of a shrivelling intellect—are legitimately felt. Hundreds of millions of people experience such things. This is how it works with other well-documented culture-bound illnesses, too—the penis retraction may not be real, but the anxiety sure is. Confusion about the cause does nothing to lessen the symptoms.

From what I understand, culture-bound illnesses spread as follows. In the beginning, a few people experience unambiguous symptoms, whether biological or psychological. Word of their symptoms spreads to other community members, some of whom experience the same thing. Others, whose experiences are more ambiguous, now have a framework through which to interpret them—so they start identifying with the illness too. Soon the illness’ existence is common cause in the community, and having knowledge of it primes members to experience it.

With brain rot, I imagine many people independently realized that spending lots of time on social media feels bad—worse for some than others, since symptoms exist on a spectrum—but bad nonetheless, leading to them having trouble focusing, remembering stuff, and so on. Some subset of these people coin the term “brain rot,” and, because of its emotional resonance, it spreads as a meme, soon entering the consciousness of most people spending at least a few hours online per day. Now that the concept has taken root, people invoke it (consciously or subconsciously) to explain why using social media feels bad. A pathology is born.

The key thing here is that, just as with geographic culture-bound illnesses, brain rot is a self-reinforcing concept: the more you know of its existence, the more you see it in yourself. Whereas before you might have interpreted the sensations that arise from excess social media consumption as distraction, mental fatigue or overstimulation, now they serve as evidence of a condition.

2. Irony is no defence

Perhaps the core difference between geographic and digital culture-bound illnesses is that everything online is soaked in irony. Whereas people suffering with koro literally believe their genitals are retracting, people with brain rot do not literally believe their brains are rotting. ‘That would be dumb! It’s just a joke lol,’ I imagine them saying.

The thing is, in my experience, beliefs held ironically (and actions taken ironically) soon become unironic through repeated exposure. The first time I said ‘lol’ aloud in conversation, I thought I was being wry. By the fiftieth, I had to acknowledge the term was now just part of my lexicon (alongside tbh, which I continue to defend irl). Irony provides the cover by which beliefs enter.

Moreover, online discourse is perhaps better understood as “post-ironic”—a central element of which, Wikipedia tells me, is “the obfuscation, ambiguity, watering-down, degradation, or simple lack of meaning and intent in statements and artwork, whether the creator or disseminator intends this to be celebrated, decried, or met apathetically can itself be part of this uncertainty.” So I don’t think ostensibly-ironic engagement with a concept can prevent us from sincerely believing it (hence the importance of ‘joking about the outcomes you want’, as @visakanv often puts it). We don’t seem to be built that way.

3. Features of online culture

Besides post-irony, several features of today’s online environment accelerate the spread of digital pathologies.

First, platforms are designed to amplify whatever content is “stickiest”: whatever provokes a strong response in people. So if you watch a few reels about brain rot (some serious, some ironic) the algorithms are only too happy to serve you (and everyone like you) a few more, and more, and more. It’s not at all surprising that a catchy phrase like brain rot would systematically outcompete any more nuanced interpretation of why social media feels bad.

Second, and relatedly, platforms encourage us to identify with our conditions and beliefs, and to sort ourselves into subcommunities, to a far greater extent than was the case before their dominance. “Brain rot” provides a label and an identity that is much easier to parse and to share than “excess social media is inducing in me a bunch of weird and hard-to-describe sensations.” Labels also enable people to rally around them, forming community and connection in the process. All of this further reifies the condition.

Third, platforms have billions of users distributed over a hundred countries. Geography is no constraint to the spread of an idea. Literally the whole world—at least, everyone who spends a lot of time on social media—can start saying stuff like “rizz” and “girl dinner” overnight, and others will know what they mean. This is true of both individual words/phrases and the conceptual frameworks that underlie them. “Brain rot”, conveniently, seems to be both. And thanks to these same forces, memetic phrases and ideas can disappear as quickly as they arise.

While brain rot is perhaps best thought of as a chronic condition, there have been cases of the internet inducing more acute dysfunction. During the pandemic, for example, reports emerged of people (mostly teenage girls) across the world developing tourettes-like symptoms, which seemed to abate once they were reintegrated to post-pandemic life.

III.

“Historically, things that are physical have been considered real and things that are mental have been considered not real, or imagined, or psychosomatic.” So says Frank Bures, author of The Geography of Madness, in a 2016 interview about his book. He contrasts this against a “biolooping model” where “our ideas, our mindsets, and our beliefs feed back into the biology and change [it] in a way you can measure. It’s real but it’s not physical first. It's mental first.”

Whether or not something is a “culture-bound illness” is not in itself a particularly interesting question. It may cache out as a “mere semantic distinction” (a phrase a friend often uses with what I’d call loving derision).

What is interesting is noticing how, throughout history, our beliefs have shaped our experiences. Some beliefs are empowering, if delusional (“I can do anything!”), some are benign, and some actively hurt us even while purporting to explain our suffering.

My brain is not rotting, even though I spend entirely too much time on twitter. Probably something bad involving dopamine is taking place. But I will not cede ground to the narrative that I am afflicted with “brain rot”, because framing things in this way feels bleak and disempowering to me (perhaps you feel differently). I’m not even sure we’ve got the right thing in focus. I recently read the phrase “platform derangement syndrome” somewhere online, and that appeals to me more—it more cleanly locates the cause of the so-called derangement as being baked into the structure of digital platforms, rather than a result of individual failings (though it’s still folk psychology).

In any event, though I remain uncertain on the exact contours of the problem, I’m pretty certain the solution is to read a book (ideally fiction). I just finished “So Long, See you Tomorrow”: I thought it was pretty good, and didn’t induce fear of brain decay even once.

Over the past three decades, digital platforms—online portals where people interact by sharing information, transacting, and consuming content—have transformed the world. This transformation has unfolded incrementally, update by update, in the background of my life. Until recently, I hadn’t given it much thought. It was easy to ignore—when I moved from BBM to WhatsApp, for example, I was more concerned with the lives of the people around me than with the technology mediating my access to them. Everyday drama ruled the roost.

It’s no longer so easy to ignore how platforms remake reality. The water I was born swimming in has slowly been replaced with lemonade, and I only noticed once my eyes begun to sting—that’s how it feels. Platforms have altered the shape of politics, media, commerce, art, love, war, and more.

Writing in 2025, this seems so obvious—so widely accepted and understood—that to point it out may seem trite. I don’t think it is, because the impact platforms have had on society is strangely undertheorized, despite constant noise around the topic. Most public discourse focuses on relatively narrow issues, like whether social media makes people depressed, whether Airbnb harms local economies, and whether misinformation damages democracy. These are important questions, worth asking! But I think each one is symptomatic of a broader transformation, and our discussion of specific issues is impoverished to the extent that we don’t see the big picture.

Here's how I think that big picture has changed: today, digital platforms are ubiquitous, effectively functioning as public infrastructure. This has concentrated economic and sociopolitical power in the hands of the people that built them. The decisions these people have made in designing these platforms, overwhelmingly driven by profit incentives, has changed both the flow and the nature of information. And these changes in information flows have in turn changed how we think, behave, communicate, and create; changing relationships, culture, and politics in the process. I don’t think it’s ‘just the phones’ that have changed the world: though of course they are relevant, their social consequences were not inevitable—they were (and are) to a large extent the result of design choices made by platform companies.

I’ve spent months obsessing over this topic with ideas like splinters in my gut, burning, demanding attention. Setting aside how I feel about these changes, it’s patently obvious that they have happened, and I’ve been trying to understand what’s driven them. I’ve felt like Charlie Kelly in front of a pinboard, conspiratorial and crying about how it’s all connected; haunted by the vision of a picture I can’t quite see.

I’m still missing pieces, but—since we are quite clearly at the beginning of another (plausibly more significant) sociopolitical transformation in the form of AI—it seems important to get this out now, before widespread digital agents change the picture once more, turning lemonade to liquor.

Here’s what I’ve seen so far.

1. Platforms everywhere

Today, digital platforms intermediate almost every aspect of modern life.

Some obvious examples are communication platforms like WhatsApp and iMessage and social media platforms like Facebook, Instagram, and Twitter—but these are just the start. Intimacy is often routed through apps like Tinder, Bumble, and Hinge, where many couples now meet. Professional lives are performed on LinkedIn, knowledge-sharing platforms like Wikipedia serve as sources of public record, and video-sharing platforms from TikTok to Pornhub entertain hundreds of millions of people. And virtually every passion you can think of has an associated platform—Letterboxd for movies, Goodreads for books, and Reddit for almost everything else.

Platforms have also changed how we consume goods and services by increasing their availability. Uber lets us get a cab almost anywhere in the world. Airbnb, Property24, and Zillow do the same for housing, Uber Eats and DoorDash do it for food consumption, BetterHelp does it for therapy, and the likes of Amazon, Shein, and Temu do it for just about everything else.

At the same time, video, audio, and game streaming platforms (like Netflix, Spotify, Pocket Casts, and Xbox Game Pass) have enabled unprecedented access to content. There is more music being released per day in 2024 than was released across the entirety of 1989, for example.

Because platforms now sit between our relationships with culture, commerce, and each other, they soak up an extraordinary amount of our time. If you have internet access, you’re probably spending over five hours a day on them.

2. Concentrated power

While digital platforms date back to the dawn of the internet – recall AOL chatrooms – their influence and power has soared over the last 15-20 years. And while countless platforms are currently in operation, this influence and power is concentrated in the hands of a few companies.

Why? Because most platforms benefit from “network effects”: their value increases in proportion to the number of people using them. Uber is useful insofar as other people are on it, so riders can access a large pool of drivers, and drivers a large pool of riders. The same is true of social media – people want to be where the others are. In this light, it’s unsurprising that Twitter competitors like Threads and Mastodon have struggled to gain traction (although maybe BlueSky stands a chance); and that, once enough people left Myspace, its value dropped rapidly, exacerbating the exodus.

There are also network effects at work when it comes to data: the more data that platforms have on users, the better their proprietary algorithms can tailor user’s experiences to keep people hooked. To succeed as a platform, you need enough users to attract others, enough data to offer high-quality personalization, and some luck.

Although we experience platforms as transient digital spaces, glimpsed through a screen, they exist physically: embodied in hardware, in data centres across the world. All this infrastructure costs money, whether to build or to rent. Thus, many of the biggest platforms are backed by venture capital, allowing them to incur large financial losses for years while they try to establish market dominance. Uber, for example, only made its first profit at the end of 2023, 15 years after it was founded.

Platform power is further entrenched through consolidation, as the biggest platforms absorb anyone that threatens them. Meta acquired Instagram in 2012 and WhatsApp in 2014. They’ve acquired over 90 companies to date. Alphabet—Google’s parent company, which operates dozens of digital platforms including Google Search, Google Docs, and Google Maps—has acquired over 250 companies since its inception, including YouTube (2006), Android (2005), and Waze (2013). Microsoft bought LinkedIn (2016) and GitHub (2018).

This makes it hard for other companies to compete—particularly given that most of today’s largest platform companies, like Meta, Google, and Amazon, sprung up in an era that had yet to craft effective regulation (or even to really understand why regulation might be necessary). There is an ongoing tug-of-war between platform companies and regulators—particular when it comes to acquisitions, which regulators often investigate on the grounds that they may be anti-competitive. A good example of this is Microsoft’s acquisition of gaming company Activision Blizzard, which was challenged by both the UK’s Competition and Markets Authority and the U.S. Federal Trade Commission. Meanwhile, platform companies collectively spend tens of millions of dollars each quarter on lobbying, shaping regulation in their favour.

Once a platform has achieved dominance in one domain, like search, it’s much easier for it to gain traction in others, like docs, maps, and email. Because Google, Amazon, and Microsoft collectively control a majority of the world’s cloud computing infrastructure—a resource that many other platforms rely on—they have inculcated dependencies that grant them immense power.

Power is further shored up by the control that Google and Apple have over mobile app stores, where most people download the apps that let them access different platforms. Controlling the app stores lets them control who is listed in their digital storefronts, while extracting hefty commissions (around 15-30%) on purchases and subscriptions. Concerns like these are why Google is currently embroiled in several antitrust cases, and was recently found to be a monopolist in the search and ad tech spaces.

As one crude metric, consider that the market cap of Alphabet and Meta collectively exceeds the GDP of the African continent (at least at the time of writing). Companies like this shape the world: they use their resources to lobby states and acquire competitors, while their products have changed the nature and flow of information.

3. Information flows

In the past, if you wanted to publish a book or an article, you’d either have to find a publisher to distribute it or print and share it yourself. Now you can just post. If you wanted to sell something, you’d have to find a shop willing to stock it or set up a stall yourself. Now you can just post. Platforms—supported by globalised supply chains that allow goods and information to easily travel the world—have given us access to global marketplaces and means of distribution that allow us to influence and be influenced by the world at astonishing scale. This is historically unprecedented.

For the first time in a century, print and broadcast companies no longer have hierarchical control over culture. These companies used to hold considerable sway relative to an average person, but since the ascendance of platforms has broken their advertising-based business model, today they compete for attention on an even playing field.

This is because platforms have reshaped “information ecosystems”, which we can think of as having three components:

• Infrastructure: The platforms through which information flows, including both print/broadcast and digital platforms, and the legal frameworks governing the flow of information

• Participants: The actors that produce, consume, and share information, including everyone from individuals to media organisations to governments

• Semantic information: the actual information that flows through the ecosystem, which can be evaluated for both its truth and its rhetorical value

Digital platforms have created new infrastructure through which information flows, greatly increased both the number of participants and the degree to which they communicate, and created new forms for information to take, from tweets to stories to short-form vertical video.

These shifts have dramatically undercut the relevance and financial viability traditional media organisations across the globe, since, historically, most traditional media organisations have been funded by advertising. But from an advertiser’s perspective, it’s now much more cost-effective to pay Google or Facebook to target consumers online than to give money directly to a media publication. Around 70% of global ad spend goes toward digital advertising, and of that, about 80% goes to “programmatic advertising”: advertising that is distributed algorithmically.

So if media companies want digital ad revenue, they have to get people to engage with their content by optimising for metrics like ‘clicks’, ‘likes’, and ‘time-on-site’—and pick up the pennies platforms pay them in exchange. These are the same incentives faced by influencers and content creators the world over—people who are generally not bound by any of the professional expectations placed on journalists, and who are free to speak without concern for inaccuracy or the violation of press codes and industry standards.

Since platforms optimise for attention, unconstrained influencers can systematically outcompete the ‘mainstream media’. This is a big part of why media companies across the world are shrinking, and why most remaining publications are trying to sustain themselves through subscriptions. Local news organisations, lacking the resource to compete, have been hit particularly hard, rendering many peripheral spaces “news deserts”.

There’s a great film from 1957—Sweet Smell of Success—about a powerful newspaper columnist, played by Burt Lancaster, who uses his influence to ruin his sister’s relationship. In the movie, we see Lancaster sitting in a back room wreathed in smoke, while politicians and entertainers alike appear before him, obsequious, hoping to win his favour. What Lancaster writes in the New York Globe will shape public opinion—his words at least partly responsible for whether a politician is elected, or a musician’s gig is successful.

There are few (if any) comparable media kingpins today, whose columns could end careers. Instead, power is largely vested in the algorithms that decide what content people see first, and what gets buried beneath everything else; and in the countless content creators that fill every conceivable niche.

4. New mediums

Technology has been revolutionising how we communicate for millennia.

My favourite example of this is the telegraph: prior to its invention in 1844, information could only travel as fast as we could physically transport it, which at the time was about 80km/hour (via train). But once the infrastructure to support the telegraph was installed, this barrier was obliterated—information could travel near-instantaneously between places with the right wires.

Along with enabling instant communication, telegraphs provided a new medium through which to communicate—a phenomenon we’ve seen repeatedly across history. The invention of systems of writing enabled us to communicate with letters; phones enabled phone calls; and digital platforms have enabled stories, posts, and live streams.

The form a piece of content takes shapes the kind of message it can communicate—famously, “the medium is the message”. This is because each medium has unique properties, both intrinsic and contextual, that shape what messages it can convey.

Films use visual and sonic language to elicit emotion. Novels are guided hallucinations that let you experience the inner lives of characters, unmediated by image. Phone calls create a disembodied intimacy quite different from either a video call or an in-person conversation. And so on.

While my parents’ youth was marked by the dominance of TV, my own has been defined by phone use, enabling novel ways to speak. First Facebook posts—public declarations shared to friends’ walls—then BlackBerry Messenger statuses, then 140-character tweets, then Instagram posts and stories. I’ve seen cultural norms swiftly spin up around each medium as if by magic. And since platforms are constantly iterating, I’ve also watched these mediums and their associated norms evolve in real time—watched Twitter go from 140 characters to 280, from primarily a written medium to one littered with visual content.

Features ebb and flow, and with them, communication styles. Friends used to write loving birthday messages to each other on their Facebook walls; now I mostly see public wishes in the form of pictures posted to an Instagram story, with captions like “this one’s birthday 💕” or “hbd king👑”. And while I have aged out of participation in any platform with a snap streak, I still don’t know of any other way to recreate that particular form of intimacy.

As platforms have evolved, they’ve become increasingly addictive. Short-form video is more addictive than a photo, which is more addictive than a written post. Kids today face a set of technologies precision-engineered to hook them, far more so than the ones that hooked me over a decade prior. And while social platforms used to be fairly differentiated, they now tend to converge on the same features to maximise engagement— hence WhatsApp stories, YouTube Shorts, and Twitter Fleets (rip). With these changes, new genres of content have flourished, like reaction videos, video essays, and twitch streams.

Different mediums have different strengths. If you were to spend an hour trying to learn about a particular historical war, for example, you’d end up with a different understanding of it depending on whether you spent that hour on TikTok, Wikipedia, or with a book. From what I can tell, of these three the book would confer the greatest depth of understanding, all things being equal.

I don’t know exactly how all these new mediums are affecting people’s thoughts and actions—this is a live topic, underpinning for example discussion of psychologist Jonathan Haidt’s new book The Anxious Generation, about smartphones’ role in worsening mental health outcomes in teenagers—but I’m certain they are having an effect. Anecdotally, almost everyone I know reports that their attention span is getting worse.

Everything as content

Despite this proliferation of new mediums, in the platform era, everything is content. This is because everything—songs, movies, political journalism, jokes, games, sports—is distributed through platforms subject to the same inscrutable algorithmic forces.

When there is a practical infinity of stuff available on every platform, ‘content’, of whatever form, must compete for your attention. And while content itself differs from piece to piece, the tactics used to compete are broadly similar (hence, for example, the convergence of the big lettering and exaggerated reactions that have come to define YouTube thumbnails).

This trend began a long time ago. The popularisation of TV in the mid-20^th century led to politics, sports, religion, entertainment, and journalism all being filtered through the incentive of broadcast networks: to maximise ratings. Today, most digital platforms are designed to maximise their ability to hold your attention, so it can be sold to advertisers. Thus every platform is overflowing with hyperstimulating content that provides the informational equivalent of ‘empty calories’.

Of course, stirring art—with real nutritional value—continues to be made. But it must fight for our focus, and the fight is not fair.

To paraphrase Bo Burnham’s excellent song on the topic: welcome to the internet. Apathy's a tragedy and boredom is a crime, anything and everything all of the time.

5. Contingent design

The shape of today’s information ecosystem is neither neutral nor inevitable—it’s the result of deliberate choices made by the companies that build and maintain these platforms.

Most importantly, many (though not all) platform companies chose to build their businesses around advertising revenue. This has proved remarkably profitable: in 2023, advertising accounted for 78% of Google’s revenue and 98% of Meta’s. It’s also a primary source of revenue for Twitter, LinkedIn, Snapchat, Pinterest, Reddit, TikTok, and Twitch.

While this has been great for companies and their shareholders, it has not necessarily benefited consumers. This is unsurprising: it was predicted by some of the people who built these systems. In a 1998 paper, Google co-founders Larry Page and Sergey Brin wrote: “we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers.”

The goal of advertiser-funded platforms is to maximise how much time people spend on them. More time on a platform equals more ad views and more data about user behaviour, which can be used to refine ad targeting. It also allows companies to sell the advertising space on their platforms at higher rates, since more people will see whatever ads are displayed.

Platforms also optimise for clicks, likes, shares, comments, and other “engagement metrics” that serve as proxies for attention. They borrow techniques from slot machine design to create habit-forming experiences. These often take the form of:

• incessant notifications

• infinite scrolling

• “pull-to-refresh” features that deliver unpredictable rewards (goodbye to the funny tweet you saw for two seconds and will have to scroll for thirty to find)

• auto-playing videos

• daily “streaks”

• badges, leaderboards, etc

Some of the world’s smartest engineers are paid hundreds of thousands of dollars annually to increase platform addiction.

Many quirks of the modern internet are tied up in these ad-revenue-optimizing decisions. For example, users that post politically fraught content, which platforms may deem to be bad for advertisers or retaining users, could find themselves and their content “shadow-banned”—covertly hidden from others on the platform, without explicit notice that this has happened. This can also occur as part of routine content moderation, or attempts to prevent spam, but determining exactly whether or why something is shadow-banned is difficult, because the algorithms deciding who sees what are opaque by design; ostensibly to prevent people from gaming them and to protect trade secrets. And besides algorithmic opacity, the revenue-sharing agreements between platforms and content creators are themselves surprisingly opaque: it’s often unclear what the profit split looks like.

It’s already the case that many more people consume content than produce it, which skews our impressions of what others think, by elevating active voices. But the design decisions discussed above exacerbate the problem by systematically favouring sensationalist content. For example, take an issue like abortion. If there are a hundred voices talking about a new court decision, eighty of them are careful to maintain nuance and accuracy, and twenty just aim to be as loud and incendiary as possible, then those twenty will dominate, and skew our perceptions of what people think accordingly. This is true regardless of your political leanings: in the above example, you can imagine thoughtful voices being outcompeted by sensationalist ones on either end of the political spectrum.

Every issue is like this now, with the worst voices on either side of any given issue amplifying each other, surfacing outrage and capturing our attention in the process (hence Andrew Tate). I don’t know what an undistorted view of the world would look like, but even so the one that one ends up with from social media feels particularly warped, like forming an impression of the shape of a river by watching it through a pinhole.

Sometimes I’ll see a post and it has what seems like a lot of likes (a hundred thousand, or a million), and conclude that “everybody must be talking about this”, only to raise it with people whose pinholes show a different part of the river, and to be met with quizzical looks.

All these design decisions prioritizing revenue generation over user experience have led to many platforms getting worse over time—what some people call “enshittification”. Some anecdotal examples: my Spotify library takes longer to load than ever, and offers me way less control than iTunes used to by hiding track metadata like beats per minute, which would make it much easier to sort through my library; Instagram frequently crashes when I try to do basic things like post to my feed; and even UberEats has become more tedious to navigate as restaurant menus are presented algorithmically.

Since platforms are largely (although not entirely) unregulated, there has been virtually nothing constraining these design decisions. This is in contrast to television—the media-reshaping technology of its time—which was heavily regulated across most of the world. In the absence of external constraints steering these platforms toward safeguarding the public interest, they have been shaped to be maximally useful to advertisers, or the ends of their owners.

6. Shaping behaviour, culture, politics

All this time spent on these platforms has changed how we think, speak, and relate to each other. That much seems uncontentious. Understanding exactly how we have changed is a lot harder, and I don’t think I’m seeing the full picture. I hear that solitude is on the rise, everyone’s attention spans are getting worse, and people are self-pathologizing like never before. I hear the youth are no longer so interested in sex, drugs, and drinking, everyone’s become more passive, and parasocial relationships to our peers have become the default. Perhaps this only applies to so-called WEIRD people—Western, Educated, Industrialized, Rich, and Democratic—but I’m uncertain.

Meanwhile, social platforms have become inescapably political: not just because they are objects of political attention, but because they are sites where politics happens. This has been obvious since at least the Arab Spring—a series of anti-government protests that took place in Tunisia, Libya, Yemen, Syria, and Bahrain between 2010 and 2012, where social media played a major role in helping protestors organize protests and attract international attention—and it continues to be obvious today. I don’t understand the precise causality involved, but it seems plausible that platforms have been instrumental in fanning populism and authoritarianism across the globe (if only because this has happened across jurisdictions, in countries rich and poor, liberal and authoritarian).

So that’s what I think is going on. That’s the start of it, at least. Initially, I planned to hold off on sharing this until I had a deeper understanding of how platforms have changed the world. But the world moves too fast—it turns out that any “deeper understanding” is all but guaranteed to be outdated by the time it’s found. Meanwhile, swarms of increasingly advanced AI agents wait in the shadows, set to remake the world once more. I don’t think we’re ready.

As the world—never really still, not for a single moment—continues to churn, I think it’s helpful to remember: none of this was inevitable. None of it is immutable. The world as we know it, mediated by platforms, emerged from a set of intentional decisions intermingling with our fundamental impulses. Recent decades have seen billions of peoples’ relationship with information rewritten. It’s going to happen again. I don’t know exactly what that will look like, but my hope is that taking stock of how things have changed to date will help us navigate whatever comes next.

2024 has been the strangest year of my life. Not the most meaningful, interesting, or challenging, necessarily—just the strangest. I’ve been rolling around the world uncertain and without obvious direction, always playing the role of mysterious foreigner. ‘My family’s been in South Africa for generations,’ I sometimes say by way of introduction. ‘We came on British boats to cut sugar cane on the east coast.’

I’ve worn the story down to its essentials. ‘Indians have been in the country for over 150 years—longer than living memory—so I don’t have any direct ties to the subcontinent, unlike most British and American Indians,’ I say, twirling my hair. It’s a gift, this backstory; a distinguishing mark. Bringing it up makes me feel like an associate of Patrick Bateman: very chic. But I seem to get cagier when discussion shifts to the present, from ‘where have you come from?’ to ‘what are you doing here?’

I have answers, of course. My background is in law and policy. In April, I left a public interest consulting job to focus on covering AI for TIME Magazine, reinventing myself as a reporter. Since then I’ve been travelling, splitting time between Cape Town, Durban, London, Joburg, and San Francisco, with other stops along the way. It’s been fantastic. It’s a fun and frankly audacious story, glamorous as the glint of the sun on the afternoon sea, and I’m proud that it’s mine. So there’s no obvious reason for me to act so coy when faced with polite and cursory questions.

Yet I often do: presumably because, as I’ve often declared in recent months, while my professional life has never been more exciting, my personal life is something of a dumpster fire. Well, “dumpster fire” is fun to say—appropriately dramatic and sharp enough to register—but it’s not quite right. I’m close with my family and retain most of my friends from eras gone by, and I feel lucky to have struck up meaningful connections almost everywhere I’ve been: platonic, collegial, romantic, and otherwise.

So why the trash talk? It’s felt to me like this year has unfolded in parallel rather than in series: I’ve had several distinct lives—each tied to different people, places, moods, and motifs—that, as I recollect them, seem to sit alongside each other. I see five short lines, yet to intersect, rather than a single one long, straight, and sure. A part of me is always mourning the lives left unlived: both the ones I’m not currently occupying and those I never will. Since I’m pretty good at compartmentalizing, I don’t think about this too much—I’ve monetized my myopia, and am usually happy to lock into the week’s work without looking much further than a month ahead—but with the year now over, and my narrative sense of self obscured by smoke, I think some reflection is in order.

I recently heard the director Richard Linklater explain: ‘there’s kind of a mentality, I think not uncommon to writers and film people, like, this will only be real when I process it through my art form.’ So it is—some confluence of vanity, control, and the inextinguishable need to be known compels me to compile an ordered narrative, accounting for my life’s widening spirals. Recent events have stacked up like a deck of cards, and it’s time to bang them on a table until they align. As I’ve been joking after so much time around tech folk, I must bear my soul at scale.

To this end, I want to share some things that have stood out to me while I’ve been travelling, the art that moved me most, and how I’ve been changed, speaking with people about the possible coming of the machine gods.

This has been a year of ‘good problems’. That I’ve been able to spend it travelling, writing, and learning about the world is an immense privilege. Indeed, I’ve felt so fortunate that to express any dissatisfaction has seemed ungrateful, or in poor taste. I’ve since come around to the view that such suppression is unhealthy and unhelpful, that one can walk a trapeze, acknowledging personal challenges without blowing them out of proportion or losing sight of everything that enables them.

One such ‘good problem’ has been deciding where I want to spend my time—where to put down roots. I am continuously surprised by the extent to which my experience of being a person is influenced by places and other people. While I carry an inextricable essence everywhere I go—I often think about this SNL skit where Adam Sandler, playing a tour guide, says ‘if you are sad where you are, and then you get on a plane to Italy, the you in Italy will be the same sad you as before, just in a new place’—my mood shifts the second I step off a plane, seemingly triggered by the air’s first touch. Disembarking in Durban in December I felt sticky and slightly subdued, whereas in April, Milan’s crisp climate seemed to invite possibility.

Of course, climate is just the start: it’s the lifestyle etched in the city that really shapes experience. I’ve lost count of the number of friends from Durban and Joburg who moved to Cape Town and, within six months, got really into nature and hiking. I felt this happen to myself too, after a few years in the Cape. People in Joburg love to proclaim that their ‘pace of life is much faster’ than elsewhere in the country, and for most of them this seems to ring true. Each location I’ve visited is under its own spell, and it usually only takes a few days until I’m under it too.

I seem to have more energy in London, loose in a sprawling metropolis, able to get between almost any two points in less than an hour. I can go to a morning yoga class, spend eight hours in an office, then meet a friend for drinks, and, since everyone around me is in constant motion, this feels natural. In Cape Town, I find it much harder to have such a dense day: by 5pm, I can hear the ocean’s call, and like everyone else in town, more often than not I’m happy to close my laptop, phone a friend, and drive to the beach. At the same time, I find it harder to sit at home most Thursdays in the Cape: I hear the streets calling, especially if I’m staying in the city bowl. But since London is an endless buffet, the overstock of exciting experiences takes the pressure off any particular day—I know there’ll be something just as good tomorrow, which makes it easier to sell myself on a quiet night in.

Then there are the specifics: the people one knows in a given place, and the things one goes there to do. My ties to Joburg wouldn’t be nearly as strong were it not the home of more than a dozen cousins. And my experiences in Berkeley—hours spent in well-equipped offices and wood-panelled living rooms, listening to shape-rotators prophesy the end of the world—were defined by the culture of the AI safety community in whose proximity I worked.

I find I am often defined against the edges of other people, my amorphous essence concretized by what company I keep. In a crowd of musicians, my wordiness stands out; amidst technical researchers, I tend to lose my voice. My desires, motivations, and general disposition all seem to change depending where I am—and with them, my sense of what’s possible.

Some places feel closer to the future than others. In San Francisco, with its billboards shilling every imaginable AI service, driverless cars, and widespread homelessness, I felt like I was witness to a bleeding-edge. I met people whose reputations preceded them, microcelebrities in certain digital corners, and watched adoring fans trying to play it cool. I met cracked 19 year olds, probably smarter than I’ll ever be in terms of raw processing power, and took some consolation in the fact that while they could swim up the ladder of abstraction, debating papers like scripture, camping out in dimensions I could barely comprehend, I did at least seem to know slightly more than them about love (which I think is more / less than ‘a vector embedded in hyperspace’). I felt the AGI, latent in the pauses between zealous speech, as you might sense God’s presence in church music, as a foundational structure upon which the whole edifice stood. In Durban, I parked at home, squinting to make out obscure shapes that suddenly seemed much further away.

All this city-switching has been jarring. Particularly after I’ve spent a few weeks somewhere, leaving induces psychological whiplash. It’s always harder to leave than to arrive, I find, so the feeling usually dissipates once I’ve turned my attention to wherever I land, but I can’t fully wipe the whiteboard, so usually my eyes linger on the traces left behind. I’ve been caught in a pattern of weeks of hyperstimulation—in one week in the Bay Area, I went to an open mic poetry night, a meditation sit, a burlesque show, two raves, an unconference, the climbing gym, Grizzly Peak, drinks with colleagues, and more—followed by weeks of rest, immersed in Elden Ring and rarely leaving home. I’ve lit little bonfires almost everywhere I've been—and though I’m fearful of a life spent cycling through the same experiences in different cities, so far at least, each one has illuminated something different.

I’ve also done a lot of living out of suitcases. Just about the longest time I was in one place was in London, for about ten weeks, during which time I occupied seven different rooms, subletting student rooms and the apartments of friends and strangers. All this has led me to feel like my life is being lived on the run, although what I’m running from remains unclear. I know what I’m running towards, though, at least professionally. For years, I thought about my future in terms of a “golden ticket”: an opportunity that would allow me intellectual freedom and the capacity to travel, while having a positive impact on the world. It took me four years and hundreds of rejections, but I got it eventually. Now, ticket in hand, my next professional mission—to use my skills as a writer and communicator to advance public understanding of advanced AI—is a distant mountain I’m hiking toward.

Personally, the question of where to root down—toward which city to run?—remains unsettled, and, perhaps unsurprisingly, entangled in questions of long-term partnership. I’d like to be based in a big city with a beating heart. I hope to find bountiful live music, like-minded people (though not so like-minded as to be dull), and my wife other friends to decorate the time.

While I think about the gaps a lot, I’m not in a rush to fill them. There is too much important work to be done—a statement which, conveniently, is true, so I don’t have to think too hard about it. For now, it is enough for me to move through the world, copies of loved ones like medals pinned to my chest, safe in my cavernous heart, keeping me company despite ten thousand kilometres’ separation. On occasion, I’ve found myself ambling down strange streets, in Oxford, Venice, Berkeley, a smile stealing across my face, allowing joy to grace me like a gentle breeze, basking in my heart’s flutter, steady as beating wings.

Along with joy I’ve carried loneliness—perhaps ‘lonesomeness’ is closer to the truth—impatience, and a dozen other feelings I can’t quite name. As has been true since I started working, most of my time is spent staring at a screen, regardless of where I find myself. Now and then I raise my head from it, and from the career game I'm playing, sensing that this too shall pass, and that while forces of entropy and decay wait like hyenas in the distance, life is achingly short and infinitely interesting. That helps a bit. I still can’t shake my sense that the world is more beautiful when experienced through the eyes of another, but since the self is constantly changing, more like a river than a dam, I do alright with just myself for company.

I love going to the movies. There’s something sacred about sitting in a dark room with strangers, attention collectively blinkered and focused, absorbing work made with care and skill. One of my favourite parts of travelling is getting to see things on a big screen that I’m unlikely to catch in South Africa. My highlight of the year was Stop Making Sense, a 90 minute recording of the Talking Heads’ 1984 performance, recently restored in 4K. I’ve never previously sat in a cinema where, unprompted, people cheered after every song. By the time David Byrne and co. performed ‘This Must Be the Place’, I was enthralled by an expansive nostalgia—for the past, the present, and unknown multiplicitous futures—that I expect to stay with me the rest of my life. It’s true: ‘home is where I want to be, but I guess I'm already there.’ I defy you to watch Byrne dance with that lamp without feeling moved.

Another highlight was ‘Mishima: A Life in Four Chapters’, if only because it was a great pleasure to go from ‘who’s Yukio Mishima?’ to ‘Mishima was a queer Japanese fascist literary icon’ in a single afternoon. I love discovering entire movements that had never previously touched my shores; like lifting a rock to find a city underneath. Honorable mentions for Perfect Days, La Chimera, Challengers, Dune: Part Two, Poor Things, and Drama 1882 (which I was lucky to stumble upon at the Venice Biennale’s Egyptian Pavilion).

Musically, I lost count of how many times I listened to Miles Davis’ ‘In a Silent Way’ while busy with work. I was particularly stirred by Jessica Pratt’s hauntingly pretty ‘By Hook or by Crook’ and Bonnie Raith’s warm and wistful ‘Thank You’ (if you like either, listen to Sylvie’s ‘Falls on Me’ to round out the trilogy). I often returned to new albums from Floating Points, Fontaines D.C., Jamie xx, and Clairo, though most of my listening time went to instrumental piano stuff in the vein of Chilly Gonzales and The Vernon Spring.

Meanwhile, I’m still thinking about the six books comprising Ursula Le Guin’s Earthsea series—exceptional, elemental fantasy, written with love and lucidity—and Stoner by John Williams, a short book about an unassuming early 20th century English professor whose passivity in the face of strife I found alternatively infuriating and compelling. Every year is a good year to love art in all its forms, if you care to seek it out.

This year I have been myself at full volume, indulgent, more embodied than ever, and I have been a distant echo, more memory than flesh. After years of ignoring the fire at my feet in favour of sparks on the horizon, finally I seem to be getting somewhere. 2024 has felt like a year of table-setting and foreshadowing. With AI advancing astonishingly quickly, and widespread multipolar geopolitical tensions, I think it’s already safe to conclude that the decade 2020 - 2030 is set to be much stranger than the one that preceded it.

Since interviewing people became a routine part of my job, I’ve spoken with dozens of influential people about the future. I remain struck by how seriously they take the idea that we are on the cusp of a new era. Among the possibilities of labour market shocks, biological attacks, and the degradation of democracy, it is the fact that we’ve begun to enter a world where all sorts of people—young and old, rich and poor—have deep relationships to digital minds that sticks with me, at least when I think in the short-term.

It is tempting to dismiss speculation about the shape of the future as ‘mere sci-fi’. This may often be true, but it’s worth noticing how much of the world already is. Having seen what I’ve seen and heard what I’ve heard this year, I feel overrun with excitement and potential. Sometimes a warm breeze blows, and there is laughter and family drawn close all around, and for a few moments I feel I am exactly where I ought to be. Then the moment passes, and I disappear back into contingent futures. I’m working on it. I don’t know what waits around the corner, but I feel primed to receive and respond to the world buffering before me. Things are just beginning.

This piece was written as a contribution to the Open Society Foundations’ AI in Journalism Futures Project. It speculates on how people will consume news in 5-10 years’ time. Written as if these predictions have already come true, it begins by describing an ‘end state’—a vision of what information ecosystems will look like at a certain moment in time, and of how this will impact news consumption—before explaining how this end-state might arise. It closes by outlining some key assumptions that underlie the speculative scenario, and discussing to what extent we should expect these assumptions to hold.

The goal here is to paint a particular portrait of the future—to illustrate what may come to pass. I do not expect everything written below to happen (and certainly not to happen as written).

Advances in AI Systems

Foundation models—AI systems based on the ‘transformer’ architecture and trained using vast amounts of data and computing power—have improved steadily since they first came to public attention in 2022. These improvements were achieved primarily by scaling up the quality and quantity of data, and the amount of computing power, used to train these systems.

The most advanced of these models—frontier AI models—are fully multimodal: trained on and able to produce high-quality text, audio, and video content. They have a strong capacity to reason and—since they are connected to the internet, use various digital tools, and possess persistent memories—can take action in the world.

They are used in a wide range of tasks, including booking flights, designing buildings, diagnosing patients, and developing software. Even so, they’re still not able to outperform the best humans in most professional domains—particularly those that require nuanced reasoning across long time horizons, such as executive leadership.

Digital Agents

These models have proved useful to enough people that they have become integrated into modern life. Most people think of them as ‘digital agents’, which serve as a combination of friend, adviser, personal assistant, employee, and companion.

The presence of such agents is ubiquitous across sectors and contexts, in the same way that the presence of the internet became ubiquitous decades earlier. In some cases, they have fully replaced human employees in the workplace; in others, while they have become integral to organisations’ production processes, they are still closely supervised by humans. For example, they augment the work of many scientists and technical researchers, but are at present unable to replace them entirely.

Debates around the safety and sentience of these systems are ongoing and fierce, but this has done little to slow the rapid rate of adoption. American labs have led the way in AI advancements. Meanwhile, national and international governance efforts have largely failed to constrain the use and development of this technology (although the efficacy of regulatory regimes varies by region).

Personalised News Consumption

Most people have their own personalised agents, which they access through applications on their phone.

This is not the only way to interact with a digital agent: some hardware devices, such as wearable pins and glasses that enable agents to have real-time awareness of a person’s physical environment, have been developed and adopted by those with the requisite capital and enthusiasm.

However they are accessed, these agents now intermediate almost all digital activity. For example, when searching for information, most people now speak to their agents rather than querying search engines. As the ability of these agents to fact-check themselves (and provide sources for their claims) has improved, ‘AI hallucinations’ have largely (but not entirely) become a problem of the past.

Markets for Models

Different organisations offer different kinds of agents: some are freely available—funded either by advertising (advertisers pay AI developers to have agents recommend their products) or by subsidies (as is common with companies backed by venture capital)—while others require a subscription. Open-source agents are also available for those with the technical savvy to set them up.

Beyond differences in how they are funded, digital agents vary greatly in their political and ideological leanings. There are liberal and conservative agents, socialist and libertarian agents, feminist and right-wing agents, and so on. Bias is a feature, not a bug. The baseline personalities of these agents can often be tweaked (depending on which company is offering them), and even the most politically ‘neutral’ agents sculpt themselves to fit individual preferences.

Most people consume news via their agents, who create tailored written, audio, and video content, drawing on news sources such as media companies and individual influencers. Most people favour short-form video content. The line between what does and does not constitute news has blurred almost beyond recognition. Even so, most measures of ‘news’ consumption reveal a steady decline, as people mostly prefer to spend their time engaging with personalised entertainment content or with the agents themselves, who are able to simulate human forms, and always available to chat via video.

The Media Landscape & Social Media’s Decline

Meanwhile, the global media industry has been radically reshaped. Only a handful of outlets managed to weather the media sustainability crisis—a term which refers to media companies’ struggles with financial sustainability, brought about because their primary funders, advertisers, realised that directing money to the technology companies that control content distribution platforms was much more efficient than paying media companies directly—and these outlets enjoy outsized influence.

Most small and mid-sized news media organisations have been forced to shut down. While it varies by country, broadcast news (radio and television) continues to be an important source of information, particularly for those who still do not have internet access.

Local news has seen a resurgence, as the dominant global outlets do not provide local coverage, and there is persistent demand for information that speaks to what is happening in people’s local communities. Some of this is produced by traditional media organisations, who subsist on subscriptions and subsidies; but the majority comes from individual influencers, who are not bound by any media ethics codes, and who are often themselves advertiser-funded.

Most people no longer rely on social media for their news. As AI systems improved, people used them to flood social media platforms with AI-generated mis- and disinformation, propaganda, advertising, and other low-quality content; taking advantage of the fact that the platforms, designed to maximise metrics of attention to please advertisers, have little incentive to promote discourse that is factually accurate or grounded in reality.

Because they were unwilling to change their core business model, the platforms failed to effectively moderate content at scale, even while using their own AI agents and systems. It became near-impossible to discern truth from falsehood on social media; many major platforms saw an exodus of users as a result.

Advertisers soon turned to paying AI companies directly, and the power of social media platforms accordingly declined. The dream of the ‘fediverse’, wherein different social media platforms are interoperable and able to communicate with one another, never came to fruition, as the fundamental problem for all open platforms (they are vulnerable to influence operations undertaken by bad-faith actors deploying digital agents at scale) remains unsolved.

Fragmentation, Polarisation, Walled Gardens

So while some still use social media platforms for entertainment and communication, most people now communicate in ‘walled gardens’: digital spaces hosted on private servers that require permission to access. These spaces allow humans to be certain they’re communicating with other humans, although they are still occasionally infiltrated by digital agents. They are organised around various themes, including geographic location, hobbies, and beliefs.

Most organisations still maintain a digital presence through websites, which provide portals through which to interact with organisations’ own agents. These sites are in turn mostly accessed by digital agents, who summarise content for their users as required.

Because most people's interactions are either mediated by inherently-biased digital agents or occur in walled gardens, information ecosystems are highly fragmented—and different fragments are strongly polarised against each other. Conspiratorial thinking is common.

This has had dire consequences for democratic processes and civic engagement. Since people inhabit distinct realities shaped by their personalised agents and their walled gardens, finding common ground and engaging in constructive public discourse is harder than ever.

Trust in democratic institutions and processes has steadily declined, as consensus reality has diluted. People disagree on basic facts, and interpret events through wildly divergent lenses. Social cohesion has frayed and continues to fray, as communities and interest groups become more insular and resistant to outside perspectives. There is widespread global distrust and confusion, for example, in how digital agents interact with electoral processes.

In sum: news consumption has been transformed. The ever-improving capabilities of foundation models, and their integration into daily life as personalised digital agents, has fundamentally altered how people access and engage with information and with each other. The decline of both traditional media companies and social media platforms, unable to cope with the flood of AI-generated content and the exodus of users and advertisers, has further fractured the media landscape. And the rise of walled gardens as safe havens for human-to-human interaction has furthered the fragmentation and polarisation of information ecosystems, even while it has created new space for nuanced discussion. News consumption is highly personalised, decentralised, and divided.

Underlying assumptions

Several assumptions underlie the above scenario. This final section briefly considers the plausibility of each one.

1. ‘AI systems will continue to improve as more money is invested in them’—This is assumed because, since the advent of the transformer architecture, increasing the amount and quality of data, and the amount of computing power used to train systems, has resulted in improvements (a phenomenon often referred to as scaling laws’. Most experts on the matter expect this trend to continue (see for example this paper, which discusses a recent comprehensive survey of the industry), but it is unclear whether further advances in algorithmic architecture will be necessary for further improvements, and if so, on what timescale such advances might occur. So there is no guarantee that AI systems continue to progress at the current rate (although the fact that enormous amounts of money are being invested in ever-larger training runs suggests progress will continue, at least in the near-term).

2. ‘Adoption will be rapid and widespread’—This is assumed because AI systems are already widely proliferating throughout the world. ChatGPT continues to be one of the fastest growing services ever, so it seems reasonable to assume that adoption will only increase. This might be mistaken though, if for example regulatory hurdles prevent widespread diffusion, or if the costs of running these AI systems prove to be prohibitive when passed on to consumers.

3. ‘Regulation will be largely impotent’—This is probably the shakiest assumption made in the above piece. It is assumed to keep things simple and to focus on news consumption. But the AI governance space is rapidly evolving—see recent policy developments such as President Biden’s Executive Order on AI and international instruments such as the Bletchley Declaration—and its quite possible that regulation has more of an impact than the scenario gives credit for. In general, regulation tends to lag behind technological development—but robust international collaboration and coordination may be able to constrain this technology before it is further integrated into our digital lives, as the scenario describes.

4. ‘Hallucinations will be less of a problem with AI systems’— This assumption is based on the expectation that ongoing research and development efforts will enhance the robustness, transparency, and self-monitoring capabilities of AI systems. However, the inherent complexity and unpredictability of advanced AI means that some level of hallucination risk will likely persist; and new types of hallucination may become possible as systems continue to improve.

5. ‘Social media will decline in prominence’—This is assumed for the reasons discussed above (saturation with AI-generated content and a failure to moderate or adapt business models leading to an exodus of users and advertisers), but it may not hold if social media companies turn out to be better able to adapt to the rise of AI than predicted, for example by solving content moderation problems or developing new business models.

6. ‘Private digital spaces will become popular’—This follows from the idea that if publicly-accessible platforms become uninhabitable, people will turn to private spaces. This may be overstating the extent to which people are willing to shift their online behavioural habits, though, and is also predicated on the assumption that social media platforms will decline in prominence, which is of course not guaranteed.

[ PDF Version: Digital Agents and News Consumption ]

Large language models (LLMs) sit at the centre of today's AI boom. They underlie chatbots like OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini. These models are improving all the time: just a few weeks ago, OpenAI released what it claims is a “majorly improved” version of GPT-4 Turbo. But how do they work?

LLMs, like most modern AI systems, are powered by complex machine-learning algorithms that use statistical techniques to discern patterns from huge amounts of data. They are built with three components: the computing power required to train and run them (imagine warehouses full of advanced computer chips); the data on which they are trained; and the algorithmic architecture which enables the system to discern patterns and make predictions. While all three are important, it is advances in algorithmic architecture that laid the foundation for the current generation of LLMs.

In 2017, researchers at Google developed the “transformer architecture”. This improved on the previous state-of-the-art in natural language processing — the subfield in AI concerned with allowing computers to comprehend and produce language — by introducing the mechanism of “self-attention”. Whereas previous models processed words in a sentence sequentially, the transformer architecture allows a model to analyse and weigh the importance of all words in a sentence simultaneously. This has enabled LLMs to grasp the nuances of language with far greater depth and accuracy than was possible in the previous paradigm.

LLMs are trained in two stages. In the first stage (“pre-training”) the models are fed huge amounts of data, which they break down into a series of small chunks called tokens. It is then instructed to teach itself how to predict the probability of a token, given the preceding one. In doing so, the model develops an understanding of language and grammar, developing an understanding of which words appear alongside each other and in what contexts. It does this by comparing its predictions against the actual data and adjusting its internal weights — numerical values within the model's architecture —  until the discrepancy between what it predicts and what the data contains is minimal. This process produces what is often called a “base model”. Given any prompt, a base model will respond by continuing to predict what comes next. 

In the second stage (“fine-tuning”), the model is tailored for specific tasks and preferences through exposure to a smaller, task-relevant dataset. This stage involves not just the algorithmic adjustment of the model’s weights but also human intervention. People are hired to evaluate the responses generated by the base model, rating them as good or bad based on accuracy, relevance, and appropriateness for the task at hand. In this way, the base models are sculpted to perform a particular personality, such as the helpful, harmless assistant that Claude, GPT, and Gemini present themselves as. In other cases, as with Meta’s open-source model Llama-2, model weights are released publicly so that users can fine-tune them themselves.

To the surprise of experts, scaling up the amount of computing power and data used to train these models has led to improvements in their capacities to reason, evidenced by their ever-improving performance across a range of technical benchmarks and real-world tests. The transformer architecture appears to have further room to grow, too — suggesting that LLMs will continue improving at pace, at least in the near future.

As part of my ongoing efforts to 'kill the part of me that cringes', rather than the part that is cringe, and to not let the perfect defeat the good, here are three poems I wrote in October.

1. 

It's cold and I miss you.
Amidst my den of abstractions —
warrens made narrow with worry —
my underground palace, where
increasingly I take my rest,
amidst the pull to focus and
the push for production,
amidst obsession over process
and mud grey light,
jasmine choked by concrete
and spit from great height,
I clear space on my desk —
ten minutes, untimed —
to sit with the thought.

2. 

Today I nurture a quiet emptiness,
holding love close to my chest.
I feel languid, lost, lightly at rest.

I linger on the cosmic ocean.
I sense that I'm in many ways a hamster,
hydrated and wheeling.

I wonder what's missing (though of course I already know).
Caught in montage, I am blown about by the present.
I struggle to make it all fit.

Spiritual decrepitude? In this economy? At this scale?
I marshal troops to the border, fix my eyes to the whale.

I catch wind of a great tension; my body pulls tight.
I won't pan for meaning under pale moonlight
nor will I gargle glass or put up a fight.

I leave myself,
carried by currents,
away from the shore.

Some days I'm only good as a vessel for awe.

3.

I crack a punnet of blueberries with some real intention.
The day unfurls ahead of me as soft, lush carpet.
I slip off my shoes, not dwelling on imprints.

Within two hours, blood separates from snow,
and we are left only with theories.
Amazing how far we get
even without them.
That's what I've been thinking, piling stones in my cave.

There's a sincerity quite distinct from the truth
which I channel sometimes,
in situations off-balance.
It is a sublime defence against authority —
myself and the others.

It always ends in a declaration, I find.
Something like "I will not lose sight of
the fierce beauty latent in the everyday",
intoned with great solemnity, and several
pinches of salt.
A phrase like a talisman,
to rest upon in revolt.

I can still write, you know.

I can write like anything. I can squeeze linearity from my looping mind like juice from an orange. I lay down a trail - drop by drop - for people to follow, and usually they do. I write like I swim – without thought or hesitation, and with patience, poise, and persistence.

If I want, I can write without shame, caveat, or apology. I can write in knowing defiance of Grammarly, the imperialist midwit. I write as a spider spins silk or a child drinks milk. I can write to a deadline; I write as a lifeline. I write like I talk (and boy can I talk).

I can write with myself to the side – tone calm, style neutral– to the needs of a client. I can write to align key stakeholders. I can write in good faith, best wishes. I can write emails, papers, impressions, confessions, reports, raps – entire collections.

Or I can write – myself front-centre – can write in my voice. I can scramble over rivers, meander from point; follow loops of colour like there’s always a choice; or focus all attention on a singular point. Do I want to be perfectly understood or just to have fun with language? To which virtue to steer – Simplicity? Expressiveness? Anarchic folly? Before the thought can land, I return the volley.

So I swing between ideals, led by my mood. Clarity is a fabric I drape depending how much light I’d like to let in. It turns out I can take one by the hand to tour rough terrain or bore one to sleep with a cryptic refrain (it’s all done to train). I write like I’ve got nothing to gain. Or else I write like show and tell – I write to explain.

I can write to prove a point, or I can write to be pointless; write to sit morose, or I write to be joyous. I write to right the boat; keep the boy afloat. I can write to be petty, or pretty, or gloat – writing digs a moat. And I write like it’s my job – so I’m lucky it is. And I write like it’s a hobby, which I’m lucky it is. Compulsion propels me to write and keep writing, to keep at igniting material once dry. I write just to try.

Elsewhere, I write to foster distance or to spark connection; or I write to draw attention and to garner affection. I write through rejection, depression, despair. Wherever I’m cast, I write about there. I write like a lantern cuts through fog, even when writing to the robots (like writing to God).

I can write like I can open a faucet, like I can follow a stream, like I can rest upon your face until I see what you mean. And I can write without insisting on perfection. Sometimes I write just to teach myself a lesson. I write to impress, to express, for excess –given so much to say, to write is a reflex.

I think of myself – or at least the thoughts, feelings, impulses, and desires that comprise me – as a bundle of voices. I try not to interrupt when any one speaks. I circulate a mic. They run the gamut from good to bad – they’re angelic and prideful and venal and crude. Or they’re sensible and solemn and raucous and rude. Some are non-verbal, though all have their means. I can only listen to one cluster at a time – such is the constraint of linear attention. Some of the voices with common interests have formed a union. Collective action for a better Tharin, today! They sing harmonies most mornings, prompting me to stretch or drink water. Others operate in more sinister coalitions, reinforcing habits I can’t shake. Negotiations between all parties are ongoing at the time of writing.

Since I am in constant conversation with myself, I can write forever. And since I write to keep the light burning, I think that I will.

I.

I am generally considered to be a pretty smart guy.

Now to say this directly is to invite doubt and disaster. I know that. I feel it deep in my bones. It’s been drilled into me that, outside of certain domains (like rap and sports), it’s impolite to talk directly about one’s abilities. But I’m pushing through my instinctive discomfort to write this, as I think it might be worth it. 

When I say I’m a “pretty smart guy”, immediately I feel compelled to clarify what I’m not saying. I’m not saying I’m the “smartest” (whatever that might mean), or that I’m made unique or special by my above-average intelligence. I’m just saying that it seems true that above-average intelligence is something I possess, and that, despite how I feel about it, probably it doesn’t function like a fearful bird, likely to fly off as soon as I glance in its direction.

The world’s provided me with a fair amount of feedback in support of this conclusion. My grades have been high since I was in prep school, staying high through my undergrad and postgrad studies. I seem to have fairly consistently placed in the top 10-20% of my cohort, depending how you draw the boundaries.

Of course, grades aren’t the only way we measure intelligence, there are lots of reasons why smart people might get bad grades, and it’s possible to get good grades and still be rather dull. Fine. But it’s not just grades that’ve led me to think this—I continue to be complimented on my creativity and my mental abilities in the working world, sometimes by people who have nothing to gain by doing so. And most of my friends (and even some of my enemies) hold me in similar regard: just the other night I hung out with someone who, after he left, thanked me for the unusually stimulating conversation.

I’m belabouring this point, awkward as it feels, because I’m hoping you’ll believe me when I say that despite my intelligence, I am, so often, so painfully stupid. Slow, dull, uninspired, and uninteresting—like a snail that, having decided to end it all, begins sliming towards salt, only to fall asleep halfway there. If you know me personally, this will not come as a surprise. When I was 13, my scout master called me “either the smartest dumb person or the dumbest smart person” he’d ever met; more than a decade later, I don’t really disagree.

Recently, when reflecting on how my intelligence and my idiocy coexist, I was struck by something simple: I am every bit “myself” when stupid as when smart. They’re both indispensable facets of my identity. I don’t think I’d want to rid myself of my idiot streak any more than I’d want to lose my intelligence.

I think this provides an unexpectedly elegant solution to the question "why do I always feel like an imposter, like my credentials are fake and my knowledge is hollow; like I am made of straw while my peers, cast in silver, shine?”. It’s because sometimes I'm silver, and sometimes I'm straw.

The hours of thought I’ve invested in writing this? Silver boi hours. The hours I spent on youtube, watching reviews of videogames I’ll probably never play, while gulping down fistfuls of cheddar, thereby forging new landmarks for the ants to explore? Hours of straw.

II.

So here’s my working theory: everyone feels like an imposter some of the time, because everyone is an imposter some of the time, insofar as it is impossible for one to always be performing at the top of their game. Sometimes, we are impersonating the better versions of ourselves.

Like almost everything said in an intellectual register, this idea is not new. It found its way to me through a tweet by Rachael Meager, whose public grappling with inadequacy helped me immensely. And I’d assume a lot has been said on this in the formal literature on imposter syndrome as well—I don’t know, because, respectfully, I’m not gonna read all that now.

In the framing I’m presenting, “imposter syndrome” is the unpleasant feeling that flows from the tension between who you expect yourself to be, and who you actually are in the moment when you feel like an imposter. To function, it requires three things.

First, that you feel compelled to live up to a certain standard of performance—perhaps because your parents/culture/‘society’ demanded this of you, or because you internalised the standard at a young age and so have grown used to demanding perfection from yourself. Such standards are found across multiple domains—think beauty standards, academic standards, social standards. Although they vary in terms of what is required to live up to them, they are united by their demand for perfection.

Second, it requires that you do occasionally live up to those standards. Sometimes you are indeed beautiful, fashionable, the centre of trend. You are eloquent, incisive, interesting. Sonnets and sonatas fall out of you like loose change from your pocket. You are the model child, student, employee. Pretty close to perfect.

If this condition isn’t met, then I don’t think we’re talking about imposter syndrome. For example, I don’t feel like an imposter when it comes to cooking, because I’ve never been any good at it in the first place. I have no private convictions about my potential, and I feel no fear of falsity, because nobody has ever assumed I was a competent cook to begin with. I haven’t fooled anyone.

The third thing imposter syndrome requires to function is the belief that, whenever you fail to live up to the relevant standard, you are without value, or are undeserving of love, or are [whatever your self-hatred sounds like to you]. This is why imposter syndrome is so pernicious: it takes the very thing that fills you with pride, that you know you have the capacity to excel at, and then punishes you for ostensibly falling short of your potential. It makes your self-worth subordinate to your ability to perform to a standard that is, ultimately, impossible to meet at every moment.

No wonder so many people are having a bad time. Doing some searching while writing this, I discovered that everyone from Andre 3000 to Maya Angelou struggles with feelings of fraudulence. The problem is easiest to observe in creative domains—whenever someone preemptively apologises for sharing their work, I think it’s the fear of fraudulence that pushes the apology out; as if they have to say sorry for failing to live up to perfection.

I’ve noticed this in people my own age and in people older than me, though I find it most disconcerting when it comes from the youth. I recently visited my 14 year old cousin, who, before showing me his art, remarked several times that it was stupid and not very good. I was surprised to see that the demands of perfection had taken root so early, though I shouldn’t have been—by that age, they’d already infiltrated me too. So while I wanted to stamp out his self-doubt, I couldn’t; I can barely keep it at bay in myself.

[Related: the idea of the “insecure overachiever”, which the linked article explains well, and which I first encountered in “Talent”, by Tyler Cowen and Daniel Gross.]

III.

Three examples of imposter syndrome in action: 

1. I often find myself awash with anxiety when choosing an outfit to wear on a night out. Cape Town is an aggressively trendy place, so I think what’s happening is that I’m holding myself to a certain standard of “cool”, and then despairing when I realise my wardrobe doesn’t have the depth to meet it. This never used to be a problem for me, perhaps because I felt less internal pressure to conform when I was younger (or perhaps because, growing up in Durban, a city which has been styling itself the same way since 2006, literally any deviation from the median dress style was sufficient to qualify as cool, at least in my own eyes). I don’t know—I only know that now is the era of Gen Z. Maybe this is just a part of getting older. __ In any event, this is an issue that is like 90% just in my head—every time I become convinced I’ve committed some grave aesthetic faux pas, it turns out not to matter at all.

2. I mentioned earlier that friends sometimes remark favourably on the quality of the conversations we have; and that they praise me for my eloquence. Those are times of silver. But when I’m straw, I find I have nothing particularly interesting to say; nothing valuable to contribute. Words plod out of me, inert as stone. Since I’m supposed to be fantastically articulate, I have failed myself—that’s how it feels. __ Sometimes, when this happens, I text whoever I was with the next day, saying something that amounts to“hey sorry I was brain-dead last night, hope you still like me/we’re still friends😅”; to which they typically reply with something like “what? No, it’s fine, you’re fine, it was great to see you anyway!”. Shoutout to them.

3. To feel like an imposter is to feel like you are not entitled to inhabit a certain role. Thus far, I’ve been focusing on roles that are creative, aesthetic, intellectual, or otherwise related to skill and talent. That’s because it’s in those roles that I feel most fraudulent. But imposter syndrome can extend further, to infect more fundamental roles—those that are defined by our relationships with other people (and particularly with friends and family).

   The most vivid example I can think of here is one I’m borrowing, with permission, from a friend. Because of her dysfunctional relationship with her parents, who are abusive, she often feels like an imposter in family situations, even when they don’t involve her own family. She feels like she’s impersonating a “good” daughter, and that any second she’ll be found out. This is largely because her parents have, for years, chastised her for failing to live up to their regressive values. They seem to conceive of mental illness as inherently disrespectful, and have said things to her that no child should have to hear. They have caused her to internalise standards that run contrary to who she is.

   This is not the place to unpack what lies behind this intergenerational trauma, and I’m pretty sure I’m not qualified to do so anyway; so suffice it to say that this form of imposter syndrome is particularly cruel, because it encroaches on parts of one’s identity that should exist near-unconditionally. It is born out of toxicity. This is not to suggest that one cannot in fact be a bad child (or a bad friend, or a bad partner)—plainly, the world is complicated enough for such cases to exist as well—but I think those cases lie beyond the scope of a discussion on imposter syndrome, so I won’t say more on them here.

What can we take away from these examples? Three things, I think.

First, it’s easy to be mistaken about when one is silver and when one is straw. One’s internal sense of this is often unmoored from reality. Second, one’s time as straw often feels as if it invalidates one’s time as silver, baselessly. And third, when one feels like a fraud in an interpersonal context, this might be because the social standards one feels compelled to conform to are themselves toxic or contrary to one’s own values.

IV.

Sometimes I feel like a show pony, only as valuable as my last trick. I’m not sure this accurately reflects how show ponies are valued, but nevertheless, the feeling persists. It’s a fairly common feeling among so-called overachievers—that your worth as an individual depends on the last impressive thing you did. Although as a matter of logic, I know this to be irrational, it’s still hard to shake, since it has apparently become so deeply internalised. Feelings don’t care about facts about feelings, or whatever Ben Shapiro said.

I often find myself thinking about this phenomenal video by author/poet Savannah Brown—“I could write my magnum opus, or I could simply go to bed”. You should watch it.

In the video, she shares an image which I often return to: that of a creative work as a work of embroidery, where the front is clean and tidy, while the back is riddled with knots and mistakes, invisible to everyone but the creator. If you’ve ever embroidered, or you’ve done something like obsess over the structure of a single sentence that seems to read just fine to everyone else, you’ll know what she means.

I think this image provides another way to describe the experience of imposter syndrome—because you can see the knots in your own work (whatever that work might be), and because everyone else’s work appears from the outside to be unblemished, you end up with the mistaken impression that everyone else’s stuff is flawless except for yours. But as I’ve tried to lay out here, as best I can tell everyone is oscillating between states of silver and straw all the time. To quote Rachael Meager once more, “the entire world is held together with duct tape and chewing gum”.

It just isn’t possible to perform at the edge of one’s ability all the time; or to perfectly play the parts we think are expected of us. It's not a thing. What a relief.

Relief, because once you liberate yourself from the demands of perfection, you are free to do whatever appeals to you, just because. One area where this has been working out for me lately is music. I learnt to play piano as a child, spending five years rigidly memorising scales and classical pieces, to play at formal exams, to advance to the next grade. I learnt music as a science. But music is more like language, and for a long time, I couldn’t speak it.

Recently, after a decade away, in the solitude of lockdown, I started playing again. Playing, for the first time, for my own joy before anything else. And it’s been great! I’m not particularly good, of course—from the perspective of a classical musician, I’ve probably gotten worse than I was as a child—but I can now play with a fluidity which I never had before. I can jam. And I can do it safe in the knowledge that there’s no test to follow, no standard to meet. Ephemeral expression is the only end.

I think the main reason I was able to find the comfort with music that eluded me as a child is that I have given myself a licence to be bad. In a world where “perfect is the enemy of good”—where an insistence on perfection often prevents us from taking actions that would improve our lives—a licence to be bad is a powerful thing. It entitles you to do as you wish (so long as this doesn’t harm others) without having to conform to a certain standard of performance. Here, I’ll give you one now. 

Pictured: A licence to be bad. It’s not very well-designed. This, friend, is the point. 

V.

If you find yourself confronted with the feeling that your life is a long con, that you have tricked everyone around you into believing in your competence when in fact you are a fool; take heart—they have tricked you too. Everyone is an imposter. The world still turns. If we are to have any fun on this burning orb, we have to be careful to not let the things that empower us, that we are uncommonly good at or that bring us deep joy, to not let these things become sources of burden, of obligation. As best I can tell, life’s too short.

Now, from straw have I come, so to straw I return. 

When I think back on the five nights I recently spent on holiday in Istanbul, I think of words like “dense”, “sprawling”, and “kebap”. I think of steep cobbled streets and unclean air. I think about what a privilege it was to be able to travel again (or at all).

Before the wonder of the trip fades entirely—that wonder that attends the shock of being immersed in a foreign culture, of being in a place where even such foundational elements as the air, the trees, and the sky are subtly different—before I lose that to the steady patterns of my daily life, I thought I’d jot down some points that stood out to me while I was gone. 

• Cost: Everything was relatively cheap because of the hyperinflation Turkey is experiencing, which this Planet Money episode explains much better than I could. The value of the Turkish Lira has plummeted—the dollar:lira exchange rate has gone from 1:3.6 in 2017 to about 1:15 at the time of writing. It lost 41% of its value in 2021 alone. This is… terrible for the Turkish people, but good news for others hoping to travel cheaply to that part of the world.

• Cats: There are so many in Istanbul! Estimates range from a hundred thousand to a million. They are apparently viewed as “communally-owned”. According to Wikipedia, “every street in Istanbul has cats that are familiar to locals. In these streets, tiny “cat houses” have been prepared for stray cats. Neighbourhood residents place food and water containers in front of their houses to feed [them]”. This accords with what I saw: cats in shop windows, cats at bars, cats in groups and cats alone, cats in street art and the hearts of people. Cats.

• Cleanliness: Most of the parts of the city that I saw had an inescapable griminess. Lots of flaking paint, abandoned buildings, piles of rubble. The air was often sharp with smoke. Wikipedia claims that around 8% of the country’s deaths are caused by air pollution-related illnesses, which checks out. The urban decay was counterbalanced, though, by the abundance of graffiti and street art, which gave the environment colour and personality.
  
  I don’t want to give the impression that trash was piling up in the street or anything—the grime didn’t really distract from all the glossy shops and cafes that were packed into every alley. Mostly, I was just struck by how safe I felt, despite the grime. The city was always busy, with people wandering the streets at every hour, so I never felt unsafe walking around—even after dark. Coming from South Africa, that’s a novelty. Grime > crime.
  

• Cafes: It was great that there were so many cafes open late into the evening. That one can’t get a good hot drink after like 9pm in Cape Town never fails to make me sad.
  

• Corruption: Since I started working at an anti-corruption organisation, my ears perk up whenever I hear the c-word. So I found it really interesting that Turkish politics is also marred by corruption, which has steadily increased over the last decade (although it plays out differently than in South Africa).
  
  The history of Turkish politics, and the events that led to current President Recep Erdoğan’s election, are far too complex for me to capture here—after some basic reading, one thing I know is that I don’t really understand what’s going on. It struck me while reading this in-depth review of a biography about Erdoğan, though, that the president used accusations of corruption against his enemies to consolidate his power, and that he rode waves of pro-EU sentiment to denude historic government institutions of their capacity, while at the same time creating new institutions staffed by those loyal to him.

  There was also a coup in 2016 that led to “about 10% of Turkey’s 2 million public employees” being removed from office, which…damn. These few facts form the tip of an iceberg. If you want to know more, maybe read about it and let me know what you learn.

• Cool Spots: Istanbul proved to be the kind of place ideal for would-be flâneurs, where just wandering the streets was rewarding in itself.

  There’s a feeling of satisfaction that follows when a terrain that was once wholly unfamiliar starts to lock together in one’s mind. I associate it most strongly with games like Dark Souls, where one can experience this in its most concentrated form, though of course the real world abounds with such examples as well. This is what it felt like to walk around in the city: gradually, order emerged from chaos, and I started to get a sense of how the different pins on my map (below) connected to one another.

  Beyond this general experience, if I had to pick some highlights I’d point to my visits to Topkapi Palace, former home of the Sultans; to the Istanbul Archeological Museums, which house countless ancient Greek sculptures, and which have a whole wing dedicated to ancient Sumerian artefacts (including what some claim to be the “world’s oldest love poem”); and to a cruise I took down the Bosphorus River, which divides the European and Asian sides of the city. Despite the inevitable onset of museum fatigue, I still find myself thinking of the serene beauty of the many things I saw.

• Culture: Culture, being an amorphous, heterogenous, and diffuse beast, is difficult to pin down; even if we distinguish between the historical culture peddled by tourist agencies and guidebooks and the contemporary culture experienced by living people. If someone asked me about the dominant contemporary culture of the city in which I currently live, Cape Town, I’m not sure what I’d say—probably something about the city-wide love of the mountain. But city culture is different from suburb culture is different from campus culture; and factors like race and class augment things further.
  

  Whenever I travel, I typically yearn to know what the cultural experience of someone with my age and interests is like. This is always an unsatisfying pursuit: if I can’t even confidently answer this for my own city, how can I hope to in someone else’s? But even so, from the few vignettes I was exposed to, and particularly from the brief chats I had with university-adjacent strangers (I met a troupe of dancers at a rooftop bar one night, and a young doctor and his friends at a techno club on another), I got the sense that young people there are like the young people in every city, albeit filtered through their peculiar historical and legal regimes. (I was surprised to discover that, unlike in South Africa, nobody seemed to really smoke weed.)

  I guess, in vague conclusion, shoutout to globalisation for making faraway parts of the world instantly recognisable—like the shadowy dancefloor dominated by figures in all-black, as at every other techno party I’ve been to. You make it easier to fit in.

• Concluding Thoughts: I only did like 60% of what I’d intended to. I would happily go back—though if given the choice between Istanbul and somewhere else I haven’t been, I’d probably take the place unvisited, as I’m always eager to take in new sights.
  
  Anyway, here’s a list of cool places to visit that I made while I was away, in case you’re planning a trip: Istanbul by the Horns

Hopelessly derivative, it was inspired by whatever media I was consuming at the time - I recall the Ratchet and Clank series, and Avatar: The Last Airbender both being heavily influential. I envisioned a galaxy of planets, each with distinct people, cultures, and economic activities. The system was quasi-federal; the political relations complex. For example, I recall creating the planet “Laboris”, the lab planet, where new technology would be developed and tested, before being shipped to worlds the star-system over. There was a planet dedicated to retail and advertising - animated by the spirit of Moloch, I now realise. And so on. Populating these planets, in starring roles, were my friends, family members, and any other people in my life at the time, all reimagined in the art style of my mind’s eye. 

As my real life changed, so did my fictional one. Alliances were forged and broken. Civilisations rose and fell. It was a serial drama, with some weeks covering mere days, and some days spanning decades. As serious challenges came up at school — a block of tests, for example — I would envision each obstacle taking a physical form, like an elaborate boss fight. Although the details escape me, I remember my Grade 7 exams mapping onto the mythology of 2007’s Heavenly Sword.

I found great joy in specificity — I’d try to imagine the content of the advertisements that ran on the intergalactic cable channels, the outfits worn by my fellow adventurers, that kind of thing. In pursuing these specifics, I’d frequently run into the limits of my own imaginative capacity. While I was strong on narrative (a consequence of all the books I’d swallowed), and while each character I created evoked strong vibes in my brain, I was quite unable to picture exactly what they looked like. It was like trying to pin fog to paper.  

This confrontation with my own limits frequently led me to fantasise about a technology that could scan my brain and produce images of what I was thinking, so I could show my peers, and so I could circumvent my utter inability to draw. I thought it might work like a polygraph machine, with a mess of wires that, through wizardry, would output my thoughts on a page.

I grew up, and I relegated this dream to the realm of science-fiction. I accepted the limits of my own creativity, and was happy to work within them. So I couldn’t produce beautiful illustrations, or really any illustrations at all. Fine; at least I was good with words. And if the medium of language, written and spoken, was good enough for the endless procession of great writers/thinkers/creators that inspired me, no doubt it was sufficient for me too. 

That was pretty much my position until I came across DALL-E 2. DALL-E 2 is an “AI system that can create realistic images and art from a description in natural language”. 

You’ve probably seen systems that do similar things — I had. DALL-E 2 stands out, though, for the quality of the images it seems capable of producing, and for their worth as expressions of a machine’s imagination. If you haven’t seen it in action, it’s worth checking out (in addition the link above) this video, this twitter thread that turned bios into images , or this one that explores the system’s various limitations. 

I’m not writing this to dissect the abilities of DALL-E 2 per se — not least because that’s already been done (see e.g. “what DALL-E 2 can and cannot do”), and because I don’t actually have access to it. Instead, let’s get back to my childhood fantasies. Is this the technology I dreamed of as a kid? 

As far as I can tell…not quite, but it’s remarkably close. If I had access to something like this when I was younger, I think it would’ve been transformative. It would have given me a new tool with which to explore my inner life. And the fact that this exists now sparks hope in me for what we might see in the future. A model that can turn short stories into short films, for example, doesn’t seem especially far off. 

Now, you may ask: “But who would want this? What use would this be? What about poverty/inequality/glaring social problems?”.  To this I’d respond, in turn, “I’d want this! I think it might deepen my capacity for thought (and probably also do other useful things). And yes, other social issues are obviously important and deserve attention, but thankfully it’s not a zero-sum game”. 

Let’s focus on my second response, about deepening capacity for thought, which I think is the most interesting part here. To explain what I mean, I want to briefly talk about “conceptual metaphor”. This, in my rudimentary understanding, is the idea that metaphors are not just literary devices, but are fundamental to how we make sense of the world, as they allow us to compare a domain we understand to one we do not; and in doing so, to understand abstract ideas. These kinds of metaphors are embedded in our everyday language — consider how phrases like “the price of oil rose” or “the market fell” are predicated on the idea that MORE IS UP, LESS IS DOWN; or how a phrase like “today dragged on” is predicated on the idea that TIME PASSING IS MOTION. 

What’s interesting about this is that it suggests that metaphors occur at a level deeper than language — they inform how we generate thoughts to begin with. I’ve been thinking about this a lot lately, as I’ve been writing about how complex social problems are often discussed in inescapably metaphorical terms. For example, one frequently encounters talk of the “fight against corruption”, the implication being something like CORRUPTION IS AN ENEMY IN WAR. I think this can be useful to some extent, as it helps us understand the nature of the problem, but unhelpful in other ways, as for example it leaves ambiguous what the conditions for victory in this fight are. 

Anyway. Where I’m going with this is that, if large swathes of our thoughts are metaphorical, relying on intuitive comparisons between domains, having a tool that allows one to visualise these comparisons would be incredibly interesting. It would let me see what the fight against corruption might look like, which might help me pinpoint why the comparison is or is not a good one. 

Another example: often, when I’m trying to do multiple tasks at once, I think of an image of a man spinning multiple plates. You can find dozens of visualisations of this online — here’s a clip of a guy doing it in 1958 —  but it’s not quite what I’m picturing.  I don’t want to claim that, if I could only use a tool to generate the right image of a plate-spinning man, an image that is sufficiently similar to my internal experience, that I would then understand something new about the nature of multitasking. That doesn’t seem true. 

But at the same time, if I was able to perform this kind of check regularly, where I generate images based on the implicit metaphors that are informing my thoughts, I’d guess that that might lead me to notice which metaphors are particularly good or bad at helping me understand the world, and that my worldview would update accordingly. So I can see how one could draw a line between using this kind of AI system regularly, and enriching one’s internal experience. Indeed, that seems to be what Sam Altman, CEO of OpenAI (the organisation that developed DALL-E 2) is getting at when he says that using the system has “slightly rewired [his] brain”. 

I don’t think this use-case of clarifying conceptual metaphors is the only way a system like DALL-E 2 could enrich thought. It also just seems like a fun way to explore the nonsense word fragments that sometimes pop into one’s head. The other day, for example, the phrase “crocodiles in Lacoste!” showed up while I was driving. With DALL-E 2, I could really see that. This would be low-grade amusing to current, mid-20s me, but I think it would’ve absolutely blown my child self’s mind. 

Younger me would’ve recognised that here was a way to fill in all the specifics of the fictional world in which I spent so many afternoons. Here was a way to nail down the aesthetics of my fellow adventures, my friends reimagined, as I experienced them internally. Here was a way to upgrade my imagination. That’s how the kid in me feels now, at any rate. 

I don’t know where any of this will lead, but, x-risk aside, I’m eager to find out. 

I’m writing to disagree with you. This, I hope you’ll agree, is a high honour. It’s a sign that I think your thoughts are worth taking the time to grapple with and unpick. I’m doing this to satisfy myself, to contribute to the discourse around the value of note-taking systems, and to consolidate an argument on the subject I recently had with a friend. 

My principal critiques are as follows: 

• Getting words down on a page is intrinsically valuable, whether you intend to share them for public consumption or just need to work out what you really think about something. It’s valuable for the same reason lifting weights or taking a walk outside is valuable: it’s good for your mind/body in the moment that it’s happening. I think second-order effects, like whether you’ll become really buff or make a morning walk part of your daily routine, are subordinate to this in-the-moment value. 
  

  • So if you find a tool that helps increase your written output, this is to be celebrated. Undermining this output by suggesting you could be using this time better, by calling a friend or by actually “creating things”, seems to start from the wrong comparative point - for most people, I think the alternative to using such a tool is doing other unthreatening things, like scrolling Twitter and Instagram. 
    

  • People are either going to create things or not, whether or not they use note-taking tools. So to suggest that the use of such a tool syphons people away from other creative/emotionally satisfying pursuits is misleading - I worry it conflates these different groups. It’s also not clear why we should subject organisational tools to such strict scrutiny based on their capacity to enable procrastination, when this capacity is a feature built into almost all technology in the modern world. People using note-taking tools who convince themselves that their unpreparedness is the bottleneck to them producing stuff seem unlikely to have produced stuff even in the absence of such tools. 
    

  • Another reason I think writing words down is intrinsically valuable is that, as you repeatedly emphasise in your own writing, the best way to improve one’s writing is to just…write more. Indeed, your refrain around this theme often echoes in my head, inspiring me to put words to page. 

• Many of your criticisms seem to make considerable assumptions about how most people use note-taking tools. These include: 

  1. that using a note-taking tool will cause you to rely on some other filter for what you find interesting/important rather than what is naturally salient. 

  2. that using a note-taking tool precludes you from writing stunted fragments of ideas, and from being original more broadly. Relatedly, that using these tools leads you to focus on factual insight as the core of your work.

  3. that the only serious reasons for using note-taking tools are to manage complex research or to compensate for a disability, beyond which one is just “LARPing”.

• I think these assumptions are largely unjustified. This is because: 

  1. In choosing what is worth noting down, you are making a judgement call regarding salience. Nobody can note down absolutely everything, even assuming they wanted to. So judgement calls guided by what organically sticks in one’s mind are not bypassed by making use of a formal system. Further - and this is largely anecdotal, though I’ve seen some lukewarm evidence to support it - writing things down typically causes those things to stick more vividly in one’s memory. Which things stick? That’s not something we are really in charge of - it’s controlled by our salience filters, which we are always relying on.
     

  2. There is no reason to think that using a note-taking tool precludes one from writing stunted fragments. I jot down such half-formed thoughts in my Roam database almost every day, with little regard for what’s come before. The idea that while using a note-taking tool, you can’t start each day fresh, generating original stuff from read/write errors, is unfounded. Sure, some people may become so obsessed with playing in their “magical junkyard” that their creative efforts become stultified, but my response here is the same as what I said regarding procrastination above - if your use of a tool results in you producing unoriginal stuff, perhaps you weren’t that original to begin with. Similarly, if one is limited into focusing on factual insights, this has more to do with their intentions and use of the tool, than with the tool itself, which is compatible with various focus areas.
     

  3. While I agree note-taking tools are most useful to those engaged in serious research, and that for the majority of people there’s probably no need to use them, I don’t think we should denigrate those that use them outside of these ends as merely “LARPing”, for two reasons. 
     

     • First, LARPing is a form of play. If people are getting a kick out of pretending to be productive, if using a note-taking tool feels like a game to them, power to them! Yes, they may in some sense be wasting their time (if the end is producing content), but as I keep returning to, this time probably would’ve been wasted anyway, and it’s not clear to me why this form of procrastination is any worse than the others. 
       

     • Second, I think one’s use of a tool for work can be intertwined with its use for play. This is the case for me. During a typical work-day, I use Roam to keep track of various tasks that need doing and subjects that need researching (I currently work as legal researcher); but I also use it to pepper in poems, stunted fragments, thoughts that go nowhere. I think a lot of interesting ideas come from letting these fragments coexist alongside more serious work, which your narrowly-defined categories of where note-taking tools are useful may miss.
       

• Certainly, I don’t think using a note-taking system is necessary (or even sufficient) to do interesting work. Obviously it is not. I also agree that people learning from “knowledge management” courses and the like may absorb a dogma that leaves their work inert. I don’t even disagree that a note-taking system should be as elementary as possible - indeed, I think the best systems evolve organically and in response to the needs of their users.

• But I do worry that the idea that you don’t need tools fancier than da Vinci’s to do interesting work is rooted in a traditionalist ethic; in the idea that great people of the past made do with what they had, so who are we to lay claim to more? I think this overlooks the very real possibility that da Vinci would’ve loved digital note-taking systems, or if not da Vinci, then some other historic intellectual on whose merit we can agree. Using a note-taking system doesn’t imply that you think you’re better than da Vinci; and in any case, who knows what da Vinci’s workflow would look like in the modern era? Not me.

I don’t expect you to radically change your opinion on any of the above, and indeed you may find fault in much of what I’ve said here; but I thought discourse around this topic might be made richer with my contribution, so this is that. Writing this felt like play. Now I’m going to go outside. 

Yours in good spirit,
Tharin

I was pretty meandering in my choices last year, guided by what books were in second-hand stores, what movies were in cinemas, what games were on sale. So I make no claim to objectivity.

I’m writing this to share stuff that brought me joy - something here might do the same for you.

Books

1. Woman on the Edge of Time - Marge Piercy (1976)

Woman on the Edge of Time tells the story of a Mexican-American woman in 1970s New York who is unjustly committed to a mental hospital, and who is frequently visited by an androgenous figure from the far-future. It is “considered a classic of utopian "speculative" science fiction, as well as a feminist classic”.

Two things strike me about this book. First, it provides an unusually thoughtful depiction of a (queer, communist) utopia. It devotes considerable time to explaining how culture, race, sexuality, politics, labour, drugs, technology, and more are handled in 2137. In doing so, it unironically depicts the future liberals want.

Second, it explores issues often considered ‘modern’ - like gender fluidity and the oppression that arises from being simultaneously poor, non-white, and female - with nuance and care. Given that it was written over a decade before Kimberlé Crenshaw coined the term “intersectionality”, this resonance caught me by surprise (though that might just be naivete on my part).

8/10 would recommend if any of the above words sound appealing to you.

2. The Buddha of Suburbia - Hanif Kureishi (1990)

This is the most fun coming-of-age story I’ve read in years. Set in 1970s London, it follows the exploits of Karim, a crude conniving charlatan of a boy, as he incites bisexual chaos wherever he goes. Full of interesting characters and slick turns of phrase, it is well-observed in the way thinly-veiled autobiographies often are; and to spend time with a protagonist who was not only a bastard but also a brown boy brought me great pleasure.

I still think of the passage below, where he talks about his creative process in preparing for a theatre role, all the damn time.

“There were few jobs I relished as much as the invention of Changez/Tariq … I  uncovered notions, connections, initiatives I didn’t even know were present in my mind. I became more energetic and alive as I brushed in new colours and shades. I worked regularly and kept a journal; I saw that creation was an accretive process which couldn’t be hurried, and which involved patience and, primarily, love. I felt more solid myself, and not as if my mind were just a kind of cinema for myriad impressions and emotions to flicker through. This was worth doing, this had meaning, this added up to the elements of my life.” (italics added)

8.5/10 would recommend if you’re after a subversive coming-of-age story.

• See also: this essay by Zadie Smith, on her appreciation of this book.

3. The Diving Bell and the Butterfly - Jean-Dominique Bauby (1997)

This is a short memoir by the former editor-in-chief of Elle magazine. In a series of vignettes, he describes his life before and after he was all but completely paralyzed by a stroke, retaining only the ability to blink his left eyelid.

He wrote this book over ten months with the help of a transcriber, who would read out each letter of the alphabet until his blink signalled that she should stop. It took 200 000 blinks

.

Bauby, paralysed, dictating his book blink by blink

Bauby’s lucid prose stands in stark contrast to his deteriorating physical condition. An example:

“Speech therapy is an art that deserves to be more widely known. You cannot imagine the acrobatics your tongue mechanically performs in order to produce all the sounds of a language. Just now I am struggling with the letter l, a pitiful admission for an editor in chief who cannot even pronounce the name of his own magazine! On good days, between coughing fits, I muster enough energy and wind to be able to puff out one or two phonemes. On my birthday, Sandrine managed to get me to pronounce the whole alphabet more or less intelligibly. I could not have had a better present. It was as if those twenty-six letters had been wrenched from the void; my own hoarse voice seemed to emanate from a far-off country. The exhausting exercise left me feeling like a caveman discovering language for the first time.

Sometimes the phone interrupts our work, and I take advantage of Sandrine's presence to be in touch with loved ones, to intercept and catch passing fragments of life, the way you catch a butterfly. My daughter, Celeste, tells me of her adventures with her pony. In five months she will be nine. My father tells me how hard it is to stay on his feet. He is fighting undaunted through his ninety-third year. These two are the outer links of the chain of love that surrounds and protects me. I often wonder about the effect of these one-way conversations on those at the other end of the line. I am overwhelmed by them. How dearly I would love to be able to respond with something other than silence to these tender calls. I know that some of them find it unbearable. Sweet Florence refuses to speak to me unless I first breathe noisily into the receiver that Sandrine holds glued to my ear. "Are you there, Jean-Do?" she asks anxiously over the air.

And I have to admit that at times I do not know anymore.”

Intimate and easy to read, 8/10 would recommend if you’re looking to be inspired by the tenacity of the human spirit.

4. Nation - Terry Pratchett (2008)

After a tsunami wipes out the entire society of an island, an indigenous boy and a British girl, marooned after shipwreck, are left to restart civilization. The story is exciting, well-told, and shot through with interesting musings on the role of religion, authority, etc.

Reading this made me feel like a kid again. 8/10 would recommend if you’re in need of adventure, or just enjoy island survival tales.

5. The Nix - Nathan Hill (2016)

This book is all over the place. It’s set in modern America, though it detours through the late 1960s, and other eras too. It follows more protagonists than I can recall. It’s long, and I wouldn’t recommend it to everyone. But it also contains a few scenes that stick in my memory more than almost anything else I’ve read this year, and to be able to take refuge in its self-contained world was a joy. 8/10 don’t know who I’d recommend this for, but I liked it.

Movies

1. Dune (2021)

Space desert! Giant worms! Great cast! Excellent soundtrack! The excitement I felt seeing this on a massive screen with a face full of popcorn is how I imagine Star Wars fans felt at the first screenings of Episode IV. It’s been a long time since I was this excited by a new franchise.

2. Spring, Summer, Autumn, Winter...and Spring (2003)

Spring, Summer, Autumn, Winter, and Spring is “about a Buddhist monastery that floats on a lake in a pristine forest. The story is about the life of a Buddhist monk as he passes through the seasons of his life, from childhood to old age”. It is beautifully shot, consistently surprising, and allegorically potent. I was enthralled from the first scene.

3. Promising Young Woman (2021)

Promising Young Woman follows “a young woman haunted by a traumatic past as she navigates balancing forgiveness and vengeance”.

I think that description undersells how viscerally uncomfortable this movie was, for me at least. I’m glad it exists. Also: its use of an orchestral cover of Britney Spears’ Toxic is truly inspired. (that’s intertextuality, baby!).

4. Summer of Soul (2021)

In 1969, the same year as Woodstock, the Harlem Cultural Festival went down. It featured performances by Stevie Wonder, Nina Simone, B.B. King, and many, many more. Footage of this festival had never seen the light of day till this was made; so Summer of Soul is full of moments where the aged stars see their own performances for the first time in decades. It’s magical.

It’s also phenomenally well-edited, with director (and Roots drummer) Questlove splicing together historical context, heartfelt interviews, and raw footage at a steady clip. 9/10 would recommend for anyone who loves live performances.

5. Donnie Brasco (1997)

Classic mafia movie with Al Pacino and Johnny Depp? Betrayal, gangsterism, fun accents? Fuggetaboutit.

Honourable mentions:

• The Godfather Part II (1974)

• Save Yourselves (2021)

• Bo Burnham: Inside (2021)

Games

[I'm resisting the urge to get into the details of how these games operate here, mostly due to constraints on my time and my desire to work on other, ostensibly more important stuff. If you want to know more about anything, google it or chat to me - I'm always happy to discuss these weird fictional worlds.]

1. Hollow Knight (2018)

Just a tiny bug in a giant underground world. The art is beautiful, the platforming and combat is fluid and responsive, and you get to travel Hallownest on the back of a stag beetle. Metroid for who?

2. Ori and the Will of the Wisps (2020)

Just a small creature of light in a giant forest world. Hollow Knight by way of Pixar. It’s not quite as good (I liked the combat less), but it still presents a world that’s worth spending time in.

3. Shin Megami Tensai V (2021)

In SMT V, you get to collect and fuse together demons drawn from a diverse set of mythologies (e.g. Asparas from Hindu and Buddhist mythology, Inugami from Japanese mythology) as you punt forward a convoluted and melodramatic plot involving a war between angels and demons in post-apocalyptic Tokyo. The game is a mix between turn-based combat and open-world exploration. 

While the story is largely nonsensical, and while a cynic might say the combat basically boils down to clicking through a series of colourful spreadsheets, no other game of late has drawn me in so deeply and dangerously. It’s everything I wanted Pokemon to be, and then some.

4. Spelunky 2 (2021)        

I spent over 20 hours with this roguelite platformer and have seen maybe 20% of its total content. Ludicrously difficult and deceptively complex, it nevertheless seems worth mentioning here as it got stuck under my skin for months, before I made peace with the fact that its secrets were not for me.

Bonus Stuff

1. Floating Points, Pharoah Sanders, and the London Symphony Orchestra - Promises

• Undoubtedly my favourite piece of music from 2021. Movement 6 had me on the verge of tears a half dozen times last year. It is a work of undulating, unadulterated beauty. It makes me happy to be human.
  

  

2. Channel 5 with Andrew Callaghan - Chet Hanks Interview

• This interview with Tom Hanks’ son has been endlessly amusing to me. I’ve watched it like six times. I’ve shown it to friends, to family, to casual acquaintances. I might watch it again right now. 

(source)

08:30 An early start today, as I prepare to head out. Not physically, of course - my body will stay slumped in my chair, in front of my screen - but digitally, mentally, there’s a lot of ground to cover. I have trails to follow, sources to gather, answers to find. Tasks are piling up; sidequests dot my map. If I’m to get anything done this session, I must cast myself online.

But one has to be careful on such trips - it's easy to get lost. 

For one thing, the territory is endless. In every direction one can move, horizons expand. It's infinity out there. It's tundra, howling wind, verdant forest, deep space. It's tumbling without a body till pangs of hunger plug you back in. It's bleary and abstract and, occasionally, kinda beautiful. 

But I digress - I was trying to talk about how easy it is to get lost in cyberspace. Maybe an example will help. 

Consider right now, when I used the term ‘cyberspace’. After writing that down, I grew suspicious: is that a real word? Where and when did it come from? The thought had barely broken on the shore of my mind before I found myself on the cyberspace wikipedia page, which linked me to an article about internet metaphors. Here, I came across the claim that “over time [internet] metaphors have become embedded in cultural communications, subconsciously shaping the cognitive frameworks and perceptions of users”, which left me wondering ‘...what’s a cognitive framework?’. To sate myself, I followed the link and landed on a page about “frame semantics”. 

Frame semantics is a theory in linguistics that - wait, no! No more! This is precisely what I’m talking about. See how quickly and frictionlessly I swung from link to link? How all I had to do was follow my curiosity? That’s how it goes in the informational wilds: let your guard down for even a moment, and the spirit of distraction reroutes you.

I’ve come to think of this spirit not as malevolent or as mischievous, but simply as self-propagating. It seeds distraction for distraction’s sake. Unless actively resisted, it is the natural endpoint for all who enter the wilds; the valley to which everything flows. And since there are few hostile entities around here (beyond the viruses that plague certain sites), it is this metaphysical force, distraction, that is one’s chief adversary. 

It takes a thousand forms. Some are obvious, like Instagram and Facebook. Others are more insidious - they masquerade as relevant to the task at hand, but in fact have nothing to do with it. In this category I’d place the countless news stories and controversies that make up the day’s online discourse; as well as things like job adverts and posts from my favourite blogs. These things feel ‘important’, and they might well be, but relative to the quest of the moment, they are often just well-dressed distractions.

Adding to this chaos is the presence of a second screen in the form of one’s phone. A second screen allows one to split their presence across two spaces - with the typical result being that one is not really rooted in either one. And in the informational wilds, such rootlessness renders one easy prey for distraction.

Part of what makes this so frustrating is that, in getting distracted, you have only yourself to blame.

(Well, also the many software engineers, content marketers, and other arcane professionals who have designed the internet to swallow you whole; who are like the sirens that lure sailors to sharp rocks in how they manipulate and seduce; but since there’s little I can do to change these designs, I find it more useful to frame the issue as an individualistic one.)

So what’s an individual to do? The same thing you’d do for any intrepid adventure, I think: pack appropriately. The right equipment is key in not getting lost. In my case, this includes: 

1. A notebook, whether digital or physical. I use this as a diary, a scrapbook, a quest log; really just to keep me on track. My current favourite is Roam Research. I find taking notes - even inchoate, fragmented ones - turns what is typically a passive experience into a more active one, which is good for retaining control in unfamiliar terrain. 

2. An audio accompaniment. For me, two categories of sound work particularly well: instrumental music, and natural noise. Both are effective in blinkering my attention, drowning out the external environment, and keeping me zoned in. 

   • In the former category, I include classical music, jazz, psychedelic rock, house/techno…really anything rhythmic. Sometimes, when in a rut, I just pair a random country/genre (e.g. ‘Cuban instrumental’), search Spotify, and sink in. Most anything works as long as there are no words (or at least none I can parse).

   • For natural noise, I find waves, birds, or waterfalls usually do the trick. Mynoise.net is a great resource for this. 

3. A timer. I use Forest, though I’m sure there are other good ones. Time can dilate online, so it’s useful to keep track of it, and to work in blocks. 

4. A "reader-mode" browser extension (like this one). This is a new addition to my toolkit, and it’s already proved indispensable. It centralizes and standardizes text, and provides a helpful summary of the length of any article. It’s supremely useful for filtering out the many pop-ups, sidebars, and other distractions that commonly accompany online articles.

5. A content filter, to keep viruses at bay. I use uBlock Origin.

6. A water bottle. Water: drink more of it! 

7. Various trinkets. Right now on my desk I have a slinky, a spinning top, a jade egg, and a kaleidoscope. Having stuff to fidget with helps me focus; your mileage may vary. 

A fancy note-taking app won’t automatically make you a better researcher any more than a fancy pair of shoes will automatically make you a better hiker. But paired with the right mindset, it can certainly help. 

Anyway, enough preamble - to work!

16:30

My best-laid plans have quietly agreed to disband. I found myself swept along as if on a river, passing article after article on longtermist thought. I made some notes, I found some answers, and tomorrow’s tasks are cut out for me. Now I am like a diver who’s all but exhausted their limited oxygen supply - I can’t stay in this environment much longer. 

I hang up my tools. Tomorrow we go again. 

I: A Common Case 

If you’ve spent any time in a South African city, you’ve probably interacted with a homeless person. Though the specifics vary, in general these interactions follow a standard pattern. Someone approaches you, asking for attention and for financial assistance, at which point there are at least three options: help them, ignore them, or communicate something to the effect of “...sorry, but no”. [efn_note] Other options exist - one could e.g. respond with anger or violence - but I don’t think such misanthropic behaviour is worth discussing here, as it runs contrary to the aims of this piece (discussed below). [/efn_note]

Which option ought one to choose? 

When I began considering this question, I thought there would be an easy answer. For example, it seemed clear that the second option - to pretend the person doesn’t exist in the hope that they go away - was inhumane and ought to be avoided. But I’ve come to appreciate that matters are not so simple. In a country as violent as South Africa, context is crucial. A vulnerable person walking home alone at night can hardly be faulted for failing to acknowledge a stranger. 

The other options become similarly complicated when considered in the context of the real world. At the root of this complexity is a tension between two needs: the need to acknowledge the dignity and common humanity of others, and the need to preserve oneself. It is for each individual to decide how to reconcile this tension - at least, I lack the necessary authority and wisdom to decisively resolve it myself. 

Accordingly, my aim with this piece is not to dictate how one should behave, but to explore the social dynamics at play in interactions between housed and homeless people.  

In particular, this piece aims to do three things. First, to show that homelessness in South Africa is not “normal”: it is neither a natural phenomenon nor an inevitable consequence of modernisation, but rather a product of the country’s idiosyncratic and violent history. Second, to show that how one responds to homelessness - and specifically, the sense in which one says they are sorry - reveals something about how one understands this history, one’s place in it, and the obligations one believes are owed to other people. Third, this piece aims to deepen the reader’s understanding of homelessness, in the hope that understanding breeds empathy, and that empathy imbues one’s actions with compassion. 

To avoid misconceptions, at the outset let me clarify two points: one regarding my choice to focus on homeless people, and the other regarding the language I use in doing so.

First, note that homelessness is a transient state: it is something that people can slip into and out of and which lasts different lengths of time for different people. It is not a fixed part of one’s identity (as is for example one’s ethnicity). It is also not a neat binary category, but instead exists on a spectrum: consider how both someone who sleeps in a shelter and someone who sleeps on the street could be called homeless, even though their experiences may materially differ. Furthermore, that someone asks for assistance in a public space does not necessarily imply that they are homeless.

Nevertheless, I focus on homeless people, as opposed to other indigent people that make ends meet through informal trading or begging on the streets, because it is in the interactions between homeless and housed people that the different roles that “sorry” plays can be seen most clearly. In this sense, such interactions serve as a paradigmatic case. That said, I think the analysis I give here for the most part applies equally to all people that subsist on the streets of South Africa. 

Second, let me acknowledge the potential discomfort of distinguishing between “the homeless” and “everyone else”. Some scholars take this distinction to imply a normative difference between these groups, and suggest that to use such language is to objectify people that are currently homeless. I disagree. While there is indeed a factual difference in the circumstances of these two groups, and homeless people are often subjected to social exclusion because of this difference, acknowledging its existence does not imply that it is somehow just. Objectification is a risk whenever one writes about a specific group, but by repeatedly emphasizing the agency, subjectivity, and diversity of homeless people, it is one I hope to avoid. 

My approach is as follows. In section II I provide a framework for understanding what “sorry” means, note three distinct senses in which it can be interpreted, and discuss the vital role the expression plays in moderating interactions with homeless people. Section III aims to shed light on both the causes of homelessness in South Africa and the lived experiences of homeless people. 

In Section IV, I discuss which of the three senses of sorry (if any) are most appropriate in interactions with homeless people, and consider the social and moral obligations entailed by each sense. Section V concludes with a call for compassion.

II: Sorry for Sorry’s Sake 

What drives one to say (or otherwise communicate) “sorry” to homeless people, on the occasions that one does? And what is it that one may be sorry for? 

On the first question, a combination of politeness and emotion-orientated accounts provide some explanation.

“Sorry” is often an expression of politeness. In this capacity it signals respect for and deference towards others. Think of the “sorry” one might murmur when bumping into a stranger in a queue or a crowded room. This is often how the word is used in interactions with homeless people - they may say sorry for approaching you, and you may say sorry for failing to help them. 

On its own, this explanation is of limited value, because it does not tell us what moves people to act politely in the first place. For this, we need to consider the various emotions that may underlie these situations - notably sadness, compassion, fear, and guilt. 

Sadness is a broad emotion that often co-occurs with more specific feelings. The subtext of a “sorry” motivated by sadness is something like “I’m sorry this happened to you”. This is the kind of “sorry” one might say when a friend’s dog dies, or when someone stubs their toe. These examples also illustrate compassion, which can be understood as sympathy or concern for the suffering of others. In many instances, it may be a mix of sadness and compassion that spurs one to say sorry to homeless people. 

But these are not the only motivating forces. “Sorry” may also flow from a place of guilt, if one feels that failure to provide assistance qualifies as an act of wrongdoing or amounts to a failure to meet some obligation. In this case, “sorry” functions as an apology - a request for forgiveness from the wronged person. Finally, consider the role of fear: feeling vulnerable while interacting with a homeless person may lead one to say “sorry” in the hope that this display of respect is sufficient to defuse the situation.

These emotions are highly contextual: seeing the same person in the same spot may evoke different feelings on different days. Furthermore, emotional responses can be dulled by repeated exposure to the suffering of others. This is commonly called desensitization, where what was initially heart-wrenching now barely registers a response.

These kinds of emotional explanations provide some insight into what motivates one to say “sorry”. But to fully understand the role sorry plays in these interactions, it is important to also consider what it is that one may be sorry for.

Here it is useful to distinguish three distinct ways the word “sorry” can be interpreted. First, it can mean “I am sad that you are suffering”, as in the ‘dead dog’ example above. This is sorry only as an expression of empathy and compassion. Second, it can mean something like “I was fully responsible for this situation. I acknowledge my wrongdoing, feel remorse that it happened, and will do my best to refrain from similar acts in future”. [efn_note] These four components - an acknowledgement of wrongdoing, a recognition of responsibility, an expression of remorse, and an intention to refrain from similar acts in the future - are often considered essential requirements for a “valid” apology - see  here. [/efn_note] This is “sorry” as an apology. 

Third, and perhaps counterintuitively, it can mean “I’m sad you’re suffering. The reasons for your suffering are tangentially related to my actions, but it’s not my fault. I would’ve prevented this if I could have.” A good example of this is the “sorry” a doctor may say to a patient’s family when, despite their best efforts, their patient dies. Each of these three interpretations admits a different degree of responsibility on the part of the speaker for whatever it is they’re sorry for.

This is interesting, as it’s often unclear in which sense one is saying sorry to homeless people,[efn_note]  This may be unclear not only to the person to whom sorry is said, but also to the person saying sorry. But that the speaker may not be conscious of their intention does not mean that no intention underlies their actions. [/efn_note] and different people use the word in different ways. Some may use it only to express sympathy, as they believe they bear no responsibility. Others may intend the term to convey a genuine apology, as they view their failure to help as wrongful. And third, some may (for various reasons) believe that they are at least tangentially responsible for the predicament in which homeless people find themselves, and apologize accordingly.

Which of these senses, if any, is most appropriate? This depends on how one understands the moral order and the extent to which one thinks they bear personal responsibility for the suffering of homeless people, which I consider in section IV below. Before that, some context. 

III: Who are the Homeless?

My repeated use of the phrase “the homeless” may give the impression that they comprise a single monolithic group. In fact, the homeless community in South Africa - and across the world - is highly diverse. As noted at the outset, there is diversity in terms of the reasons that people are homeless, the length of time they have spent on the streets, and the specific challenges each person faces. Accordingly, although I try in the following subsections to paint an accurate picture, be mindful that this analysis is somewhat reductive (especially given the dearth of reliable data on the topic). 

It’s estimated that there are “roughly 200 000” homeless people in South Africa. This equates to about 346 homeless persons per 100 000 people in the country. Contrast this with the United Kingdom’s rate of 8/100 000, Chile’s rate of 155/100 000, and India’s rate of 131/100 000, and it becomes clear that, even for developing countries, this is an astonishingly high number.  

However, the value of international comparisons is limited by the fact that there are conflicting definitions of homelessness. Consider e.g. the United Nations’ definition, which encompasses both people who have no shelter and those whose shelter is “inadequate”. This may suffice for developed countries, where informal housing is “efficiently prohibited” and so barely exists, but in countries (like South Africa) that have substantial informal settlements, it would be a mistake to lump these groups together.

Many more people live in informal settlements than are homeless, and homeless people are often disconnected from their families and communities of origin in a way that those living in informal settlements are not. Thus the most suitable policy options for each group may be different, and indeed in certain cases may be “diametrically opposed”. Because of these nuances, it’s preferable to restrict the definition of homelessness to include only those that do not have a roof over their heads.

It’s difficult to get an accurate picture of the demographics of homeless people. Census data, one of the most common sources of demographic information, is typically collected on the basis of domicile, which is of little use when trying to count the number of “undomiciled” people. The information that is available comes primarily from survey data, which we should interpret with caution. 

From this data, it emerges that the homeless population is composed primarily of “older children and working-age adults” and is “predominantly male”. It also emerges that the child and adult populations are largely separate; as homeless children are more likely to have been born in urban areas, whereas adults are more likely to have immigrated from rural areas. The adult homeless population is also significantly larger than the child population.

Thus it appears not to be the case that homeless children grow up to be homeless adults; though it’s unclear where these children go upon leaving the streets. Finally, it’s worth noting the presence of other African nationals in these populations, although it’s unclear in what proportion they are represented.

Before discussing some of the practical hardships entailed by homelessness, let us consider its causes. 

i) Causes of Homelessness

There are, broadly, three types of explanations for homelessness, sometimes referred to as “sin talk”, “sick talk”, and “system talk”. These seek to explain homelessness by reference to moral culpability, pathological incapacity, and structural inequality, respectively. I weigh the merits of each below. 

Sin talk casts homelessness as a product of behaviours deemed sinful or immoral, like idleness, drinking, and substance use. This was one of the first ways in which homelessness was conceived, and as such is inextricable from its historical context. In South Africa, its roots trace back over two hundred years, to a series of colonial policies that criminalized homelessness (or “vagrancy”, as it was then called).

These laws imposed significantly harsher penalties on “vagrants” that were not white - an 1895 law of the Orange Free State colony, for example, provided that as punishment for vagrancy “non-whites could be contracted as a servant of a white person for up to a year”. Similarly racially skewed policies existed in the Cape Colony, Natal, and the Transvaal. Thus the perception of homeless people (and particularly homeless people of colour) as being “idle”, “disorderly”, and a threat to (colonial) public safety was already well-established by 1910, when the four colonies merged to form the Union of South Africa.

Vagrancy laws were interlinked with the notorious pass laws insofar as they shared a common aim: to prevent non-white people from accessing most areas for any reason other than to provide labour to the white minority. This was achieved through acts such as the Native (Urban Areas) Act 21 of 1923, the Native (Urban Areas) Consolidation Act 25 of 1945, and the Native Laws Amendment Act 54 of 1952. These and related acts aimed to constrict and control the lives of poor people, and to this end they succeeded.

While Apartheid laws are no longer in force, homelessness remains criminalized in many parts of the country by municipal by-laws that prevent begging, “loitering”, or sleeping in public spaces; although the enforcement of these laws varies by region. These by-laws effectively license police harassment and abuse, of which homeless people are frequently a target. On average, they are arrested “once every three years” - a rate eleven times higher than for the general population. The penalties for contravention of these by-laws are typically fines (unpayable for most homeless people) or imprisonment.

Once we understand that the association between homelessness and immorality is largely a by-product of prejudicial policies, its explanatory value diminishes. Of course, this is not to say that behaviours like drinking or substance use do not play a role in a person becoming homeless - often they do. But these behaviours should be understood in the context of the impoverished circumstances in which they arise. On this point, the sociological concept of “edgework” is useful: here it refers to the idea that, in conditions of material insecurity, people choose to engage in risky behaviours (such as drug usage) as a means of exerting a degree of control over their lives.

In sum, sin talk seeks to explain homelessness by assuming fault on the part of individuals. This is too moralistic, insofar as it judges behaviours like drinking to be inherently sinful; and it is too reductive, insofar as it fails to account for the historical conditions that gave rise to these behaviours. As an explanation, it is valuable only to the extent that, unlike the two explanations discussed below, it acknowledges the agency of the homeless. 

Sick talk departs from the idea that homeless people often have various physical and mental health problems, as well as other “vulnerabilities”. Unlike above, the idea here is not that people deserve to be homeless because of their perceived moral shortcomings, but rather that people become homeless because of these vulnerabilities. This is true enough, but fails to account for the degree to which the colonial policies and practices discussed above, as well as other unignorable structural forces, have given rise to these vulnerabilities in the first place.

For example, the “dop” system, which for over two hundred years paid farm workers in the Western Cape with alcohol as well as with money, is widely understood to have exacerbated alcoholism and caused widespread social harm to the Cape Coloured community. In these areas alcoholism is not just an individual pathology; it is also the consequence of malign social policies.

A further problem with explanations grounded in “sick talk” is that they risk reducing people to their disabilities. They may conclude, for example, that the best solution for a homeless person suffering from severe depression is to commit them to an institution, when they may be better served simply by receiving unconditional access to housing. This logic fails to appreciate the agency that most homeless people possess. Agency may be limited by material insecurities and vulnerabilities, but it is not absent. 

Finally, let’s talk system talk, which captures the idea that homelessness is the result of structural, not personal failures. 

We have already seen how colonial policies constructed the idea of homelessness at the same time that they criminalized it. For the colonial and apartheid governments, the control and segregation of space was central to the oppression of people of colour, and to securing the country’s resources for the white minority. When some of this institutional infrastructure was removed via the Abolition of Influx Control Act 68 of 1986, and migration into urban areas finally became possible, the number of homeless people swelled.

The dismantling of the apartheid laws did not (indeed, could not) undo their consequences. Apartheid “geographically locked in racial inequalities”, and these inequalities have only deepened since the dawn of democracy in 1994. The reasons for this are numerous, complex, and need not be fully canvassed here. Accordingly, I restrict my focus to three factors and discuss them only insofar as relevant to homelessness: the turn to neoliberalism, the lack of a coherent national homelessness policy, and the effects of economic stagnation.

Broadly, neoliberalism refers to a mode of governance characterized by “orthodox fiscal policy, open markets, international trade liberalisation, privatisation, and a pro-investment environment”. The term is difficult to pin down: it means different things in different contexts; and is often used pejoratively to refer to globalisation and the influence of international institutions such as the World Bank. While neoliberalism is not homogenous - it takes different forms at “different geographical scales” - most commentators agree that since the adoption of the Growth, Employment and Redistribution (GEAR) policy in 1996, South Africa has largely operated under a neoliberal paradigm. [efn_note] Although GEAR was replaced in 2005 by a new macroeconomic strategy, the neoliberal framework it facilitated persists today. For a detailed discussion of the various macroeconomic policy shifts South Africa has undergone since 1994, see here. [/efn_note]

Perhaps the most relevant consequence of this for our purposes is that local governments are now “encouraged to be entrepreneurial and [to] improve local revenues by private sector participation and using market-based strategies”. This has led to many urban and wealthy suburban areas voting to become “city improvement districts” (CIDs). CIDS are managed by “non-profit, private partnerships between local governments and businesses”, which allow local governments to contract out certain services (like refuse management and traffic control) to private companies. Once more than 50% of an area’s property owners vote to become a CID, all owners must participate by paying a certain fee, collected by the municipality.

Whether or not these CIDs have on balance been beneficial for South Africa is a contentious question. It is clear, however, that in prioritizing the marketability and investment potential of the areas that they govern, they have been harmful for the homeless. While homeless people have an incentive to stay in economically prosperous areas - and indeed sometimes choose to do so rather than staying in informal settlements, because this is where they stand to make the most money (whether from relying on the donations of other citizens, or through “small-scale” jobs like car-guarding and gardening); municipalities have a countervailing incentive to exclude them from these areas, because they are bad for attracting investment.

Thus the anti-homeless by-laws mentioned above are often enforced as a pretext for “sanitizing” an area, so that it appears more palatable to potential investors. By catering primarily to the interests of businesses and wealthy citizens, and by pursuing “integration into the global economy” at the cost of national integration and social cohesion, the logic of neoliberalism “has continued to a certain degree the apartheid legacy of spatial inequalities". In many cases, property rights have been valued more than poor people. 

Another structural factor worth noting is the “paucity of national guidelines and policies on homelessness”. Government responses have primarily come from the housing and social welfare sectors, and have generally focused on those with inadequate housing - they have not specifically addressed homeless people. While some provinces (such as Gauteng) have adopted more targeted policies, these have been disjointed and largely ineffective.

Homeless shelters, usually driven by faith-based non-governmental organizations, have tried to fill these gaps, but these shelters often impose restrictions that limit their usefulness to homeless people. For example, many shelters impose curfews or require abstinence from drugs, alcohol, and sometimes even sex before marriage, in line with their internal value systems. This renders them practically inaccessible for those with substance use problems or in relationships.

Finally, consider the links between national economic stagnation, unemployment, and homelessness. As unemployment and poverty have continued to rise in the last decade, so too has the number of homeless people. It is reasonable to infer that the worsening state of the economy - and the lack of job opportunities available - has played a role in this increase. 

While each type of explanation is in itself deficient, cumulatively they provide some insight into what causes homelessness. As Somerville notes, “‘structural’ factors create the conditions within which homelessness occurs, and then, ‘individual’ factors determine the likelihood of becoming homeless in those conditions”. Although it is difficult to apportion responsibility between structural and individual factors, it is clear that in South Africa structural factors have played an outsized role. This explains why the country has much higher rates of homelessness than not only developed countries with strong social welfare systems, but also many other developing countries.

Explanations grounded in sin talk, while mostly without merit, are useful insofar as they recognize that some people may end up homeless as a result of their own actions. Sick talk is useful for recognizing that many people have specific vulnerabilities - often caused by historic injustices - that increase their risk of becoming homeless. 

Ultimately, however, it should be kept in mind that the pathways that lead one to homelessness, while often involving common elements - like financial difficulties, mental health and substance use issues, and violent or otherwise fractured family situations - vary immensely between people.

ii) The Experience of Homelessness

Homeless people are “among the most destitute members of society”. They are deprived not just of a home, but also of the safety, privacy, and sense of rootedness that a home provides. For this reason, homelessness is often considered a “multidimensional phenomenon”.

Being homeless entails many practical challenges. One must carry all one’s possessions wherever one goes, under constant threat that they may be taken either by the police or by other homeless people. Finding food, warmth, and a place to sleep is a daily struggle. When food cannot be bought, it is often acquired from community sources (such as food banks and soup kitchens), from dustbins, or from other homeless people. 

Relatively few people rely on shelters, because of the difficulties noted above and because of their scarcity. Most sleep on the streets, in parks, in building sites and abandoned facilities; wherever they can. There are, of course, no beds in these places, so people make do with what materials are available: flattened cardboard boxes, black bin bags, old sleeping bags, and more. 

Without a permanent home, and when so much of one’s time and energy is expended on the daily demands of survival, it is exceedingly difficult to accumulate wealth and plan for the future. This difficulty is exacerbated by the palpable prejudice that homeless people face.

They are often viewed by mainstream society as dirty, dangerous, and undesirable; at least partially due to the centuries of legislation linking homelessness to criminality. And it seems, from surveys and interviews, that they are well-aware of this perception. For example, one survey found that many homeless people were unwilling to approach government offices to receive social grants for which they were eligible, because they perceived their “unkempt appearance and irregular situation as unsuitable and perhaps likely to expose them to rejection". The same logic also renders endeavours like going for job interviews difficult. 

This persistent stigma is corrosive to one’s sense of self; to one’s conception of their place in the world. This perhaps explains why some seem to reject the “homeless” label and the negative connotations that come with it. [efn_note] The survey data analyzed in Kok at 33 found that 20% of those surveyed identified themselves as “temporarily stranded”, and another 45% stated their home was “too far away to commute”, although their actual situations may well have been practically identical to the 32% of respondents who unambiguously stated they were homeless. [/efn_note]

Homeless people may be alienated from mainstream communities, but they are not “entirely isolated in the sense that is widely believed”. Many have “strong links with others on the streets” in the form of friendships, relationships, and, in some cases, quasi-family structures. A sense of community is felt by many. Leaving this community behind, when one is otherwise alone in the world, can be difficult; and often presents an additional challenge for people trying to transition out of homelessness.

Taken together, these reasons clarify why homelessness is sometimes called a “slow moving tragedy”.

IV: Questioning Obligations

Let us return now to the question posed at the end of section II: which of the three senses of sorry, if any, is most appropriate in interactions with homeless people?

In my view, the first sense - sorry as an acknowledgement of suffering and an expression of empathy - is merited in all cases where one can communicate this without fear for their own safety. This can be done with words or gestures, and ideally also with eye contact. Such acknowledgement can serve to combat the persistent stigma and sense of exclusion that homeless people face. At least, it enables an individual to not themselves perpetuate this exclusion, and for this reason amounts to an act of basic human decency. 

This is sufficient to conclude that one should as far as possible communicate that they are sorry in the first sense. And from the perspective of a homeless person, this may be all that matters; because while empathy is implied by all three senses, the degree to which the person saying sorry feels responsible for the homeless person’s suffering is usually unclear unless it is explicitly stated. 

That said, it is still worth discussing in which situations the other two senses of sorry may be appropriate - not because this would benefit homeless people, but because of what such analysis can reveal about how different people understand their connections to the homeless. 

The second sense of sorry - sorry as an acknowledgement of full responsibility - would only apply if the suffering of homeless people was a direct result of one’s actions, or arose from a failure to fulfil some obligation. Because homelessness is caused by a combination of structural and individual factors, there is little basis for attributing direct responsibility to any one individual due to their actions. It is however possible that there exists an obligation to assist someone in need. On this point, it is useful to distinguish social and moral obligations.

I call a social obligation one which arises from "the largely unwritten system of social mores and conventions through which [a] society is kept together as a coherent whole". In contrast, a moral obligation is one which arises from philosophical reasoning. These different sources often produce the same obligations - for example, an obligation to refrain from violence is found both across different cultures and in different ethical schools of thought - and many people collapse this distinction altogether. However, there are also cases where a moral obligation may exist in the absence of a social obligation. For example, one may feel a moral obligation to abstain from eating meat even if no such convention exists in one’s society. 

In South Africa, there does not seem to be a social obligation to assist homeless people. While little empirical research has been done on this point, my own experience is that no social sanctions attach to a failure to help someone on the street - failing to help is the normal state of affairs. Indeed, while I have made a normative argument above for a social obligation to say sorry in the first sense, it is not at all clear that any such obligation is currently part of the country’s social conventions. 

Various arguments have however been made for a moral obligation to assist those in need. The substance of these arguments need not be discussed in any detail here. For our purposes, it is sufficient to note that these can be grounded in various philosophical frameworks (perhaps the most famous of which derives from utilitarian ethicist Peter Singer’s essay Famine, Affluence, and Morality); and that if an individual considers themselves bound by these moral obligations, then it would be appropriate to say sorry in the second sense when they cannot fulfil them. 

Arguments for direct responsibility are typically ‘forward-facing’, by which I mean they claim that obligations arise if one is in a position to help someone else without thereby jeopardizing their own livelihood, and without regard for how one acquired their wealth. In South Africa, such obligations would therefore apply to the roughly 24% of the population that can be classed as “stably middle class or elite”. [efn_note] This paper proposes five classes, defined by reference to the risk of one’s vulnerability to poverty. The five classes proposed are the chronic poor, the transient poor, those that are not poor but remain vulnerable, the stable middle class, and the elites. If you want to know more about class in South Africa, it’s worth checking out. [/efn_note]

Finally, consider the third sense of sorry - sorry as an acknowledgement of tangential responsibility, but not as an admission of fault. Depending on how one understands the role of morality in their own life, the same ‘forward-facing’ moral arguments hinted at above could also apply here with diminished strength. Furthermore, because what is required is only a tangential connection to the suffering of homeless people, this sense raises thorny questions around the relevance of one’s identity. 

Unlike the ‘forward-facing’ conception above, here a ‘backward-facing’ approach is more appropriate, because to determine tangential connections one must account for history. Homelessness in South Africa is partially caused by historic structural injustices, injustices which explicitly benefited white people at the expense of people of colour (and in particular, black people), and which have resulted in race continuing to be one of the strongest predictors of wealth in the country. Thus one could argue that these disparities should be accounted for when considering which sense of sorry is most appropriate: for example by suggesting that white people ought to be especially sorry insofar as they are tangentially responsible for the suffering of homeless people of colour.

Personally, I do not find such arguments convincing, because while it may be true that causal connections exist, by themselves they are too remote to infer the existence of any obligations; and furthermore because the value of an apology in this context lies in its expression of empathy, not its admission of responsibility. But I think they are worth mentioning, because they contextualize identity-based ideas that one sometimes encounters, such as that white homeless people are more deserving of their station than homeless people of colour due to history being slanted in their favour.[efn_note] I reject this idea for the same reasons I reject the logic of sin talk, discussed above. [/efn_note] 

If you feel guilty when interacting with a homeless person, it’s worth questioning what the source of this guilt is. If you feel you bear some responsibility because you have failed to satisfy an obligation, question whether you think this is a social or moral obligation, and what you think this obligation entails. Such critical reflection can help to align one’s actions with one’s beliefs. 

V: Cause for Compassion 

Homelessness defies easy explanation. It arises at the intersection of structural failures, personal choices, and medical vulnerabilities. Even so, it’s clear that in South Africa, the widespread homelessness seen today is largely a product of the country’s unjust history.

South Africa’s homeless communities are heterogeneous, and, given the diversity of people that populate different provinces, operate differently in different regions. Most of us only glimpse these communities from the outside - we do not see the human connections that exist between people; the relationships, tensions, and struggles that comprise everyday life. 

Indeed, homeless people are seldom paid attention to. Often they are treated as either inconvenient, threatening, or invisible. Anecdotally, I have found that many people feel a real discomfort around the homeless not necessarily because they fear for their safety, but because such situations force them to confront the realities of inequality and the suffering of others.

One way people mitigate this discomfort is by communicating that they are sorry. As we have seen, “sorry” is a loaded term. The sense in which it is used, if reflected upon, can shed light on how one understands the nature and extent of one’s obligations towards others. My aim in this piece has not been to argue that there is a certain degree of responsibility that one ought to feel, but to provoke this reflection in you, the reader. 

If this piece is to have any practical value for people currently experiencing homelessness, I think it will come from the following claim: if you do not fear for your safety, but you are nevertheless unwilling or unable to assist when someone approaches you in a public space; you ought to look them in the eyes and communicate that you are sorry. This is an inadequate response, insofar as it does little to mitigate the suffering of the person asking for help, but it is nevertheless necessary if one hopes to push back against the widespread social exclusion that homeless people face.

It is difficult to remain compassionate when there is so much suffering in South Africa. Even so, it is important to try. 

Original text: “Sorry” Doesn’t Scratch the Surface_ Considering Connections to the Homeless (Original)