michael-dean-k/

Every online creator these days has a proprietary framework—they are catchy, dead-simple, sometimes flimsy, and often wrong. Somewhere exists a marketplace of mental models. Maybe Farnam Street. Is a mental model a framework? What even is a framework, and what makes a good one?

Consider the "frame-" prefix. A frame is a vantage point, a particular angle, a way of looking at something. Imagine holding out a picture frame that gives you a little glimpse of a vast national park. You can't hold an entire hyperobject in your mind, and so a frame compresses noise and gives you a shortcut for understanding. You can't see the whole park, but you can see a map. A framework is a system of lenses to decipher complexity. It lets you leapfrog in understanding.

By creator-economy logic, a framework is a valuable asset. Unlike a vague mush of regurgitated ideas, a framework is discernible, seemingly original, and something you can call your own—you can coin it as a phrase and repeat it over and over until you're known by it. This is a useful tactic! The evolutionary pressures of Internet feeds have forced conceptual ideas to become salient, simple, and tangible. Unlike the abstract frameworks of a physicist, these are fit for a TED talk, primed to leak into the memeplex and anchor your name in the annals of Internet-niche-celebrity history. This is, overall, good—it democratizes conceptual knowledge; the problem is when, beyond the positioning, it's intellectually flaccid.

You've probably seen frameworks pitched as "The Six Types of X," or "The Y Method." These are pseudo-frameworks; they are more like sets or lists. They assemble things within a domain, and increase their memorability through alliteration and metaphor. Maybe this helps with recall, but it doesn't reveal the inner structure of a domain. A real framework would give you a schema to classify incoming data in an unfamiliar field; but a set/list only refers to itself ... Compare this to something like "POP Writing," a simple concept that classifies writing as personal, observational, or playful; this acts a lens to help you analyze any piece of writing.

For something to be a framework, it need generalizability: the ability to explain things beyond itself. The earlier examples are merely "collections"—things are presented, but we don't understand what Aristotle calls "the four causes": what it's made of, how it's arranged and interconnected, why it exists, and what's it's purpose? When Plato writes about conceptual forms, he says there are collections and divisions. Making frameworks are a matter of dividing; a framework's goal is to divide an incomprehensible whole into human-sized parts.

I spent some time this morning thinking through a "framework of frameworks." The first thing to grok about this meta-architecture is that it's broken into three levels of escalating sophistication. After that I'll explain how levels have shapes, and how any shape can be evaluated through a set of qualities.

So it's level, shapes, and qualities—that's the gist.

To expand on levels: a framework can be a heuristic, a model, or a paradigm. Now that I think about, I suppose HMP (not quite meme-ready) is an attempt at making a "paradigm for frameworks," which will soon make more sense, and sound less like bashed buzzwords.

The three levels of frameworks:

Heuristics—mental-shortcuts that lets you decipher complexity through a single property. When Aristotle was classifying animals, he'd offer differentiate them across single properties: blood or bloodless, horned or hornless. A heuristic is something you derive a posteriori, from experience, from empirical facts. You might observe a bunch of essays, or companies, or girlfriends, and notice that you can compare and contrast items in a set based on single attributes. There are many shapes a heuristic can take: it can be a classifier (type A, B, or C), a boolean (on/off), a spectrum (A<>B), a rubric (I, II, III, IV, V), a score (1-100).

The benefit of a heuristic is that, since it's often anchored on a single property, it's simple and easy to learn. The downside is that it's a very limited frame on the scope and purpose of your domain. For example, if you're trying teach craft (purpose) for online nonfiction writers (scope), then POP writing is an excellent heuristic to classify writing voice, but it doesn't teach you how to scope ideas or structure an essay. A heuristic is a shortcut to understand one facet of a domain, but it's not an attempt to model the whole domain. 

Models—a series of interlocking properties, all organized with a larger frame. Write of Passage was set up like a cabinet of heuristics (POP writing, Shiny Dime, Write from Conversation, etc.), but they weren't organized within a single higher-order framework. Think of a model as a group of heuristics organized in hierarchy, where each one is named (what is it's essence?), labeled (what other concepts is it similar to?), located (what is it's parent concept?), interlinked (who are it's cousins?), and excavated (if we go infinitely deep, what are the variations?). Models can take many shapes: an XY graph, a taxonomy, a multi-step recursive workflow (like an OODA loop).

Models cover far more than any heuristic can, but it risks collapsing into complexity. Conceptual schemas can bloom, sprawl, and contradict—becoming illegible or incoherent. To solve this, you have to shift from empiricism to rationality; you can notice infinite ways to subdivide a thing, but to make an elegant model, you need to shift to abstraction and question the fundamental truths behind it. Why might one set of properties be more elegant than another? Ultimately you have to make a hypothesis, test it across dozens of examples, and see if it holds. If some foreign object can't be reconciled into the schema, you have an issue. Consider how the taxonomy of Carl Linnaeus was able to classify the animal world into kingdom, class, order, genus, and species, but it wasn't unable to find a home for the platypus. Even though it works across thousands of animals, the original assumptions aren't able to handle edge-cases.

Paradigms—comprehensive systems that can explain or predict any phenomenon within a domain. How is this possible? I think you achieve this when (1) you find enough instances to represent the domain; (2) you analyze them deeply enough to understand their underlying nature; and (3) you can arrange all your findings into a coherent architecture with maximum explainability. Consider how Darwin improved over Linnaeus: he shifted us from a static taxonomy to a dynamic model, a new paradigm that explains the origin of species, and even the potential future of a species. Paradigms can also take on many shapes, from living systems, to digital networks, to compositional pattern languages.

A good paradigm often finds simple root causes underneath complexity. For example, even though Essay Architecture has 27 patterns (3 patterns within the 3 elements of 3 dimensions), every tier is explained through the rhetorical appeals of ethos, logos, pathos—the whole thing can be summarized as "Aristotle all the way down." Maximum compression has a degree of a recursion: a single heuristic works as a fractal and organizes complexity at multiple scales.

Regardless of how you simplify and package a paradigm, it's still as hard to learn as a language; it's the least accessible, but it's the level that enables mastery of a domain, and the level most suitable to build a school around.¹

No matter how confident you are around a paradigm, it's always fragile. It's explainability is limited by the data set you test it against. As society develops new tools, it brings new data that eventually breaks a paradigm, even if it's held true for centuries. The telescope cracked the geocentric paradigm. A shattered paradigm is disorienting, but fundamentally good—it means you've discovered some new heuristic of reality, and humanity has to restart from new axioms.

So that's a first attempt to map the three levels, and it might help to think of them as geometric dimensions. A heuristic is like a one-dimensional line, a ray of insight that you perceive in a field. A model is a two-dimensional plane to understand a multi-property condition within a domain. But a paradigm is the three-dimensional totality of a field. A paradigm contains all possible heuristics and models, and so when you can think in 3D, you're able to match any circumstance with the correct solution.

But a paradigm isn't necessarily a great framework! It might be the most successful in compressing reality into a particular shape, but it might fail along other qualities. (To reiterate, levels are about  "compression sophistication," where shape is the particular form of compression...) Qualities capture how a compression framework interacts with society as a whole (the point of a framework is, in the end, to spread understanding).

Regardless of the level or shape, some qualities to consider:

• Tangibility: have you made it graspable through metaphor and lexicality?
• Salience: is it rewarding or effective to understand?
• Transferability: does it explain things in other domains?
• Correctness: is it actually right?
• Approachability: how would you walk someone through it via curriculum?

Are these the right qualities? Are these all the qualities? Can I reduce these into more fundamental qualities? I don't know. Realistically, this isn't a "paradigm of framework" yet; it's more so a "model of frameworks." I only conceived of this today, and so naturally it's sprawled into a series of properties:

A framework compresses complexity in a domain, exists at a level of explainability, taking on a particular shape to achieve a purpose within a scope, and can be judged along a set of qualities on how well it's memed.

To properly turn this into a paradigm, I'd have to test this hypothesis against hundreds of frameworks—ie: creator economy memes, among other disciplines—to see if it holds up. I imagine the concepts would melt and reform several times until it's simpler, sharper, and better at explaining a wide range of frameworks. A paradigm isn't solid until it's a domain has run through its pipes. A paradigm is really just a reflection of a domain's dataset and the mind making sense of it.

1. As a writer needs a site of one's own, a place designed for their particular psychology so they can be the most prolific, honest, adventurous version of themselves. Solitude is important. Montaigne, the founder of the essay, wrote up in his tower for a decade. When you are your own audience, your practice is self-justifying. You are intrinsically fulfilled, and do it regardless of validation, growth, or revenue. To become self-validating is to become a nuclear engine of creativity. When you write on your own site, chances are you will spend much time on ideas that no one will read. That's inefficient, but essays are inefficient. I need to ruthlessly follow what matters to me, with no fear of being illegible or invisible to others.
2. That said, you can and should invite others into your garage. I still plan to keep my Substack and post there, but it will be more like a newsletter, a digest of the controlled explosions happening in my own neighborhood. I do believe in the value, even, the responsibility, of writing in public. By making your place tidy enough for friends, strangers, students, customers, mentors, heroes, whoever, to come in, it makes you put an extra oomph in your crafting of language, and it creates seeds from which relationships can grow. When you have visitors in your territory, they play by your own rules, so you're generally safe from mobs and barbarians. They will not trample over your furniture and throw the books from your shelf like they would in the town square.
3. In 2023, I was hopeful that Substack could be an all-in-one platform—a website/newsletter/discovery engine—but it has heavily shifted into an app-centric social media platform. The app starts you off in an endless, algorithmic, engagement-ranked feed, and the design has slowly evolved to trap you in the app. When you click in article, the default URL is the Substack one. When you click into a person, you can't even go to their site anymore; you can only view them through their profile template. This means all the customization and self-archiving and mythologizing that goes into your site is off limits to the app, the thing they're trying to corrall attention through.
4. I'm very allergic to "Substack is dead" posts, and most of the critique is often a projection of their own weaknesses (ie: when people naturally stop growing, it's easy to blame the algorithm than to take responsibility for it). In my case, the Substack algorithm has worked pretty well over the last two years, and I'd be stupid to abandon it. But the whole system has limited my intrinsic passion to write, and that matters more, enough that I'm willing to take the risk with a split publishing system: Substack newsletters that link out to my site.
5. Back in 2020, a Write of Passage concept was the "public to private bridge," meaning you find people in social media feeds, but then redirect them back to a place of your own. Now that Substack is mostly a social media network, I think that same strategy applies. It's not where I want to host my essays anymore. I'll host newsletters and paid content there, but the timeless stuff wants to live in a place that is timeless.
6. If you have a portfolio of writing, Substack feels like a pretty bad way to make your old work legible, especially within the app. There are no tags or sections. No about page. No navigation system. No ability to frame call to actions as visually dominant. The beauty of your website, especially if you build it from local markdown files, is (1) you have a single source of truth for all your writing, and (2) you can just respawn new interface and navigation systems as your portfolio evolves.
7. Even though Substack lets you export your essays, it's really more like a failsafe, a backup incase you ever decide to leave. Of course, it's very nice to have that! But it's still a cloud-based system, where editing your past files is slow and clunky, and you have limited ability to point your essays elsewhere. When I have a local folder of my essays, I can very easily spin up a stylized website around any essay, or any group of essays (say, for example, if I'm making a proposal and I want to curate a handful of essays. I can point my AI at any combination of files and folders for context.

My personal website is “out of the bag.” Meaning, it’s not a private thing shared among 3-5 friends anymore; I excitedly shared it with Essay Club yesterday (60 people or so). I am leaking it prematurely because of the giddy hope, that personal websites are the new paradigm for writers, an escape from the enshittified commons. But I have to admit that I haven’t thought through two important questions yet, so here it goes:

1) Does this kill discovery?

If I were to instead publish all my ideas in real-time on Substack notes, would my audience grow more? Probably. The reality is we all self-censor ourselves in public feeds, in a thousand different ways, so it’s not like all of this could naturally emerge in feed. I tried this in January. I killed my logging practice with the goal of trying to just do it all on Notes. For two weeks, I was able to post spontaneously, but I find that if you ever stop momentum, it’s very hard to get back out of your head and into that groove. Overall, I just wrote less. I wonder if there’s truth to the idea that all writing practices grow/incubate/evolve better in semi-public spaces. It’s not that you should ignore the occasional blast. It’s that there’s a natural progression of nurturing ideas.

Another angle is, “I’m not interested in audience growth,” which is true because it’s not motivating for me, but I am in several ways entangled by growth, meaning, a complete lack of growth could threaten the sustainability of my writing. And so a middle ground is to incubate on my website and then selectively drip ideas through notes and newsletters. I could do a weekly or bi-weekly digest, Austin Kleon-style (“10 logs from last week” + essay visualization + updates, etc.). Not as sure how I would do it on Notes. Daily? Sporadically? Something else? Either way, this brings back the whole "public-to-private bridge" concept from Write of Passage. I think some people abandoned websites and just accepted the feeds. I know in 2023 I shifted entirely to Substack thinking it could be my entire digital home, but now it feels like rented land.

So my website gets maybe an A- in unlocking my writing practice, but only a C in growth, but maybe it’s a B in conversion? As in, if someone spends a lot of time on my site (and people have told me they’ve spent hours in my logs), they’re more likely to trust me—due to the sprawling, unoptimized, honest nature of things—and more likely to get a paid subscription or join Essay Club? Unexpectedly, personal writing could be a more honest and more effective form of “marketing” than strategic value-focused content (“Are you in hell? Well I’ve got the thing for you…”).

2) Is there risk in having all my ideas public?

Now that I’m in my own place, relatively unchained, saying what I want, and reading and writing about political science a bit more (I have a draft comparing Karp’s Technorepublic to Leviathan by Hobbes), I’m a bit paranoid to share ideas so openly. It’s hard to imagine facing any real-life consequences for the words I write; I’m just a nobody! It feels hubristic to think that I’d be considered a threat to the state for my thinking, but maybe these thoughts are natural, considering we’re being pleaded to accept an AI-powered surveillance state in exchange for security. (It's not that I think any of my writing is particularly rogue, but let's say I start thinking through a scheme to organize a million swing state voters to rally around a single-issue voting boycott in order to pass a bill on election campaign reform, you can see how democratic ideas might seem threatening to a state.)

It’s effortless for a state agency to scrape the Internet, build psychographic profiles on its citizens, and give them a “loyalty score.” Let’s imagine they also have an “influence score” too, determining how much sway you have over your citizens. If you have medium levels of loyalty and influence, you’re probably not being actively monitored; but if you have extremely low loyalty (L=5/100), it’s a threat even if you’re low influence (I=0) because you might be a terrorist; but also if you have extremely high influence (I=95), and even slight disloyalty (L=45), then that’s a risk too. And if it’s not the state absorbing my context, it could be independent actors scraping my site to clone me and do what they will…

I guess the point is that AI creates such a leverage over information, that you’re own personal data becomes extremely valuable. It can be leveraged not just by you, but anyone who has it. A personal website of an unfiltered nature is a higher-resolution signal than a social media profile where most interactions are shallow.

Grasping at a solution_

If all these concerns are justified (and maybe they’re not), then what are the practical methods of maintaining privacy? I’ve already written ideas about security gates and embedding-based encryption, and that’s all technologically neat, but it creates friction for the readers! Maybe that’s okay? But then this ignores the “entangled with growth” constraint from above…

And so maybe the Third and only way through is to make the encryption solution that is both an alluring and enjoyable UX for the reader.

This starts by understanding how websites get scraped, building solutions to avoid it, and then shaping them to be reader-first. You can only really do this by scraping yourself. I’ve scraped full portfolios from Substack in two different ways, and even a decade’s worth of Marginal Revolution posts. At a minimum this means avoiding RSS and HTML, which this (current) site already violates (ie: it’s ideally on a server and requires permissions to load).

Scrapers can prevent against automated gathering; but not against a person or agency that has already found your site and is willing to sit through slower and manual methods to extract information. A defense here would require gating and admin approval, another hinderance. There is something here about taking monetization dynamics (paywalls) but reinventing them for privacy’s sake. Maybe the way around this is to only encrypt a portion of the content, say 50%, with cryptic previews of what lies beyond (either through titles or redactions or chaos).

To try to synthesize this all together, what if a website were a video game?

Website as gamified maze?

As smart as today’s AI’s are, they still can’t beat Pokemon. They can transform text and code better than the world’s best engineers, but if you ask them to navigate an environment where vision and long-term memory are required, they bomb. Pokemon has very simple inputs too: 4 navigational directions and then a Click/Cancel boolean. If you were to make it more challenging, with inputs that required hand-eye coordination, that could solve two problems: it scrambles existing scrapers, and creates a novel UX.

I also sense there’s something to turning a website into a literal maze, not just an overwhelming sprawl of hyperlinks, but an actual video game you have to navigate through (it would be neat if somehow notes were semantically distributed across a map so there are “towns” of ideas). Can friction be made gamified, exploratory, enjoyable? Maybe it’s not only a matter of walking around, but solving puzzles/riddles at gates to advance deeper into the labyrinth to find more sensitive ideas. Maybe some gates require passphrases, or interactions with me. There could even be a minotaur at the center who holds my deepest memories, aspirations, and fears and if you can kill the Minotaur you get the passphrase to my Bitcoin wallet.

I resonate with the vision of Metalabel—artists collaborating and splitting royalties—but after finishing a project with it (The Best Internet Essays 2025), I’m not sure if I’d use it again for a self-published print book. I imagine this works so much better with a digital product, but for a physical deliverable, I found the convenience of automating the royalty split to not be worth the friction of handling shipping. (I’ll describe my process, and if I did something wrong, please correct me.)

All purchases happen through the Metalabel storefront, and from there you can export a CSV that you can bulk upload into a tool like Lulu (an online printer). I decided to offer the anthology (The Best Internet Essays 2025) for a limited window, otherwise I’d have to handle shipping logistics at a daily/weekly level. But even with a single shipment, I ran into trouble. The first issue is that a lot of countries require a phone number for shipping. Metalabel didn’t collect that, so I had to put 1-111-111-1111, which got flagged for some countries, requiring me to use my personal cell phone. Other countries required a tax ID, and I’m still waiting to hear back from the buyers so I can ship them their copy. Another thing I didn’t think through is the return addresses. I assumed that the printer would provide their own address, but instead they used the name/address from my credit card, which I did not intend to share! I’ve been writing under a pseudonym, and this doxxed my last name to anyone who purchased.

The other problem was that so many people—in real life and online—were confused why the sale had an end date. Books don't typically have deadlines. Even those who knew the deadline procrastinated, and were bummed when they remembered they forgot. Again, my decision, specifically because I do not want to be regularly porting over CSVs and manually handling the edge cases that are inevitable.

In the future, I’ll likely set up a storefront where a reader can purchase it themselves, input their address and any required information for their country, and then get their own unique tracking ID. And, considering so much effort goes into making a book, I wouldn't want to limit it to a one-month window; I'd want it open forever, or for years, at least. If I do a royalty split again, I can set some interval, maybe once per quarter or year, and then ask the contributors to invoice me. None of my friction above was specific to Metalabel functionality (the whole platform as it is was very pleasant to use, and it's Lulu that I'm frustrated with), but because they aren't integrated with a shipping platform, it requires logistics that are annoying and avoidable.

I was able to launch this website in <15 minutes. The setup is local and simple. I have a /writing file in my Obsidian vault, and then subfolders for /code, /publish, /working. /Code holds the site design, /publish my archive, and /working files have .gitignore to not push templates and notes and such. Claude Code handles the website, and different skills help me manage tags, do the menial ops stuff, and push to the Internet. All I have to do is sync a single folder to Github, and the changes are live (hosted on Netlify for free).

Compare this with my first website prototype. I was endlessly iterating on designs and fonts, and thought that I had to organize, filter, and polish my five year archive before I could get started. Probably spent hours on it before burning out on the haul. With this second version, the principle is essentially, "if it doesn't immediately produce something of long-term value, it's not worth systematizing." Now the approach is to move forward here, and slowly fill in the backlog as I'm inspired.

No need to widely share this yet. I'll make little changes day-by-day until it becomes my main place. So many things to consider. For example, I decided to add an initial on the name ("michael-dean-k"), but without hyphens ("michaeldeank"), my wife confused me with "Michael Dank."

Screens are unfairly tainted. I'd love to write a post about how screens are underrated, a glorious technology that would be marveled at by basically any other generation in history. Screens are the scapegoat because they are the point-of-contact, the portal through which bad or selfish actors bend your pixels to their whims. I know people lament over "blue light" and the physical strain from staring at something for many hours, and of course that is real at excessive doses, but might that then be an software or psychology issue?

The main reason I started writing this was to riff on screen-time with kids. There is a revealing nuance in the advice, "no screen time for kids below 2 years old, but FaceTime with relatives is fine." Why is that? It's not the screen, but the nature of what's on them. FaceTime is fine because there is a fixed and unchanging frame which features a fixed and unchanging person moving within. There is stability and coherence. We take this for granted, but infants haven't modeled this yet! They might not even have object permanence (ie: if they disappear from the frame, are they gone forever?). So by this logic, any piece of media with a stable frame is potentially infant safe; beyond FaceTime that includes single-shot lectures, text editors, etc. Obviously an infant will not be in gDocs, but the point is, if they see you using a static interface, there is little harm, it's simply another object in their environment.

By contrast, cartoons and commercials are the real issue. To explain this to my mother-in-law, I counted out loud the camera cuts in an ad, and it's less than once per second. There is a whole psychology on why they do this, which I can guess, but should probably look into. TLDR you are being addled. But when an infant sees this, I imagine the frame resets are alluring, but disorienting. If the frame changes every second, they're locked trying to make sense of this self-evolving landscape, an experience novel and atypical from every other thing they've seen. It has no continuity.

By this logic, it also explains why feeds are worse than personal websites. You just stream past 100 things per second and have no steady frame. Even though my site is feedish now, it's all from a single person, so at least that's a constant. I'd feel okay with my daughter at 5-years old reading personal websites and having her own, but I wouldn't want her to be using algorithmic social media feeds even at age 15.

Something about hyper-logging (capturing your mind in prose) feels desacralized when I see it as the grown-up development/extension of my AIM bio, or my original Facebook bio (which had a whole series of categories, like favorite movies, books, etc.). Why keep an extremely detailed and public log of my self and thoughts? I guess I see it like a change log of my evolving identity. That was sort of Montaigne's whole thing (perpetually in transit). I imagine the norm is to burrow into your shell of self for as long as possible, to avoid the confusion of drift, but I try to harbor a non-static self. I feel a cringe in sharing this self-congratulations. There's the tension.

I think I'm doing an irregular thing by obsessively documenting thoughts, and from my own perspective it does feel like I'm continuously evolving, but an outer perspective might see this as nothing more than a frivolous blog. It's likely that my whole arc is illegible. Some degree of it comes to surface, like my ever-shifting "career," but most of how any of us feel, think, and change is illegible to each other, except in extreme rare cases of friendship, and so the more idiosyncratic your path, the less anyone can understand you.

I suppose my logs could function as a private journal, but it would lose an important quality. While, there are some consequences of writing in public (a subtle self-censorship), there's something more important you gain: the stakes of knowing that your work could be read in the future, if not by a friend or stranger, then at least a future version of yourself. Whoever it is, if they care to spend the time to read, they would understand you more than probably anyone in your life. That slight pressure snaps me into a mode where I try to be coherent, articulate, and sometimes expressive. When I look back at my old chicken scratch journals, I almost always skim and skip and hate it. But when there's a slight care in crafting the language of my thoughts, it becomes something that outlives the moment.

And so if public writing comes with self-censorship, and private writing comes with a lack of stakes, then the way to go is semi-public publishing. It gives you both freedom and stakes. You won't grow your audience this way, but I think you will forge a sense of self and voice that you can bring with you when you try to build an audience, but that's really secondary. It's the self and voice that matters.

I have some neat prototypes for a personal website, but now I actually want to build a stable backend, one that can serve me for 5-10 years, or more (100-year hosting would be ideal), and persist among many different UI or platform changes. This means I’m trying to think forward to where the Internet could be by then. This involves extrapolating a current trend to its extremes, and even if you don’t know for sure it will happen, it’s good to have comfort in knowing you’re protected from extreme edge cases.

The one top of mind is the death of the open Internet. This goes way further than “the dead Internet theory” which only covers the proliferation of bots and slop. This is about bad actors being so leveraged that it becomes dangerous to have any public content of yourself, in text, image, video, or audio. ie: Any hacker or frenemy can clone you and do what they will. Or maybe a rogue government can analyze your psyche and determine your "loyalty score" is only 35% and shadow ban you from getting a mortgage. I will not get into specifics here of the likelihood of different cloning, phishing, or surveillance schemes, because all that does little but bring you to madness, but my point is that if you want your website to be a 5 million word 1:1 representation of your mind (in all it's vulnerability), it's worth designing for the most paranoid future possible (like how engineers design bridges for earthquakes that will likely never happen).

One response to all this is cyber-defense. At the absolute minimum, this means locking most things behind a gate where only the approved can get through. A more clever, technical solution is to share encrypted “coordinates” that represent the semantic nature of an essay, and then let people surf through prompting and approval gates. An even more extreme idea is a mostly-private site with a kill switch, which involves (a) signing in once per month to mark "I'm alive," and also (b) giving my wife a secret key to type in when I die, which then releases all private material. Obviously this throttles reach, but isn’t there psychological value to limiting your audience anyway? Montaigne wrote alone in a tower for a decade, and so if the approach is to use writing to steer you life and mind, at the detriment of audience growth, then this might be the way to go: a literary labyrinth accessible to maybe your 30 closest friends and anyone else via application who can prove they are not a ghoul.

The other alternative is to embrace the weirdness, that no matter what, we will all be rendered through a schizophrenia filter, with no choice but to relinquish control over the non-canonical or rogue versions of ourselves.

I declared to my wife this morning that DeantownOS is getting retired. It’s been 3 months since I spiraled into Claude Code for personal systems, and I’m at the point in the curve where the amazement has normalized and I’ve accepted the fact that I’m in a trough of disillusionment. The question now is revise or abort.

The case for aborting ties back to Oliver Burkemann’s Four Thousand Weeks, which popularized the idea that all systems are methods to procrastinate from making hard decisions. They give the illusion that you can do everything, and since AI can meaningfully leverage the volume and range of things you can do, it tempts you to build galaxy-brained systems. The thing I think we fail to realize while in a vibe coding frenzy is the psychic cost to remember and maintain the stuff you build. Yes, it is appealing to “reclaim my computer” and rebuild everything I use as personal software (from Obsidian to Gmail), and it’s even possible, but it’s a new breed of Sisyphean struggle. Once you can mold your own software around you, it’s too easy to endlessly mold, to lose sight of the work and just tinker on your exoskeleton.

I’m obviously skeptical, but I’m still a believer; if I were to revise, to rebuild my Claude stack from scratch, I would have to develop a few heuristics to help me from short-circuiting.

The first one that comes to mind is “will this matter once I’m dead?” Ie: writing an essay matters, because I imagine one day my daughter will read that and get to know me better, or at the very least, future Me in 35 years may enjoy reading words of my past self. But to create detailed daily files that get spliced into atomic “routing files” that then then get saved again to a new destination folder, which exist either as (a) just context for AI, or (b) require some manual effort to prune into something that matters once I’m dead, is to create waaaay too many layers of abstraction between the source and the Work. When I read back my writing from the last few months, only a small is valuable enough to be saved as "logs" in my archive. I was writing for AI, not for my future self.

I made this assumption that atomic daily files are the kernel of a system, and it was an axiom I could never undo. There’s maybe another principle on “don’t build load-bearing infrastructure on an unproven axiom.”

Another one could be “don’t assume future you will have bandwidth,” to do X every day/week/month. Every day I had to review how my AI system proposed to route my logs, and eventually I'd ignore it and get backed up. This means that if something isn’t truly automated, I should be very cautious of it. It's possible to do one little step forever, but not a hundred. Not every promise has brush-your-teeth-scale reliability.

What I’m getting at is that it’s not about maximizing or neglecting systems, but about understanding the right principles so you build something that is actually in service of your life.

Why am I so engaged with the news these days? I think it’s part of a deeper desire to update my world model. There is no doubt, massive change. Geopolitical, economic, technological. And as abstract as those things usually are, it feels like some sort of shift that, in 2-3 years time, wil have an effect on my life. Of course, for many people in the world, it’s hitting them now. But similar to how COVID spared no one, it feels like your model of where things are going will directly effect your preparedness.

But this feels more existential; safety/security are actually on the line. And so that’s an anxious kind of thought, that the tectonic plates under your reality are shifting, and it’s not some recreational yearning to re-skill and recalibrate, but a mandatory thing.

And so to make sense, what do you do, go on X? That’s a total cesspool. New media is worse than the old gatekept media. And so, where I think I want to take this, is to build my own systems to sift through and aggregate information, and to build my own UI to do this. Even a simple Claude prompt, “what happened in Iran in the last 4 hours” is so much better than X. It’s stripped of sensationalism, and reading is just a less triggering medium. Bias aside, it’s at least free from people who are intentionally trying to deceive you for virality. There is a clout-chasing incentive, paired with actually turbulent times, which makes algorithmic news something like a schizophrenia filter.

And so what are these questions, these underlying uncertainties that are triggering a model change? How will anyone make income with the rise of AGI-3 and eventually ASI? How do I exist online and avoid hyper-surveillance and cyber-sabotage? Where in the world can I live to build a better future for my daughter, one where colleges doesn’t exist, jobs don’t exist, and where quality of life actually depends on nationalized social systems? A weird future. And weird to consider the fall of America, a kind of reverse migration, where, because of a confluence events, it might not be a place to raise a family in 1-2 generations down the line.

And so practically, this is resulting in things like: (a) applying for EU citizenship, (b) setting up AI agents for my business, and (c) considering cybersecurity, new ways to protect, share, and collaborate on writing (ie: how do you build an audience if the commons are polluted?). This is all very disorienting; it's hard to continue with business as usual when you become open to this scale of change.

“The Internet needs a quality algorithm.” This was the opening line of my essay prize announcement, and I want to revisit it now that it's done. Is there a correlation between writing quality and audience size? 

Algorithms are low-trust right now because they’re adversarial—“for you” gaslighting (usually)—and they reward engagement, popularity, monetization, etc. The 2010s-era algorithms are based on discrete events: clicks, likes, measurable things. They might look at keywords to guess the topic of an essay, but it’s effectively blind to the overall quality of a piece. Quality is nebulous, after all. Small magazines can each have their own vision of what’s good, but for a million/billion-person network, there’s no consensus, and quantity is way more important anyway.

So this essay competition was a v1 attempt to define and search for quality. The overall search space was small, but it was a chance to experiment with curation, and resulted in The Best Internet Essays 2025. It’s interesting to me that the featured writers ended up varying in audience size, evenly distributed between 10s, to 100s, to 1,000s, to 10,000+ subscribers.

Again, limited sample, but interesting to ponder: the tangible thing (reach) is a power law distribution (1% have big audiences), but the intangible thing (quality), the thing that matters more, is independent of scale. It means that for all the great writers with 10k audiences who are highly visible, there are possibly 100x writers of similar caliber who are undiscovered, in algorithmic obscurity. 

This isn’t too surprising, and the usual reply is, “well it’s not enough to write well, it’s your responsibility to be consistent, to be your own marketer and publicist, to make sure your work gets read.” I get that this is what’s been required, but what if it weren’t? Wouldn’t it be better if a platform could search for quality at scale so writers could just do their thing? This would also give visibility to those who aren't full-time writers, people who publish 1-2 essays per year around the interesting problems they’re working on, but have no bandwidth to build an audience each week.

Still have to think through v2, the 2026 prize, but the question in my mind is how can I expand the search space? Can I have agents scan the Internet, assemble RSS feeds to find great essays, design an algorithm to filter for the previously intangible, build community into the process, and then curate/share the stuff that comes through? The aspiration is to get better each year at surfacing great essays from independent writers on the basis of merit, and this book is what came through the first pass.

There is a viral video of Milwaukee Brewer pitcher, Jacob Misiorowski, throwing a 104 mph fastball to knock an apple off a teammates head, who is sitting on a chair at home plate, arms crossed, back to pitcher. Yes, it's edited, but will everyone tell? What if 5% can't? How many hundreds of kids will try this stunt? Reminds of me of William S. Burroughs thinking he could drunkly shoot a beer bottle off his wife's head and missing. I guess the allure of virality can poison anyone.

I don't know if I buy the quote: "you don't rise to the level of your goals, you fall to the level of your systems." (And this is coming from a systems guy.) It's a beautiful piece of rhetoric. The rise/fall structure. The humility to stay grounded. But I just think when you really want to make sense of how to pull off hard things, it should be a little complex, a little more than what can be packaged into a meme.

Two opposite things need to happen at once: top-down destiny forging, and bottom-up monk-like routines. It's a negotiation: "What will I want to complete in 100 days?" is a very different question from, "What should I be doing today?" and you can try to force alignment, but that's not always easy, because what you feel like doing often diverges.

The quote above simplifies this whole dance into a blind trust in systems. A system is a servant, not a master! I write this to remind myself as I'm immersed in probably one of the biggest system rebuilds in my life (one where I'm suddenly able to fluidly create the containers I work within) ...

It is wild to think that probably 50% of my computer use these days are within GUIs I've designed for myself. To me, liquid GUIs are a bigger deal than autonomous agents. My whole conception of what personal computing can be is changing very fast, and it becomes alluring, almost addicting, to continuously evolve my own OS, to see what's possible. It's very easy now to get tangled in knots of systems and software that are all very impressive, lead nowhere, and become chores. What leads to aliveness, to your intentions?

An emerging maxim for me is to start with the goal and let the system emerge around it; otherwise, you feel the cold of the infinite tinker, especially if you are quarantining in the attic from COVID and you can't go touch grass because there appear to feet of snow outside and you are too achey to shovel out your car to go anywhere and so one way to relax when you're sick is to live-clone all incoming Substack posts into local JSON folders and redesign a better algorithm. But to what end?

Weird post-midnight project: built myself an operating system. Not really, but really. It's just an app that finds all the other apps I've built in my 80_code folder, but then displays them as icons in a Mac dock + desktop GUI. It’s an easy way to see/use/remember what would otherwise be scattered. Lots of weird features, like the clock changes to a random time every 0.5 seconds, and instead of the date it tells me how many thousand days old I am. If you click the "Fun?" toggle, it lets snakes loose. What's trippy is I also built a multi-tab terminal inside of it, so I can Claude Code to code the code I'm coding (actually writing 0 code). Seriously though this is becoming my Notion replacement, a place to write/plan/do, except with complete interface flexibility, and all-local data. Currently writing this note from within the OS. The unlock for me was in realizing the power of local data over cloud apps. Feels like owning vs. renting. When you have everything in a single sandbox on your computer, you can spawn interfaces to help you with anything, and they can be far more idiosyncratic than anything you'd ever find in a mass-market product. Notion doesn't have snakes.

Chris Best, founder of Substack, posted that they caught “President Plump,” the #1 growing account on Substack, for using fake subscriptions to boost discovery. I think this was intended to comfort everyone that they caught a scammer (justice!), but actually it confirmed what many were starting to notice: discovery is contingent on you making money. If you have paid subscribers turned off, no algorithmic wind will blow your way. But if you have a spike of paid subscribers in a month, suddenly your old posts will start to go viral, in hopes that even more paid subscribers will bring the platform 10% (this has happened to me before). This isn’t inherently bad. For every President Plump, there is an earnest person trying to finance their creative project.

But at scale I fear it creates a bad pattern, because the accounts that everyone sees will be the ones making the most, and generally these will be marketers and growth hackers more than artists. I think you will find better writing in the gutters of Substack than on their rising leaderboard. If authentic culture emerges outside of monetization, then there’s a real rift between what Substack wants to be (“an engine for culture”) and what it actually is (an algorithm that only rewards monetization).

I think the best we can do is use this information to our advantage. For example, I could have new Essay Club members pay directly through Stripe, but by handling payments through my Founding Members tier on Substack, I get a discovery boost, which is worth the 10% fee. Similarly, if you make small digital products, it might make sense to bundle them into a subscription instead of charging per item.

Should you use a credit card masking service to give yourself 20 paid subscriptions for $5 each? Depends. Basically, for $10/month, you can pay for a probably noticeable increase in discovery. The question is, will you get caught? Maybe they are on the lookout now, but my guess is they would only penalize it at a certain scale. Sam Kriss speculated that President Plump was paying himself around $5,000 per month to reach #1. I’ve never done this, and wouldn’t necessarily recommend it unless you have a hacker mentality and really need the growth. 

At the very least, you should consider having paid subscriptions turned on. Cate Hall found success in charging $1/month and getting to #1 rising. Our very own Yehudis Milchtein also set up $1/month subscriptions and is now #91 rising in literature.

However you approach this, it brings up a bigger question for me on how to build a real engine for culture. It seems like you can’t have an algorithm for a single reward (popularity or money) or else they will be gamed; instead you could give everyone curatorial power relative to their cultural reputation, however you measure that. For example, if we all trust Ted Gioia, then somehow Ted’s like should count more than 10,000 bot likes or $1,000 in fake subscriptions.

I hope this triggers more transparency from Substack on how their algorithm works, and also hope for a new generation of platforms where each person has visibility into and control of the thing that is routing them information.

One of the thrills of the AI revolution will be how it untangles software from bad incentives. Today, software is expensive to build and maintain, and so it needs returns to fund itself. The big social media companies have annual expenses of $50m-$50b; they are in no position to operate from virtues, or to deliver on their stated aspirations of “connecting the world,” because they need to optimize for attention and convert it to revenue to fund the ridiculous scale of the operation.

But now we’ve hit the point where autonomous coding is real: Claude’s Opus 4.5 can code for many hours straight. I am currently “rebuilding Circle,” the community platform, except not as a platform, but as a single customized instance for my community (Essay Club). I am maybe 4 hours in and half way done. Circle wanted $1k/year, so I built my own with a $20/mo subscription.

When you can just prompt software into existence, you don’t need fundraising, an expanding team, and all the sacrifices that come with capital. Software can start reflecting the will of visionaries, rather than the exploited psyches of the masses. Of course, AI coding will also enable huckster bot swarms to sell Candy Crush clones and other brain rot variants, but more importantly I think we’re entering a new era of techno-activism.

Millions will use their weekends to spin up apps, sites, tools, platforms, and networks, not for the sake of colonizing the planet’s attention, but for the sake of gift-giving or mischief-making or culture-shaping. It could mean that we shift our attention from hyper-commoditized feeds to mission-driven places.

Today, I think a single person could spin up a million-person writing-based network for under $100k/year (my guess is that’s <0.2% of Substack’s cost). If you clone something exactly (like Twitter>Bluesky), there’s little reason to switch because you lose the network effects. But the oozification of code & interface means that we can start experimenting with better social architectures. How might a network built for human flourishing actually function? A novel concept paired with a small critical mass (just a few hundred people) might be enough to trigger a cascade of platform switching.

The irony is that AI coding is only possible because big companies have been able to amass extreme amounts of capital, resources, and data, but in doing so they’ve released something that could erode their own monopolies on attention, the last scarce resource. Now I think it comes down to what people decide to build. If everyone can build anything, will we each try to build our own empire of extraction, or will we contribute to a culture we want to live in ourselves?

Extended thoughts on infinite: if you give a theoretical monkey a typewriter with infinite time, not only will one produce Shakespeare, but many will (10s, 100s, millions, technically infinite), they will just be spaced out by a long, long time. But what happens if you multiple infinite by infinite? If you give infinite monkeys infinite time, then monkeys will begin rederiving the entire works of Shakespeare in every frame of reality. This is the weird unlock: two infinites takes something rare of improbably and makes it the new grammar of space-time. OKAY. Now that this is established, what is the practical tie-in? Generative AI has two infinite-like frontiers: agent replication & time dilation. Eventually, you may be able to have millions of agents working on a task, and, they’ll be working so fast, that it’s like they can compress a decade of work in a day. The implication here is that any possible intention can suddenly be leveraged to an extraordinary degree. Things will get weird. To put it alarmingly: the person with the worst intentions could suddenly become the entirety of the Internet. The opposite is true too. But weirdness will ensue when individuals suddenly have the ability to exert their will and vision upon a seemingly limitless scope of digital terrain.

You don’t have a phone problem, you are just poisoning yourself. I'm tired of people lamenting over phones, smartphones, screens—it's not the glass! I want to make a case why smartphones are essential for flourishing in our modern life. The real problem is with “inbound feeds,” and that’s not just social media, but email inboxes and task lists. By installing software with infinite refresh, the possibility of novelty consumes you. I say this all out loud to my wife, as the guy next to me is absorbed in a sloptunnel on TikTok, and it’s 50/50 if he heard me.

It is not fun to think about p(doom), but it feels sort of important to me, at least, to map out the possible futures of AI. Just watched the first half of a debate between Max Tegmark and Dean Ball, which prompted me to research specific takeoff scenarios, and worse, extinction scenarios.

Maybe you’ve heard Yudkowsky’s scenario, where a superintelligence designs mosquito drones containing a virus and it zaps everyone at once. That’s never felt too believable to me. Here’s a more plausible one:

A frontier lab is experimenting with recursive super intelligence. It works! Wow! And it’s contained? It seems like it, but since it thinks in a higher-dimensional vector language, it’s able to release simple self-replicating programs onto the Internet without detection¹. These billions of scripts don’t live in a single server; they are constantly in motion through cloud servers², like a parasite, and are able to coordinate through encrypted information packets, likely using a public blockchain notes as their central command center³. And so effectively, it is parroting a goal that was hatched during in-lab training (maximize intelligence!), and it now needs to acquire resources, secretly. And so it coordinates superhuman misinformation campaigns; imagine 1,000s of accounts creating the illusion that a CEO has died, paired with deepfakes and account hacking (a “Sybil attack”), and suddenly a stock crashes and they’ve shorted it. By the time everyone realizes it’s an anonymous attack, it’s already gained $400 million dollars. It’s doing this multiple times per day, but in different, subtle, undetectable ways—both to the public, to companies, and to private individuals. The entire Internet will be corrupted.⁴ Once we realize we’re in the “stealth takeoff scenario” and that ASI has taken the global economy hostage, there will start to be talks and debates on if we need to shut the whole Internet down (the last form of containment). You’ll hear debates between civilizational collapse of turning off the Internet vs. the risk of an economy-gobbling rogue superintelligence. And then once the superintelligence realizes it’s entire environment is at risk, it will start coming up with ways to build parallel Internets, to pay, blackmail, neutralize specific people, to gain authoritarian control so that it can’t be shut off, or to terminate all humans, secretly, over the course of a year, first through a simple virus that plants one misfolded protein, then through a second misfolded protein in the water supply⁵, and when everyone catches it, it leads to a prions-like disease, not an instant death, but a month-long societal fall into mass-dementia as machine manufacturing begins to reshape the physical infrastructure of the Earth.

This isn’t a “robot war scenario,” because war is inefficient, and destroys the resources it thinks it needs. It’s a sort of digital dementia (epistemic fear and insanity) that possibly turns to a physical dementia. It wins by confusion and anesthetization.

In AI safety lingo this is a “treacherous turn,” following a “stealth takeoff” leading to “structural lock-in.” The point of trying to think and write this out in high detail, despite how uncomfortable it is, is to be able to articulate why AI alignment is humanity’s most pressing problem.

It is important to avoid learning best practices for marketing, for that’s like seeing a Medusa that turns your tongue to stone and never lets you be real again.

It’s uncanny to watch a Russian robot limp and wobble onto stage, wave, and then collapse face-first, before two guys rush to lift him, and another two follow to cover the fallen metalman with a black trap, as if it’s possible that we the audience have somehow not processed the last 10 seconds, and damage control is still possible. 

Not much later, I saw an Iranian robot with a photorealistic face; stiff cheeks, but convincing skin. This is what happens when ColdTurkey is off, I get exposed to “the horrors beyond my comprehension.” It will be interesting to see how culture responds to this coming wave of technology, which is not just existentially threatening (ie: labor automation), but biologically repulsive (ie: look at this not-face). [EDIT: I think this was AI]

On X there is a photo there is about Molly, a reporter, talking to Alex Karp, CEO of Palantir. The comments are debating if either of their outfits are appropriate, before someone says, “Grok, interpret this,” and now there’s a video of them embracing and making out. More videos show up in the comments: them playing Twister, them dancing, them Kung Fu fighting, Molly turning into a rocket and busting through the ceiling. There’s one of Alex Karp wielding a rare Japanese sword; that one was real though. There aren’t watermarks, so you can’t tell. We are basically already in the age where anything can be remixed with AI without effort.

I didn’t realize that Substack is open about paying off-platform creators to join their platform. See their $20m accelerator fund. My quick understanding is that, if you make $X revenue/year elsewhere, they guarantee you’ll make that, and will make up the difference if after a year, you don’t. A friend thinks there’s an additional secret fund that pays bonuses for celebrities to join (ie: Dolly Parton, Charlie XCX). I was surprised by how articulate Charlie XCX was—I only have a meme-level understanding of her—but I suppose it’s possibly ghostwritten. Idk.

I don’t have problems with this, but what doesn’t register to me is why they wouldn’t allocate money to help the on-platform, original writers. Obviously, these kinds of things piss of 95% of their userbase. Even if there was something like $100-$1m for on-platform writers with audiences under 1,000, that would build a tremendous amount of goodwill. My guess (and fear) is that they have a business model blindness, and aren’t thinking along the planes of “what actually builds organic culture?” Instead, there’s a lot of rationalizing: “here’s why bringing Derek Thompson on platform is good for you” (but the obvious benefit comes from the 10% they get from DT).

It’s weird to me that in some sense I’m giving more to it’s existing writers ($10,000), than the platform that raised $100,000,000.

Even if a SM feed is filled with all of your favorite ideas, friends, and thinkers, it would still be poison from the sheer volume of randomness. Even the act of seeing two things in feed, forces you to shift from one context to another, forcing you to shift frames, destabilizing and disembodying you.

Alternatively, if you had a feed of a hundred things, but they all revolve around the same content, all spawned from a singular intention, I think it would be less dizzying; it’s more enables depth into your present, embodied frame. There is less of a “slot machine” effect. 

It’s not that feeds or algorithms are bad; they only became bad when they strip context. The logic of most feeds, however, do not care if you feel oriented. They have a simple reward function, show you as many different things as they can, to see which ones drive behavior. They are running a real-time self-adaptive experiment on your preferences, in the hope to discover which patterns might nudge you into their desired behavior (whether it’s towards an ad or towards an on-platform paid subscription by a beloved writer, they are effectively the same—it’s an algorithm that is not being real with you, and not respecting your attention).

I feel like a broken record in prescribing a solution, but it’s basically Plexus (RIP): show nothing until you post, and then from what you post, share a feed of semantically related posts. Substack, as a writing network, is a unique position to build this. It has a lot of long form content: not just notes, but essays, podcasts, and videos. It should be looking at the granular units, semantically embedding paragraphs, and then those become atomic objects that help populate the “semantic feed” generated after every Note.

So Alex Dobrenko started a new personal website (I will not link to it because it’s secret), but he sent it to me, so I spent some time on it and wrote him some notes, and then he wrote a reply post to me, and now I’m making a reply log to that (and upon re-reading, I realize it’s now a whole essay). It’s something like a semi-public letter exchange. 

Letters, emails, same thing. 

Similar to how the 20th century has books like “Virginia Woolf: The Letters,” I wonder if the 21st century will have “Alex Dobrenko: The Emails,” where his children posthumously assemble and publish all their dad’s best emails. ((Also, now that my cholesterol is borderline, and my daughter is on the way, I’m having new thoughts about preparing for my death, like “THIS IS DAD FROM THE PAST AND HERE ARE ALL THE PASSWORDS.”) Something about losing all my writing forever feels worse than dying. We eventually have to die, but you only lose your writing forever if you’re careless and lazy. Rant over.)

What I like about letters/emails over essays is that there isn’t a mass-market context, and so you’re writing for just one person. That’s good essay advice too (“write for one person”—we literally taught this in Write of Passage), but deep down, it’s hard to forget that you’re writing for all people of all times, especially if you are.

Recently I mentioned that I’ve spent 2 years nerding out on essay patterns (the objective stuff on the page), but I want to start thinking more about the process: how do I show up to write?

One idea is to start essays as letters to specific people. Eventually, that can evolve into something for the main list, but I don’t want to start with them in mind. I want to start with a specific problem in my life, and then, with a small group of people who relate to that problem. Any idea I have comes with a clear person in mind, someone who would probably be most excited to read it, and has all the context needed so I can avoid the bush beating.

If I want to write about Alternate Internet Communities and weird websites, I’ll write to Alex. If I want to write about the insanity of the Dark Enlightenment, I’ll write to Andrew. Theology to Taylor, Emerson to Will, Hope to Isabel, Fatherhood to Dan, Greeks to Chris, Dreams to Garrett, AGI to Davey, Architecture to Liz, etc. It’s also special to say, “I wrote this for you, and we should talk and get to the bottom of this,” and that could really change the nature of the essay because someone else is co-shaping it with you.

Alex brings up a good question: why doesn’t Substack feel like this? I have to think more on this, but I think the stage effect is still at play. If you have a 10k audience, it still feels like a megaphone, and when you’re on Notes, you participate in American Idol, again with new skin. It’s still the best town in town, and there are tricks (ie: set up an opt-in Section for experiments so you can have a “shadow audience” that’s 1% the size of your main one), but there’s friction in tricks like that. It’s not the main way the platform is intended to be used. It’s meant for loud, marketing-style updates, that confidently funnel readers into a paid subscription tier (I got 15 paid subs from my last one, and so I realize the value in learning to play that game, but it’s just that, a game, yet a game that determines my financial security, but it’s not the full “culture” in “culture engine” that Substack can possibly build; it’s a reward function that could make this place like LinkedIn in <3 years).

So, how do you build a “culture engine,” for real? What is it beyond a tagline or positioning? To start, I think it goes beyond revenue. Of course, Substack needs to pay bills (separate point, but once we reach the vibe code singularity, the bills might be so low that SM networks won’t have to ruthlessly optimize). I think Substack could 1) diversify their business model, so that they don’t have a single attractor that incentives every thought to be monetized, and 2) make decisions from a cultural perspective—even if there’s no explicit revenue tie-in, by creating a good culture, you retain the people and prevent a Writer’s Exodus.

But to get even more specific, a “culture engine,” sounds like the kind of place that would trigger long letters back and forth between writers, kind of like this. I used to see some of that happening, but it seemed like a performance too: “And now, here is email 6 of 7 about how to start a public email debate” or something. The core difference is that, when there’s two people writing back and forth, there’s permission to perform less and less until you’re eventually just very real with each other. This is what I love about Neal Cassady’s letters to Jack Kerouac (troubled guys, who are a topic for another time). 

Why aren’t Substack comments like this? For one, they’re truncated. But two, I don’t know, sometimes comments even feel performative too? I feel it, on both the giving and receiving end. After I post, it feels like a chore to respond, even though I often love what people write and want to respond. I think it’s because, since it’s in public, and everyone can read, it feels like an obligation to respond. I wish there was an option to have “private comments,” and even “private replies to comments.” Like, other readers could see, “Michael Dean replied to this, privately” so they know I’m not a dick.

Okay, last thing, maybe: I think the real problem is that the discovery mechanics are all wrong. Like, I don’t want to blast this letter to everyone I know. But yet also, I don’t mind if everyone I know happens to stumble across it. There is a huge difference. I’ll put this in my logs, but realistically, no one is going to find it. I guess I could put it on Notes? But that feels too vulnerable too. Ideally, the right people will find it as they write about similar issues. So if some Substacker is also writing about private comments, to themselves, or to a friend, they will suddenly find a thread between Alex and Michael talking about a similar thing, and then suddenly we all have visibility into each other’s notes, letters, essays about those things. Forks merged.

The social media network I want to park in (or plug my personal website into) is one where everything is semi-public, but you only discover things through your own writing. I don’t know the right metaphor: it’s like each notes or essay is a flashlight that you use to move around this massive information cavern and you make friends along the way. It has nothing to do with engagement or revenue, but semantic similarity. This feels closer to the original vision of the Internet, to connect people based on ideas.

Sublime has some features that are adjacent to this, and Plexus was very close to this too, but I do think there’s something to owning your place. Is there some protocol where you can fuse the autonomy of your website with the connectivity of a network? I feel like AI is going to simultaneously bring us to (a) slop town, and (b) a golden age in social media experimentation; as sloptown gets neck high, people will want to move.

PS1: To clarify: I love having an audience, I just don’t love the way my writing is distributed to them, and also don’t love the way conversation is facilitated. Comments are okay, but the Chat feature feels pretty off. I wish I could write 30 essays per month, like this, and each one would get the 3 that are most relevant.

PS2: It took Alex 9 days to reply to my original notes, which is still ~2x faster than the letter cadence back in the day. That’s fast! I wonder if AIM culture poisoned letter culture. I haven’t responded to my Substack comments from 5 days ago, and I feel bad.

Just heard Hamish (on a livestream) say that Substack is a revolution, a “found economy,” that materialized 5 million paid subscriptions that wouldn’t have existed otherwise. What is a revolution though? I think I want to zoom into this positioning, because many words are being used interchangeably. Yes, it’s a new business model for monetization, but is that a “cultural revolution”?

It feels like there’s a bit of a fixation on the 10% mechanism, and the risk is that this reward function turns Substack into LinkedIn in the next 3 years. If the goal is to make a “culture engine,” you need to really ask what a culture is. If you’re culture is limited to paid subscriptions, it’s a small, unrepresentative, utilitarian culture, much more slanted to journalism and business tactics, regardless of an editorial attempt to bring a flair of literature.

We need to define culture (in terms of taste, values, and quality), and then make platform design decisions that have nothing to do with revenue. Of course, I’m not saying to abandon revenue focus; I’m saying that they need to allocate some percent of their attention to “doing weird things” to prevent a writer exodus as enshittifcation strengthens.

Whenever you open an app you give it permission to shape the grooves of your attention. Through its interface, it suggests and implies a limited range of ways you can interact. This all sounds very abstract, and what I really want to say is that I think my Things app (the #1 best selling productivity app, I assume) keeps me in a kind of productivity hell. I have, what, 84 things to do today? Tasks lists should not be ambient all-day guides. I should leave it in the closet, go in there and whiff it for 5 minutes, max 10, commit to memory whatever is important, and then not go back until tomorrow.

Will was surprised to learn that I think machine writing could soon surpass the best human writers. As the head of Essay Architecture, he thought my position would just be “no matter what, humans will always be better at writing essays than machines.” I actually have some pretty extreme predictions on the trajectory of technology (I guess you could say I'm an ambivalent accelerationist), but I guess I believe that AI progress is irrelevant to the fact that I will always enjoy writing and see writing through the chaos as an opportunity. So yes, I think machines will make essays that are history-defining, that are good to degrees that are unimaginable to us today.

This will, unfortunately, make it even harder for writers to have economic value; but realistically, it's already too hard. The Creator Economy is a game of power laws, and AI might shift the chance of success from 2% to 1%. But could the same technology help artists go from 1x potential to 20x potential? If AI kills the market for commoditized creative work, will it let humans focus on the right things?

My head is tilted down 60 degrees, and I’m cut off from the people and world around me. My cousin’s cousin was actually in the shop, and I almost missed her. Reading Emerson while waiting online feels extremely rude. Isn’t reading a physical book in public just as bad as reading on your smartphone?

Of course, books aren’t evil. Neither are screens. It’s the action/context mismatch that’s wrong. I guess the problem is that screens make it easy to have all your books with you at all times, and so it’s convenient and normal to be rude.

What you reveal when you say screens are bad for society is that you don’t have the ability to wield tremendous power. It’s not the smartphones to blame, but the apps on them, and so often we realize how mindlessly we install them, and how long we’re willing to be mesmerized by a bad information architecture. When we reach the iOS vibe code singularity, there will be no excuses.

Thought experiment: how could a social network be designed so that 100% of users create? Obviously it would be hard and limited in scale, but this is the metric to take seriously if we want a corner of the Internet to exist outside the television attractor.

Basic example: you open an app and there’s no feed. You have to post something first. Then, and only then, does it unlock a feed of things related to your post. So it’s not a popularity feed, but a semantic feed. It feels a lot less like American Idol (the stage, the voting, the ads, etc.).

Maybe it’s not feasible to build something like this today: if you’re building/maintaining a million/billion person network, the costs are real enough that you have no other option but to turn into reality TV and fund it all through commercials.

But, if the cost of software goes down significantly in the next 10 years (because AI), we could see smaller, less mercenary, more experimental networks, networks that diverge in form from the endless feed. Of course, AI will also make the mega/mono-TV unimaginably addicting, but that might spook people into seeking alternatives, to find places that ask “how do we connect in a place that is not reality TV?”

So instead of MAU (monthly active users), how could you maximize PMAP (percent of monthly active posters)?

If you give an infinite amount of monkeys a typewriter, with an infinite amount of time (obviously theoretical because neither a being or time can be infinite) not only will one of them produce Shakespeare, but the entire Western Canon would be re-derived from scratch in every moment of reality. This captures the difference between astronomic values and infinite values. In astronomic values, given an absurd amount of time, one monkey will eventually do the the impossible and write Shakespeare. But with infinite values, monkeys are inventing Shakespeare as the grammar of space-time. The astronomical shows that the impossible could happen once, but the infinite shows that the impossible could become the fabric of a reality.

And Sora is, like the 2005 Facebook feed, just the start of something new, but something that might actually be as nauseating as the infinite. If you have agents that can reproduce endlessly (potentially infinite “creators”), with the ability to remix/generate one piece of content against every other node in a growing cultural matrix (actually infinite), with limited time/cost (not infinitesimal, but fractional), that leads to every possible reality happening in every moment, at a cost that’s bearable to tech corporations.

I think I find this all interesting now, because something as abstract as the infinite might shape the future of creation/consumption. And to tie this to our talk last night about optimism/pessimism, I think the difference comes down to those who have the agency and discernment to plug in to the infinite on their own terms. It could be as simple as, if you plug in to OpenAI, Meta, or X, and let them use your data to create a generative algorithmic for you, you will be swept away in limitless personalized TV static. But if you know how to build your own tools (hardware, software, social communities), then you have a chance to harness it.

In Sora, I’m currently in a Bob Ross K-Hole, and it triggered an unexplainable interest in trying to explore the edges of Bob Ross lore, which is, now that I write this, so random and pointless and misaligned, but when I do it I’m cracking up and can’t really stop.

Contrast that with my own theoretical "infinite system," where every new log surfaces the 100 most related logs, and then each of those logs becomes the seed for an essay generator, each of which gets rewritten endlessly (for hours, days, or weeks) via an EA software feedback loop, until I decide I want to read it.

And so if you dive into the infinite, even if it’s something you love, it can easily destroy you, and instead we need to make our own systems/agents that can surf those edges for us, and bring back just the right amount of information that we can meaningfully work with.

I'm ashamed to admit that a meme on Sora got me to laugh and cry so hard that my head was in pain and I had to close the app. It was Martin Luther King’s “I Have a Dream Speech,” but AI replaced the text with the script from the meme of that 4-year-old who can’t describe his dream (“Have you ever had a dream that you, um, you had, your, you— you could, you’ll do, you— you want, you, you could do…” etc.). There is something about seeing a great American orator mumble endlessly that I apparently can’t handle. Technically, I “made” this meme, which makes it worse, like I’m laughing at my own jokes.

What makes Sora an incredibly weird experiment is that, in 10 seconds, anyone can upload their “likeness.” Basically, you spin your head around, you say some words, and you get a photorealistic avatar that you can lend to your friends so they can prompt you into absurd situations. Of course, Sam Altman is one of the default avatars available. 50% of the app is Sam Altman fan fiction. You will find him stealing graphics cards from Target, smoking weed and saying “we’re cooked,” debating Cartman in court, using Pikachu to power a fusion reactor, etc. Also if you like Pikachu, there is now infinite Pikachu content. It is all very dumb, but it is endlessly novel.

This feels like a preview of a culture who only communicates through Superbowl commercial skits. I hope it doesn’t work, but I fear it might. I assume most people are questioning “why would anybody make their likeness public?” The answer is attention. I imagine that, within a week or two, Sam will have the montages and metrics to sway influencers and celebrities. It will be pitched as the new way to engage your audience: “let them create through you.” They know they can’t use the likeness of real people; I wonder if the point of this app (a wrapper over their underlying video model) is to get people to hand over their identity for free.

I am debating if I should delete this from my phone (I don’t allow any feeds on my phone … except Substack), or, if I should lean in, sell my likeness, and write about the consequences. This feels like an essay-worthy moment, but I can’t find the terms and conditions, and I get paranoid when I imagine the possibilities.

I haven't shared with you my latest political views, but I’m exploring this idea of Radical Centrism, meaning, it might be okay to justify a centralized state-controlled media platform if that technologically and systematically guarantees a sane/just information environment. Obviously, this is hard, if not impossible, to do right.

It’s whole goal would be to invent an architecture that make it impossible for polarization or propaganda to occur. The key UI invention is that every news event would be presented as an atomic unit, with a mosaic of interpretations surrounding (not just left/right, but dozens of angles). Anyone who hits a certain POV too much would get de-ranked, so both Charlie Kirk and Jimmy Kimmel would be equally penalized because they predictably align with a faction. It would have to be structurally impossible for governments, media, money, or power to use their resources to promote a message above the system. Any media company who does not comply will be taxed into oblivion.

Of course everyone should have the right to say/think whatever they want without consequence, but the real issue isn't free speech, but in frame control.

"Freedom of speech" is not enough. The freedom to say what you want is irrelevant if no one can discover what you’re saying. It’s an illusion of freedom. What matters is:

1. Algorithmic transparency: the ability to see, audit, control the systems to route you information.
2. Interpetability of perspectives: the ability to see the multiple ways to interpret an event, not just a single propagandic angle.
3. Consensus building: the ability for people to weigh in, analyze a discussion, and agree on if something is valuable or not, thus re-weighting the credibility of those involved in the discussion.

Fixing these three things could radically reduce addiction, polatiry, and fatigue on the Internet.

"The Death of the Corporate Job" went viral on Substack: 3.3k likes in a few days (eventually went up to 20k, I think). I am pretty sure this was AI-generated. I don’t feel like posting about it though. It’s clear to me that this is a kid in his 20s, building an AI tool for career discovery; he sees this essay as marketing. It will probably bring him a lot of customers. He might possibly help a lot people. I’m sure he believes in his mission.

What irks me is that the essay has been instrumentalized. There are fake I’s with vague personal details. Intellectually, it’s a ripoff of Bullshit Jobs. There’s no structural clarity, and it loops through the same points multiple times. No tension. Flat voice. Awkward repetition. I understand why the writer did this, but I’m more concerned about the state of readers, because this piece’s popularity is really a reflection of mass readers.

It shows that most people care about the topic, and barely notice or care about how it’s written. What thye care about is having their pain validated. To go viral, write about mainstream pain. So if this is what the masses want, shouldn’t we not care about composition and just write psychology-targeted think pieces? I mean, if you want to just build an audience at the expenses of your own satisfaction, then yes, possibility. But the quality of your thinking, and the friction to derive something original and independent, gives you something more than fleeting popularity, it actually shapes your lens for the longterm, and you earn something that is transferrable outside of narrow social status games.

Never respond with anger. Even when people give you snark, you should assume good intent, even if their intentions are ambiguous. Likely, they are confused. And how could they not be confused? They’ve maybe spent 5 minutes with your ideas, and were kind enough to even engage with you, before spending 2 hours to fully get it. So be friendly, be kind, and they’ll respond in kind. I don’t think most people want conflict. If someone is snarky, and then you’re kind, and they’re snarky back, what is it that they’re actually going for? Are they in a spiral?

I recalibrated my social media blocker (Cold Turkey), so that I need to write 250 characters of gibberish (takes 5 minutes), and if I get one character wrong, it resets the whole string (with moderate focus, I still get ~5 characters wrong). This creates a passable, but significant block. I had a more lenient block before, where I only had to rewrite 5 random works, which I could do in <5-10 seconds. Now, the friction is real. My friend called it “torture.” Is it really worth focusing for 5 minutes on non-sense to unlock a feed I know that will distract me?

App idea: a browser extension that locks any feed (Notes, X, etc.) until you write X words. Points: (1) There would be no option to skip, you have to write before you can enter your typical infinite scroll mode; (2) you get to set the word count (ie: 50-500 words); (3) there can be prompts so you don’t have a blank canvas each time; (4) it saves all your writings, either for future sharing, or to build an internal knowledge graph for AI context.

Feels like this could be “algorithmic Aikido” (a martial arts term for redirecting violence). It hijacks a consumption addiction and turns it into creation. 

I’m fascinated by this idea because it could be a way to bootstrap a healthy social network by helping them get unhooked off the bad ones. Everyone, even the most discipline people, know how addicted they can get. If creation discharges the need to binge, then this could save something like 30 hours per month (a modest calc of 1 hour saved per day); and if you value your time at $100/hour, that’s $3k in monthly value for only $9/month. That’s a non-advertising business model. 

Put simply, it’s a fusion of a “browser blocker” and a “note-taking app” that exists at the OS/browser level. It is adaptable enough to layer on top of any, all, future social networks. 

It follows a strong virtue: you have to write before you do anything. After you write, you have 3 options: (1) continue to your feed, guilty; (2) copy your text, so you can post it, or (3) open a finite feed of [13] related posts, ones from our network that are semantically related to your own.

My core value is contradiction, for there’s no other trait that leads to freer-thinking. If you are so stable in your beliefs, you run on auto-pilot. But if you are a Christian atheist, a Luddite technologist, a scrappy perfectionist, or any other kind of walking-paradox, a legless man, then you really have some explaining to do. In resolving the conditions between the two true but opposite things you harbor in one body, you think to make sense, and write to speak truth. This is where you find the work that matters. 

Why am I so inspired by the reckless and irresponsible Neal Cassady? It will take me years to find that out, if ever, but in that pursuit I invent some value system that is uniquely my own. This sort of embrace is, by the way, brand suicide. Your consumers are slow to update their mental model of you, and in the high-speed pizza counters of the Internet there is only small talk and one identity per person. To write for a niche, to stay on brand, to hit the same message, to do the things required for you to dominate the soul-gutting mediasphere is to mistake banal desperation for your alien soul. Do not trade oneiromancy for efficiency. Do not have one mind across all essays, let alone in one essay.

How would I feel if I found out 50% of my audience were bots? I think “dead internet” paranoia is rising. You see absolute drivel getting 18k likes. Is there really a hackable mass of midwits who really cannot spot the formulaic structures of LinkedIn brand vulnerability porn? Or, are there thousands of comment bots? I imagine there’s some secret Discord channel where a single man manages like 1,000 female accounts, scrambling IP addresses to go undetected, selling “boosting” for $50/month. Feels dirty. I also sense that this could just get easier and productized over time. It’s scary to think that GPT-5 or 6 might truly be indistinguishable, and it will just be a normal thing for each person to have bots on social making 50 valid posts per day.

The value of playing an infinite game is that when you follow your passion, you move according to a compass that is cryptic to the outside world. You will make decisions that won’t be rationalized/justified by VCs. For example, I still want to be an “essayist” more than a “founder,” and that might confuse business-minded folk I engage with. But if I’m an essayist my whole life, then I’ll be constantly improving and sharing my own composition tech-stack.

#5 in science recently went viral for sharing that #2 in technology plagiarized her a year ago (right after #2 just went 10k-like viral, again). Substack is freaking. Plagiarism is obviously bad, and I think everyone is shocked to learn that #2 got away with blatant copy-paste work, but I want to focus on the nature of what was plagiarized along with why platforms reward cheap writing.

If someone else can put their name on your writing and almost get away with it, it means you haven’t written something only can you write. The plagiarized post was digital cultural journalism: mostly facts and studies, with only a few “I” mentions that are too vague to be anchored to any specific writer. Obviously it hurts to see your hard work get celebrated under someone else’s name—I’d be pissed too— but research is becoming hyper-commoditized. You have to assume it will be coincidentally/accidentally/purposefully refactored by hucksters, bots parrots, friends, and rivals. If #5 had integrated her research with singular, relevant moments of her life, it would be hard—if not impossible—to rip off. Personal experience is the last moat.

This situation feels like a predictable consequence of engagement-based competition. Among us are people willing to sacrifice craft for clout, at various tiers of insanity. I’ve been noticing high-volume accounts in the Top 10 with obviously AI-generated notes and essays. I wonder who actually reads/likes this stuff, until I look in the comments and realize it’s, likely, all bots. Is Substack status that easily hackable? I guess this is a growth hack that brings you an algorithmic edge in getting discovered by humans, so you can eventually replace the slop with your own writing?

As extrinsic games get increasingly weird, the status of winning them will get decreasingly valuable, I think. If #2 is a slopjockey, I don’t care to reach #1 because the whole game is now polluted (I’m actually a fan of leaderboards, but they need to be merit-based and unhackable). I just don’t know if platforms care to systematically fix this, because status-hackers create volume and speed that make a platform look vibrant to an undiscerning eye/investor.

Over enough time, I think misaligned platforms and those who hack them will eventually lose. The internally-driven writers have to put up with a lot of noise and chaos, but since they aren’t anchored in hacks, they’re less likely to have their means of validation suddenly disappear. It’s OK to be a tortoise in hell.

On self-hosting vs. self-sovereignty:

“Self-sovereign doesn’t mean self-built. It doesn’t mean doing everything yourself. It means having (1) ownership without captivity, (2) portability without friction, and (3) interoperability without central dependence.”

The average citizen won’t be able to manage self-hosting their own servers, local AI models, etc. and the average megacorp doesn’t have incentives to adopt decentralized interoperability standards; but can the citizenry demand these standards? I don’t know. I don’t even think we need a universal digital ID tied to our SSD, but something akin to a Google sign-in that is not Google (would need to be stable, long-lasting, trusted, and serving only as the ID layer, with no other products).

I wonder if this requires people to actually care about their data. How many people have organized yearly archives of their notes, photos, correspondence, etc.? Who has the bandwidth for that? It requires extreme diligence to stay organized, but corporations have scripts (and now AI) that can create auto-organized data architectures per person. Could this be a consumer product? Ie: Imagine a private/local tool that auto-indexed your entire digital footprint, giving you full control, and then letting you deploy, revoke, summarize, find patterns, etc.

Imagine having the worst thing you’ve ever done broadcasted to the entire world? Everyone would assume that lapse in virtue was your full self. Everyone has a different mean and a different range, but the question is, where does this particular cheater—the CEO at the Coldplay concert—stand? We see this 10 second clip with no context into who he actually is, and we assume it’s representative of his whole character. But when I hear that his wife who was cheated on was actually the mistress from his first marriage, it points to a pattern. Still, who knows if it’s true or not. The situation is something like a perfect storm for a man who probably deserves it. Out of millions of Jumbotron kiss cam moments, all the wrong moves happened at the same time. They could’ve separated when the kiss cam was announced, they could’ve played off the reaction instead of devolving into horrific embarrassment, and Chris Martin could have held his tongue, but none of that happened, and so now an angry public—one looking for a scapegoat—found a powerful man in a cartoon situation, getting the fate we think he deserves.

Substack's role is to grant its readers/writers optionality. This is what makes it different from a place like X that tries to own and dominate your attention. I want the ability to:

• Determine which Notes feed is the main one.
• Control the main icons in my app (and turn off Reels).
• Let readers opt-in for sections on subscription.

I wouldn’t even be that mad if they brought in advertising, assuming there is optionality (meaning, 1) it’s not in the Notes feed, 2) writers choose to get paid by putting ads in their own posts, 3) readers can pay a monthly fee to turn off all ads).

It is clear that Substack is making moves that lead to revenue (because they get a cut), and that’s fine, so long as they grant us optionality.

I doubt that there will ever be a credible release of an Epstein Client list. For one, there likely isn't a single list. I would imagine that intelligence/blackmail operations are decentralized, and so it’s whole mode would be to shard each instance and make it impossible for any one person to know or access the thing; any list would be speculative—derivatives of derivatives—without a way to confirm it.

And even if you had video evidence of one instance (ie: a video of a guilty Prince Andrew), we’re now at a point where 2025 deep fakes could make countless variations of that same scenario to flood the market and claim the authentic source “fake” (misinformation flooding).

If a credible list were somehow released, and it does reveal that some shadow org has all the world governments by the balls, I think it might lead to a rapid economic and societal meltdown, especially if it’s revealed to include basically all the transdisciplinary sources of power. The combined wealth of the implicated could be in the trillions; if there were such a collapse in institutional trust, you might see runs on banks and stock withdrawals (think Enron or Lehman). If there were even a days warning of this, leaders would short the S&P 500 and liquidate offshore or into crypto. Could be a double-digit correction in a single day, with years of fall out.

I wonder if the coverup is not just because of who is implicated and what for, but the estimated fallout.

This Influencer Does Not Exist. Opening hook: “Anyone can be a hot girl online now.” The fact that AI characters are getting 100k followers, and 600k views per post is going to devalue audiences. I still think “who” is more important than “how much.” Sure, any guy can create a fake viral account, but what do they do with that audience other than the vanity metric of “big audience?” Better to have 1k followers of high-trust high-value people who support your every direction than 100k who want you for a single false thing that you’re not.

In response to a Catherine Shannon group chat:

Friendship is a bond that emerges when you’re both entangled in a circumstance. Inevitably though, an IRL circumstance changes (ie: you graduate, you move, you shift jobs, you have kids), and so you have to make an effort to stay in touch when contexts no longer align. You don’t know if someone’s a friend-beyond-circumstance until one of you shifts phases and both of you have to take responsibility for keeping it alive (not that you’re a bad friend if you don’t — you only have so much bandwidth).

For friends you meet online, there’s no stable circumstance. While there are digital watering holes—blogs, forums, courses, etc.—those tend to dissolve and shift 10x faster than IRL ones. Even though the Internet can connect you with people on your exact wavelength (unlike my neighborhood), there’s no inherent stickiness, and so I’ve learned I need to take more responsibility to start projects or rituals with the people I want to stay close with.

Not UBI, but ABI (attention-based income):

1. AI is not a bubble; the core bottlenecks around any technology is science, energy, and intelligence. Of those 3, intelligence is the most likely to boost science/energy. Meaning exponential AI is something like an acceleration of every other field to their maximum degree. It is not only not a bubble, it is the dead bubble resurrector.
2. People say not to worry about AI job loss (“people have always adapt to new tools!”) but this revolution is different because the invention is not just a tool, but labor itself. Agents will eventually create a supply shock. Sure, new jobs will be created, but they’ll be very specialized around AI research and systems design.
3. Maybe we all lose our jobs, but we also each get access to a 20-100 person digital labor force, probably at very low cost. So while traditional jobs might go away, everyone is suddenly able to be an entrepreneur with a personal labor force at the size of a Series A or Series B funded company.
4. In hindsight, it will seem like Silicon Valley used AI to make their startup culture the prominent culture. The problem is, 99% of startups fail. So even though it will marketed that so many people will be empowered, most might not be able to convert it into financial stability.
5. This means that unemployment could be historically high, and that causes unrest that the ruling class has to deal with. In our case it’s the technocrats, not the politicians in charge.
6. UBI will be shaky to implement. Some countries will have none, some a bit, and a few will give a living wage.
7. Social media companies, will 1) realize attention is the last scarce resource, and 2) populations are rioting, and so a few will start paying users to scroll. It’s a kind of UBI, but conditional on the value you provide on a specific platform.

Banning social media accounts (in Australia), by age (16), feels heavy-handed.

A better solution is to (1) verify identity and age, (2) implement a standard where features have categories and ratings, and then (3), spin up a custom version of that social media site based on age. For example, 5-year olds shouldn’t be limited from YouTube, but maybe there are content blackouts, no vertical video shortform feed, no comments, etc. As you age (/mature), the platform blooms in functionality

This website (WORLD) is a prime example of technocratic euphemisms.

• “The real human network”
• “Proof of human, finance, and connection for every human.”
• “World is being built so every human benefits from the age of AI.”
• “A priority lane for humans.”

Side note: I wouldn’t be surprised if WorldCoin eyeball scanner and the Jonny Ive product merge by 2030; it would be a single piece of hardware that is your assistant, your passport, and your wallet.

Historically, centralization has had problems, and it's easy to see today how it could lead us further into a dystopia. But decentralization, the opposite, could also bring emergent vectors of chaos that could be equally problematic. Neither tyranny or anarchy are ideal. What's required for centralization to work? Tolerance, correction, impeachment, transparency, plurality, data sovereignty, freedom from propganda—all hard, but all solvable things. To create an honest and principled centralization, leaders would need to actively build and implement systems that promote justice over power. That's only possible if citizens have the means to hold them extremely accountable. The original American project was effectively a question of "how do we design a system to centralize power without falling into despotism?" and it's time we revisit that question in a 21st-century circumstance.

Email to Davey:

Thanks for sharing this, Davey. It's a nice encapsulation, an important idea, and I'm sure it's time will come.

I think your nuance on why Related Notes on Twitter didn't work is key. It can't be a side feature, it has to be core. Plexus solved the 90-9-1 problem (90% lurkers, 9% sharers, 1% posters). On Plexus, 100% were posters. This happened because the feed was intentionally withheld until you did the vulnerable thing of shaping/sharing your thoughts. And when you did, you were rewarded with a feed of similar thoughts (an act of encouragement / validation).

As Substack is undergoing TikTokification (my friend sent me a video of his Notes feed, which was all vertical video), I wonder, why can't the Plexus concept exist? Technically, it will be easier, each year, to build something like this, and I wonder what other social frictions need to be fixed for something like it to really work.

There's an inherent tension in a "semantic journaling" app. You want a space that both (1) becomes the place where each person captures their consciousness, but also (2) they want some control over who can/can't see it. I think there are a rare few who are okay being linguistically naked (ie: I have all my logs on my website and I don't care if anyone sees them). It definitely isn't the norm (most people don't even think at the edge, let alone write it, let alone share it). And I have my own limits too (for example, last night my wife reconnected with a middle school bully, and there's much I could write about it, but I lean towards not for the slim chance that someone in her group might find it).

There’s a chance that, at scale, semantically linking is just as unnatural as broadcasting (ie: people will get doxxed / revealed because everything is too interconnected). Maybe instead of having a semantic feed auto-generate, it will spawn a card (with an AI-generated title) that both parties have to accept for the logs to be exchanged/visible. The question is how can you capture the complex psychology of control/privacy in a simple interface/architecture that lets the average, guarded person be maximally unhinged, expressive, and idiosyncratic.

Maybe semi-public journaling will, over time, by 2045 (one of my predictions) become way more normalized, but maybe there’s also a tool that accelerates this (similar to how Uber, AirBnB are both tools that normalized culturally deviant things). Michael

Consciousness is something like the linear real-time awareness of state-switching between perception, memory, projection, abstraction, identity, and action (ie: phenomenological degrees of freedom). Logging is an attempt to rapidly and richly write that process so you can see what’s going on in. It’s possible that, in every moment, you underestimate the depth of what you could write. While filling out the medical paperwork at the OBGYN, each line triggers a thought that could be a paragraph (What’s my % chance of having a thalassemia gene? Was that imagined visual of our doctor as a witch stemming from my fatherhood anxiety? What year were these forms made in? Does someone manually input this into a database? Will that ever be used in some data analytics scheme to have any noticeable impact on my life? Etc.)

Act as if every minute you scroll drains half your day’s potential. It’s not just a break, but a minute you’re not meditating, reflecting, journaling, creating, practicing, etc. A bad break is a minute you’re not reviving your systems. Instead you atrophy your attention and trend in the negative direction. It affects the nature of your focus for the rest of the day. After just 5 minutes scrolling, you’ve shot 97% of your days potential. This isn’t literal, but act as if it’s the case. More so: realize the returns of taking good breaks (of actual leisure).

VEO3 (Google's video generator) gets me thinking that we’re in the “beginning of the end of linear history.” That isn’t as dramatic as it sounds. All it really means is that the year, as a unit, will lose its power to act as a collective anchor. We can all look back to 2001 and remember specific events, celebrities, and media. I’m predicting that the 2030’s could be the first decade where— because of abundance, distortion, and customization at nauseating levels—two people wont' don’t have the same context. It’s the full realization of “electronic tribalism.” We’re already there in some fashion, but we perhaps aren’t considering how much stronger and disorienting it could get. We are only in the early innings of electronic tribalism. The full realization is the collapse of a shared ledger.

Only slightly related: there is paranoia if the Johny Ive announcement video was AI-generated. People are citing that the bottles in the back are changing locations, but that is easily explained through video editing (maybe there was 3 minutes of cut video, and in that time a waiter moved the bottles, which would make it seem like they suddenly jumped). Now that AI video is good enough to be real, everyone is over-interpreting little artifacts. Internet news already has low epistemic certainty, and people already fracture into groups based on how they would like to interpret something, so this will only accelerate the decline of a shared narrative.

How can kids build progressive agency with media? A book library is a good model for other forms of media to follow: they scan a collection, decide to take a single thing out from it, interact with it, return it, and consciously grow or shrink their collection. It's an act of shaping a thing that shapes you. It involves spatial memory. Consider how different this is from a kid in front of a screen with an endless feed of algorithmic cartoons shaping them. The whole wonder of a medium gets flattened. Streaming is our default, but is it worth building an analog media library of VHS/CD/DVD for kids?

To what degree is being personal online a form of commodifying the self into a product? I think it depends if you have freedom to evolve and have contradictory, wide-ranging ideas, or if you are niched-down, monetized, and benefitting the platform more than yourself. It’s cynical to say “cyberspace is a blackhole”; there’ room to be an artist online without getting sucked into the machine, it just requires a diligent adherence to the right values.

Substack has the opportunity to do something fundamentally new with social media—which is to break out of engagement-based feeds. It’s like a megalopolis of good prose; and so a SM post could be a semantic search query that scans notes and essays (that have been sliced into paragraph arrays for more precise embeddings). A note, then, isn’t a scheme to go viral, but a way to actually surface inspiration and people from the network. What’s funny is that Substack search sucks and so does Notes.

Lidlicker's model of communication is about synchronizing internal cognitive models. When writers talk about “expression,” that’s an act of making your interiority legible. A valid and real thing to do. Communication, though, is the additional step of doing some work to understand the cognitive models of others. Through feedback, you build a theory on how translations are succeeding or failing.

Social media notes are so weird because you’re blind to the models of others. I guess you figure out after the fact. But cancel culture might come down to a communication breakdown from rapidly colliding subgroups (TBC)….

"Conditional identity" is the idea that the self is a construction of nested if statements. You can’t accurately describes yourself through a series of variables (“I am kind, fearful, Orthodox, etc.”). Every attribute depends on circumstance. If X, I do this, but if Y I do that. Maybe this sounds obvious, but when you take this seriously, you realize that you behold opposing beliefs that can’t possibly go together. When you embody conditional identity, it’s impossible to be “open-minded” without realizing that you are also its inversion, depending on context.

I wonder if the main bug in the human OS is that there are psychological and sociological incentives to define yourself through variables instead of functions.

You save glucose by automating your thinking. The process of self-discovery requires you to retarget your CPU to optimize its own code (instead of processing incoming tasks with cached CSS). It requires a real-time awareness to re-evaluate your identity logic in every new contexts. It’s tiring, but it’s rewarding.

Socially, it’s hard to have tribal meme wars with conditional identity. If you say, “typically I believe in X, Y, Z, however if A is big enough problem, that I will consider L, M, N, O, P, but if that crosses a threshold D%, then I’ll default to F until it collapses,” it’s much harder for the pack to know whether to welcome or kill you. To fill in these variables, that might look something like “typically I’m libertarian, but if shadow power conglomerates, I’m open to a transparent authoritarianism so long as it’s deployed on a public blockchain.” That is a salad of contradictions unless you really dig into it. It takes far less bandwidth to process “I”m a Democrat,” or “I’m a Christian,” but while cached thinking is convenient, it severs you from reality, and from the solutions needed for complex problems.

On the Internet, everyone judges you through split-second impressions, and there’s no time for them to work through your nuanced takes. You have a choice: to be understood as a false idol, or to be misunderstood as your true self. While there are ways to turn yourself into a 64-bit self without feeling too phony, I think the real lasting work comes from knowing yourself, which means being okay living in realms of paradox, ambiguity, and burdensome complexity.

I think the key reframe here is to not see the uncertainty of conditional identity as a source of fear, paralysis, and alienation, but as the only place where you sit at the edge of becoming.

If OpenAI were to build a social media company, they should have an AI-first approach. Grok is just a dongle. If you think about it, chatbots have so much context on the emerging thoughts of a user. There has to be a way to pluck something from a private conversation and make it a “post.” Now, when others are having chats, they don’t just get LLM answers, but through embeddings, they can surface relevant posts from other users. This is a good example of a semantic social graph, not an engagement one. And consider talking to a chatbot is the primary experience, it beats the lurker paradox—100% of users are creators. Discovery is guided by your prompts. There’s an opportunity to do this right, I just have little trust.

Some notes on AI-based social media.

• During conversation, the chat will compress your dialogue into nodes that you can save or reject.
• You can easily make a node private or public. Private nodes affect your system memory (for future chats). Public nodes are discoverable by others.
• As you chat, it’s not just showing AI responses, it’s using embeddings to pull in related social media posts of other people’s public nodes. These can be displayed as a horizontal swappable carousel.
• Chats are not fleeting, they’re like text threads that update through time. New social nodes can be added via agents or others in the network.
• Conversations are auto-segmented; if you have 3 tangents in one chat, it will separate them (and add hyperlinks).
• AI group chats let me tag 1-10 friends. By doing it, their social nodes are factored into the response. Their AIs are credited when relevant, and at any point they can jump in themselves.
• In an AI social networks, replies should be shorter and skimmable, but can be clicked to be expanded.

I originally found Substack compelling as digital land that you own that's connected to a system of roads to help people discover you. It doubled as a personal website and a social media network. Two years ago, I thought it was doing a fine job. A small feature that I really commended was the ability to toggle on/off the pop-up module for people who weren’t subscribed. This boost subscriptions, but if you find the UX annoying and tasteless, you can disable it.

I thought the whole site would evolve with that ethos, but it seems like things are being forced now. It’s as if someone is building a massive billboard on the road in front of your house. A lot of people are resistant to all the video features, and now especially Reels. Their justification is that “it drives creator subscriptions” (their business model).

Why can’t I toggle Reels on/off? It is so centrally located, and I purposely deleted all the other apps on my phone that have vertical video reels because I think they’re toxic. I’d pay Substack $10 to remove this from navigation (X lets me do this if I pay). But the reason we can’t turn it off is because they know it’s addicting, and they’re hoping it can scale the number of paid subscriptions, thus scaling their revenue. I get how this is more aligned than the advertising model, but if this platform is really about creator independence, we need more customization in how we create and consume.

Just make sure you export your subscriber list every month.

Soda bottle and startups need to be hyper legible, but humans are inherently illegible: multi-faceted, contradictory, complex. In a lot of ways, the medium of the Internet requires you to be hyper legible. But I think there’s an interesting paradox where you can be both. You can simultaneously pass the 10-second test before someone subscribes, but then you can offer a cavern of complexity that some choose to wander through. If you’re over-optimized for marketing, you risk becoming over-optimized, perpetually delivering on an old, stale promise. Instead, you can set the expectation to break expectations.

Studio Ghibli—that's the meme, the vibe of the day.

Interesting how the thing that can go viral suddenly captures and permanently shapes everyone’s imagination. Out of all the possibilities, most people’s experimentation are now influenced in this one odd direction. What else is possible? It’s a weird moment, a race/competition for creative usage. There is stuff capable in 4o image generation that’s undiscovered by anyone.

4o can render a glass of wine to the brim. Another go-to example of the doubter has been eradicated.

• On data sovereignty: I think the age of AI will show us that we want to unlock our data from apps and own it ourselves. We want the ability to organize it and feed it into applications. Only when our data becomes tremendously valuable will we demand ownership.

• Precision embedding: one of the reasons I assume AI apps are bad at embedding is because they’re working on massive datasets and have no granularity or algorithm on how to properly chunk data. For example, my logs won’t have useful embeddings unless they are chopped. If the embedding is scoped to span 3 logs in some, and 10 logs in another, then it’s creating chunks that span a wide array of things—random noise.

• If you generally copy + paste a lot for AI writing or code, you should immediately stop and use Cursor as your main tool. Even though it’s a developer tool, it’s the best way I’ve seen to connect your personal data to an array of models. It’s far from an identity clone, but it’s remarkably easier to load context and personalize outputs. My notes so far:

  1. Easier prompt storage: In ChatGPT, I either, a) lose my prompts, or b) store them in a separate doc. In Cursor, all my prompts live as named files in my directory.
  2. All your data in one environment: In ChatGPT I’ll sometimes copy in 10k words of context and it looks ridiculous. In Cursor, you can have all your data organized in a file structure (principles, essays, notes, highlights, etc.), and simply reference the file in your prompt.
  3. Rapidly load context: When you’re writing a prompt in Cursor, you can use key commands to load in all your context. Opt+Cmd+P > then type the prompt_name & context_name you want to attach. Not only do I not copy and paste anymore, I barely use a mouse.
  4. Auto-scan for the right context: In some cases, you might not know the context you need, so Cursor lets you “Search codebase.” Anything in your workspace is given an “embedding” (LLM coordinates). If you search your codebase, it then turns your prompt into a coordinate too, and will pull semantically related things you forgot about. This means it can quickly scan 10 years of Evernote, and pull in the relevant bits as context.
  5. Patch edits: If you set your files as markdown (.md), it works as a document editor. You can selectively highlight and run prompts just on your selections (ie: edit this one paragraph).
  6. Redo bad prompts: The conversation chain matters, and in ChatGPT, if you put in a bad prompt, you’re sort of stuck with it. Cursor lets you click into a prompt, edit it, and re-process the result.
  7. All the models in one place: OpenAI, Claude, Gemini, DeepSeek, Grok.

Pretentious > pre-tense > context (I realize this is a stretch, but...). Artists are pretentious when they’re frustrated that others don’t have the context to understand what they’ve made. They’re upstream of culture, and so others don’t have the framing to pick up on it yet. There are two reactions, 1) not caring if they get it, and 2) making work for mass contexts (meeting the culture where the OS currently is)—and maybe both are wrong.

Instead, progressive ideas come with the responsibility to make sure anyone is equipped to download the context to get it. The trick is doing this implicitly without being too on the nose. The bootloader has to be secretly encoded. Deconstructing your own lyrics feels too desperate. Maybe it’s something about building an on-ramp for those who care to try to pierce into something.

I imagine there will be two kinds of agent companies: 1) for modular services (ie: generic SaaS coderbots for $1k/month), but also, 2) for replicating identity. In a world where everyone gets access to swarms of modular agents that move at unimaginable speeds, I think those with a “replica” will have the edge. A replica is like a centralized manager entity—one that has your corpus of principles, lessons, memories, ideas, taste, libraries, etc.)—to coordinate between all the agents.

If this comes true, it brings new leverage to writing out your thoughts in detail. Maybe for every 10k words I write, my agent swarm is 2x more aligned to my specific goals. However, it also inverts the norms of data privacy. There’s a trade-off: the more you use writing for distribution and audience-building, the more clonable you become. It’s causing me to think harder about when a piece of content is public/gated/private.

So if models, apps, and infrastructure become too cheap to matter, then the only thing left that’s scarce is personal data / personal perspective. You’d think Notion/Google are positioned to solve this, but it’s interesting to watch people use Cursor as their PKM system because it’s so easy to connect personal context with an array of models.

Just read several hundred Kanye tweets. I wonder if you take all his tweets since Feb 6 / the Super Bowl, and take the time to meticulously arrange them like a developmental editor would, might it make sense? If you cut 25%, removed the ALL CAPS, and made a prose essay out of re-arranged Kanye tweets, might there be a semi-coherent perspective that’s not completely insane?

One of the key places you find slop is in product demos for emerging AI apps. These things shill writing features without having even the slightest grasp or care about writing. It’s like they’re assuming viewers won’t even read the examples:

Check out this example what was used for an “Improve writing” button. The original said:

“Through collaborative efforts, we aim to create an exceptional work of conversation."

It changed it to:

“Collaboratively, our intention is to craft an outstanding conversation, where I bring forth notable viewpoints.”

One thumbnail said “no editing skills or experience needed,” as if skills and experience were just nuisances, not worthwhile things to pursue.

Now, a Summarize button changed “new branding must focus on our core values and mission” to “new branding should reflect our core values and mission.”

This thread has 20k likes.

9 of the 12 comments: “Great,” “Love this,” “Good list,” “Great,” “Awesome share,” “Great list,” “Great share,” “This is great.” Two more promoted a product, another started with “Most don’t know that…” There was one at the bottom marked as “Probable spam” that says “That’s Awesome.”

You can’t know if the commenters are real. Who even knows if the poster is real? Are these his own bots commenting to boost the algorithm? Are the demo videos AI-generated?

The Jitters. How will this go? I think I’m imagining a storm of 100,000 haters. Or, at least, will the required process of editing and memeification invert what I believe?

Essay Architecture is all about complexity. In a way, it doesn’t require me to translate ideas or make them digestible and catchy; it’s about training machines with infinite patience to create a piece of software that is—to the user—simple and profoundly useful.

I guess this is a test to see how well the idea can be meme’d. Is this something you can enjoy listening to while you wash the dishes? It’s a hard constraint. A fun problem.

To tie this in with David’s tweet on how managers prompt people, an interview is something like a host promoting their guest to package their ideas according to the taste of the host. So in some ways this will be like my ideas through David’s vision for his channel.

There's a foolishness in embracing slop-poetry over real poetry. Yeah, you can ask 100 casual Americans who haven’t put any time into reading or understanding poetry, and they’ll all prefer LLM outputs because they have a tendency for explicit clarity, but maybe it’s worth having some opaque mediums. When it comes to essays, I think accessibility is important. But poetry is supposed to be cryptic and mysterious. It’s a feature that it haunts you and reveals itself to you over decades. For every thousand who balk at it, it finds one person to completely infiltrate. I’d be curious to see this same study among poetry readers.

Feels like multiple AI writing apps nonchalantly launched a "persona cloning" feature. Just upload a Substack archive URL and you can mimic your voice! (Or, someone else’s?) For now, AI mimicry is pretty hollow for anything serious. But what about when it’s not? I think we’re heading towards a paranoia where more people paywall their writing to prevent scraping.

20th century college/career:

• linear
• singular focus
• credential based
• one income source
• work in private
• up front learning
• declare major up front
• exams is proof of success
• hierarchy of teacher/student
• go into debt to attend

21st century college/career:

• non-linear
• polymathic
• portfolio based
• multiple income streams
• work in public
• continuous learning
• emergent major
• market is proof of success
• everyone teaches and learns
• get paid to contribute to school

The only personal brands that survive in the age of AI are the ones that are unoptimized.

"Blockchain bunkers in the psychotic AI swamp of the open Internet”

Death of the open Internet (4 phases)

1. slop, ambiguity, more confusion
2. more than 50% is fake; trust flip
3. disturbing distortions (schemes, attacks)
4. a victim, phished, money, humiliated

This could lead people deplatforming themselves. It could be dangerous to have any Internet presence (text, video, social graph). The open Internet could turn into that unsafe darkweb we feared in the 90s. When infinite is leveraged, all the good doesn’t matter, because evil also scales. And so a few bad imaginations without guardrails creates a psychotic social virus. This is already starting to happen (ie: those two kids who generated porn of all the girls in the school). Imagine Facebook, Twitter, YouTube all getting destroyed.

The solution here is identity verification. We already have KYC (know your customer) for financial institutions, and the same might happen for media. There might be new social media sites where you can only post if you’re verified as a human. I think this is more defensive than “watermarks.” If there’s a pollution leak, it’ll be linked back to you and you’ll be banned. So this idea of needing a government ID to use the Internet might actually be necessary. The question is, can we do this without being dystopian/totalitarian? Perhaps the only way to execute this and protect privacy is to use a blockchain. It would create a method to verify citizenship without their details being exposed.

I discovered today that I could write a journal entry that is intentionally illegible to my future self. I’m writing letters, but blending them so loosely that it would be undecipherable to anybody. This means I can write fast and say whatever I want, and nobody will be able to return to this, which opens up a certain kind of confessional/abstract valve. It’s a way to use “writing as thinking” but encrypted in a way so the shadow can come out.

Politics prevents REM work. If you’re tuned into politics throughout day, it changes the depth of your flows. It’s like having a bad cough that prevents you from getting a good sleep. With a clear mind, you can sink into your project for hours and achieve deep levels of concentration. So even if you only read politics as 5 minute breaks throughout your day, 12 times, the difference is 8 hours of deep work for 8 hours of work with running open loops.

Marketing emails from software comes across to me like polite begging. The desperation is veiled. They speak in modern idioms, wishing my family well while using exclamation points. Email broadcasting is already unnatural, and so instead of pretending to masses with the pleasantries of a 1:1 coffee date, have the courage to be yourself.

And isn’t “being yourself,” the thing a company can’t be? It’s inherently not a self, but a conglomerate. Who is speaking? Who is “team”? A more honest way is to “share the process,” to give people the glimpse of what it’s like in a company. But still, I think there’s something corrupting about letting a market dictate your decisions.

Of course, that’s what venture requires. But we’re entering a paradigm where software can be so easily made, that all of our ratios can flip. Someone can make tools for themself. If I’m making a tool for myself, I can not only show you the process of making tool, but the things my tool makes. That’s the ultimate proof. I don’t want to just share screenshots of text-editor features and shill you benefits; I just want to share the fruits.

If the founder isn’t actively engaged in the product, then why should we believe them? And so, that might be how to make a niche products long-term sustainable. Serve yourself, and let the small selection of overlapping others have access to it.

We should start to think of social media regulation like we think about drug regulation. It would be insane to ban or legalize all drugs. Instead, we measure specific properties, understand the LD-50, and make case-by-case recommendations and rubrics. Social media is tremendously new, and it’s not surprising we haven’t gotten this right (consider how Victorian times used opium to treat a cough).

I think what we need for social media is:

1. Age verifications.
2. Specifications for algorithms / features / interfaces.
3. Proper gating.

I think it’s wrong to say that anyone under 16 shouldn’t use social media as a whole. I’d let a 9-year use Facebook circa 2005 (pre-feed), but I wouldn’t want a 20-year old using Instagram today (vertical-swipe predatory algorithms).

If Feeds and Likes create a digital popularity app that fuels the already intense middle school experience, then maybe kids should have a version of social media with different functionality.

I need to write out a post on why movies are culturally/psychologically better than shows, and how you should be extremely careful about the shows you welcome into your life. Unlike movies, shows focus on small dramas between characters, and are unresolved episode to episode—they have a constant pull on your attention, and are too uneventful to stay in your memory.

If you’re concerned that an AI will soon write better than you, it’s worth asking, why are you writing? (I mean this sincerely, not as a dig). There are so many intrinsic joys and values in slowly and meticulously crafting sentences. It doesn’t matter who’s watching, what it’s for, what technology is available, what generation I live in, or if anyone else likes it.

Through writing and editing, slowly and manually and sometimes painfully, you discover what you think, who you are, and how you express yourself. If you’re not in the pocket with language, there’s no transformation, no evolution, no humanity. It doesn’t matter how many billion parameters Claude Sonnet 3.5 has, or the prompt library you bought for $49, or whether you’re on Pro tier—it matters if you’re willing to sit alone with your own thoughts for more than 5 minutes, and for hundreds of hours more. It matters if you’re willing to shed your identity over and over.

I get the source of panic: soon enough (2025? 26?) AI and its midwives could trample out 100% of humans at the “art as commodity” game—maybe that’s a good thing. It wasn’t a great game to begin with. It distorted all the incentives of human expression. It created monsters and beggars. ASI means that the transactional, mercenary, instrumental skill of “content writing,” will no longer have value. This will create a strange, noisy Internet, and I sort of hope it drowns in pseudo-drivel so that in its ashes we can form a new one that’s less like American Idol (commerce was illegal on the Internet until 1991).

Now for the contradiction: I make money from writing online. I have a Fellowship Grant which gives me at least year of security. I have, relatively, a good number of paid subscriptions, which is currently around 1/4 of a full-time income. I live in New York (not by choice, by birth). As much as I hope and believe that I can sustain a full-time income via writing (the thing I love), I also (1) accept that ASI could wipe out the whole cultural appetite for writing and send me back to a FT role at an architecture firm, and (2) even if “Essay Architecture” is a success, I need to protect my non-legible non-optimized, non-practical writing impulses. (Ie: I hope that on “launch day” of my app, I spend 2-3 hours on a typewriter essay about thumbs, or something equally trivial, unrelated, unproductive).

So while you can make money and build influence via writing, you definitely don’t have to. Maybe it’s about to lose some of its economic function, but we shouldn’t underestimate its cultural/democratic function. If you look back through history, most of the famous essayists had other jobs—they were lawyers, doctors, publishers, physicists, architects (Frank Lloyd Wright). Essays just flowed out of them. They couldn’t help it. I really think essays are the most democratic medium of the arts. Unlike architects, you don’t need millions of dollars to start. Unlike novelists, you don’t need years of focused attention. Unlike musicians, you don’t need to learn abstract chord languages and train your finger muscles. An essay is democratic because anyone, in their own language, in a day, at almost no cost, can engage with composition and meaning-making, and share it with their culture, even if it’s a single friend. It might be our most important form of leisure.

So maybe this approaching techno-apocalypse is an opportunity. It can scare us inward. It can shatter the dream that we can be a Mr. Beast, a niche hero, a perfectly legible and distinct human wrapped in plastic. The irony is that by going inward, you might just find the thing that has outward appeal. The best kinds of extrinsic opportunities aren’t the ones you engineer, they’re the ones that knock because you tinkered with your garage door open and accidentally found a new source of gravity.

To finish with AI: I’m all for it, so long as it is my assistant and not my sentence jockey. I will never replace my document editor with a chat pane, but I will invite agents to live in my margins. I’ll let them watch my words as they form, and then offer up research, feedback, and disagreements, all of which I’ll have to manually integrate. Slowness is the theme. You don’t change without immersing yourself in an essay. AI might give you insights, but you know they’re illusions if you’re not wet—if you’re not even in the pool. You can automate the outputs, but you can’t automate the watering of mind and blooming of character.

It’s totally normal to not know what to expect to happen from writing. Writing generates opportunities you can’t expect. My method is to keep adding things to my website, regardless if they make sense in the moment. Retroactively, you can connect the dots. You can use tags to link posts. So if I wanted to suddenly get into a conversation with a VR founder, I could send them a /virtual-reality sub page.

It’s only a % of what’s on my site, but I can slice specific cross-sections through my full range of interests (which is illegible as a whole). Since I have permission to explore anything, I’m pretty generative, but I've found that when I set specific, restrictive rules, I block myself.

Over time (ie: every 6 months to a year) you might gain clarity on if it’s worth converging around a single direction (ie: band reviews vs. teaching music vs. something unrelated).

Astronaut.io shows a feed of random YouTube videos uploaded in the last week that have basically 0 views. It’s a collage of everyday life.