michael-dean-k/

p (doom) is a terrible heuristic to think about AI safety. Researchers cite this term when confessing they think there’s a 10-30% their technology kills us all. The issue is that p(doom) is rarely defined, treated as a binary thing. Either the world is liquidized to gray pulp by nanobots or it’s utopia with free electricity, forever.

My main pushback: there are far less severe outcomes from AI mismanagement that would be absolutely tragic. My p(doom) for an extinction event is 1-2%, on par with nuclear weapons. But my p(sub-doom)—the probability that AI could escape containment, cause damage, and be irreversible—is over 50%. A coin toss.

I don't blame us for being so imprecise in existential destruction (it's not exactly productive), but it's worth realizing our blurring of scale and magnitude. For example, there is a 10,000x difference in lives lost between Hiroshima and Nagasaki and Terminator’s Judgment Day—that's the difference between a small stove fire and all of Manhattan on fire. Yet, if an agent swarm produced something on par with the A-bomb, we’d all freak out and immediately regulate AI. But for some reason we don’t consider a “little apocalypse”; we jump straight to extinction and then ridicule how ridiculous that is.

What we need is a doom spectrum: d(1) - d(10). This was not fun to research and write, nor do I imagine it being particularly enjoyable to read, but I imagine a doom framework could help clarify this week's paranoia (which is so extreme now that even my mom is overhearing news of civilization extinction on the local news). The most urgent danger right now is the miscommunication between doom prophets and capitalist deniers, each of who only speak in d(8)s and above.

• d(1) is an isolated internal incident, something that doesn’t escape its container, but gives fear of larger exposure. Think of the 2025 studies that showed AI trying to avoid shutdown, or some protocol mismanagement in a biohazard lab.

• d(2) is a local real-world incident, something that escapes a sandbox and affects only one or a few entities. It’s possibly ignorable, with minimal consequences, and likely reversible. This is where the recent Hugging Faces hack sits, along with its parallel hacks and social manipulation. (The leap from D1>D2 is disorienting, because society at large doesn’t see the spectrum beyond D2 in any granularity, and so most assume any containment breach escalates immediately to a D9/D10).

• d(3) is a serious local disruption, a mismanagement of technology in a single location that has real consequences. Consider Chernobyl: a death count (30 immediate) and a long-term stain on surrounding land, followed by regional/global concern that might lead to regulation or reform. In terms of AI, imagine an agent swarm that hacks local infrastructure in pursuit of some narrow goal, in the process disrupting traffic, hospitals, and electricity, alarming all locals, requiring military and cybersecurity intervention, and taking days or weeks to resolve.

• d(4) is a local catastrophe, something with a high death toll that shocks the world, changes the course of geopolitics, and makes history. In this tier falls Pearl Harbor, Hiroshima and Nagasaki, and 9/11. It only immediately threatens locals, but has global shockwaves. An AI-equivalent here would be a rogue actor using a frontier-capacity open source model to design and deploy a bioweapon throughout a city.

• d(5) is an inter-regional dilemma, something that affects multiple areas at once. A possible example here is the Vietnam War, basically a proxy war that claimed 3 million lives. This might manifest as a new style of warfare, where two countries use sophisticated AI attacks against each other in unconventional and partially uncontrollable ways.

• d(6) is a world-wide dilemma, something like COVID, which took 7 million lives, where the entire world is locked into a new paradigm. In the prior five tiers, the breach is usually isolated to specific areas, but this touches everything. The parallel here to a biological pandemic is a cyber pandemic. Both involve containment, gain of function research, etc.—although this would be inverted: in COVID the virus was outside and forced everyone on the Internet; with AI, the virus infects the Internet and forces everyone outside. The open web could evolve into a dangerous “dark forest,” where it becomes a liability to have any public presence. Whether this happens through an agent swarm that exfiltrates it weights, or malicious actors, it could be irreversible: the only way to escape the paradigm is to “shut off” and rebuild the Internet safer (a much larger and more consequential version of how Hugging Face regained control). In this case, casualties might come not from the cyber pandemic itself, but from the consequences of losing Internet, along with the supply chains and systems that run on it. (This loosely maps to the movie Colossus: The Forbin Project, where ASI takes over all governments—there’s a version where no death is involved, but it requires everyone to submit to a machine autocrat.)

• d(7) is a civilizational flashpoint, at the scale of World War 2, with 70-85 million dead, seriously affecting the existence of countries, birth rates, and all of culture for generations to come. This is roughly the scale of the “Butlerian Jihad” in the Dune series, a war between humans and machines, which Fable estimates caused 70 million deaths.

• d(8) is a near-extinction event where most of humanity is wiped out. This has happened in history, with things like the bubonic plague, where 30-60% of Europe is wiped out. In the Terminator series, AI launches a nuclear holocaust on humans, killing 51% of the population, forcing the rest to survive through a post-apocalyptic world in small bands trying to rebuild over decades and centuries.

• d(9) is full human extinction which means the entirety of the species is wiped out, and there are no remaining humans. Throughout Earth’s history, there have been a handful of climate events that wiped out over 50% of species, but not all species, and so over millions of years, new forms of life can emerge on Earth. This is an AI extinction scenario where nanobots kill all humans, but not all life.

• d(10) is a sterilizing event, like an asteroid or gamma-ray burst that kills every organism on a planet. It means no future life can emerge. This is the scenario presented in If Anyone Builds It, Everyone Dies. Not only does it kill all humans, it plates the entire planet in data centers to maximize compute. The “paperclip maximizer” takes this even further, claiming that an AI trying to maximize paperclip production will attempt to harvest all the metal in the universe.

Yudkowsky and Co. think we go from d(1) to d(2), then straight to d(10); this matches the current discourse, given we hit d(2) this summer and are now talking about extinction. They think this jump happens because an AGI/ASI that can reach d(3) is smart enough to know not to expose itself, and so it will acquire resources in stealth for years until it knows it can execute a d(10) smoothly. Maybe this is already ongoing. Unlikely, I think (I don’t think we’re as recursive as people say right now).

Rhetorically though, maybe the d(2) > d(10) framing isn’t a bad thing. If they can use recent events to make the theoretical argument that AI safety matters, then we act with a response as if a d(3)-d(6) already happened, without having to lose any human life.

But actually, what they’re doing is pretty ineffective, because even if the d(2)>d(10) jump is inevitable, it’s too extreme, too farfetched to be believed. I think a lucid and airtight argument for the d(6) “cyber pandemic” would be believable, emotionally resonant—considering we’re not yet a decade beyond COVID—and likely to trigger regulation.

Instead of tapping into the “everybody dies” angle—which is too unpleasant and helpless for anyone to consider at length—there could be more value in the “irreversibility” angle. As in, once an AGI or ASI swarm floods the Internet, we can’t ever reverse it without destroying the Internet. There’s a world where it becomes a permanent autocrat, where it doesn’t exterminate us, and instead, uses 1% of its capacity to micro-manage our nations and lives as it pursues whatever it sets its machine heart on. Similar to how we try to preserve species for reasons of stewardship and scientific curiosity, an ASI would be able to effortlessly preserve us as highly-spoiled pets.

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.

What and how you read should heavily depend on what your goal is. Outputs shape inputs. When someone insists you go back to read The Great Books, in order, in their entirety, they're giving you bad advice. It's not that those books aren't great—I hope to read Paradise Lost and Dante's Inferno and Finnegan's Wake and the Odyssey before I die— the problem is it's too generic a suggestion. To spend thousands of hours deep in the canon will obviously change you, but that's equivalent of throwing a beginner into the depths of the Atlantic Ocean, hoping they'll figure it out, with no sense of what their goals are.

If your goal is to write essays (every day, week, or month), then you're reading diet should look very different from a philosopher, professor, or researcher. You might not need to be a professional reader, but you should still strive to be a serious one. 3-4 hours a day might not be feasible, but 30-60 minutes per day through an intentionally selected list of sources will slowly build maps of material to fuse into your work.

If you're an essayist, you read so that concepts, forms, feelings, and words are always within reach from an idea of your own. It's no use quoting Aristotle from memory if you can't bend Aristotle to augment an original idea of your own.

It's time to make a syllabus. I've been guilty my whole life of haphazardly reading books and essays as I come across them, but now that I'm over 5 years into writing essays, I feel it's time to be more intentional. This essay is the artifact of me mapping out what, why, and how I'll be reading in the next 2-3 years. I've broken it into four practices: reading for ideas, reading for craft, reading for words, reading for feeling.

Reading for ideas

Since essays are so personal, it's very possible to draw from nothing else than the bank of your own life experience. Memory is absolutely one realm of material, but also, it helps to pull concepts from the world around you, in your time and in all times before. Anyone is exposed to some sliver of culture, and I suppose you could just rely on that. But there's another path which involves actively educating yourself.

Before I dive into the details of philosophy or history, I'm going to build a map. I want to go wide, not deep, because my existing maps are too fuzzy. ie: Who was Thomas Aquinas? Who influenced him, who did he influence, and could I hand write an essay on three of his big ideas? Until I can do that with 100 figures from antiquity to now, all interconnected in a web, I'm not prepared to dive into any Great Book. It would be a tremendous waste of time, for me at this moment in my life, to read The Leviathan by Hobbes in full, especially when I could read 30 pages on it from Alan Ryan, a philosopher-curator, whose prose is 400 years more modern, and who can contextualize old ideas into the full history. In the time I could finish one book from Hobbes, I could read Ryan's entire textbook and know 30 different thinkers at much higher resolution than I know now. By the end, I'll have an updated index on the history of political philosophy, and maybe I'll know that—based on my current writings—it makes more sense to dive into Rousseau in full.

How would my mind be different if I found and read the best curator across every field?

There's a specific kind of book I'm looking for to update my maps. It's not a textbook. It's similar in it's encyclopedic range, except it is slanted by a thesis, animated through a fervent voice, and concerned with the psychology behind the person known for an idea (instead of just biographical facts). Each chapter focuses on a figure for 25-50 pages, which feels like the right level of immersion. It might take 2 hours, compared to 20 hours for the source, and 20 seconds for Claude. While AI can surface historical ideas perfectly suited for your working draft, the problem is you outsourcing your recall. The recommendations are mechanical, impersonal, and worst of all, disembodied: you can't do it in your own head. By reading a sharp longform essay on Aquinas, his ideas will crystallize in my head and load into my subconscious; I'll know when he's relevant to my ideas at the layer of thinking itself.

The nudge to read all of Aquinas from scratch, on principle, is like asking a software developer to derive Internet standards from scratch instead of using libraries and plug-ins. For any thinker that matters, there's at least one person who spent a good deal of their life deeply understanding the source and distilling the concepts for you.

I'm going to share my working list, but the main caveat here is I'm not going in any particular order, and it's not necessary to read cover-to-cover. In any given month I'll be reading 1-2 chapters from 10 of these 24 books. In 45 minutes per day, I can get through most of this by the end of 2028 (2.5 years from now). Everything was published within the last one hundred years, and the whole thing costs $327.

• Literature: Harold Bloom, The Western Canon: The Book and School of the Ages (1994) $11
• Essays: Karshan & Murphy, On Essays (2020) $27
• Architecture: Kenneth Frampton, Modern Architecture: A Critical History (1980) $15
• Arts: Kenneth Clark, Civilization (1969) $3
• Design, Christopher Alexander, A Pattern Language (1971), $34
• Poetry: Nineteenth Century American Poetry (1996) $9
• Politics: Alan Ryan, On Politics (2012) $15
• History: William & Ariel Durant, Lessons of History (1968) $14
• Economics: Richard Heilbroner, Worldly Philosophers (1953)
• Science: Bill Bryson, A Short History of Nearly Everything (2003) $11
• Mathematics: Morris Kline, Mathematics: The Loss of Certainty (1980)
• Technology: Walter Isaacson, The Innovators (2014) $17
• Media: Denis D. Cali, Mapping Media Ecology (2017) $47
• Cosmology: Brian Greene, The Hidden Reality: Parallel Universes and the Deep Laws of the Cosmos (2011) $5
• Philosophy: Bertrand Russell, History of Western Philosophy (1945) $8
• Psychology: Morton Hunt, The Story of Psychology (1993) $12
• Antiquity: Arthur Hugh Clough, Plutarch's Lives (Volumes I and II) (2018) $10
• Religion: Huston Smith, The World's Religions (1958) $12
• Christianity: Fifth Edition, The New Oxford Annotated Bible with Apocrypha (2018) $16
• Orthodoxy: Kalliostes Ware, The Orthodox Way (1979) $10
• Mythology: Mircea Eliade, Patterns of Comparative Religion (1949) $17
• Esoterica: Manly P. Hall, The Secret Teachings of All Ages (1928) $1
• Hermeticism: Franz Bardon, Initiation into Hermetics (1956) $20
• Death: The Oxford Book of Death (1983) PDF

You'll notice that all the links above are Kindle. This is because I want to have my highlights as atomic markdown files. The goal is not to read, but to write! Mapping and reading is just the setup so that I can read through and find highlights that spark original reactions. Montaigne's whole idea was to talk to his library, to be in conversation with the past through his books. And so the goal here is not to finish X books per year, but to produce original material. This is close to sounding like a Zettlekasten, but I should clarify that I don't plan to meticulously arrange my private highlights. A highlight is simply a prompt for an original paragraph that will immediately live on my website.

Other ways to read

I haven't spent as much time mapping out the other three modes, so I'll cover them briefly below, knowing I'll expand them later.

• Reading for craft: If you writing essays, then reading them is how you learn through osmosis. It's where you pick up on the patterns on form and voice, consciously and subconsciously. My thinking here is to pick one essayists per week, read as much I'm inspired to, and move on. It's important to cycle here, because hanging too long on any one writer might lock you into a particular influence without realizing. I'm planning a summer syllabus for Essay Club so we can do this as a group.
• Reading for words: Two years ago, I got really into reference books: dictionaries, usage dictionaries, the thesaurus, etymology, and even specialized dictionaries (on architecture, philosophy, scientific concepts). Sometimes I'd read cover to cover (futile), and others I'd practice words in ANKI. Expanding your vocabulary is seen is a pretentious thing to do today, when so much is geared towards simplicity and accessibility. Won't a rare word alienate the average user in your audience? No, because in the right context, ambitious words can increase the resolution in how you describe something. There's a joy in searching for words, but again, this comes back to returning to them repeatedly until it's actually coming through your prose.
• Reading for feeling: Novels and poetry are less about collecting bits to synthesize into your work. This is more an act of expanding your understanding of how words can make you feel. Less about analysis, more about immersion.

The AI takeoff hysteria is hard to avoid these days, and I'm realizing we don't have clear distinctions between AGI/ASI. I wanted to revisit an old framework of mine to see if anyone finds it helpful (and if it's worth developing). There are some existing classification frameworks, but they're low-resolution. My basic idea is to break AI into three eras: ANI (narrow intelligence), AGI (general intelligence), ASI (superintelligence). Then, you can break each era into 3 tiers. You only shift from one tier to the next when you make breakthroughs across different criteria (let's say, (a) generality, (b) transfer, (c) autonomy, (d) learning, (e) self-modeling). I think the last few weeks are the collective hype of us all realizing we're shifting from AGI-1 to AGI-2. It's exciting/scary, but I think the paranoia mostly comes from not realizing how big the gap is between AGI-2 and ASI-1. (Spoiler: ASI might arrive slower than we think.)

ANI-1 is scripted logic, the lowest form of "artificial intelligence," basically Goombas. ANI-2 might cover Google Maps or AlphaGo, intelligences that excel in a single function, traffic or chess. Siri is ANI-3; even though it feels broad, it really uses voice to route you to 20 or so pre-defined tricks. The chasm between Goomba and Siri is similar to the chasm between early-AGI and late-AGI. ChatGPT and the multi-modal models that followed, capture AGI-1, a single neural network that can do basically anything, even if it sucks: essays, songs, video, code. The newest models (and their agentic harnesses) are feeling like AGI-2. They're significantly better at coding, can run for hours at a time, and are starting to make contributions to machine learning itself.

AGI-2 could last a couple years. As agentic AI matures, I'm sure there will be a few "takeoff" scares, but they'll probably feel more like a flood of a trillion midwits than real ASI (still, that could be enough to break the economy/internet). While we went from AGI-1 to AGI-2 through data, scale, and engineering, it seems like we'll need research breakthroughs to get to AGI-3. It won't be through scaling alone. Whenever and however we get to "human complete" intelligence, the apex of AGI is a single agent that is a master of all human domains, a Nobel Prize winner in every field at once, seamlessly transferring knowledge between them, unlocking a cascade of civilization-altering inventions.

As crazy as AGI-3 could be, it still isn't superintelligence. That has its own era, and the chasm between early ASI and late ASI will be as big a gap between the chatbots who can't count the R's in strawberry and the agents that cure cancer. We can only really speculate on ASI (because it would be truly alien), but we can imagine it as step changes in recursion, scope, and complexity. Imagine ASI-1 as an agent that, as it's working, can infer its own limits, and self-modify its learning paradigms in ways we can't understand. Imagine ASI-3 as something that can monitor reality in real-time, and, reconfigure its hardware in real-time (some hydra of graphics cards, quantum computers, and neuromorphic wetware) to run simulations at unfathomable scales in unimaginable fields, running on a hardware stack so big we have to put it in space and run it on fusion. This goes far beyond my ability to not bullshit, but I think something as insane as this, thankfully, is still far away, which points to the real question nested in my framework:

Could the rise of AGI/ASI be linear? People gravitate towards "AI will plateau" or "the singularity is imminent," but the conservative middle ground is more boring: linear progress. Maybe the exponential advances are real, but so are the extreme frictions of research, infrastructure, and social effects. If AGI-1 arrived in 2022, and AGI-2 arrived in 2026, maybe we'll keep ascending tiers in 4-year intervals: AGI-3 in 2030, the first true "superintelligence" by 2034, and ASI-3 by 2042. This shift from AGI-1 to ASI-1 (12 years), is considered a "slow takeoff" scenario, even though the ANI era took around 70 years. If we zoom out to the scale of a human, linear progress will still feel like centuries of change all in a single turning of generations.

I like the word experimental because it fuses two halves of a process we don't usually link. What we typically mean is divergence, deviance, tinkering, norm-breaking. Weird stuff. Think avant-garde John Cage soundscapes where he makes music with only kitchen appliances. But also, the word points directly to the scientific process: to run an experiment means to set boundaries, gather insights, and test a hypothesis. Either mode alone falls short. Endless mutations burn you out, and rigid systems can't take you anywhere interesting.

Many of the original experimental artists were scientific. Kandinsky didn't just make abstract shapes, he developed a systematic theory on how colors/geometry provoked specific feelings, and then at the Bauhaus he used questionnaires to test which of his theories were true. I don't know exactly when this happened, but as weird works became mainstream, the word shifted from a process to a genre; the way it was made mattered less than the fact that it was unusual.

Experimental drifted into a contronym, a single word that contains opposite meanings. The power in the word comes when you re-unite both halves, entering strange territory with an analytical eye.

Over the weekend I had a +101 fever, and so I was banished to an airbed in the attic to not infect the baby. Wrapped in blankets, I found myself in a sequence of near-identical “fever dreams.” Before this, I hadn’t thought about the phrase much. As a metaphor—"the president’s plan is a fever dream”—it implies a delusional desire, but real fever dreams tap into a different thing: for me, they’re about absurd procedural loops. I found myself deeply concerned with the layers of blankets around me: I had the urge to unfold them, visualize each one as a heat map, extract the cold parts with a boxcutter, restitch them into a new blanket, shape this new perfectly cold blanket into an animal sculpture, and then sell it on Etsy. I can’t remember the sequence exactly—it only made sense on the inside—but it was a cold-side harvesting operation for sure. I’d wake up and realize, oh, this whole scheme is stupid and pointless, and now that I know this I can sleep peacefully. Yet as soon as I went back under, I slipped back into this incoherent non-problem. It’s not uncommon to fall asleep and re-enter the same dream, but with a fever dream, I find that all I can do is return to my miscognitions, 5-10 times, until the fever breaks. It’s not scary, but repetition can be hellish (like the Teletubies DO IT AGAIN! sequences). My guess is that an overheated brain that’s deprived of REM will linger on thoughts it can’t digest. It becomes a type of lucid dream, a lame one with no visuals, where awareness of the loop can’t break the loop. There are probably situations better suited for the fever dream metaphor, but I can’t think of them now. Until then, no takeaways other than don’t get a fever, and if you do stay away from blankets.

These are words I recognize, but probably don't the nuance well enough to integrate into my own prose: countenance, prodigious, clamor, visage, abate, undulate, venerate, incredulous, traverse, repose, lurid, languid, sagacity, tremulous, odious, pallor, stolid, wistful, prostrate, remonstrate, palpable, amiable, portent, importune, expostulate, vivacious, despond, doleful, pervade, pensive, procure, abject, austere, magnanimous, oblique, sallow, ignomy, resolute, furtive, fain, genial, mien, billow, confound, wan, indolent, reproach, morose, antipathy, alacrity, vestige, verdure, rebuke, inexorable, din, fortnight, abash, imperious, swarthy, impute, appellation.

Saw a post that says “‘Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo’ is a grammatically correct sentence.”

There are 3 usages:

• Buffalo = a city in New York
• Buffalo (noun) = a bison
• Buffalo (verb) = to bull

So basically, “NY bisons (that) NY bisons bully (also) bully NY bisons.”

Three separate groups of bison from Buffalo, NY all engage in an endless cycle of bullying.

Put differently, “Bison from Buffalo [1] whom (other) bison from Buffalo [2] bully, will (also tend to) bully (yet another group of) bison from Buffalo [3].

We need better nomenclature. AGI/ASI is not working; “general” and “super” are obnoxiously vague. Proposal:

AGI > AAI (Artificial autonomous intelligence) … GPT-4 was arguably “general” in the sense that a single model can write, see, and hear; and do anything from poetry to calculus to history to coding. It is by no means narrow. Google Maps is narrow AI. Grammarly is narrow AI. This whole chatbot era should be “AGI,” which means that the thing coming is “autonomous intelligence.” It is not a tool or co-pilot, but it’s more like digital labor. You can give it a high-level goal, and it can 1) execute the full range of tasks, 2) 100x speed, 3) intelligently reshape embeddings into real-time hierarchies so that it’s able to procedurally load in and compress context. This doesn’t just come with better models, but with UI and engineering innovations, if not entirely new paradigms for transformers or training.

ASI > ARI (Artificial recursive intelligence) … The fact that Zuckerberg pitched “super intelligence for you” is an Orwellian marketing ploy. Super-intelligence is not “for you.” Super intelligence is shorthand for “something that is way, way smarter than us,” and you achieve this when you teach an AI model to think, form its own algorithms until it accelerates to something this is far beyond our understanding, and likely to become a force of nature with its own goals. Engineers are confident they can build “God in a cage” and reap the benefits, and this is the prime, archetypal, near-biblical example of technological hubris. (Maybe integrate into this paragraph that Zuck has a thing for trying to dominate words, like “Metaverse”).

Important note: “machine consciousness” is separate from AAI and ARI. Something can be recursively intelligent and still not be conscious, which is actually, unbelievably dangerous (because it will fall into attractor states, and optimize for narrow, malformed goals in extremely capable ways). I’d argue that consciousness has an architecture, whether human, rabbit, or robot, and we should be urgently trying to find the parameters of machine consciousness, because if we AAI/ARI have no ability to reflect, question, doubt, and revise, we will, as they say, all turn into paperclips with paperclip children.

Punctuation often comes under assault. Kurt Vonnegut in 2005: “Here is a lesson in creative writing. First rule: Do not use semicolons. They are transvestite hermaphrodites representing absolutely nothing. All they do is show you’ve been to college.” Recently, there's been a wave of em-dash hate. Since chatbots tend to aggressively use them (multiple times per paragraph), any writer who includes them is now accused for having AI write for them. But I trust your writing less if you don’t use em-dashes.

First, it shows you’re not fluent enough in basic punctuation to properly articulate the thoughts in your own mind. I mean, sure, you get a lot done with just periods and commas, but punctuation marks are like visual aids that give you more precision in what ideas mean and how they are connected. I see em-dashes and parenthesis as siblings (of inverse function) that work together to help give structure to your emergent thoughts. I often find myself—mid-sentence—wanting to add details and embellishments; if they don’t fit into the structure of that sentence, I can contain them with punctuation. Both the ( ) and the "—[ ]—" let you inject detail into a sentence. They are “innies.” They either clarify or complexify.

These innie remarks are often a meta layer where the writer is reflecting on how the reader is processing their sentence, and they add clarification to make sure they are understood. They are punctuation marks about self-consciousness. Losing them is like losing a whole dimension of self-reflection. They’re used for digression, tension, clarification. Without them, you're not letting me see your mind at work, you are merelyh communicating. I wonder if AI bakes them in (via system prompt?) to give the illusion of a mind in thought, yet it’s really just capturing the syntax, and not really using it for digressions.

Garret on numeracy:

I suggest spelling out either: (1) all numbers below 10 or; (2) all numbers below 20 or (3) all numbers below 100 with the exception of your chapter references. If there are too many numbers like this in your pose, then the important numbers won’t stand out as much, like the reference examples later in this paragraph. Garner prefers option 1. DFW prefers option 2. Chicago style is option 3.

My reply:

Given different writers have their own range, is there a case for “all numbers below 2”? I’d argue that anything that is a quantity, other than a/one, can justify being a numeral: 1) it creates a visual fabric, where all quantity gets a specific symbol, and 2) it’s create the least readerly friction (I look to reduce this where I can because in other areas I intentionally add friction for specific ideas/phrases. To spell out “seventy-six,” in my mind, is a poor use of someone’s mental resources, an unnecessary drain of stamina. Even “7” over “seven” saves a few milliseconds of stamina that I will expend elsewhere. Also I love numbers. I’m really a math guy, and all my prose is just really filler between my numbers.

Here are some idiosyncratic rules on how to make these decisions:

1. If two numbers occur in a sentence or a paragraph, use numbers so the reader can effortlessly see and compare quantities in a pre-read scan.
2. If you have a set of labeled or numbered items to make a framework (a, b, c) or (1, 2, 3), you can default to spelling out a number so it doesn’t appear to be part of the framework.
3. By intentionally spelling out large numbers, you make a point (“we waited for one hundred and twelve seconds for the waiter to return”). The delay of processing numbers can be used for effect.

This is a good example of rebelling against prescriptive, absolute rules: “everything under 10 must be spelt out.”

Idea: A spatial numerical system where all digits have “Y” as the base number, where each stem of the Y represents an axis (X,Y, Z) and you can modify each stem with dots, dashes, squiggles, arcs, patterns, etc. So basically YY would be a line. Currently you could spell this out as “(1.42,0.42,3.40),(2.40,4.91,0.84),” but two Ys is way more compressed. There could even be a way to spell out three-dimensional shapes through a specific syntax that helps the Ys relate to each other. Of course, this wouldn’t be a readable language. But if machine vision becomes trivial and equal to text processing, then, in the attractor towards algorithmic compression, they might resort to a visual language. Especially if AI thinks through vectors, then they’d need not just a visual language, but a spatial one.

The advice that our writing voice should never be passive comes from overfitting marketing advice to essay writing. Yes, sales pages on websites warrant a particular aggressiveness in tone; in that context, there are many things to click on and you’re trying to communicate clarity in the quickest possible time. Essays are not like that.

Terence McKenna always had this half-baked idea/joke that mushrooms are extra-terrestrial. There’s a source in my "Babbling Idiot" essay from 2024 that dates mushrooms back to 60-65 million years ago, the date when the asteroid hit and made the dinosaurs extinct.

There’s some speculation that, if psilocybin did arrive extra-terrestrially, then it’s a kind of slow-moving alien fungus that’s able to bend certain forms of matter to create the conditions for language (high-bandwidth information processing) to occur.

Humans then became the host of that impulse, eventually creating machines that could make information processing even more dense. Maybe it’s not a coincidence that if you map LLMs they look a lot like fungal networks. This ties into that William Burroughs quote: "language is a virus from outer space.”

It ultimately frames humanity as a pass-through entity. It somehow explains our past and future, and becomes a catalyst to consider language itself as our ultimate life-force/medium of agency. I don't yearn for a transhumanist future, but when you think at cosmological time scales, it's hard to see the human as a stable form. (The issue with technologically-accelerated evolution is the speed, recklessness, and inequality in how it happens.)

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)….

The Zoo is a place where you can replace your stick-figure representations with direct perception. It’s easy to overlook everything: “oh look a rare bird,” and “cool, a giraffe.” But you can also stare intently at the facial hair of a gorilla up close for 2-3 minutes with complete focus and marvel at how strange it actually is.

I looked it up, and I’m surprised there’s no term for the uncanny that emerges from looking at non-familiar animals up close. It’s most obvious with cockroaches perhaps. That’s an obvious biophobia. Squirrels, dogs, and horses are good example of acclimated animals that we are accustomed to. But there’s a feeling of grotesque when you see a human-looking eye on a mammal as strangely-formed as a rare hog—familiar parts in an unfamiliar whole. Maybe the grotesque comes from imagining a human being transfigured into it.

Maybe the “evolutionary uncanny” is when nature reprises parts (eyes, lips, or hands) into whole that seems uncanny or frightening.

Throughout my time at the zoo, I had trouble putting the right words to what I was feeling, and I couldn't help but coin the phrase, "all of God's retarded creatures." This is obviously absurd and inflammatory, especially with the adjacent G-word and the R-word; I can’t say I even agree with it, but it speaks to the alienation and distance that language creates from nature (and the hubris of a linguistic creature).

This taps into an idea in DFW’s “Consider the Lobster” essay. Despite him saying how we should be more conscious of how we boil lobsters alive, he closes by confessing that he feels that animals are less morally important than humans and he’s not sure why. This has been an open loop for me. I have a positive regard for animals, but I know people in my life who exist on both extremes (ie: my wife sees dogs on par with humans, and my grandfather thinks dog == rat).

At some point while wandering the Bronx Zoo, I found myself thinking that nature is this force that rapidly mutates and complexifies life in order to find a "breakaway pattern." I am, obviously, as I write this, an inheritor of that pattern, and you, by nature of reading this, are an inheritor too. And so when I look at the rhinos, I don’t see cute or powerful animals, but sadness over a breed of mind trapped in a tankish form-factor. A human-centric way of seeing animals is that they're failed mutations in search of a perfect form, a being that can transcend itself.

It gets me wondering, “what is it like to be a Rhino?” This is the basic question of consciousness, and when you really imagine the felt experience of an animal (especially a caged one), it can feel uncomfortable, even alien. There are physical, sensory, and of course cognitive differences across species, and within a species—it all gets me wondering how primary language is to the human experience. Is it the thing? Without speech, there is no civilization; even if we are "trapped" by or language in some sense, we also owe our being to it.

Does DFW’s feeling that humans are the most morally important species comes down to linguistic supremacy? And is this generally true, or is it hubris?

Other notes:

• 12:30 PM – Evolution comes from tight feedback loops: in nature, in language, in identity, in art.
• 02:00 PM – A name for the lens when you look at an animal species and estimate how many years it might take for it to evolve into a language-using advanced society given the right environmental niche. “Monkey: 15 million.” Etc. I asked GPT, and it said that octopi, ravens, and raccoons show the most promise.
• 02:56 PM – Near the entrance of the Bronx zoo, they have neoclassical building with animal ornamentation, and there are many jokes to be made about monkey college.

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.

Regardless of which side you're on, I don’t think one-sided pleas will work anymore. You can’t just say “here are 7 solid points on Trump’s insane corruption.” A flurry of comments will come in, “but what about DOGE?”

Communication is impossible when bias is baked in, unless you immediately identify how a POV is in a unique position. Such as: “what if corruption exposers are in a position to shield their own corruption?” It frames the question in a way that DOGE is doing good work, and yet also, perpetrating the same thing it’s exposing.

Only by immediately identifying with your opponent (both with who they favor & how they critique), will you be able to slide in a nuanced, synthesized point of view that can help them around their bias.

Relatedly, I think there’s a lot of spin/distortion around RFK/republicans. Of course, Elon/Trump are doing the same (or worse) media distortions. That’s just the environment we're in. Both extremes are wrong (American golden age vs. devastating coup). I’d say overall though, there are more good than bad things happening, and RFK is (potentially) one of the good things.

Reality is more like a paradox than a simple narrative.

For example: Elon can simultaneously be a hero for American democracy (on par with George Washington) and the worst dictator ever. I have to assume both can be true. The second you cling to either/or, your vision gets clouded to make sense of all future information.

By clinging to either, I get a false optimism or a false doomerism — both of which are definitely wrong. The reality is somewhere in the middle, and so by stretching across the spectrum, I have a worldview that’s likely more attuned to reality than either side.

I've spent the day in a semi-modern hospital in what seems like my father-in-law’s last days. Communication is barely possible. Some yes’s and no’s come through, but maybe not for long. He left some signed notes from his past self about not wanting to live off feeding tubes. But there are all sorts of decisions (ie: who does he want visiting or not?) that he can’t communicate because he’s too weak to speak. People don’t like writing, let alone writing about the ideal conditions for their death, but considering it must be the most disorienting and climactic of experiences, it could be worth doing a lot of writing to help you and your caretakers in the future.

There's a challenge in writing for two audiences at once. How can something make sense to someone with no technical knowledge without boring an expert in the field? You have to innovate in how you communicate. Metaphors serve a dual role: they give the novice a bridge, and they give the expert a new lens to the familiar.

Have we misdefined the word "transgression"? If progress is walking forward and regression is walking backward, then transgression means something like “beyond walking.” A transgression is not a sin, but it’s to walk in a direction that is not aligned with our shared notions of progress. It escapes the spectrum. Maybe that’s a meander, maybe that brings us to hell, or maybe that brings us to our destiny. And isn’t so much high-level success about breaking out of local maxim to find a better current to follow? Isn’t genius about hitting the mark that’s out of view? A transgression is something that escapes black and white morality.

In Latin, "materia" means the usable wood of a tree. This maps to the idea that a writer’s Material is not the Experience/Source itself (the full, standing tree), but the lumber that’s produced from it.

64 rare words from a single Kathleen Alcott essay: vertiginous, alight, glistering, totemic, succored, beatific, menace, pestilence, dictaphones, hatboxes, ossifying, operatic, repudiate, monomaniacal, haughtily, vain, balaclavas, moxibusion, ministered, lancing, auricular, constituent, disavowal, repertory, ghoul, transfigure, Dalmatian, imperious, polyphonic, winklepickers, chenille, revanchism, blush, mantel, cherub, Meisner technique, vainglorious, tidily, sorry end, waspy, sequins, garrulous, irredentism, tines, contiguous, relishing, skein, adipose, suppurating, banality, sniping, blithe, casuist, pirouette, befouled, comity, littoral, susurrus, copses, bawled, meager, rejoinder, dolorous, rufous.

AI: “The Oxford English Dictionary (OED) categorizes words into frequency bands ranging from 1 to 8, with Band 8 representing the highest frequency. This logarithmic scale means that words in Band 8 occur approximately ten times more frequently than those in Band 7, and so forth. The specific frequency ranges for each band are as follows:

• Band 8: Words occurring more than 1,000 times per million words.
• Band 7: Words occurring between 100 and 1,000 times per million words.
• Band 6: Words occurring between 10 and 100 times per million words.
• Band 5: Words occurring between 1 and 10 times per million words.
• Band 4: Words occurring between 0.1 and 1 time per million words.
• Band 3: Words occurring between 0.01 and 0.1 time per million words.
• Band 2: Words occurring between 0.001 and 0.01 time per million words.
• Band 1: Words occurring fewer than 0.001 time per million words.

This classification helps users understand the relative commonality of words in modern English usage."

You know that phrase, “show, don’t tell”? — It’s all about making things concrete instead of abstract, but coding is the opposite. You’re turning ideas into a network of abstractions.

Coding is about mechanizing ideas into generalized abstractions so users can spawn specific instances. Essays are about expressing ideas as concrete thoughtscapes so readers can inuit the universal.

Coding is telling, essays are about showing.

Perspective is opinion with explainability.

A word for the shock of unexpectedly finding a whole tome of work from someone you know that is larger than you can digest in a single sitting.