Surprising but Inevitable: The First Six Months
The Linchpin at six months: highlights, themes, and what's next
Six months after launch, The Linchpin hit 1,000 subscribers and ~100,000 readers. Here are my personal favorite posts from each month, the themes that keep showing up, and what I’m working on next.
Happy New Year, everybody! I’m excited to announce that exactly six months after starting my blog on July 2nd of last year, we reached 1000 subscribers!
In the last 6 months, I’ve written 21 posts on The Linchpin (15 “serious posts” if you exclude video recordings of interviews and announcement/summary posts like this one). I covered a broad range of topics that interest me, from anthropic reasoning to the game theory of war, from aeronautical engineering to zoology.
Along the way, I’ve had many great conversations, built better models of the world, made several new close friends, and managed to share some early ideas on the most interesting and important questions facing humanity.
Without further ado, let’s dive in (“delve into?”) the most important recurring themes that surfaced in my writing, my favorite posts from each month, and what’s next for 2026.
Recurring themes
At 6 months old, the blog is still quite young, and I’m still trying to find my voice and (mostly) just writing where my curiosity leads me. Still, here are some themes that have surfaced in my writings that I’m happy about:
Surprising but inevitable
On a good day, I want my posts to come across as “surprising but inevitable.” Let’s consider both halves of that statement.
I aim for my posts to be “highly surprising” recursively:
If you know some facts about me and the topics of my previous posts, you shouldn’t be able to guess the topic(s) of my next post. It’s important to me that I talk about a broad and surprising range of topics.
If you know my general disposition and hear the topic of my post, you shouldn’t be able to guess my thesis statement. I want my judgments to be interesting and genuinely thought-out, rather than have my posts justify a foregone conclusion.
If you know my topic and thesis statement, you shouldn’t be able to guess my actual arguments. Good, high-quality arguments shouldn’t look like they’re just “going through the motions.”
If you know my thesis and arguments, you shouldn’t be able to guess my specific examples. This one is more of an aesthetic choice, but it’s important to me that my examples are interesting and genuine, and independently thought-out. To the best of my ability, I’m not rehashing old arguments and reusing old examples.
The main exception is endings. For most nonfiction essays, if you carefully read the rest of the essay, the conclusion ought not be particularly wild or surprising. The “surprising but inevitable” dictum was originally from Aristotle about good story endings. But for nonfiction, the opposite is true: the body should be surprising but inevitable; the ending can just flow naturally.
On the inevitable half:
Merely being surprising is not enough. You probably wouldn’t read me if I merely write surprising but false things, or surprising and true but irrelevant things. Instead, the aesthetic of a good explanation ought to be surprising before you read it but obvious in retrospect, whether we’re talking about unknown known technical concepts, the embarrassingly simple geometry of stealth, or why playing the Nash equilibrium is not always optimal for rock-paper-scissors.
Or at least, that’s one ideal. Reality is complex and under no obligations to come across as inevitable to human minds, and I would not sacrifice truth for the purposes of a cleaner story.
Making the abstract visceral
I honestly believe, and try to show, why abstract ideas matter, through both direct real-world examples and charming hypothetical ones. The rising premium of life isn’t just an interesting economic fact or statistical oddity, it’s a key trend that underlies much of modern culture and society. Epistemology is conveyed not through abstract philosophy but through a tier list. The counterintuitive bargaining theory implications against war aren’t just interesting academic theory, but a live realization that most wars are mistakes.
Consilience
Rather than take a single methodological approach to all issues (like Bayesianism, or the scientific method), I’m more confident in a result when a consilience of different methods all point in the same direction. Basically the approach explicitly advocated for here, which is the backbone for much of my work.
Concretely, the rising premium of life reasons not just from economics and psychology but also from history, media, and even evolutionary biology. The honey post weaves biology, philosophy, and economics. The Puzzle of War primarily combines bargaining theory with international relations, but also looks at history, uses fantastical examples, and is implicitly driven by moral force.
Also on an aesthetic level I just really enjoy unexpected juxtapositions and surprising connections! I find it more fun.
Selection effects as everyday reasoning
Selection effects rule everything around us. I want readers to think like I do, and understand how selection effects shape the world we live in, and ourselves. In my humble opinion, the two most interesting selection effects are anthropics and evolution by natural selection, but I’d be keen to discuss other selection effects in future posts as well.
In particular, I want readers to think of anthropic reasoning not just in terms of weighty questions like cosmology and doomsday paradoxes, but also in terms of everyday events like waiting lines at phones and the relative crowdedness of lecture halls, so people don’t treat anthropics as its own special domain. I also want evolutionary thinking, particularly on non-politicized topics, to feel naturally integrated in people’s thinking and less ad hoc. That way, readers can reason better and be appropriately skeptical about just-so evolutionary stories of others.
Curiosity, playfulness, and awe
I’m deeply curious about both the realm of ideas and the physical world we live in! I feel a deep sense of awe about all the subjects I discuss, and I never feel like I’m “too good” for a subject, whether I’m discussing a children’s game like rock-paper-scissors or the underlying causes of aging and war.
On the flip side, I also don’t think any subject I discuss is “too good” for me. I try not to let my sense of awe and reverence overcomplicate things, make things “deeper” than they have to be, fabricate a sense of nuance where there isn’t one, etc. I try to be intentionally playful and irreverent, even about deeply important subjects.
Do you think there are other important themes of my blog that I’m missing in the summary here? Are there other themes you wished my writings had more of? Tell me in the comments!
Best of Each Month
If you are pressed for time, here are my candidates for the best posts from each month for you to read or reread.
July: Why Are We All Cowards? (The Rising Premium of Life)
Over the last century, something profound has shifted in how humanity values life and death. Americans now value a statistical life at over $11 million, likely an order of magnitude higher than a century ago in inflation-adjusted terms. We shut down the global economy to prevent COVID deaths; parents who once roamed miles as children won’t let their kids leave the front yard; young people take fewer risks than any generation in history. The post traces four forces behind this shift: wealth effects (richer people both likely have better lives and can afford to buy more life), safety-risk aversion feedback loops (each generation raised in greater safety develops lower tolerance for risk, similar to how bats evolved longer maximum lifespans than mice because flight made them harder to kill), secularization, and smaller families. The result is a civilization increasingly organized around death as the ultimate evil, which is mostly positive but likely comes at the cost of dynamism and the willingness to take risks that matter.
Runners-Up: Eating Honey is (Probably) Fine, Actually, beefing with Bentham’s Bulldog on a range of considerations one ought to think about on questions like ecology and insect welfare, and Why Reality has a Well-Known Math Bias, on the anthropic reasoning/evolutionary explanation for the unreasonable effectiveness of mathematics in the natural sciences.
August: Ted Chiang: The Secret Third Thing
I tried to write the best book review of Ted Chiang. I think I succeeded.
Chiang is an amazing science fiction writer, but many readers, even enthusiastic ones, miss what makes him distinctive. Critics often categorize science fiction as “hard” (scientifically rigorous) or “soft” (science as window dressing), but Chiang frequently writes neither: he creates universes where the principles of science themselves differ from ours while remaining internally consistent. More remarkably, Chiang makes philosophical positions felt: his time-travel stories don’t just explain compatibilism (free will coexisting with determinism) but convey what it would be like to live it. The post also identifies Chiang’s blindspots: his weakness at portraying how entire societies interface with transformative technologies, and his tendency to hit one philosophical note (compatibilism) repeatedly rather than attempting equally deep treatments of other ideas.1
This post is my second most popular post to date, and still high on the front page of Google search results for “Ted Chiang review.”
September: Nine Intellectual Jokes
What separates a joke about intellectuals from a joke that actually teaches an idea? This post collects examples of the latter: jokes that use humor as a vehicle for genuine insight, and in a coda I tried to analyze what makes intellectual jokes work. I also rewrote every joke for clarity, precision, and comedic timing. Hopefully it’s a hoot! 2.
Runner-up: The Puzzle of War, a significantly more ambitious undertaking where I tried to solve war, or at least summarize and extend the literature on the intersection between bargaining theory and international relations.
October: A Field Guide to Writing Styles
Most writing advice tells you what to do (be clear, use active voice) without explaining the underlying choices that define a coherent style. Drawing on Thomas and Turner’s Clear and Simple as the Truth, this post presents eight distinct writing styles (classic, plain, practical, reflexive, contemplative, romantic, prophetic, and oratorical), each defined by its stance on fundamental questions: What is truth? What’s the relationship between writer and reader? Is thinking complete before writing begins, or does writing constitute the thinking?
Rather than advocating for one style, the essay inhabits each in turn, demonstrating their textures through sections written in those styles. My overall takeaway from the book is that style isn’t personality or tips; it’s a principled choice on a small number of deep issues, maintained consistently throughout a piece. For internet writers specifically, I encourage people to try to experiment with different styles and understand their contours, rather than just stick with a single style you’re used to.
November: Rock Paper Scissors is Not Solved, In Practice
For November I joined Inkhaven, a writing residency where I wrote a blog post every day! Naturally the quality of my blogposts went down, so I published everything on a separate blog so as to not pollute The Linchpin with lower-quality ramblings. Nonetheless, I had a few successes.
Rock Paper Scissors has a trivially simple Nash equilibrium: play each option with equal probability and you can’t be exploited. End of story? Nah, RPS bot tournaments reveal staggering strategic depth. The post traces an escalating arms race of strategies: from pure-rock bots (dominated) to pure-random (unexploitable but can’t exploit others) to pattern-detectors that find regularities in opponent history, to meta-strategies that model which pattern-detection algorithm their opponent is running, and so forth.
The central tension between “predict and exploit” and “don’t be exploitable” generalizes far beyond RPS: to poker, prediction markets, and any domain where you must simultaneously model others while avoiding being modeled yourself.
Runners-up: Aging Has No Root Cause is probably more important and novel, arguing that most anti-aging research and development is operating on a flawed theoretical foundation. How to Win Board Games is probably more directly relevant to my reader’s lives, offering a surprisingly simple strategy that many readers have already told me elevated their gameplays significantly. You can also read an earlier review of my best November posts here.
December: Unknown Knowns: Five Ideas You Can’t Unsee
Some concepts feel less like things you learned and more like features of reality that were always there: invisible until acquired, impossible to unsee afterward.
This holiday post presents five such “unknown knowns”: the Intermediate Value Theorem 3(tipping points must exist; your single vote can matter), Net Present Value (infinite future streams have finite present value; thus, “solving homelessness forever” isn’t infinitely expensive), differentiable functions are locally linear (with long-ranging implications like why altruists should be risk-seeking with their donations, and you shouldn’t buy extended warranties), Grice’s maxims (the hidden rules of cooperative communication), and Theory of Mind (the reminder that enemies, allies, and fictional characters are agents with their own goals, not mere NPCs in your glorious self-narrative).
I think these ideas are simple, surprisingly subtle, and yet important, and I hope readers can learn and appreciate them!
What’s next for The Linchpin?
How much I write will depend significantly on what job(s) I work on this year. What I write about will likely be shaped by that too, though in nonlinear ways. If my day job involves AI, I might blog more about it, or less (to preserve the blog for other interests). We’ll see.
Posts in the making
I’ve finished significant sections of drafts for the following posts, though I am somewhat stuck.
The simplest possible case for AI danger
Every single argument for AI danger/risk/safety I’ve seen seems to overcomplicate things. Either they have too many extraneous details, or they appeal to overly complex analogies, or they seem to spend much of their time responding to insider debates.
I want to write the simplest possible argument that is still rigorous and clear, without being trapped by common pitfalls.
Skunk Works review
As promised in the Stealth post, I’ve been working on an in-depth review of Skunk Works by Ben R. Rich. Key sections include the ideas-to-weapons pipeline, the making of a moonshot research program, the strategic OODA loop, and the ethics and tradeoffs in arms manufacture. I’ve finished about 70% of the first draft already, but the remaining 30% currently eludes me.
Theory of Mind
As introduced here and here, I’ve long been interested in Theory of Mind (related concepts might include strategic modeling, level-k-thinking, intentional stance, decentering, stakeholder analysis, (non-) hope chess, sonder, alterity, or polyphony.). I suspect many people miss significant parts of it, and it is central to many confusions and avoidable human suffering. It is even a plausible answer to the Puzzle of War introduced earlier.
So I’ve been working on a post laying out the basic case for why many adults have insufficient Theory of Mind, tentative ideas for how it can be improved across the board, and sketching out the consequences of what significantly better ToM across the board might look like.
You shouldn’t read people just because they’re smarter than you
Simple argument, with a few nuances. About a third of the argument is buried here in the comments to Bentham’s Bulldog’s post. The other two-thirds rely on a) questioning the empirical correlation between intelligence and accuracy for people you’re likely to read at all, and b) suggesting a better holistic framework and/or proxies for who you should read instead.
Bifocal thinking
English doesn’t have a word for it. That thing where two perspectives on the same problem both have merit and you need to think from each angle, first one then the other, to deeply understand what’s going on. In psychology, we might think of biases vs heuristics. In statistics and machine learning, we might think of bias vs variance. In macro-history, we might think of historical inevitability vs contingency. I still don’t have a good term for it, but friend-of-the-blog Katelynn Bennett suggested “bifocal thinking” as a stand-in.
I’ve been working on this off and on for three months, and it’s been difficult. Ideally I want to write an elegant explanation of both the overall concept and ~12 interesting and important bifocals to look at the world. But I worry it's too ambitious!
Future Topics
These are general topics that I haven’t written much about in 2025 but I’d like to spend significant effort on in 2026.
Mapping the unknown
See here and here for earlier notes.
It’s a tautology that you don’t know what you don’t know.
But contrary to popular belief, it’s often possible to get additional information about what you don’t know, and start to map the unknown.
I suspect we can do much better than just say “it’s often possible.” I’d like to figure out how to systematically map gaps in your knowledge. Or at the very least, be able to find a specific subarea of tractable knowledge gaps, and systematically explore how to map something we are collectively ignorant about.
Not sure if I could make significant progress on this question in 2026, but I’d love to try!
Practical ethics
For a large fraction of my adult life, I’ve been concerned about various questions in practical ethics. Roughly, how can I — a limited, often dumb, always imperfect, agent — still try to do good in the real world and act in ways I endorse?
I think I've learned some useful things, and I'd like to figure out how to share them on Substack. Plus, many of my friends on Substack, like Ozy Brennan and Bentham’s Bulldog, seem to have been pretty effective with their practical ethics posts!
But I also don’t want to come across as too preachy or certain, so it’s a fine balance4. See the following comments as examples of ideas that I might want to one day turn into full posts.
Artificial Intelligence
AI is likely the biggest deal this century. I think it’s incredibly important and dangerous, and I’d like to address it more seriously head-on.
Thank You
To the 1000 people who subscribed to my blog, and the others who read and enjoyed my blog, thank you!
Thank you for reading. Thank you for the comments, the pushback, the DMs, the kind compliments, and the detailed technical rebuttals on cultural evolution and anthropic reasoning.
I’ve long believed that good thinking is a highly social process, for humans. So it’s great to have so many smart, competent, well-intentioned people read my very long posts. The blog is better because of you.
See y’all in 2026. I’m not sure what I’ll write about, but I promise to keep it interesting and surprising.
Like my writings? Hate a specific post with the burning fury of a thousand suns? Have a specific topic you want me to cover over the next year? Tell us in the comments!
Also if you’re a fan, consider sharing your favorite Linchpin essay with at least one friend!
I tried to avoid spoilers in the summary, but note that the book review itself has significant spoilers.
I intentionally wanted to write something more popular than my usual fare with this post, and it looks like I succeeded! The post is responsible for ~⅓ of my total views. And Daily Nous, the world’s most popular philosophy blog, shared it!
I think I used the math here more analogously and less precisely than I’d ideally want, unfortunately.
Not that I’m terminally against being preachy; if me being preachy causes good things in the world like lives saved, I wouldn’t get too mad at myself.