This November I attended Inkhaven1, a writing residency where 40 of us posted daily, workshopped each other’s pieces, and received feedback from more experienced professional bloggers and editors. A month of writing under pressure was challenging, but I’m overall glad I did it.
I was worried about the quality and frequency of my posts there, which is why I segmented my Inkhaven posts to a different blog. If you’ve previously noticed a lack of The Linchpin posts/emails in November, now you know why! :)
Anyway, without further ado, here are my best Inkhaven posts out of the ~30 I’ve written.
The lovely Lighthaven campus, where I spent much of my waking hours in November Source
Well-written
On a technical level, I consider these posts to be well-written and executed, with a clear through-line.
A post about board game strategy. The key concept is that you should understand the win condition, and then aim towards it. The post goes into a lot of detail about how best to apply this principle and gives a bunch of specific examples. But the basic idea is very simple and I think it explains a large fraction of the difference between good novice play and mediocre or bad plays.
Of course, you should not expect to be able to use such a simple strategy to win against very strong and experienced players. However, I do think it generalizes quite far, and many people who think of themselves as strong players (and indeed might even be so according to objective metrics like win-loss records against average players) could stand to learn from it.
The post was overall well-received, with many commenters who either find it helpful or endorse the strategy based on results (whether from themselves or others).
Rock Paper Scissors is Not Solved, In Practice
A deep dive into Rock Paper Scissors (RPS) strategy, particularly in the context of bot tournaments. RPS strategies have two goals in constant tension: predict and exploit your opponent’s moves, and don’t be exploitable yourself.
I think lessons here can maybe generalize significantly to other arenas of adversarial reasoning, though it takes some skill/time to figure out how to apply them precisely.
While I tried to illustrate each RPS-specific strategy through an approximately increasing order of complexity (pure rock -> pure random -> String Finder -> Henny -> Iocaine Powder -> Strategy Selection), I also tried to illustrate other general principles and ideas on the side. As an obvious example, that pure random is a mixed-strategy Nash Equilibrium. But also the reason you don’t always want to play the Nash Equilibrium strategy is due to less sophisticated agents/bots in the pool, which generalizes to other contexts like prediction markets and trading more broadly.
But the main thing I wanted to illustrate is just that an extremely simple game has almost unlimited strategic range in practice, which I found fascinating. Many of my readers agreed!
This post made zero splash when it first came out, but I’ve gotten a steady stream of new readers for the post since, and now it’s my most-liked post from Inkhaven!
How to Write Fast, Weird, and Well
A post on all the advice on writing (for myself and others) I could think of that’s important, non-trivial, and not previously covered in my earlier post on writing styles.
Key points include writing a lot, getting lots of feedback, and saying surprising (but true!) things other people don’t expect you to say.
This was my first Inkhaven post. It didn’t have many views or likes, but was surprisingly well-received by many Substackers who I consider to be good writers.
It was not that popular elsewhere, which is unsurprising. Relative to their respective audiences, writers really like writing about writing, Hollywood directors really like making movies about Hollywood, and composers really like songs about musicals, and so forth.
People at Inkhaven took the program very seriously! Source: https://jenn.site/inkhaven-photodiary/
Conceptually Original
The well-written posts above are more self-indulgent (about writing, games) and less important. They also served as better explorations/extensions/explanations of ideas first discovered by others, rather than me making a truly original case of my own.
The following posts are of lower writing and execution quality, and therefore messier, but they have ideas that I think are more original. As far as I know, I came up with the ideas myself. So if others had the same idea, it’s more likely due to independent convergence.
Skip Traditional Phase 3 Trials for High-Burden Vaccines
During major pandemics, policymakers should skip Phase 3 trials for vaccines. Instead, give people the vaccine right away while continuing to study whether it works, and pull it if problems emerge. Having a process in place for rapid deployment of vaccines during major pandemics, and also for new vaccines for ongoing high-burden diseases (malaria, TB) can save at minimum hundreds of lives, and as many as hundreds of thousands of lives per vaccine hastened.
I think this is incredibly important! At least in theory. I hope scientists and public health professionals will take more efforts to make this happen, or at least more productively explore this idea so society as a whole can be more confident rejecting it.
It got a moderate amount of interest in the EA Forum but not elsewhere. I hope someday (ideally someday soon), somebody with greater domain expertise can champion this idea and make a stronger case than mine.
I present a case that:
- a predominant position in philosophical arguments against aging (as articulated by, e.g., Michael Huemer) and a predominant model in modern anti-aging biotech research and development are based on the idea that modern medicine is implicitly a “whack-a-mole” approach. We should instead attack the root causes of aging (telomeres, cellular senescence, and so forth)
- I argue that this is wrong. The “target the root cause of aging” model directly contradicts the most plausible scientific theories for why aging actually happens, on evolutionary grounds.
This does not mean that #1 is necessarily wrong. Maybe the main scientific theories for aging are wrong. Maybe the main scientific theories are partially real and the “root causes” explain some but not all of the story. Maybe my argument is wrong and there’s a clever way that the contradiction isn’t real. But I think this tension is a really big deal, and I wish antiaging advocates, scientists, and especially the companies seeking billions of dollars in investments and public funding would publicly grapple with these theoretical challenges.
This post got some attention but the core tension is still essentially absent from public discourse on aging research. I hope to revise it one day, improve, enhance and develop the post overall, and maybe publish it in a magazine somewhere.
The maximum lifespan of bats vs mice is directly related to my argument above. Why? Read the post to find out!
Presenting my case that people are underrating how improvements in conceptual technology and how we formulate ideas allows people to meaningfully think about deeper problems than our ancestors were able to grapple with, despite relatively low if any change in our base intelligence/hardware.
Relatedly, ideas that are extremely, blindingly, obvious in retrospect are hard-fought and hard-won. And we moderns who fully integrated those ideas don’t understand how radical, surprising, or confusing those ideas were when they first emerged on the scene. Examples I gave elsewhere included Intermediate Value Theorem, Net Present Value, Differentiable functions are locally linear, Theory of mind, and Grice’s maxims. But these are specifically chosen as ideas that are currently hard for some people to understand. No intellectual alive really disputes the idea of “zero.”
I’m worried that ideas about ideas and writing about ideas would come across as too navel-gazey and uninteresting, even to “normal” nerds. If I ever have a better angle on conveying these ideas (sharper imagery, better examples, more clear direct payoffs and practical applications), I’d love to revisit this idea and do it justice.
As it is, other writers are welcome to take their own shot at addressing this concept!
Honorable Mentions
Here are 4 posts that I think were neither particularly well-executed by my lights nor had as strong conceptual interestingness or originality, but still had strong things going for them:
Legible AI Safety Problems That Don't Gate Deployment
Wei Dai had a very sharp observation on legible vs illegible AI safety problems. I tried to understand his position and extend it. Wei Dai argued that legible AI safety problems (ones obvious to leaders) will gate deployment anyway, so working on them just speeds up timelines. Instead, we should focus on illegible problems instead.
I think this is directionally correct, but conflates “legible” with “actually gates deployment.” AI psychosis is highly legible but companies keep deploying anyway. The deeper issue is that illegibility to lab leaders may often be motivated rather than epistemic: “it’s difficult to get a man to understand something when his salary depends on not understanding it.” This suggests we might sometimes be better off making problems legible to less biased audiences (journalists, policymakers, the public) rather than assuming the bottleneck is technical sophistication.
I like this post because AI safety is very important, Wei Dai’s observation is sharp, and I think my comment positively contributed to the conversation. So I’m glad to be able to make my own, if limited, contribution.
Middlemen Are Eating the World (And That's Good, Actually)
Many people despise “middlemen” “bullshit jobs” and prefer “real jobs” that can be done by a pig wearing clothes in a children’s book, like pork butchering.
I argue this intuition is completely backwards! Middlemen (and the generalized idea, roughly people who help others coordinate better) are extremely important, and in modern societies, often more important than direct/object-level work.
My most popular Inkhaven post by views (almost 6k?). Higher than median (though not average) post on my main blog, tbh. I think it’s a pretty obvious idea. Certainly not original to me.
The article is in a bit of an awkward middle spot. For an academic piece, it was light on citations. For a populist (anti-populist?) piece, I think the examples could’ve been more emotionally motivating. I doubt it’d convince anybody actually on the other side, but I think it’s a decent piece of inoculation for high-schoolers/college freshmen and other people new to the ideas who have not previously heard clear articulations from either side. At least, I think my intro is better than you’d usually get in introductory economics classes.
Anyway, in general I thought it was a fine piece, and unsurprising that it’s popular given the topic.
Building Without Apology: My a16z Investment Thesis (Guest Post)
A creative writing exercise where I made up fake evil startups in order to lampoon the immorality of Andreessen Horowitz for funding all their real evil startups that tear apart the social fabric. Ozy Brennan and Georgia Ray contributed some of the ideas/jokes.
I enjoyed writing it and I legit think it’s quite funny, but I think the jokes were sometimes a tad too cerebral and overall weren’t sharp enough to go viral. Alas.
Five Books That Actually Changed My Life (Not The Ones I Wished Had)
A description of five books that actually changed my life, with concrete examples of why, and then a list of 25 other books that I liked and hope other people might like too. To be clear, this is different from my favorite books, books I might enjoy the most, books that I consider of the highest literary merit, etc.
This post was surprisingly quite popular (even though my personal posts usually perform worse). I’m not sure why. One hypothesis is that people just really like lists. Another possibility is that my most life-changing books (and other books I thought were good) also positively correlated with other people’s life-changing or otherwise good books. So they like it more and want to share more when people say nice things about books they like, similar to my Ted Chiang review.
Posts by Other Inkhaveners
People who enjoy my Inkhaven blog may also enjoy the blog posts of other Inkhaveners:
Reflections
Now that it’s been over two weeks since Inkhaven ended, what do I think of my experience there?
During Inkhaven, and in the days immediately afterwards, I was profoundly disappointed in myself. I like the social aspect of meeting other writers, and enjoyed many of my conversations. I also liked the food, snacks, and environment. But my output wasn’t the best, and I was constantly saddened by my productivity.
Concretely, my hope before starting the program was that my typical post would look like How to Win Board Games, with an idea that’s not original to me but surprising to the vast majority of my audience members, a clear throughline, competent execution, a clear reason why (some) readers might be interested, and generally a solid-but-not-stellar blogpost overall. I hoped I’d have 3-5 high-quality blog posts with the above but also genuinely original ideas, beautiful writing, clever analogies and anecdotes, and a wealth of unexpected connections, akin to Why Reality has a Well-Known Math Bias or Ted Chiang: The Secret Third Thing. The hope, too, was that I could crosspost my better posts to my bigger/more serious blog The Linchpin.
Instead, How to Win Board Games was closer to my peak of writing at Inkhaven. None of my posts in November combined original ideas with what I think as genuinely competent execution. Alas.
But now that two weeks have passed and I’m a bit more removed, I feel better about my output! Partially because I have some more distance and I can look at everything more objectively. But honestly part of it is because people are still liking and sharing my old posts from Inkhaven, suggesting that at least some of the posts might stand the test of time and be more than just a flash-in-a-pan phenomenon. I also think upon rereading my posts, the quality standards for my better posts were decent for blog posts in general, not just for blog posts written in a hurry in 24 hours. So that’s good.
I’m also of course really glad to have experienced the amazing venue at Inkhaven, and the chance to talk to amazing mentors and fellow writers. The experience overall was solid and I’m glad to have learned from them.
Would I ever want to do something similar to Inkhaven again? Unclear, but I’d seriously consider it!
My Inkhaven experiences also entailed falling in quicksand and then saving a kitten that night. So that was pretty cool.
What’s next for this blog?
As many readers know, I’ve published my first post-Inkhaven post!
As far as I can tell, it’s the best explainer available online for the basics of stealth technology. The core idea is surprisingly simple! I’m glad to have enough time to carefully refine the post and try my best to only include what needs to be included, and no more.
My next Serious Post is a continuation of the above, a full review of Skunk Works, a memoir by Ben R. Rich, the former Director of the Advanced Research and Development Department at Lockheed that made advancements like stealth airplanes and many other critical military technologies. I intend to cover technical, organizational, geopolitical, and ethical implications.
I’ve also resumed doing sporadic interviews of other philosophy or philosophy-adjacent Substackers who interest me, including my recent 3h+ marathon chat with Ozy Brennan. Feel free to comment or DM if you have ideas for other people who you think I should interview!
Finally, I’m cooking up a short post on the theory and empiricism behind gift-giving, hopefully just in time for Christmas and New Year’s.
If you like my work, please subscribe and share your favorite article with at least one friend. I’d love for more people to see my best writings!
Subscribe here: https://linch.substack.com