2025-11-12 21:14:53
Reputedly the first-ever novel, isn't that wild?
I enjoyed this but preferred its arch rival The Pillow Book, if you're only looking to read one seminal Japanese work from 1000 AD make it that one. (At time of writing this I'm less than halfway through Genji and not sure I'll finish, this thing is 70+ hours long).
Heian era Japan is wild, from what these two books have shown me at least. [Author's note: Assume everything I'm about to say has some chance of being true, but is also likely to be based on a misunderstandings, I know incredibly little about this time and place].
Basically: it seems that sexual/social mores at the Heian court were just way more loose and liberal than I ever would have imagined could exist before, say, the 20th Century. Genji is – I'm sorry to say it – an incorrigible fuckboy, he sleeps with (and abandons) both young maidens and married women with abandon, and even though some of the affairs are considered inappropriate and would get both parties in trouble if discovered, it feels like the existence of such affairs is broadly accepted and/or encouraged, the game is just not to get caught.
There are vast numbers of conversations in the book that could 100% take place in a modern Californian polycule, with Genji berating his primary for the suspicions he assumes she has about his amorous adventures, and which (though I may be anachronizing here) seem to largely involve defensive projection.
At the exact same time, some of the gender norms at the time are extremely restrictive: women hide behind screens when talking to men, and Genji goes to crazy lengths just to get a glimpse of the women he's talking to. And yet with many of them he's also able to sneak behind their screens at night for some sexy-times, but still can't see their faces in daylight, it is truly a realm of contrasts.
The weirdest thing about The Tale is how completely it reads like self-insert fanfiction by a teenage boy: it's about a guy who's really great and handsome, and everyone says how great and handsome he is, and also how great his poems and art and music are, and all the girls want to be with him, and after they get with him they spend the rest of their lives pining for him.
This is weird specifically because it was written by a woman (imperial courtier Murasaki Shikibu).
Eventually I started to wonder if there was some level of irony or reversal I was missing: was this novel an indirect way of dissuading other women from falling for Genji-types?
For a start, the things Genji did that were described as great and wonderful did not generally strike me as particularly great or wonderful, though I may just be myopic due to time and distance. (One definitely-unfair parallel here is that the book is full of poems that the narrator insists are amazing and swoon-worthy, but at least in my translation they all seemed to be variants of "Is the something-flower really all that purple? It has dew underneath it" or "My sleeves are wet [because I am crying so much about you]" – I fully understand that they might be incredible Japanese, and if you understand the references in them, but it's funny to experience in English).
Meanwhile, all of Genji's bad behaviours are described relatively sympathetically, but strike me as really very bad. I started to wonder whether, in some way I don't fully grok, the book is meant to bury Genji, not to praise him – perhaps a sly warning to women to be wary of handsome men, who can do literally anything and will be praised by society regardless? (Handsome friends of the blog are welcome to chime in here).
One mistake I think people often make with old books, highlighted by me making this mistake with Genji, is to think "this feels strangely familiar, the past is not such a foreign country after all." I think there are two underlying mistakes here.
First, the ancient text has been pre-processed for you by modern scholars and translators far more than you might realise. In the case of Genji, I found out a very funny version of this after reading: apparently in the original none of the characters have names? According to a redditor (pre-llm era):
In Classical Japanese court prose, one almost never mentions the subject of the verb in a sentence. Instead, the subject has to be inferred from the relative formality of the verb's conjugation, which indicates the rank (and sometimes gender) of the subject of the verb in the social hierarchy. [In Genji], almost all the female character were assigned names by consensus by later scholars.
Amusingly,
Even the Author's name is unknown! Murasaki is the name of a character in the book. The author was Lord Shikibu's daughter, and scholars came to refer to her as 'Murasaki' based on the character in the book she wrote AND MURASAKI THE CHARACTER IS NOT EVEN NAMED MURASAKI IN THE BOOK!!!!!
When the Murasaki character showed up I was like "oh lol Murasaki-The-Author put in a self-insert character Murasaki" but it turns out... possibly no part of that sentence is exactly true?
Secondly, the process by which this book reached your hands is extremely non-random, both in terms of which works get preserved by history and which works you're reading today. I read this book BECAUSE it has a reputation as a relatively fun and accessible read; it's a romance novel with a straightforward plot and some truly delightful Jane Austen-y observations.
Meanwhile, it is only rare freaks like Jehan who read the great but truly difficult works of the past; there's a reason that I picked Genji off my pile and not the roughly-contemporaneous Beowulf, because hwaetever is going on with that thing is far beyond my patience these days.
Anyway, Tale of Genji: lightly recommended I guess? I'm enjoying it, and it's kind of magical that something so old can be so enjoyable, but also I'm not sure it's worth the time I'm putting into it, and now that I've started I've found it hard to stop.
2025-11-10 21:03:14
First up: yes, candidly, this guide is an excuse to hawk my game Person Do Thing (the absolute no 1 most desirable gift this Christmas), but for the sake of all of our dignities I will leave that to the end. Except for PDT, I have no connections to anything on this list and will make no money from you buying them.
In accordance with my life-philosophy I am focusing on gift recommendations that I think would have the highest counterfactual impact; that is, you don't need me to recommend things that everybody already knows about, so I'll try to recommend things you'd be less-likely to buy otherwise.
You upload a photo and they somehow carve it into a bead on a bracelet; when you shine a light through that bead it projects that image. It's like a locket for a favourite memory and I find it really beautiful; there are $100+ versions on American websites but I got mine for $4 on Temu and am delighted with the quality.
Just a very sweet gift for children (or eccentric adults). This is basically a Where's Waldo clone where you can customize the name and image of the person being searched for; rightly or wrongly, finding out that a company can profitably print completely customized Waldo books with your name and face in them was one of my most visceral "isn't modern technology amazing?" moments.
For hamstring stretches. I don't understand why this is so satisfying to stretch on but it's so great. Product-wise the price feels high at $15, even though I get far more value from it; I bet there's a way to get one cheaply from a building supply store but I don't know how.
Did you know that it's very easy to do headstands once you have something supporting your shoulders? I didn't either, and it's wild, and very fun.
A tribal artist from India comes to London, encounters planes, trains and cities for the first time, and draws his experience. A wondrous book.
Look, I'm sorry: I think it's outrageous that they're charging this much for a toothbrush, and even more outrageous that they charge $10 for replacement heads. I do not want it to be true that it's the best toothbrush I've ever used, and so much better than the last generation of electric toothbrushes. But sometimes we must face the t(r)ooth, and if you can afford this ridiculousness it's just objectively great. (H/t excellentfriend-of-the-blog N.)
Love this battery pack so much. Has saved me from many battery deaths.
Of all the dumb cheap things in my house that bring me joy, this one brings me the most disproportionate joy.
I'm sorry I'm sorry I just love standing on things and balancing.
I guess you'd have to be pretty close to give someone an $800 gift but this extendable table is so brilliant, it's a phenomenal variable-length dining table that folds up into almost-nothing. If you're in Europe you can buy it from the manufacturer for cheaper with slow shipping, which is nice.
You knew this was coming: Person Do Thing, my party game for lovers of Taboo, Charades and Monikers. If you are fun (or just a fan of fun) you will like it, I really think it's the perfect high-upside, low-downside gift for the widest possible range of friends and family.
2025-11-05 21:34:05
Sometimes I fall into patterns with other people where I keep noticing the choices they make violate my personal rules of thumb.
For example, a simple rule-of-thumb I have is "if asking someone whether they'd like to do something is costless, you should just do it rather than speculating what their answer would be."
That is: maybe you think so-and-so wouldn't do such-and-such because of thing-and-thing, but if it doesn't hurt to ask then you might as well just ask them and let them decide for themselves; any time you spend debating how likely they are to do the thing is completely wasted for everyone, you should just ask them and be done with it.
But this doesn't seem to be a common rule of thumb, and often when I'm organizing things with other people I'll say "should we ask Person?" and my interlocutor will say "Person will probably say no because of Reason," and I'm thinking sure, that might be true, but is there any reason not to ask them?
And this is not trivial, because there are situations where it is costly to ask someone about a thing, so in my head I'm like are you trying to tell me there's a cost to asking in this case? Or that you don't want them to do it, but you're pretending they wouldn't want to as an attempt to save face? Or do you just enjoy trying to figure out whether people will or won't do something, as a prediction? Or is there some mysterious other thing happening here?
And eventually, my thought is should I just explain this rule of thumb and then check whether you share it or not? But that part is costly to me, because I think people generally respond badly when you say "hey, I think you're making this entire category of decisions badly, here's an academic-sounding explanation of why there's no point in us continuing this conversation."
If anyone has advice on how to do this better I'm truly all ears. The above is just one example, it happens to me with various rules of thumb, so I'm looking for a solution in general.
Books newly in my pile, let me know if they're in yours:
Playing to Win, by David Sirlin
Reality, by Peter Kingsley (h/t reader J.)
The Design of Everyday Things, by Don Norman
This Is Your Brain On Birth Control, by Sarah Hill
Call Center Management on Fast Forward, by Brad Cleveland and Julia Mayben
Ten Nights in a Bar-Room and What I Saw There, by Timothy Shay Arthur
Natural Remote Viewing, by Jon Noble
Passage, by Connie Willis (h/t reader KL)
If your job was to bring together hunters from lots of different places, you'd probably get sick of having conversations at parties where you explain your work and then have to be like "no, I'm not that kind of hunter-gatherer...."
2025-11-03 21:35:12
Thanks for all the feedback on my recent How Weird Should This Blog Be post. My main takeaway is that (most of) you massively overestimate how much mail (most of) your favourite bloggers get; for most writers the answer is "very little," and you can meaningfully determine what someone writes about and/or how happy they feel about writing just by dropping them a line. Email your favourite blogger today!
One big thing that Silicon Valley is right about is the vital importance of tightening feedback loops. They didn't invent this, but I think a lot of businesses could benefit by adopting their cultural obsession with it. (Obvious caveat: I've never worked in Most Businesses, this is all just my view from the outside).
Ceteris paribus, in any given industry, I would bet on the business that has found a way to build an endless feedback-and-improvement loop over the ones that haven't. A lot of the time there's a way to make your product that allows you to constantly solicit and implement feedback, and many other ways that don't, and I don't think people who are not-implementing feedback loops are cognizant of how much they're missing, or even always that the high-feedback alternative exists.
For example, David Chang writes:
My first restaurant, Momofuku Noodle Bar, had an open kitchen. This wasn’t by choice—I didn’t have enough money or space to put it farther away from the diners. But cooking in front of my customers changed the way I look at food. In the early years, around 2004, we were improvising new recipes every day, and I could instantly tell what was working and what wasn’t by watching people eat.
Or Walt Hickey on Marvel comics as test loop for movies:
Marvel Comics was an R&D operation that went on for 60 years that tested, at a very cheap rate all things considered, ideas and concepts and characters, that ... were able to decisively determine what the best possible stories from the best possible characters were.
Making movies is very expensive, making comics is less-expensive, so publishing a ton of comics and seeing which ones readers respond to gives you a much tighter feedback loop than making movies blindly and waiting to see how viewers respond to those. (Even if you show the movies to test audiences before final editing and releasing, you're not getting anywhere near the amount of honing that comics gave you).
Or take my own business of board games. The creator of Settlers of Catan would supposedly playtest every single weekend with his family, stop whenever the game got bad, then change the rules for the next weekend to improve that bottleneck. I would bet on any boardgame creator who found a way to playtest every day, over one who playtests every week, over one who playtests every month. But a significant number of board game creators are only managing to test once per month.
There's a crazy element here where if your boardgame (or other product) is even moderately successful you're anticipating tens of thousands of hours of usage in the wild, which for some games is thousands of times more than they spent on testing it before they launched. This is insane! Some amount of feedback reaches the creator after publication via online reviews and angry emails, which hopefully can be incorporated into version 2, but overall there's a ridiculous disconnect and most of the useful information about the games gets lost (much like most of the information about how restaurant-dishes are received gets lost if the chef is in a closed kitchen).
A great advantage our digital-game-creator cousins have is that they can endlessly track and iterate their games over time; physical games and other physical products can't do that. But I suspect that we can get a lot closer to the ideal by somehow incentivizing more structured feedback from post-publication players, or releasing more and better prototypes with more and better feedback loops before going to market. I've experimented a little with this, but I think there's a million miles further to go, and that it's plausible I could tighten the loop by 100x and make my games 1000x better as a result. If you know how to do this, please let me know.
2025-11-01 03:00:02
A very happy 🎃 Halloween 🦇 to you and yours.
The most American Building. My grandfather slept in it while it was unfinished in between training stops in World War 2. My father took classes in it. I went to it on field trips. It’s a wonderful building.
Preach, Nabeel, Preach. I wonder why “education” rather than “age” has been what has sorted our politics 🤔.
On relationship between growth and trust, arguing that living through periods of higher GDP growth leads to higher societal trust. On one level, this squares well with the idea that trust is a mixture of competence, commitment, and character, with societies delivering growth being seen as competent. On the other hand, I would expect higher levels of trust to also unlock opportunities for faster growth.
Home field advantage in the NFL is actually real and it basically disappeared in 2020 when no fans were in the stadiums. Via Crémieux.
More than 98% of new vehicle sales in Norway were EVs in September. From Elective via Anton.
Unconfirmed but from a reliable source: Amazon drivers are paid 12 cents per package delivered.
A growing share of Americans (+13%) say Religion is gaining influence in American life according to Pew. I’m not sure how to square this with the thing I learned last week, that support for declaring the United States a Christian Nation is falling amongst Christians or that the fastest growing Catholic sects are the strictest ones. Strange things are happening!
For the first time in 35 years, no rap songs are in the top 40. Rolling Stone.
Should we care about process or outcomes? Some really successful people (see Tom Brady here ) seem more to favor the process over the result while others favor the result over the process (see Phil Knight / Nike and Sam Altman). How should I make sense of this?
What would have to change for Western Society to become less individualistic? Is it possible for Western Society to become more individualistic? What would it look like to short individualism?
Episode 3 of Dangerously Skip Permissions is next week: LLM pricing is broken, but not in the way that you think with my friend Anjali Shrivastava.
2025-10-29 20:13:23
I have conflicted feelings about how weird and experimental a blog like this should be.
One model is like a restaurant: a blog exists to serve up variations on a particular theme. When I go to an Italian restaurant I am doing it because I want to eat some kind of wheat-cheese-tomato combo, and while I'm fine with the restaurant adding some light original flourishes, I will be fundamentally betrayed if they suddenly serve me a thai green curry. And this is true even though I love green curry, and even if it's unexpectedly a good green curry.
Another model is like a snack conglomerate: there's some limitations on the variety of outputs based on the skills and resources of the proprietors, and maybe some very loose sense boundaries to what kinds of things count as "a snack," but fundamentally the output can cover a lot of ground, and its understood by everyone that there will be misses as well as hits.
Part of the issue here is that there's two different ways to interact with this (or any) blog: you can be a regular reader who gets the new installments every week, or you can find yourself reading one standalone piece that fell into your stream somehow.
I suspect the regular readers might have a relationship to their favoured blogs more like a restaurant – I know I'd feel weird if Matt Levine suddenly started publishing speculative fiction under the banner of his finance newsletter, even though I suspect he'd write incredibly good spec-fic.
Meanwhile, the drive-through readers are treating this more like a snack they picked up at a train station – even if they enjoyed it they probably don't remember the name of the piece, and certainly not the name of the publication, so they really don't mind how (in)consistent the range of offerings is.
So far I've attempted only very minimal variation in my writing on this blog – I did try a couple of fiction pieces, which got 0 explicit reaction and which I therefore assume were not-good, but mostly I think the pieces are pretty consistent within a recognizable cluster of themes.
But I contain a great deal more weirdness, and could make this blog much weirder if desired; I also contain a great deal of classic atoms-vs-bits-ness, and could keep the blog whatever-it-currently-is as well.
What would you like to see here? Some of you will immediately tell me that I should do whatever I want to do, which is kind, but assume that I will incorporate my own preferences separately, and I care deeply about your own desires: how weird do you want the blog to be?