Atoms vs BitsModify

2025-06-04 21:25:49

I sometimes go watch live standup. Here's a few thoughts:

If you go to an open mic for music, you'll usually see 1) amateur musicians 2) playing professional musicians' songs.

And this makes sense: playing for an audience is harder, and writing your own songs is hard, and if you're going to do hard things (I'd say) it's often better to practice one hard bit at a time instead of multiple hard things simultaneously.

If you go to an open mic for standup comedy, you will often see someone disintegrate on stage, but you will never see anybody telling anyone else's jokes. Using someone else's material is so taboo in standup that I'm scared to even write about it in case I get banned from all comedy clubs.

And I totally understand this in terms of undisclosed joke-telling (much like it would be terrible form to play someone else's song and pretend you wrote it!).

But surely the idea of "covers" could exist in standup? "Tonight I'm going to perform this bit by Tommy Cooper" or whatever, and then you do it.

Surely this would allow amateur comedians to practice delivery and stage presence and just not freaking out in front of an audience while using material that is provably funny to begin with?

Speaking of stolen bits, I stole the following thought: standup is the most empirical humanities discipline, and it shows. Standups get on stage again and again and get real, concrete feedback about every little part of their act. It's true Deliberate Practice – regular feedback on SMALL PARTS which can be worked on individually – in a way that I don't think exists for writers or painters or musicians.

When I was younger I dreamed of setting up a writing program that had this property, but I was too lazy and ineffectual so I never did. I'm convinced that we could improve writing by at least 2x and possibly 10x if writers trained as deliberately and effectively as standups (or, for that matter, athletes).

I think if I had been put in such a program as a teenager I would've been insanely gratified. But I'm not strong willed enough to put myself through such a program as an adult.

This guy wrote a standup comedy syllabus to teach you how to comedy. I'm not going to do it because it's long and I'm lazy, but I think it's fundamentally what good pedagogy probably looks like: lots of practice, specific questions and analysis (both for your own work and other people's), then more practice and do it again.

2025-06-02 21:52:28

While I was in kindergarten everybody had to play the recorder, when it came time for that.[^1] This is presumably because recorders are cheap to produce and easy to manufacture, they're relatively easy to learn, they're somewhat easy to make nice sounds with, your guess is as good as mine.

Personally I put down my recorder at around age 7 and haven't thought about it again for more than 2 minutes since. But there are people who hold on to their recorder-ing as adults: here's a blogpost on the life of a professional recorder-player, here's a recorder youtube video with 2.5 million views, here's an FAQ from the American Recorder Society.

Sometimes I imagine what it's like to be a passionate recorder-enthusiast. My hope and dream is that when you meet new people, and they ask what you're up to, and you say you're recording your new recorder-album, they say "oh isn't that fun" or "ah how lovely" and then let you talk for a bit about why you're still into this thing that (as other people tell you, endlessly) everyone else gave up at age 7.

That is to say: playing the recorder is a pretty niche hobby, but everyone knows it's a niche hobby, including everyone who's into it and everyone who isn't. Having a niche hobby is swell, the world is richer for having so many niches. But there's a fundamental congruence between the place of recorders in adult society and the perceived place of recorders in adult society: nobody feels badly that they're not into recorders, and nobody thinks you're judging them for not being a recorderer, because there's no cultural baggage attached to any relationship one might have (or not-have) with this instrument. Under heaven, more than is dreamed of in anyone's philosophy, there are a great many things, and recorders are one of those things, and that's great. There is a deep and satisfying harmony. The world is at peace.

When I was in high school I was into literary fiction. Much like the recorder, literary fiction is foisted on children of a certain age, and most of them don't care for it, and give it up as soon as they're not-forced-to. A very few people actually genuinely love it, and stick with it even after they don't have to. But that's not how it feels.

You can speculate on why exactly literary fiction feels so different from recorders: whether it's an artifact of our university system, or path-dependency in our media ecosystem, or the fact that for a long time writing was the only mass-media and people who like writing about things are disproportionately biased towards writing about writing. (Whatever else you can say about it, history is written by the writers.)

But my friend was once featured on the front cover of the New York Times Book Review and said that sold exactly 0 additional copies of their book; I think this is meaningful because 1) "The New Times Book Review" exists, 2) you'd think it's something that many people read, 3) either people aren't reading it, or people aren't buying the stuff it reviews, and I suspect the answer is both.

There is an entirely uncited claim going round the internet that there are 20,000 serious and consistent readers of literary fiction in the United States. I honestly, genuinely have no idea if this is true, and I don't want to go overboard relying on it given that it's a statistic and it's uncited, which is two black marks as far as I'm concerned. Still:

1) I suspect there are more than 20,000 adult recorder players in the United States.[^2]
2) Literary fiction has a kind of cultural footprint that is entirely disproportionate to the number of people who do it.

Here's a tweet I read:

I keep saying "literary fiction" but maybe what I mean is closer to "writing culture" in the tweet above. And sure we can argue about what exactly Literary Fiction means, and partly this is an artifact of the fact that if something sells enough copies we no longer consider it Literary: Colson Whitehead / Zadie Smith / Jonathan Franzen / the other David Mitchell would probably all be seen as Literary Fiction Authors if they weren't also multi-million sellers.[^3] But honestly, I think we can pull a Potter Stewart here: I know literary fiction when I see it, and almost nobody is seeing it.

I suspect that much unhappiness within people is caused by things that are not themselves bad, but by the gap between what is true and what is perceived to be true. It's fine to be bad at cooking, but if you think you're good at cooking when you're not then you might cause some issues. It's fine to be extroverted, but if everyone else thinks you're introverted you're all going to end up having trouble.

Of course, this is what I think is happening with literary fiction: it's a niche hobby whose troubles are caused by a perception (both among its adherents and its apostates) that it's much bigger than it is. There are so many fine hobbies out there in the universe: knitting, fishing, jumping rope (another thing I learned in elementary school). I don't have numbers, but I think each one of these is ~100 or ~1000 times more popular than literary fiction.[^4] But none of them have a New York Times supplement.

If you're into literary fiction, people do think you're judging them for not being into it, projecting-ly or otherwise, because people (for a certain segment of people, at least) think everyone else is reading lit-fic, and they alone are failing to. And again, I don't know-know what anyone else is doing in the privacy of their own homes, but I'm pretty sure it's mostly not-reading literature, and I think we'd be better off if we could all acknowledge that.

Again, there's nothing wrong with loving literature if that's what floats your boat: truly, poetry is as good as push-pin. I do think reading lit-fic specifically can show you interesting things about the human condition that I personally value deeply. But I suspect that so can fishing, and so can knitting, and so can playing the recorder. (So can sitting there silently and trying to perceive the true nature of sensory reality, but who has time for that?)

More than anything, I wish that you could go to a party as a literature-lover and enthuse about what literature means to you in a way that felt less like apologizing for the hegemon and more like stanning the recorder: "here is a thing I adore, among the many things in existence, and maybe you would like it too." The strange position of literature comes from the total disconnect between its seeming cultural hegemony and its actual place in the culture. (If I were better read I would here insert a reference to a novel that perfectly illustrates this tragedy, but I can't).

my biggest fear in writing this piece is that it might annoy my dear friend Henry, who I think will disagree with it; if you also disagree with it, please check out his blog.

[^1]: ok fine it was probably grade 1 or 2. But here's a poem I love:

While he was in kindergarten, everybody wanted to play
the tomtoms when it came time for that. You had to
run in order to get there first, and he would not.
So he always had a triangle.

[^2]: I emailed the American Recorder Society, who very kindly notified me that they don't know how many Americans play recorder, but that they have 2,500 members. I suspect that there are 10x more people doing any given thing than joining the membership association for that thing, but of course that's totally speculative.

[^3]: with apologies to Tom Lehrer: "to be lit / it must be it/terly without redeeming publishing fortunes."

[^4]: look, let me level with you: a problem I often have when writing blogposts is that I don't know if 100 people will read them or 100,000. My motivation to go out and get exact figures for every claim is small in the first case and large in the latter. If you (collectively) are interested enough in this piece, I will go back and put in some better facts and figures.

2025-05-30 21:58:44

1a) in most companies, how much more power does the CEO have than the next most powerful person? Is it 1.1x, or 2x, or 10x, or 100x?

Does it scale with the number of people in the organization, e.g. (1 + n/10)x where n is the number of people in the org – i.e., the bigger the org, the more that the CEO is over-powered relative to her next subordinate?

1b) same question, but for politics.

2) what % of great actors are great because they have ~no personality? Like, we imagine them as super charismatic hotties because that's what we see from their characters, but they're actually like the undercoat you put on a wall so you can paint something interesting on top?

I buy a good amount of specialty coffee, and always looked at the "tasting notes" section with total bemusement. But then I watched a video by James Hoffman about coffee-bean buying, from which I liberated the following information that may be helpful to others (the video has very little information per minute, but the info is good, hence the liberation):

• fruit words in tasting notes are often (but not always) clueing how acidic a coffee is, and you probably have a preference over how acidic your coffee is, so try to pay attention to that and buy accordingly. The rubric is:
  • citrus or berry words = high acid
  • apple or pear = some acid
  • "transformed fruit" words, e.g. jam or candied orange = low acid
  • non-fruit-words, e.g. caramel, treacle, chocolate, nuts = no acidity
• Then, separately, some specialty coffees undergo "dry" or "natural" processing (instead of the usual "washed" processing), which often gives a kind of fermented taste, and Hoffman estimates 40% of drinkers hate that taste (while others love it). You should figure out if you like it or hate it and then buy accordingly, which means:
  • a coffee that says "dry" or "natural" process on the bag will usually have this flavor, UNLESS it specifically tells you otherwise
  • a coffee described with TROPICAL fruits (e.g. mango, pineapple) is often clueing this flavor

I do personally think that having a descriptive system that reads like a cryptic crossword is not ideal! But then nobody asked me.

2025-05-28 22:09:50

Good evening. Wow. Thank you, thank you, it's so good to be here. Welcome to the opening plenary of the twentieth annual meeting of the A.H.A. [Pause for applause]. 20 years, can you believe it? Gary's showing it a bit. Just kidding Gary.

Truly, though, I can't believe how much this organization has achieved in such a short time. Let me take you back a bit to 2025. [Slide of ridiculous 2025 fashion; pause for laughter]. In 2025, Louisiana required a license to be a florist. In Florida it took six years of study to be allowed to practice as an interior designer. Tennessee required 70 days of training to shampoo hair. In Kentucky you needed a license to put things in boxes. Eight states required a license for a job called "travel agent" – back when "agent" still meant a person. [Pause for laughter].

That's right: the health and safety of the public was being protected in all these domains, and yet unlicensed coding was permitted in – as it was then – all "50" states. Literal children were not only allowed, but actively encouraged, to learn to code without restriction. Can you believe that? Human viruses are combatted by extensively trained and (just as importantly) properly credentialed doctors. But at that strange time, computer viruses could run rampant with nothing but the distributed local efforts of motivated individuals to combat them. Unconscionable.

It was a difficult battle. So many of you were there in the trenches for it – Gary's got the gray hairs to prove it. I'm kidding Gary, you look great. Some of the stories of how we exactly got that first crucial legislation passed are perhaps, ahem, best saved for a more private venue. Let's just say I'm glad that we in Silicon Valley moved on from our "teenage" relationship to Washington DC, and figured out that when technologists and government work together, we can do more than either one of us could do on our own.

If our first ten years were a struggle to get the Responsible Coding Act into law, our second decade will be remembered for the staggering success of the Coding Internship program. It's hard to believe that twenty years ago an "intern" at a software company was a college student paid a market rate in the hopes of hiring them to a full time role in future! [Pause for laughter].

Now, through the extreme selflessness and generosity of our colleagues at the five MAGMA companies – generosity that has of course extended also to sponsoring tonight's beautiful gala, thank you colleagues – we have transformed coding education and ensured that every licensed coder in the United States has spent a rigorous two-year stretch at one of the five biggest companies on earth learning how reliable, robust corporate software must be built.

To the interns here today: shouldn't you be at work? [Scattered laughter]. I'm kidding, I'm kidding. Just know that the process is making you a better coder, and that all of us are stronger for having gone through it. Well, not us oldsters who were grandfathered in, but I've heard from all six of my assistants over the last two years that they're ultimately glad they went through it.

In the process, we've truly raised the bar on what it means to be a developer. Gone are the days of some 19 year old who thinks she can drop out of college and build a software business just because she "knows how to code." Gone are the days when an unregistered 30-something can come back to coding without keeping up their continuing education credits and association membership dues.

Most importantly: gone are the days when hundreds of thousands of inexperienced amateurs can sully the MOST important profession [pause awkwardly, unclear whether for applause] with their unprofessional work; thank you to Sarah for running our legal efforts on that and making sure that violators pay a heavy price. As of last year the law is finally up to date with the times, and we can finally demand jail time for unlicensed coders, even if the software never leaves their own device. You have to squash a lot of bugs to write good software! [Uncomfortable muttering].

Yes: in the bad old days of 2025, 100k students graduated from computer science degrees, and as many as 300k more "students" "graduated" from "bootcamps." I'm proud to say that – learning from the example of our esteemed rolemodels at the AMA – we have brought those numbers down to just 50,000 qualified coders joining the job market each year today. The market results speak for themselves: while average salaries in most industries have declined over the last ten years, average salaries for coders go up by up to 5% every year. Let that sink in: that's 50 thousand good-paying jobs created by the AHA for AHA members.

So, here's to you, American Hackers: let's make the next two decades as productive as the last two.

2025-05-26 22:06:14

There's a fun little nugget that, among American men aged 20-40 and over 7 ft tall, 17% play in the NBA. I say "nugget" not "fact" because this is potentially untrue, possibly because it neglects to consider that (until 2020) the NBA measured player heights including their shoes, itself a nugget that confirms the worst suspicions of everyone who dates men and uses dating apps.

But I digress. If we stacked enough measurable traits together, we could create a category in which even 50% of its members do play in the NBA – I'm making up the details but imagine something like [American men] [aged 20-40] [and over 7 ft tall] [and able to bench press 185lb 1x20] [and able to vertical jump 20 inches] [and played high school basketball].

Now: if you find someone in this category, and they're not in the NBA, can you say "well, they could clearly play in the NBA, if they wanted to?"

No you couldn't, or rather, you shouldn't. The more obviously NBA-worthy traits someone has stacked on top of each other without also being in the NBA, the more that we should suspect that they have some hidden traits that prevent them from NBA-ing: for example, maybe they're the kind of person who always chokes under pressure.

I think this is a common category of error, but I don't know a common name for it. It's got a little bit of selection bias and a little bit of adverse selection and a little bit of your grandmother telling you "if it looks too good to be true, it probably is."

Pending better ideas, I'm going to call it the bruised peach problem. There's a famous paper in economics about the difficulty of distinguishing a "lemon" (a bad used car) from a "peach" (a good one), given that the seller has a bunch of secret information that you as a buyer don't.

My claim is that an amazing-looking second-hand car that's priced well below market and still hasn't sold is disproportionately likely to be a lemon, or, more evocatively, an internally bruised peach.

Here's some examples:

• if you meet someone implausibly attractive and also single, there's arguably a higher chance they're a bruised peach than someone who is single for very obvious reasons. The flip side of this is that if you're single and people don't understand why, it might be better to convey to them a legible reason for it.
• if you find an apartment that's priced way under market and in a great location and with lots of space (etc etc etc), you should worry that it either 1) has a hidden but terrible flaw, 2) is a scam
• I have lots of friends who seem like they "could" earn a lot more money than they do, based on their skills and pedigree. The more such skills a person has, the more I suspect they actually couldn't in practice earn more money than they do, because of some invisible internal anti-making-money-trait.

I feel a bit weird writing this post, because sometimes a peach is really a peach, and I don't want to be causally responsible for anyone missing out on a find. But mostly, if it looks too good to be true, it probably is, and specifically it probably has a hidden trait that makes up (down) for its visible ones.

2025-05-23 21:29:22

1) I often get told I end phonecalls prematurely, as in, I hang up before the other person thinks we've finished the Goodbye Ritual. I think this may be because TV has seeped into my brain too much, and on telly they don't want to "waste time" on repetition.

Supposedly the trend of people shouting into a speakerphone while hanging out in public also comes from television, where producers made reality show contestants do it in order to capture both sides of a conversation with minimal fuss, then actual humans inexplicably decided to do it too. Moral of the story: don't take your social cues from television!

2) if I send someone an email and they don't reply, there's generally a low chance that a week later I'll remember I sent it to them, and an even lower chance I'm upset that they didn't reply.

But if someone sends me an email and I haven't replied a week later, there's a high chance that I'm constantly thinking about that email and feeling terrible for not having replied yet.

Occasionally people message me to say "sorry I didn't reply to your email yet!", but (as mentioned) this is backwards because I don't care about that; people should be sending me an apology email when I haven't replied to them.

3) Amazon should do me a solid and make movie rentals 52 hours instead of 48. The current setup stresses me out, and often stops me from renting movies at all: I don't know if I'm going to finish it tonight, and if I don't finish it tonight I have to finish it tomorrow, and maybe tomorrow I won't feel like it.

I can't watch it overmorrow because 48 hours isn't enough for that, realistically: if I start at 8pm Monday I have to finish by 8pm Wednesday, which means watching the second half at 6pm or whatever, and realistically I shan't. But if they just stretched out my rental deadline a few more hours I would have three whole evenings to finish it, and boy that would be nice.

p.s. I emailed [email protected] with the above suggestion, because I read ages ago that this was a thing you could do. Only after sending the email did I remember that Jeff Bezos no longer works at Amazon. This makes no actual difference to how much/little my email will be read or acted on, but I do think it's fun that the referent and its meaning are now completely detached: in the future, perhaps, Jeff Bezos will be gone but [email protected] will still be a support address for Amazon, assuming Amazon + emails outlast Jeff Bezos.