> life in a monospace typeface

[life-in-a-monospace-typeface]

The Circle of Skill Issue

There's a concept in psychology of "Circles of Concern, Influence, and Control"

Your Circle of Concern is everything you care about, including things you can't necessarily affect, like the weather. Within that is your Circle of Influence, containing stuff you can affect but not control, like whether you get promoted at work. Within that, your Circle of Control is stuff you have full control over, like what words you say.

Importantly, the exact boundaries of the circle of control are fuzzy, movable, and usually not quite where people think they are. In particular, I like to imagine a circle that surrounds the circle of control, that I'll call "The Circle of Skill Issue". This contains things which seem like you can only influence them, or like you can't affect them at all, but which you actually could fully control, if you knew how.

Consider this newborn baby

This baby's circle of control is tiny, and doesn't contain things like basic bodily functions, or the movement of her limbs. But the thing is, her nerves are connected up the same as yours and mine - in a sense her body is already "fully under her control", she just hasn't learned how to control it yet.

This continues into adulthood. Some people can roll their tongues and some can't, some can waggle their ears or flare their nostrils and some can't. Importantly, the boundaries here are not obvious. Being unable to roll your tongue feels the same as being unable to waggle your ears - you try to make the muscle do the thing, and it doesn't work. But tongue rolling is genetic - many people will never be able to do it, it's forever outside of their circle of control, or even their circle of influence. Whereas basically anyone can learn ear waggling or nostril flaring, those are both just flexing specific muscles. If you can't control your ears and nostrils, it's not because they're literally outside your control - the nerves are all hooked up, the muscles are 'under your control', you just haven't learned how to use them yet. Such things are in your Circle of Skill Issue. These are things you can learn, and thereby move into your Circle of Control.

How though?

Learning to do this kind of thing usually looks like either spamming random muscle movement signals in the right general area and noticing if any of them cause your ears to move, or doing some conscious motion that includes moving the target muscles as a component, like smiling or raising your eyebrows, and then isolating the specific muscle from that. By carefully focusing on what mental motions cause which muscles to move, and practicing a bunch, you can steadily build up full deliberate control over these muscles.

I'm not saying you should do this for its own sake, to be clear - the reason you never learned how to use these muscles is because it's basically never useful to use them.

But the thing is, these kinds of things can be self-fulfilling prophecies. Practically speaking, you can't learn how to waggle your ears unless you believe that ear waggling can be learned. If you believe it's outside your control, you're automatically correct.

This isn't hypothetical, because... I kind of lied before, tongue rolling is not purely genetic! A study found that for 7 out of 33 pairs of identical twins, one twin could roll their tongue and the other couldn't. This suggests that there are currently large numbers of people in the world who learned in school that tongue rolling was genetic, tried to do it once and couldn't, and then never tried again, even though they could learn if they tried!

Tongue rolling isn't important, but consider the broader implications! You probably currently have things which are fully in your control, but which you think of as fully outside of it! Some of them might be very important!

If our culture holds that something is outside of our control, we don't have a reason to try to learn how to control it, so we can't tell whether its controllable or not.

Some things that we are told are outside of our control but sometimes might not be:

Taste in food - I already know this one is controllable because I've learned how to do it, and now there are basically no foods I dislike

Emotions - I've never felt emotions very strongly so I can't tell whether I'm good at controlling them, or whether I just don't have much to control. Or, for that matter, whether those might be two sides of the same coin

Hiccups/Coughing/Sneezing? - I've definitely just Decided Not To Have Hiccups Any More and had that work

This is all very low confidence, and can of course veer dangerously close to unhelpful and harmful sentiments that amount to "have you tried being normal?", which is not at all what I mean. Many things are in fact totally outside of your control, and the last thing I want to do is encourage people to give themselves a hard time about things they can't change. If you read this and think some variation on "Ah, as I suspected, my problems are all just caused by me being Bad and just Not Trying Hard Enough", then The Law of Equal and Opposite Advice is hitting you really hard and you likely need to move in the opposite direction.

But like, have you actually tried though?

If you don't know how to waggle your ears, flare your nostrils, or roll your tongue, I recommend you try learning it. (Or, if you know all of those, then learn crossing your eyes at will, or going wall-eyed at will, or de-focusing your vision at will, etc). Not because these are useful things to learn in their own right, but because it's useful to learn (or I guess to remind yourself) what it feels like to learn to control something that you previously had no control over.

If you can do that, I do think it's worth experimenting with applying the same general types of mental motions - the moves which would allow you to learn how to waggle your ears - to all kinds of other things that you probably don't have control over, just to see. Just take some time to consider, like "If I assume for a minute that this thing is actually fully under my control, and I just don't know how to work it yet, what would I need to do in order to figure it out?". Maybe you can spam all kinds of random possibly-related mental motions/commands and see if any of them affect it. Or, the next time the thing you want to do 'just happens' or happens by accident, introspect carefully on the experience, and see if you can reproduce any of the mental motions that feel connected to it.

You may have control over more than you realise.

[life-in-a-monospace-typeface]

As It Was

"A society grows great when old men plant trees in whose shade they shall never sit" - Proverb

I want to share with you my blurry memory of an experience I had more than a decade ago, that changed how I see the world.

Back when I lived in Nottingham for my PhD, I was lead singer of a little string band. We called ourselves the Nottinghack Time Band, and we mostly sang covers of random folk music. Songs about love and ships and mining (well... about heartbreak, shipwrecks and cave-ins). This was mostly for our own amusement - making music with friends is extremely fun. We only went on tour within walking distance of the Nottingham Hackspace, and that was it. But we had a decent repertoire of songs we could play from memory, we had some nice harmonies worked out, and at our best we sounded fairly good!

I don't know how we came to the attention of the Cider Club, but that's the experience I wanted to talk about.

The Nottingham Cider Club is an interesting organisation. Old-school socialist vibes - it's a long running non-profit organisation, which for a very long time has operated the same plot of communal land, on which they grow apple trees, the oldest of which were planted by the grandfathers or great grandfathers of people currently in the club. Throughout the year, they work together to tend to the trees. They prune them and cut off the dead branches and pile them up, and when the autumn comes around and it's time to harvest, they all come together to gather the apples, clean them, and press them in home-built contraptions. They ferment the juice in the way their predecessors taught them, and they end up with a large amount of home-made cider. At which point, what else? They have a party. And someone there must have heard we had a little acoustic string folk band, so we were invited.

The sun was low in the sky when we arrived to the orchard - of course the party's held in the orchard, where better? The leaves were beautiful in the golden evening light as we wandered between the trees, planted not in neat rows, but freeform like a forest. We hung out under their branches and chatted, learning about the club and the equipment and the process. I love to listen when people feel happy to pour out every detail of their obsessive expertise. And they were experts - they've been at this a long time. I'm not much of a cider drinker, but this was good stuff, and deceptively powerful. We were soon in a state where it wasn't too scary to be playing music in front of people who until recently were strangers. So we played. As the sun went down, the people collected the dead branches together and made a bonfire, which we sat around as we sang our songs. And as the cider went down, the people got up and danced to our music.

I remember, towards the end of a song, feeling a great joy about the music I was making with my friends, and how well it was being appreciated. I looked up at the stars in the cold night sky, and at the branches of these beautiful trees, lit by the light of the warming fire. I saw the same light on the joyful faces of the dancers, the grandchildren and great grandchildren of the people who planted these trees. These noble trees, that warmed us from without by the fire made from their branches, and from within by the cider made from their fruit. We were celebrating the trees, and perhaps we were celebrating simply being alive.

As the last notes of the song died away, I sensed something within me that felt incredibly old. I thought "Oh. This scene would have made perfect sense to every one of my ancestors, back to pre-history. This is what music is about. This is simply what it means to have a good time, as a human being on earth. This is what it's always been about".

I love modernity. I love having every song ever recorded available to play in high quality at any time for free. I love that we have access to more than one party's worth of cider per year. But this is a strange way to live, and it is worthwhile to sometimes connect with something more like the way things were for almost the entirety of history, when every experience of listening to music was paired with a simultaneous experience of making music. There are pockets of it still to be found, and some may yet be created.

[life-in-a-monospace-typeface]

I wish they’d made me memorise more

I had a good education, in a school system that had left behind the Bad Old Days, when teachers would hit you with a stick if you didn’t rote-memorise endless facts. I was basically never required to memorise anything by rote. I was never threatened with violence either (at least, not by teachers)^1, and I think those two things may be related. I mean, you do basically need to threaten kids to get them to do rote memorisation - even they can see that there’s no point to it. You can just look things up! What’s the use of stuffing your head full to bursting with a load of dry dusty facts?

Except, that’s not how learning works, is it? The brain doesn’t store knowledge like a hard drive, with a fixed amount of space that you can fill by learning, and if you learn too many things you have to erase other things to make room. The brain stores knowledge like a network, or a graph. You learn new things by connecting them to things you already know. New facts aren’t just inserted into the brain in the next empty space, they’re added to the existing structure, connected up to the surrounding knowledge.

If you come across a fact that has no connection to anything you know, your brain decides (usually correctly) that this fact is not useful to you - you can’t do much with it. A beautiful old stone building you walk past every day has a plaque that says “1667”. Ok, so what? If you don’t know anything else related to that, why would your brain bother to hang on to the number? It’s meaningless. The only way to get that kind of fact to stay in your head, without connecting it to your existing knowledge, is sheer repetition ^2.

But if you do have some related knowledge, like, perhaps you know that the Great Fire of London was in 1666, well, now that plaque means something to you. This building was built the year after a huge fire burned down most of the city. I wonder what was there before? Then later, you read in passing that the plague killed like 20% of Londoners in 1665, and again this date might have been meaningless, but now it has things to connect to, so instead of sliding off it, you think “Wow, what an incredibly shitty time to be in London. The vibe while they were putting up that building must have been wild”. And so on. Each thing you know gives more places for new facts to stick to, and those new facts give you yet more possible connection points. The more you know, the easier it is to learn.

And once you know enough, you can connect up the different parts of your knowledge into one huge graph that makes most new facts you come across just automatically stick in your brain without trying. It is a joy to go through life with your understanding of the world effortlessly growing and deepening each day, in a virtuous cycle.

But before you can fly, you have to get up off the floor! If new facts hit your brain one at a time at random, each fact will probably find nothing to connect to, and just fall out again, leaving no purchase for the next one. A few will stick, so eventually you will start to build out from the basic things you started off knowing, but only very slowly, because you’ve haven’t got much surface area for new knowledge to stick to. Most of the information you’re exposed to will bounce off and be wasted.

What you really want, to get this feedback loop kick-started, is a kind of low density ‘scaffolding’ of facts, that covers a wide area, so that new incoming facts of all kinds have a decent chance of finding something nearby to stick to. You could think of it like, you want to spread out a wide area of ‘nucleation points’ for new clusters of knowledge to start growing around.

For example, in order to have dates not just bounce off, you could learn the dates of major historical events: to get full coverage, you probably want to know the boundaries between periods - starts and ends of wars, reigns of the rulers, that kind of thing. To give information about places a chance to stick, you might want to just learn what countries there are, and maybe their capitals.

It doesn’t make sense to go crazy with this, you only want a small fraction of your learning to be this kind of memorisation, since it’s a much harder way to learn any given fact than just slotting it into an existing structure in a way that makes sense. But if you use memorisation efficiently to build a framework skeleton to grow the rest of your knowledge on, it helps enormously.

The problem is, I didn’t realise this kind of thing was a good idea until after my formal education was done, and to be honest I haven’t been very successful at doing it on my own, without external motivation structures. And the worst thing is, now I’ve had years and years of valuable information just bouncing off my brain, where a lot of it might have stuck if I’d only had the structure in place to catch it. I wish the value of memorisation was explained to me early enough to make the most of it.

Footnote 1: I dimly recall that one of my teachers (an older one when I was very young), still kept the stick in the classroom and would talk about how he’s not allowed to use it any more.

Footnote 2: (or mnemonic techniques, which basically all work by artificially inventing connections between the fact and things you already know).

[life-in-a-monospace-typeface]

Tumblr algorithm has developed an UNSHAKABLE CONVICTION that I am a huge fan of some game of thrones spin off show I've never heard of before. I swear I've only ever interacted with anything related to this show via the "Not interested in this post" feature, but no, it would seem I am very much IN this fandom, whether I want to be or not. Whatever I do, my dash is permanently about 50% posts about how gay everyone is for this guy with deer antlers. I'm at peace with it tbh

[life-in-a-monospace-typeface]

Reasons to be Unreasonable

Imagine you've just bought a tasty ice cream (+1 utility), but before you can start to eat it, a big guy grabs it from you and walks away. As a reasonable and rational agent, you think:

"Ok, I could fight that guy, but he's bigger than me and I'd lose. I could hurt him a bit (-2 utility to him), but he'd hurt me a lot (-10 utility to me). So I guess the best thing to do here is to let him go"

But suppose you were a lot less rational. Suppose you were the type of person who, if wronged in that way, would get implacably furious, and your anger would overwhelm your reason. You wouldn't be able to stop yourself from fighting the thief, even though you knew you'd lose and pay dearly for it. That version of you would get your ass kicked.

So the rational version ends up at -1, and the emotional one at -11.

But realistically, one of those people is going to get their ice cream stolen a lot more often than the other, aren't they? If you're prone to rage, and the thief can tell that ahead of time, they'll reason that the +1 of benefit they'd get from taking your ice cream isn't worth the -2 of injury they'd sustain while winning the ensuing fight, so they don't bother taking it in the first place, and the situation never even comes up.

In this regime, the fact that your anger is able to overwhelm your reason is a feature, not a bug. Your emotions can force you, in certain bad situations, to make obviously wrong decisions that hurt you. But by doing so, they can prevent you from being in those bad situations in the first place, so you come out ahead overall.

This means that in the more realistic setting, the irrational person is at +1 and the rational one is at -1!

This should cause us to question what we mean by the word 'rational' - surely the rational approach is whichever one wins. Perhaps we can be more rational by allowing our reason to be overwhelmed by anger? But it's not just anger, several other emotions seem to follow this pattern, including guilt, shame, and jealousy. Is the rational thing to just let these emotions drive us? That doesn't sound right.

There is, of course, a second level of rationality, where you implement the correct policy but without necessarily feeling the emotion. You think "Well, for game theoretic reasons I need to be the kind of person who would fight this thief even though I know I'll lose", and then you fight the thief. In principle, you can work out what job the emotion is doing, and implement the same function using reason.

You might even be able to implement something better! After all, emotions can be pretty error-prone - people often get really angry at times when it doesn't make sense (even taking into account this game theory argument), and also often fail to get angry at times when the exact same dynamic applies but just in a way that doesn't trigger visceral emotions. With reason, you can continue to take the actions that anger would suggest, but only when it would be correct to do so.

We can aspire to this, but it's pretty difficult in practice.

The first problem is that the soft animal of your body may not fully implement the policy you decide on. Fighting a man who's overwhelmed with rage is a different experience from fighting a man who's dispassionately decided that fighting is necessary for game theory reasons. Without the help of adrenaline and cortisol, you may not be able to inflict the necessary -2 of injury.

The second problem is that it's not enough for you to simply be the type of person who reliably imposes costs on those who wrong you such that it's not worth it to wrong you, you also have to visibly be that type of person. The consequences for fucking with you have to be predictable, and more than that they have to be predictable to the kind of person who steals ice-creams, and such people often don't have very sophisticated predictive abilities.

It's not obvious how to solve these problems within this framework. In practice, rational people will structure society such that ice-cream stealing is dealt with by a system of professional norm-enforcers whose threats are legible and credible without relying on emotionality. But this has its own issues.

[life-in-a-monospace-typeface]

You absolutely cannot expect decent tea to be available in this country, so when I'm in the USA I always carry an individually wrapped teabag. I also, for largely unrelated reasons, always carry a condom. I have now mistaken one for the other in both directions.

[life-in-a-monospace-typeface]

There Should Be More Common-Premise Short Story Collections

[Content warnings for death of all kinds, including illness, accident, murder, and suicide]

Machine of Death is a fascinating and excellent book. The premise is simple:

The machine had been invented a few years ago: a machine that could tell, from just a sample of your blood, how you were going to die. It didn’t give you a date and it didn’t give you specifics. It just spat out a piece of paper upon which were printed, in careful block letters, the words “DROWNED” or “CANCER” or “OLD AGE” or “CHOKED ON A HANDFUL OF POPCORN”. It let people know how they were going to die. [...] But the machine was frustratingly vague in its predictions: dark, and seemingly delighting in the ambiguities of language [...]

People became at once less fearful and more afraid. There's no reason not to go skydiving if you know your prediction says "BURIED ALIVE." But the realization that these predictions seemed to revel in turnabout and surprise put a damper on things. It made the predictions more sinister: Yes, skydiving should be safe if you were going to be buried alive, but what if you landed in a gravel pit? What if you were buried alive, not in dirt, but in something else? And would being caught in a collapsing building count as being buried alive?

The premise is great; I love inexorable prophecy in stories. But the thing that makes Machine of Death really special, in my opinion, is the format: The book isn't a story based on this premise, it's 34 stories, each by a different author, all based on this premise.

Each writer goes somewhere quite different with it. The high-effort parents trying to get their son into a fancy school, although his card says PRISON KNIFE FIGHT. The man whose job is to find creative workarounds so that his boss's assassinations technically count as matching the victims' predictions. The mental health effects of finding that your card says SUICIDE. The elite team of soldiers, fearless in battle because they know they all die from ALZHEIMER'S.

How hard do you fight the CANCER that you know will win in the end? How many people with plausibly plane-crash related deaths should be allowed to board the same flight? How do you conduct military strategy when you can be pretty sure which soldiers will live and which will die?

What does the world come to look like, what's the equilibrium?

The prompt raises so many questions, and trying to answer them in one setting, one continuity, one story, would be a chaotic mess. One author's perspective isn't enough. A collection of short stories is the perfect way to handle this, each story can explore an interesting angle on the prompt, and be tightly focussed on that.

I want to see more books like this! I think the Machine of Death premise is unusually rich, but I'm sure there are lots of prompts that could produce similarly wonderful collections of work. Probably some have already been compiled, but I don't know of them.

What is this format of book called, and how can we get more?

[life-in-a-monospace-typeface]

Ultra-portable Transportation is Underrated

I used an electric skateboard as my main form of transport for several years in Nottingham UK, and it was fantastic. Now I'm in Berkeley CA and it's even better (less rain). I think these types of vehicles, which I'll call "Ulta-portable Transportation" are the best option for far more people and far more situations than I see them being used, and I want to encourage people to try them.

What do I mean by "Ulta-portable"?

It's a fuzzy definition, but my categories are like this:

Not portable: You can't carry it. Car, Motorbike, Aircraft Carrier

Portable: You can carry it if you need to. Electric scooter, road bike, most electric bikes

Ultra-portable: You can carry it basically anywhere. (Electric) skateboard, OneWheel, Brompton Bike (just about)

The key tests are:

Do you need to lock it up, or can you bring it into the office/restaurant/store?

Is that still true if there's a flight of stairs?

Can you throw it in the back of a car/taxi?

What's so good about Ultra-portable Transport?

An electric skateboard scores extremely high on a metric I call "Not Having to Think About It":

Road Tax, Insurance, License? Don't have to think about any of that.

Parking is never an issue. (This is kind of true with portable transport too, but you still need to find a place to lock up, which is usually easy but not always).

Now, you can use Uber to get the benefits of a car without worrying about owning, taxing, insuring, or parking a car. But then you have to deal with the unpredictability and inflexibility that entails. You never know how long it will take to arrive, until you actually book, and even then there's variation.

You also never know how long the actual trip will take, because of Traffic. That's never an issue with portables and ultra-portables.

The combination of no waiting time, no traffic, and no looking for parking, means that Ultra-portables are extremely consistent. If it takes me 10 minutes to skate to the office, it takes 10 minutes every time. If I leave now, I'll be at my desk in 10 minutes. Maybe driving is faster overall and usually only takes 5 minutes door to door! But add in the variation from traffic, plus either variation in waiting for the Uber, or variation in time spent looking for a parking spot (and in walking from the spot to the destination), and it could take 5 minutes, but it could take 15. And if it could take 15 minutes, you have to allocate 15 minutes. So even if it's normally 5, an uncertain 5 isn't as good as a certain 10.

Most of these things are true of a normal bike though?

The thing that Ultra-portables have over mere Portables is that you can have them with you in many more situations, which gives you flexibility.

Here's a situation that often happens: I cycle to work, then at the end of the day some people are going out somewhere - a restaurant, someone's house, a climbing gym, etc. They've ordered an Uber, am I coming? Well, ok, I can either cycle there separately, which is annoying, or I can leave my bike at the office. If I do that, I either come back there at the end of the evening to cycle home, which is annoying, or I find some other way to get to work tomorrow, which is annoying. With an ultra-portable I just say "Sure!", and throw it in the back of the car. Then I can go back home using it, or rideshare with others, whichever is more convenient, and I have it ready for the next day.

The combination of public transit and ultra-portables is especially powerful. Car, bus, train, whatever, you can just bring it with you, don't need to think much about it. I can use BART to get across the bay, and then skate from there.

I only wish Google Maps had a route planning option that's like "Plan me a route by public transit, but assume I walk at 15mph".

Isn't it dangerous?

In my experience, not really! As long as you do it right - that is:

Helmet, always. I have a folding helmet that I can throw in a backpack easily. Also, wrist guards - these are little metal strips that strap to your wrists. By far the most likely kinds of accident are basically equivalent to falling over while running. You put your hands out to catch yourself, and might sprain or break a wrist, and/or scratch up your hands on the ground. Wrist guards prevent all of that, and they're small and light to bring with you.

I used a skateboard as my main transport for years and only fell off a couple of times, and it was never a big deal. And when I did fall, it was because I was doing something a little silly.

That's the main thing actually: Be sensible and cautious, and ride defensively. These things are so fun that it tempts you to do risky things, but I claim that this hugely skews the statistics, and almost all of the danger comes from people riding in a very fun but dangerous way, instead of a still pretty fun but safe way. As long as you're alert and highly risk averse, it's really not very dangerous. Some rules I try to follow:

Never pass a truck on the inside of a turn

Don't go the fastest speed you possibly can go

Actually stop for lights and signs

Never assume a driver knows you're there

Never be on the inside side of a truck when it goes around a corner

Every parked car is about to swing its door open

Every pedestrian near a curb is about to step into the road without looking

Basically everyone else on the road is always about to do the dumbest thing they could possibly do

Never be alongside a long vehicle like a truck, especially in the direction it might turn

Be patient, be calm, don't get tilted

At night, light yourself up like a christmas tree

Do your very best not to ever be on the left of a truck that's turning left, or on the right of a truck that's turning right

Seriously, if you see an article about someone who actually died on a bike or skateboard in normal road use, check to see if it could have been prevented with rule 1, 5, 9 or 12.

To be honest, cycling or skating feel a lot safer to me than driving, because I'd rather be injured or killed on the road than have to deal with knowing I killed someone else's kid on the road, and there's enormously lower risk of that on a bike or board.

[life-in-a-monospace-typeface]

Idea: Indie Game Unearthing Competition

I was watching a Jacob Geller video recently, where he’s talking about how he discovers all the fascinating obscure games he often covers in his video essays:

I do not have a magical ability to only play good indie games; I play a lot of duds you don’t hear me talk about on this channel! But that search, stepping just outside the mainstream, can be endlessly rewarding. Because the same shadows that keep these games from commercial success conceal what they’re truly capable of. Direct your attention outside the spotlight for a little while - those hundred-million dollar blockbusters will still be there next month. Explore the dark. And then, please, show me what you found!

That got me thinking about how something like this could be done on a community scale. I thought “Man, if everyone watching this video could just play one or two really obscure games, and share the good ones, that would be amazing!”.

So here’s my idea: A competition with a prize, kind of like a game jam, except the goal is to discover a game, rather than creating one.

The rules could be pretty simple. Competitors must do two things:

Submit a game

Play two submitted games they’re assigned, and say which one they think is better

The judgements are used in a tournament structure to pick a winner. However, a game is eliminated from the competition if either:

More than one person submits the same game, or,

The person assigned to judge a game has heard about it before

So the way to win is to submit the best game that nobody else submitted, and that the judges don’t know about.

The ideal strategy of where exactly on the "good <--> obscure" trade-off to aim for is kind of interesting. If you submit a game that's awesome but not that obscure, you might get lucky and randomly draw judges who haven’t heard of many games, but... consider what kind of person is likely to enter an Indie Game Discovery Competition!

Of course, one way to ensure nobody already knows your submission would be to submit a new game you made yourself. Maybe we’d want to disallow that, but maybe it would be fine, and the competition is a kind of hybrid thing. Amusingly in that case it’s almost like “Game jam except plagiarism is fine as long as the judges don’t notice”.

If we want to incentivise obscurity less, we could also set the standard to "The judge hasn't played the game before" rather than "The judge hasn't heard of the game before".

Either way, I think if an event like this were run and promoted by a person or organisation with a significant following (like Jacob Geller, in fact), it would be a big success. Unlike a game jam, it doesn’t ask that much of the competitors. A lot of people would be happy to do something that amounts to “Get recommended two video games, tell us which you prefer, and also maybe win a prize and some bragging rights by getting a bunch of attention on a game you like”.

If this were run annually, I expect it would become an institution, regularly shining a spotlight on great new indie games that deserve to reach a wider audience.

[life-in-a-monospace-typeface]

What I Learned Writing A Blog Post Every Day

This is the last day of Inkhaven, a residency program where I had to write a blog post every single day for the month of November. When I started this, I hoped to learn things about writing (and about myself) that would help me to produce more and better YouTube videos. I don't really know if that's happened, but I did learn a lot, and I'm very glad I participated. Here are 5 things I learned:

1. You Can Just Write Things

This month I wrote in taxis, at airports, and on planes. I wrote on my phone while walking to the office, and in my head while cycling. I wrote while drunk at a party, and while on LSD in the middle of nowhere. Looking back over my posts, I can't recall which ones I wrote while drunk or high, and which I wrote in a carefully controlled environment. There's no obvious correlation between writing conditions and the quality of the end result. It turns out I really do not need ideal working conditions in order to write, or even to write well.

This is a pretty important thing to realise. I don't need to wait for the right time or the right conditions to write. I can just write things. As the excellent blog at inkhaven dot tumblr dot com puts it:

"If you don't write when you don't have time for it, you won't write when you do have time for it"

This wisdom is made all the more remarkable by the fact that that blog has no afiliation with the Inkhaven Residency, and in fact hasn't been updated since that post, which was more than a decade ago. The image links to a blog post about how to develop a daily writing habit, which ends:

"The number one piece of advice that professional writers will give to you is this: write. Do it every single day. In a month’s time, you’ll start to see where you are as a writer and where you’ll want to improve."

This is not a coincidence, because nothing is ever a coincidence.

2. I Don't Need to Outline and Plan

I tend to think of my writing process in the way I describe it in The Rambling Salesman Problem - I have a big unsorted list of Points I Want To Include, and then I need to plan out how I'm going to smoothly take the audience through all of them, while making sure to explain the prerequisites to a point before the point itself, and so on. This means a lot of time working with bullet points, re-ordering things and tweaking, and thinking about possible segues and narrative flow.

This seemed necessary to me, because I felt like when you write out the whole thing as prose, then re-ordering it and chopping and changing things becomes too clunky and confusing. It's too much to keep in your head, so you need to pin down the whole structure ahead of time as bullet points.

Perhaps it's just that the things I'm writing for Inkhaven are shorter than my typical video scripts, but the fact is, I didn't do that once this month. Time was short, so I always wanted to just be writing my final prose from the start.

Any time a part of the piece started coming to me, I just started typing it out in full. If the flow petered out, I just stopped there and moved on with a different part of the piece. If the thing that was alive for me at a given moment didn't connect with what I'd already written, I just added some extra newlines to indicate the disjunction between paragraphs, and started writing the other point out in full. If I came up with a new point I wanted to add somewhere but didn't want to write in full yet, I just added it as a bullet at the end of my draft.

As I was writing and re-reading, various segues and possible structures would occur to me, and I'd make decisions on the fly, moving paragraphs around and changing things, not worrying too much about the specifics of how everything lines up.

It was not, in fact, too much to hold in my head. If you're really writing, you're in it, and the whole thing just is in your head.

I realised I was doing the thing I talked about in Thoughts on Indecisiveness, or, Thoughts on Indecision?:

"I have a friend who embodies this. She decides things (let’s say) ten times faster than I do. So in the time it takes me to make one good decision, she’s made 4 good decisions, 2 ok ones, and 2 bad ones, and then discovered the issues with the 2 bad ones and fixed them. This means that even though she’s making mistakes sometimes, she’s functionally still doing like 6 to 8 times better than I am."

In the time it would have taken me to figure out the perfect way to write something, I just wrote it three different ways and kept the one that worked best.

3. I Mostly Want to Write About Non-AI Subjects

I expected to draft a lot of AI video scripts this month, and that didn't happen. I actually only wrote about AI 3 times, which is admittedly more than 3 times more often than I normally publish longform work about AI in a month, but it's also only 10% of my output. And those were not pieces I'm particularly proud of.

The things that most strongly drove me to write were largely, I now realise, the same type of thing that drove me to write my early AI videos: Rants I often find myself going off on. I used to rant to people about The Orthogonality Thesis and Instrumental Convergence and the like, but now I've spent several years surrounded by people who know all that stuff inside out. Nowadays I find myself ranting about things like:

"Acquiring Tastes" is a skill you can easily learn, hating foods seems strictly optional and it's wild that it's apparently my job to notice this

Navigating by the sky is extremely easy and (again) it's weird that it's not standard knowledge

British money is actually really well designed

In a similar vein, another class of thing that was very easy to write is "Stories I've told many times", like:

That time I ran a game at an indie game festival, which is the thing I feel the widest range of pride and shame about

The time I got banned from Wikipedia and then totally outplayed by a guy from Thailand

That 45 minute long game of "20 Questions" about the concept of Shibari (which unexpectedly now has more than 20,000 notes on Tumblr)

Actually, this is reminding me of some other stories I maybe should write up:

The guerilla art project I accidentally helped start, that's now on every continent

The time I learned what Music is meant to be for human beings

The Evil Empire I ran at school

And the last real category of easy to write posts is "Cool projects I never finished", like:

The videogame you play using only your eyebrows

The essay planning software tool

The dating app I'll probably never build

...and actually the Thai Wikipedia thing is also one of these

Which also makes me realise there are more of these to write up:

The Battleüke, a ukulele I built which is also a real battleaxe

A stained glass window I designed to make a mathematician from 1950 lose their shit in two distinct ways

521, my webapp that makes picking songs for jam sessions really really easy

4. I Will Not Run Out of Ideas

I started Inkhaven with a list of about 10 ideas for posts to write. I wrote 30 posts, and now my list has about 31 ideas for posts to write.

Interestingly, the little sessions people ran with the goal of "Let's come up with lots of post ideas" only resulted in a few posts each for me at the time, but they helped prime me to spot new ideas, and over the course of the month, most days I wrote down a few new ideas. That adds up.

I think I could keep writing a new post every day indefinitely, if I really wanted to. But...

5. The Daily Requirement is Too Fast for Real Work

There are a lot of posts I'd like to write which I couldn't, because I had to pick topics which I knew I could write in a day. I had some ideas that seemed promising but would need several hours of research/work, and the work might reveal that the idea isn't worth writing about. Other ideas were for posts about projects, which would work best if I finished up the project first. Those types of ideas almost never got written about, because I only had a day for each post.

It's not just that I had to pick topics I could write in a day, they had to be topics I knew I could write in a day. I couldn't risk having a situation where I spend all my mental energy on an idea that it turns out can't become something worth posting without taking more time than I have left, and then I have to hurriedly switch to some unknown backup post idea, when all my excitement and momentum is still tied up in the old idea.

This is a big part of what kept me from writing about AI. I expect I could write a good explainer about a technical AI Safety research paper in a day - a paper doesn't take more than a day to read. But if the paper turns out to be too complicated, or doesn't have any nicely summarisable findings, or I can't understand it well enough in the available time, I'm kind of screwed. So the stakes are a little too high for exploratory work.

The way to deal with this, I'm told, is to quickly dash off a simple post or two to keep in your back pocket so you can spend the rest of the day working on something bigger, and pull out a backup post if you need to ("remember, switching to your backup post is always faster than reloading"). But I was never able to do this. My writing process is still too slow, and I had lots of other things going on this month, so whenever I did finish a post early, I tended to spend the day on other things rather than extra credit Inkhaven writing.

Will I Keep Blogging?

I’m told the program will continue after the end of November, but at a weekly cadence, in a thing people are unfortunately calling 'Foreverhaven'. That seems like a much better release schedule though, so I’m in!

[life-in-a-monospace-typeface]

Dating App Idea: The Council

Consensus seems to be that the dating apps are not ok.

Interactions are shallow, atomised, impersonal. People ghost and flake, conversations die. Bots and scams and spam. People put in the minimum possible effort, and when you try to put in effort yourself, people make you wish you hadn't.

Perhaps a better designed app might help. So here's my (sadly extremely heteronormative) idea:

When a man matches with a woman, he can be put into a chat, not with her, but with a Council of 3 to 5 of her most trusted advisors. If he can impress them enough that they vote to let him talk to her, he’s in.

Why is this good for women?

I expect women’s experience of using this app would be much better than with most dating apps.

Much less harassment, for one thing. Most harassing or abusive messages would never be sent in the first place, because why would you send that kind of thing to a group-chat of gatekeepers? And if it does happen, facing it as a group makes it much easier to deal with.

The general standard of one on one conversation you experience on the app would be much higher, because you’re never one-on-one with a random guy, you only talk to people who your Council has voted to send on to you. You don't have to deal with a deluge of losers and time wasters, they’re filtered out for you by your Council. In fact I expect most of them would be filtered out even before then, by being unwilling to use this app in the first place.

So, less hassle in the app, talking to higher quality guys, and thus having better dates.

But also, the dates themselves are probably better, even holding the guy constant, because your Council can ask all the questions you'd want answered early on, but which you'd find awkward to ask. You probably don’t get nearly as much information on a first date as you’d like, because, well, you want to go on a date, not conduct a job interview. But first dates often end up being kind of job-interview-y anyway. If your Council conducts the job interview for you ahead of time though, that part is mostly over with before you're even really involved, the guy has already passed the test, and the first date is free to just be a date.

“But I don’t have 3-5 people who I trust / who know me well enough!”

Honestly I would probably try to fix that before trying to date.

Why is this good for men?

Look, if you want to properly date a woman, you're going to have to win over her friends anyway, so you may as well get that out of the way at the start.

Facing down a group of judgemental gatekeepers can be intimidating, but this is easily dealt with by not being a fucking coward. Did you expect to achieve romantic success without risking social discomfort? Faint heart never won fair lady, get in there and prove yourself!

That said, I expect the experience of using the app would actually be more fun and interesting than the average dating app - certainly more exciting, and probably much less corrosive to the soul:

For one, there’s far less ambiguity and uncertainty. You won’t wait hours for a reply, not knowing if she’s ghosted you - there are like 5 people in this group chat, someone will talk to you. You’ll basically never get kinda-ambiguously-pseudo-rejected by The Council - if they vote you down, 3 to 2 or 4 to 1 or unanimously, you know where you stand.

Plus, you’ll get a lot more feedback! The Council has no reason to be afraid to tell you exactly why you didn’t make the cut. The feedback might be hard to hear, but I’d rather have uncomfortable honesty than a blank featureless wall and no guidance on how to improve.

The Council also keeps you from wasting a lot of time on women you’re not compatible with. If the relationship is going to fail, it should fail fast. And this way, it fails before you have a chance to get attached.

And it’s a two way street. There’s a lot of information you can get about a woman through her Council which would be much harder to infer from her dating app profile. If the Council’s questions are all about kids, that tells you something. If the questions are all about your income, that tells you something. If they’re all about your politics, that tells you something. If a woman is looking for something that I’m not, I want to know that as early as possible.

So being rejected by a Council on those kinds of grounds shouldn’t feel too personal. If it’s super important to a woman that you share her politics, and you really don’t, that’s just saved you both a lot of hassle.

"Jesus, I don't want to run a gauntlet and meet a girl's whole social group, I'm just trying to get laid!"

That's ok! This app is not for you.

Why is it good to be on a Council?

Aren’t you tired of your awesome friends, who deserve all the romantic success in the world, making bad choices and dating assholes and losers? Don't you wish you could help them find a suitable boy?

Does it not sound fun to have a group chat where you all judge guys who want to date your friends? Be real, do you not already kind of have one? But what if you could grill those guys yourself? And you don't have to be chill and polite, you can ask the hard questions and make sure your friend doesn't get her time wasted.

Talking to randos on dating apps is not fun. Judging randos with a group of your friends is fun as hell.

Ok, but another dating app? Would it even catch on?

The viral potential is high - each woman who signs up needs to get 3-5 other people to be on her Council. That’s extra friction of course, but it also means 3-5 people will make accounts and be using the app regularly. It's natural that some of them might then want to start using the app for themselves, and appoint their own Council, and so on.

I don’t think it will be at all difficult to persuade people to sign up to sit on a friend’s Council. I’ve been led to believe that friends tend to enjoy hearing about, and involving themselves in, one another’s love life decisions.

And a Council Seat is obviously a position of high esteem, a strong signifier of trust and closeness. People will want to be on their friends' Councils simply for the sake of having been deemed worthy of the role. I imagine there will be considerable interpersonal drama about Council Seat appointments. All of this is excellent for the app's mindshare.

Isn’t this whole idea kind of horrible?

In a way yes, but it’s importantly a very different kind of horrible than the existing dating apps.

Plus we can licence that Spice Girls song for the ads.

[life-in-a-monospace-typeface]

The Rambling Salesman Problem

Sometimes a video script or essay will come to me fully formed, in a bolt of lightning. But usually, I have a general topic or subject I want to explain, and I need to find the story within it. I have a big unsorted pile of “points I want to hit”, and I need to figure out the path that will smoothly take the audience through everything in a coherent narrative which is engaging and understandable.

Every time I try to do this, in a messy ad-hoc way, I think about whether it could be automated, or at least if I can make tools to make it easier.

This post is some of my thinking about what that might look like.

Formalising the problem

Given a set of points to hit, there are two types of constraint at play - Dependencies and Segues.

Dependencies: Point A “depends on” point B if you need to explain B before you can explain A. This is naturally transitive - if A depends on B which depends on C, then A depends on C. You can think of laying out your points in a dependency tree, though really it's a directed acyclic graph (DAG), since a point can both depend on more than one point, and have more than one point depending on it.

Segues: Point A “segues into” point B if, when you get to the end of explaining point A, there's a smooth and natural way to flow into point B. So the segues form a simple directed graph.

You could also imagine segues as a matter of degree, where each ordered pair of points has a number associated with it which you might call a "segue cost". Perhaps a zero cost segue from A to B is so smooth that A and B can basically be thought of as a single point, and a cost 100 segue is so forced that you may as well be jumping to another point at random. If you're tracking the quality of segues like this, they form a weighted directed graph.

The Software

I'm imagining a program where you throw in all the points you want to hit, as well as their dependencies on each other and your ideas for segues between them, and the program runs an algorithm to solve what I'm calling The Rambling Salesman Problem, which is basically just:

"Find the path through these points which hits every point, while using the smoothest segues possible, and also never visiting a point before every point it depends on has already been visited"

This is a constrained version of the Travelling Salesman Problem, so it's NP-hard. But there are lots of good existing algorithms that quickly find approximate solutions, and approximate solutions are fine here, so we can just modify one of those.

We run the algorithm, it spits out an ordering of points, and you can use that as your outline!

Design Considerations

Practically speaking, I don't expect this to be a one and done, "big design up front" type of process. It's not practical to write out every dependency and segue at the start. I more expect that you put the obvious ones in, have the program generate a draft outline, and then when you read that through you'll realise things like "Oh wait, we shouldn't really try to explain X until we've explained Y", so you add in an "X depends on Y" and run it again, and keep iterating like that until you get something you like.

We could have a whole GUI, showing the different graphs and letting you add and adjust dependencies and segues, but I actually think I'd prefer to use a simple plaintext format like the DOT graph description language. I suppose there's no reason you couldn't do both, with the GUI modifying an underlying text representation.

Once this format and workflow is established, you could extend it in various ways. For example you could also add weights to the points themselves, indicating how valuable it is to include each point. That way there's a natural way to express things like “Here's a point that might be nice to include, but only if there's somewhere it fits in neatly”.

Why write about this without building it?

I am under no illusions that this project would save me time on net. I know full well that if I start building it, I'll always be futzing around with my writing program instead of writing. But this way, I can write about my writing program idea, without being distracted by my writing program!

[life-in-a-monospace-typeface]

Ten Things To Be Thankful For

Today is Thanksgiving, and I'm thinking about what I'm thankful for. Many of the things I'm most thankful for - my girlfriend, my family, my community - are specific to me, and perhaps not that interesting to you.

So here are some things I'm thankful for that I think you might benefit from reading about, because they're things you might also be thankful for, on reflection.

1. Free and Open Source Software

Especially the people who maintain and support it. Sometimes I feel possessed by the urge to create a piece of software, and I often then release the source code. That's a nice thing to do, and I'm grateful to those who do it. But I have never felt possessed by the urge to track down a bug which never affects my way of using the software, or to explain basic things about it to an endless stream of clueless people. Open source maintainers put up with a lot, and I'm grateful to them.

2. The Existence of Masochists

I'm not a sadist, but there seem to be quite a few of them. You could easily imagine a world that has sadists but no masochists, and that would be disastrous. Since we have sadists, it's great that there are also masochists. I hope everyone can get what they want.

3. People Who Organise Events

I love going to parties and meetups and reading groups and picnics, and I'm so grateful to the people who put in the work to make them happen.

4. Spiders Can't Fly, and Horses aren’t Predators

Nature is scary sometimes, but it could easily be enormously more scary than it is.

5. Health

Our bodies are absurd squishy Rube Goldberg machines, and it's astonishing that they just keep working in all conditions for decades at a stretch. There's a Jewish prayer/bracha, Asher Yatzar, which you're meant to say every time you use the bathroom. It captures what I mean:

"Blessed are You, our God, King of the universe, who formed man with wisdom and created within him many openings and many cavities. ... if even one of them would be ruptured, or if even one of them would be blocked, it would be impossible to survive and to stand before You"

Amen to that. Every time you use the bathroom and everything works as it should, that's something to be hugely grateful for.

6. The Moon and The Sun are The Same Size

Have you ever seen a full solar eclipse? Holy Shit. That really didn't have to be possible; if the Moon or Sun were a little differently sized in the sky, eclipses would be much less cool.

7. Polaris

It's the brightest star within like a 30 degree radius of sky, and it's also less than one degree away from being exactly due North? That's so handy!

8. Wikipedia

Everyone is so casual about the fact that Wikipedia exists, but a priori it's a completely insane and obviously impossible project. I'm thankful that Jimmy Wales dared to dream absurdly big, and I’m thankful that the group of amazing internet weirdos he somehow corralled together actually managed to pull it off.

9. The Sciences, Each Straining in Its Own Direction, Have Hitherto Harmed Us Little

It's pretty remarkable how good science and technology have been for the world, all things considered. There don’t seem to be very strong reasons why it would have to be that way. We might have lived in a world where it turns out you can make nuclear weapons with common household ingredients, or something like that.

10. Modern Logistics and Global Supply Chains

I won't try to explain this one, just listen to Landsailor

[life-in-a-monospace-typeface]

Supersillyus: A game for two people's eyebrows.

TL;DR: I made a silly game, you can play it here

Long ago I saw a game jam, I think it was called Alt.Ctrl, where the theme was like "Anything but a game controller". People made all kinds of crazy input devices, and it got me thinking about all the different ways you could control a game, especially games with really simple input schemes like Pong or Breakout.

I'd also been playing with face detection and tracking, for my "robot that shines a laser pointer in your eye" project, and the resulting idea was pretty obvious: What if Pong, but you move your paddle by waggling your eyebrows?

You could have the game just divide the webcam image in half, find the largest face on the left and the largest on the right, and have them play against each other. I like the idea of setting it up at an event and having people wander by and suddenly be playing a game against the person next to them.

My first pass at doing this was to track the face with the webcam, and then extract the line of pixels from the eye to the forehead, and analyse that to see how the eyebrow moves.

I did get a version that worked reasonably well for me, which just used the change in brightness over that line of pixels, and tracked the movements of the low brightness area in the middle.

The problem with this, of course, is that it requires the person's eyebrows to be a very different colour from their skin. If I published it like that, anyone who didn't have a lot of contrast between their skin colour and hair colour wouldn't be able to play.

I needed a new approach.

Since the plan was to show the game at indie game festivals at a booth, I realised I could use physical props. So I thought it might be fun to have big brightly coloured eyebrow stickers to hand out to people, that they could stick over their eyebrows. Then the code wouldn’t have to care about people’s skin or hair, it could just track the fake eyebrows themselves. The colors of the eyebrows could be integrated into the game design as well, so if you had bright green eyebrows, playing against bright pink eyebrows, your paddles and score counters would be in that color too. And maybe they could have the name of the game on them as well, so people might like to wear their eyebrows around the event and advertise the game for me!

(Artist’s impression, care of ChatGPT)

This was tricky though. Experiments with Post-it notes found they wouldn’t stay on for long under the high intensity eyebrow waggling the game called for, and experiments with stickier stickers found they were painful to remove. And it would be cool if you could just play the game without physical props, so I could publish it online.

Ideally I’d find a way to track eyebrows that didn’t involve custom novelty eyebrow stickers or racism.

Thankfully, I realised I could just avoid having to do my own eyebrow tracking at all. There are existing facial landmark tracking libraries which are designed to work for all kinds of people, and I could just use one of those, like Dlib:

Then it was pretty simple to measure the distance from the eyebrow to the eye, and normalise it to the size of the face (otherwise getting further from the camera “moves your eyebrows down”).

Different people have a different range of motion for their eyebrows though, and it’s important that you’re able to cover the full height of the screen with your paddle. To deal with this, the system keeps track of the highest it's ever seen the eyebrows and the lowest it's ever seen them, and normalises the paddle location relative to those.

One problem with that approach is that sometimes the system will glitch out and give you a bogus eyebrow reading that’s too high or too low, and then your calibration is messed up from that point on. I deal with that in two ways:

Firstly, there’s some temporal smoothing when we update the highest and lowest positions. If the system sees your eyebrows higher than it’s seen them before, it doesn’t just immediately set the new maximum height to be that height, it only goes a fraction of the way there. So the max and min won’t suddenly jump in a single frame, the new extreme value needs to be there for several frames in order to meaningfully update the calibration.

And secondly, the max value slowly falls and the min value slowly rises, so if they accidentally get set too far apart, they’ll close in again over ten seconds or so. Since the game is likely to involve you covering your full range of eyebrow motion quite often, this means the max and min will keep being set back to reasonable values on a regular basis.

The end result worked pretty well! On my machine.

The problem was, it needed a bunch of libraries, and wasn’t very portable.

So today (years later), I asked a language model or two to take my old python code and convert it to something that runs entirely in a web browser. Opus 4.5 did a pretty ok job of it, and after mere hours of back and forth, the game runs!

There’s a lot of work to do before this is a real game with any polish, but it runs! Here’s me playing against myself, by leaning from side to side to be each player:

You can play Supersillius yourself here

[life-in-a-monospace-typeface]

Things I always have with me

I tend to carry a lot of stuff, and I find this improves my life quite a lot. Here's a run-down.

On my person

Obvious stuff: I always have my phone, keys, ID, and some bank and transit cards.

Heat-It: When a mosquito bites you, it injects certain proteins that your body responds to by itching like hell. These proteins can be denatured by heat. It used to be you had to do this with the back of a carefully heated spoon, which meant nobody did it, but now there are special battery powered electronic devices that heat up to the perfect temperature. The Heat-It is even better, because it just uses USB-C to power itself from your phone. It's extremely small and works excellently, I always have it on my keychain.

Earbuds: They don't make my favorite round-the-neck earbuds any more, and it's tragic because they have one feature I like which is rare, and one feature I can't live without and have never found anywhere else.

The good feature is, they clip together with magnets, and this also acts as the on/off switch. Separate them to put in your ears, and they turn on and connect to bluetooth. Take them out of your ears and let them drop, they snap together and turn off. Very neat

The great feature is, they remember the last two devices you connected to, and there's a button (a physical button, mind you), which makes them disconnect from the current device, and try to connect to the other one. This is not "Bluetooth Multipoint", which tries to connect to two devices simultaneously, and gives you an unconfigurable and confusing mishmash of both devices' audio output and media controls. This is just regular bluetooth connection, and then when I get up from my desk I just press a button and it quickly disconnects from my computer and connects to my phone instead. I don't know why this is rare, but I haven't found anything else that can do it.

Noif: I love the little Victorinox knives, I bought a bunch and usually keep one in my pocket and one in my backpack in case I lose the first one. The big feature is the scissors, but I get some use from the nail file/screwdriver, and the toothpick and tweezers are also surprisingly useful to have. I lose them often, and often they find me again because, inspired by Van Neistat, I engrave my name and number on them.

Sunglasses: I get mine from Primarni for £3, several at a time. Life is too short to care about breaking your sunglasses. These ones are particularly good because they're so thin that they fold almost completely flat and have a very low profile in a jacket pocket.

Pen and Notebook: The japanese cannot be beaten on stationery. The disposable pen has been perfected, it's the Mitsubishi UB-150, thank you, I will not be taking any questions. And Muji's slim notebooks (with the combined 5mm line and dot grid) are completely ideal for my needs.

Backpack

The Backpack itself: I really like the MilTec Assault Pack Large, I've bought several. It has a good number of pockets, plenty of MOLLE straps on the outside to attach stuff to it, and, my favourite feature, it opens all the way out flat like a suitcase.

So much room for activities! Also, the compression straps are great, they mean that a backpack you can use as your only luggage for a week can also be very small and flat when less loaded. Retroreflectors: Did you know they make black retroreflective tape? I covered every plastic clip and toggle of my backpack, and put some through the MOLLE as well, so the pack is all black everywhere, but also super visible to cars at night.

Bike lights: Keep bike lights on your bag, not on your bike where they can be stolen. They're also useful when walking at night. The bike light I use is actually a battery powered mini studio light, so it can do a range of colors and is handy for shooting videos as well.

GaN charger: Not everyone realises that power supplies recently got way better, thanks to Gallium Nitride semiconductors. My Anker Nano charger is 100W and it is tiny - chargers with that much power were three times bigger a couple of years ago.

These two chargers have the same power output!

Toothbrush and Toothpaste: I do a lot of rolling out of bed at the last possible moment, so it's nice to be able to brush my teeth on the way somewhere.

Earplugs and eyeplugs: The ability to nap comfortably any time and place is very important.

Change of shirt and underwear/socks: A single shirt, underwear and socks is really not much size and weight, but when combined with the toothbrush, earplugs etc, it means I can go on short trips with zero preparation, which is extremely freeing. I can randomly decide to stay over somewhere, and have everything I need to be totally comfortable.

Folding Hat: I've never found a baseball cap that folds down smaller/flatter/lighter than the ones provided to employees of McDonalds Australia. Seriously. Just cut the logo off.

Power Pack: Get the bigger power pack. It feels great to have a total abundance mindset on electricity, to know that you can fully recharge your phone, and your friend's phone, and your other friend's phone, and still have enough in the tank to help out that stranger at the party's who's worried their phone will die before they can book a ride home. Also, get one that can put out enough power to charge laptops.

USB Cables: I always have a loong USB-C to USB-C cable - they make them 11 feet long, and why settle for less? I also carry a short cable where one end is USB-A and the other end is a weird three-headed thing with USB-C, USB-Micro, and whatever that iPhone connector is called. Cute girls do have older iPhones and those iPhones are about to die and they would like help with that. The cable came with a spring loaded retractable spool thingy which I removed because it's more bulk than it's worth.

Drugs: Following the advice of Thomas Kwa I maintain a small pill organiser with Paracetamol, Ibuprofen, Lactase, Diphenhydramine, Aspirin, Melatonin, and a few other things of that sort. These help others more often than they help me.

Needle and thread: I carry one of those tiny sewing/mending kits hotels will give you for free. Sometimes useful, and the weight and space is rounding error. Thread is extremely small when wrapped around a piece of card rather than on a spool.

String and Tape: I usually carry some string, some paracord, and some sticky tape, which is so often useful. Tape can be made very small by wrapping it around a business card rather than on a bulky roll.

Ukulele: This one is not universal, but I love it. You never know when a jam is going to break out. I've often regretted not having an instrument, but not yet regretted having one.

Pocket microscope: Sometimes there's a cool bug! The toy microscopes from Carson have surprisingly good magnification and image quality for the size and weight, which is tiny. I'm surprised how often I find myself wanting to get a close look at some tiny thing.

That's... a lot?

You might be surprised at how little space and weight all this stuff actually amounts to! But even if it is more than usual, I think it makes sense.

At one point I realised that almost every time I left the house without my backpack, I found myself missing something from it, so now I just always bring the bag with me. I tried to keep taking things out of it to keep it small and light, swapping things in and out, and I realised that not only did I often made mistakes and need things I'd left behind, but also I was spending significant mental energy on keeping track of what I do and don't have or need for each outing. All this to make my bag a little lighter, and I realised that I'm a young and healthy man, and I can just... carry it all, very easily?

It's wonderful to just have a single bag for all situations, to be prepared for all eventualities without ever having to think at all about what I'm bringing. I just bring everything, and that lets me help the people around me and makes my life better.

[life-in-a-monospace-typeface]

It's amazing how often some weird thing I always carry will come in handy and solve some problem someone is having, and they'll say "Wow, why do you even have that?" and I'm like ??? ?? because of the exact thing that's happening right now?

[life-in-a-monospace-typeface]

Illusion Confusion

The human vision system has a tough job, figuring out complex scenes using only ambiguous sensory data, and it sometimes makes mistakes, which we call optical illusions.

For example, when you see a hollow face shape, lit from below, it really really looks like a regular face lit from above.

This is a very good optical illusion in my opinion; the effect is quite striking in person.

Father McGuire's cars

Or consider this image:

All three cars are actually the same size!

Except, that's bullshit, isn't it? The cars are obviously different sizes! That one in the back is clearly much bigger than the one in the front.

"Ah, but if you measure them, they're all the same number of pixels"

I've never seen anyone measure a car in pixels, show me your pixel tape measure!

"No, like, in the image"

Right, but you said "the cars are the same size", not "the images of the cars are the same size".

Certainly the images of the cars are the same size, but that's just because the cars are different distances from the camera. That car in the back really is bigger, it only looks the same size because it's further away.

In a way this is just semantics, but it's completely critical semantics. You can't get away with equivocating between the image and the scene, when the relationship between the image and the scene is what the illusion is meant to be about.

In fact, I'd say this isn't an optical illusion at all, it's just how vision works - probably the only way vision could work in a three dimensional world. Calling this an optical illusion is like picking up a toy car, holding it up at arms length so it looks the same size as a real car some distance away, and saying "What a crazy optical illusion, these cars are actually the same size!".

That's how it works! That's how any of this works!

Et Tu?

The problem isn't restricted to this type of illusion, and it's not restricted to low quality sources. Look that this, from our own Scott Alexander:

The caption says “Squares A and B are the same color”, which is false - the squares are obviously different colours. They only look the same colour in the image because one is in shadow and one isn't. The actual squares in the actual scene are different colours. Yes, in the image, they're the same color, but in the image they're also not squares, they're parallelograms!

Again this seems semantic but it's important. By talking about "squares", you're accepting the premise of the image, that it represents the scene it seems to represent. To say "the squares are the same color" is to accept the premise of the image in terms of 3D geometry, but reject it in terms of lighting. That is, at best, an obtuse way to think about an image.

So this is also barely an illusion in my opinion. When you look at the image, you're not actually wrong about the scene you're looking at.

Getting Pigeonholed

But actually, maybe you are wrong about the scene! What if the whole scene is actually perfectly evenly lit with no shadows, and the "shadow" is actually painted onto the checkerboard? In that case, the squares really would be the same colour! That scene would produce exactly the same image, and it would be a real illusion - you'd think the squares are different colours, and you'd be wrong.

But that's not the claim normally made about this image. And in fact, why stop there? You could be wrong about the scene in all sorts of different ways. What if the board is actually a totally uniform grey colour, but it's having a checkerboard pattern protected onto it? What if the whole "scene" is just a flat TV screen with the image on it? What if the left half is a screen but the right half is the actual scene you'd expect, lined up just right? What if part of the checkerboard is actually flipped around and off to the side, and we're looking at that bit through a perfect mirror at exactly the right angle? What if the whole scene is a huge cloud of coloured confetti and the image was captured just at the exact moment that all the pieces happened to line up perfectly to look exactly like a checkerboard with a cylinder on it?

There's a practically infinite set of possible scenes that would produce this exact image. And that's true for every other image as well!

Wait, how is it even possible to ever infer a scene from an image at all?

Priors!

Well, that confetti thing seems really unlikely, doesn't it?

In order to ever usefully interpret an image as representing a scene, you have to have a prior over scenes - you have to make assumptions about what scenes are more or less likely. A shadow from a light source is more likely than a shadow painted on the ground, which is in turn more likely than a perfectly aligned cloud of confetti.

You simply have to make assumptions about the scene, and sometimes those assumptions will be wrong.

Give your eyes a break

The hollow face effect is a real illusion - your vision system is leading you to believe false things about the scene. But it's not really making a mistake - the vision system is doing something very reasonable here! Things are lit from above much much more often than they're lit from below, and normal face shapes are tremendously more common than hollow concave face shapes. The vision system's guess here is just straightforwardly the correct guess to make. And it does have to guess, that's the only way to ever infer a scene from any image!

So some things we call "illusions" aren't really illusions at all, just unavoidable consequences of how images work, and for others we've adversarially constructed bizarre objects specifically designed to violate perfectly reasonable and unavoidable assumptions.

We should cut our visual cortex some slack! It's doing about as good a job as anything could, and I for one am very proud of it.

Honestly it's so like us, to make a load of basic errors while criticising human vision. If our logical reasoning systems worked even half as well as our vision systems, we wouldn't have this problem.

[life-in-a-monospace-typeface]

The Overworld

His grandmother still believed in 'The Overworld', still held to the old myths. Angels and magic. Every religion had its own ideas about who or what occupied the world that hung upside down in the sky, miles above us. But the closer we looked at The Overworld, the more it seemed like our own world. It took the invention of telescopes for most people to accept the truth - there was no Overworld, there was just The Mirror. It's hard to hold on to the idea that the Overworld is full of gods, when you can look up with a telescope and see only yourself looking back. As a scientist, he'd done public demonstrations, lying on his back under the university's most powerful telescope and using The Mirror to shave.

Frivolity aside, there were still big questions to answer. How did The Mirror get there? How did it stay up? His theory, which had become the current scientific consensus, was that what appeared to be a mirror was really the upper boundary of the universe itself. Light flew up there, ran out of reality to fly through, and came right back. Obviously there wasn't going to be a literal mirror up there, no matter how much people might like the idea.

In his mind though, the truth was far more magical than a physical mirror would have been. It was the boundary of the universe, the ends of the earth, the edge of reality itself! And he was a mile away and closing.

People had raised all sorts of absurd objections to the project. Like, what if he damaged the mirror? Never mind that it had been around since the dawn of time and nobody had ever discovered any sort of blemish whatsoever in the perfectly flat, perfectly clean, perfectly reflective surface. Never mind that it obviously wasn't a real mirror anyway. He was pretty sure he could simultaneously fire all the cannons ever made at the thing and have no effect whatsoever. Never mind any of that. "If you already know so much about what the mirror is like", they'd said, "why are we funding this expedition for you to study it close up?". So, at the top of the balloon, the rails of the observation platform had spring-loaded padding, to protect the mirror's surface when he reached it. He'd complained about the extra weight of course. The balloon was the first ever made to be able to achieve these altitudes, and every ounce was a careful decision. But in the end he'd had to humour them. Solving the increasingly impossible-seeming engineering challenge was much easier than trying to argue. By the time the craft was ready to fly, it even carried a set of holy artefacts from the major religions, in case he needed to pacify any gods he might meet.

He checked the altimeter again. Getting close now. He couldn't help looking out of the porthole, even knowing that there would be nothing to see. The altitude of the mirror could be determined experimentally from the ground, and other researchers had pinned that number down with more precision than an altimeter could measure. So the only thing to look at was the little brass dial, creeping up. Very close now; time to relocate. He quickly adjusted the controls to slow the balloon's ascent, clipped on his helmet, and stepped into the airlock.

This was the only really dangerous part, he thought as he climbed up the ladder on the underside of the canopy. If he'd miscalculated, and the craft was closer to the mirror than he thought, or ascending faster, the unexpected collision might knock him off the ladder. But as he climbed around to where he could see straight up, he knew he'd pulled it off perfectly. Some four hundred feet above him he saw his own balloon, upside down, gradually drifting down to him as he drifted up. That meant the mirror was around two hundred feet up, and he'd have just enough time to get up to the observation platform and use the simple controls up there to dial in the... landing? It would be close, but that was good, it meant he'd get more time up on the platform before he had to climb back down and repressurise the suit's air supply.

Away from all the equipment in the gondola, with the burners off, it was eerily still and peaceful as he climbed up the side of the balloon. The only sound was his own breathing inside the helmet, and the clanking of his boots against the ladder. He could even hear the faint echos of his own footsteps as- wait, that was it! The expedition's first really novel observation! He'd predicted exactly this, years ago, but now he knew for sure - a fact that nobody else knew - the mirror reflected sound as well as light! He let out an involuntary "Ha!", listened to it echo back from above him, and sped up his climb.

On top of the world, he tweaked the controls as the balloon closed the last few feet and the padded rails made quiet contact, bringing the craft to a gentle halt with the mirror just above his head. Up close, it was just as strange as he'd always imagined it. The perfect surface was completely invisible, apparent only from the flawless reflection. The platform's array of finely tuned scientific instruments sat untouched, temporarily forgotten, as he reached his hand up to touch The Mirror. He'd expected it, but it was still unsettling to feel the surface sliding completely frictionless against the material of the glove. The only thing he hadn't expected was his own reaction; he'd anticipated another burst of joy like the one he'd got from hearing the echo, but actually it was kind of an anticlimax. What was he going to say, back on the ground? "I have visited the heavens, and it's exactly how I thought it would be"?

No, he had to do something more dramatic. He hadn't come this far to stop short by the thickness of a few layers of fabric. He could go all the way. Couldn't not do it, now that he'd thought of it. He'd have to be quick. He'd have to repressurise right after. But he could do it. And when he returned, he'd say "I have touched the sky".

The miniature "Suit Pressure" dial inside his helmet plummeted into the red as soon as he released the latch that sealed his glove to the rest of the suit. He didn't need that dial really, he'd felt the change in his whole body. The thing to watch was the "Tank Pressure" dial, which spun terrifyingly fast for the split second it took him to clasp his hand around his wrist. It wasn't a good seal though, he was still losing a lot of air. Ignoring the spinning dial, he slowly, ceremoniously, reached for the sky. His exposed hand was quickly starting to feel numb as the blood rushed into it, and extremely cold as the sweat boiled off into the thin frigid air, but when he gently laid his palm flat on the mirror's surface, it was surprisingly warm. Actually, why was that surprising? Had he expected The Mirror to be cold, just because a regular mirror would be cold in that situation? He felt momentarily embarrassed for making the same mistake he'd criticised so many others for. Of course The Mirror reflected heat as well as sound and light, he knew that. He'd even published the calculations to prove it - if the heat hadn't been reflecting back, almost all the thermal energy in the universe would have already been lost that way. Maybe he could blame the low air pressure for his foolishness. He did feel a little lightheaded. Still, that was another prediction confirmed, and time to get back to the gondola and repressurise.

He hadn't realised how loud the hiss of escaping gas had been until he reattached the glove, and the eerie stillness returned. The suit pressure dial returned almost back to normal, but the tank pressure was in the red now. He was sure he had enough air left to make it back, but his head was spinning, his vision a little dark around the edges, his ears suffering from the pressure changes. He'd cut it fine again. He forced himself to breathe normally, his breath sounding strange in his ears as he moved carefully to the ladder. The echo of his footsteps sounded odd too, almost like the echoes were coming slightly before his footfalls. Actually, wasn't he too close for an audible echo anyway? Confused, he stood and stared up at his reflection. It was starting to climb down the ladder.