[Old] Wayfinding series

post by Gordon Seidoh Worley (gworley) · 2019-03-12T17:54:16.091Z · LW · GW · 0 comments

Contents

  finding your way
  natural wayfinding
  finding focus
  refinding thoughts
  refinding optimizations
  refinding augmentations
  finding priority
None
No comments

As I said in the intro to the "mapmaking series" [LW · GW] post, I'm reposting some content from my old blog that people found useful. They particularly found things I wrote in this series useful, and of the two series I agree that this is likely the more useful one, the former being reposted mainly because this one occasionally references it. If you are familiar with LW much you probably don't need to read the first post.

As I also said in the first post, this is old content that I no longer entirely endorse, although I think most of it remains useful and surprisingly skillful advice given how little I knew back then. I've been about 7 different people between when I wrote these posts and now, to give you some idea of the distance I'm suggesting I have from it. Nonetheless, I think learning to master the topics I discuss here is pretty essential if you want to attain to at least whatever level of mastery I've achieved at being a human, much else besides not included.

I hope these words treat you well.


finding your way

having now a good sense of maps and mapmaking, we can begin to use our maps to find our way through the territory. this requires more than the skill of map reading, which is implicit to mapmaking. you need the skill of the wayfinding: finding your way through the territory to a destination.

what sort of destination? the kinds of destinations that exist in the territory: possible future worlds states you could find yourself in. for example, hungry? find your way to eating a sandwich. thirsty? find your way to drinking a glass of water. ennui? find you way to having a life's purpose. to put it in a story:

you are trying to go to your friend's house. you have a map that shows you that to get there you need to walk down the street, though a patch of brush, past an angry dog, and then up a narrow lane with your friend's house at the end. the map tells you where to go, but just reading that map is not enough to get you there.
you know how to walk down the street just fine, but when you come to the brush just knowing you must go through it is not enough: you have to know how to clear a path through it. and if you try to pass by the angry dog he'll probably bite you, so you also have to know how to approach the dog to keep him from attacking. only then do you get to go up the narrow lane to reach your friend's house.

when put this way it may seem obvious that wayfinding is a skill apart from map reading, yet many people fail to notice this because they don't really know how to make meaning of wayfinding beyond map reading, especially when the metaphor is less clearly laid out.

suppose you want to lose weight, and further suppose, science aside, that weight loss is a simple matter of eating fewer calories than you use each day. in this case you know exactly what the map looks like: you eat less and do more so that calories in are less than calories out. great! having read the map, you have now lost the weight. only it doesn't work like that. you somehow have to use this knowledge of the map to lose weight. this is the skill of wayfinding: taking what you read in the map to actually get somewhere.

this is what makes wayfinding the first gap you must cross to move beyond mapmaking. it's not enough to read the map: you have to do things in order to move based on what you have read in your map. following the winning way requires that you not only make a map that helps you win, but also using the map to help you win.

natural wayfinding

although wayfinding is not as simple as map reading, it's still a skill you are already surprisingly good at. most of the ways in which you are not good at wayfinding are because you don't have a good map of the territory or because you find yourself in situations more complicated than those your ancestors evolved to handle. but because you already have excellent natural skills, we should look for ways to amplify them so that they work in the modern world.

this is, in essence, the whole story on wayfinding: the rest is just details.

but details are never 'just' details: they're the information you need to know. so let's begin by considering the things you are already good at in terms of wayfinding. the core of your skills boil down to these:

you may not appreciate how amazingly good you are at these processes because you regularly hit their limits, but consider that all the time every day you make millions of decisions by automatically and unconsciously prioritizing possible future actions, base those priorities on what you are thinking and what you have thought in the past, and then execute on those decisions without being constantly distracted by the hundreds of bits of sensory input you receive each second. most of the universe, and even most living things, can't do that.

why do you sometimes drink water? not ultimately, but proximally how do you come to drink water? first, your body senses a lack of water through one of many possible mechanisms, so let's suppose here that your mouth is dry. this produces a 'thirsty feeling' thought about needing to drink water. remembering there is a glass of water near you, you choose (even if it happens below the level of consciousness) to lift the glass of water and take a sip. while drinking this water you are automatically ignoring most of your sensory input to focus only on the senses you need to help you take the drink. you continue drinking until the 'thirsty feeling' thought is prioritized under some other thought, like the 'take a breath' thought, at which point the choice process has already begun again.

this story about water drinking is highly stylized: there are many details excluded, mostly because we don't even understand all the details. but the point remains that you have an incredible ability to make complex choices based on a massive amount of information. to the extent you can get better at wayfinding, we'll be looking for ways to take these natural skills and build something more powerful out of them.

finding focus

although we have marveled at the power of our natural wayfinding skills, if we are to improve upon them we must acknowledge their limits so we can build systems to do more than we could on our own. thankfully the limits of human cognition are great fodder for academic research, so we know quite a bit about the edges of attention, choice, and memory. we'll consider each in turn, so let's start with the most fundamental of all to getting things done: focus and attention.

focus is limited and heavily influenced by environmental, metabolic, and psychological factors. this is well known, but it bears repeating because many of us do a bad job of acting on this information when we need to focus.

the simplest things to deal with are the environmental aspects of focus. in terms of environment, everyone is a bit different, but i think we can summarize an environment conductive to focus as:

some might object that they focus well in chaotic environments. they feel like they get more done there than in a quiet room. i have no objection to this and see it as just another version of a distraction-free, clam, comprehensible environment. maybe quiet distracts you? fine, add some noise. maybe being alone is distracting. fine, add some people. the point is to find an environment in which you are minimally taxed by having to deal with the environment. you should feel comfortable and able to pay attention to what you want rather than what you are made to.

metabolic aspects of focus are a bit trickier, but still relatively straight forward. at the high level, the things you need to be able to focus are:

sleep probably has the most impact on attention. adults need between 4 and 12 hours per day at the extremes, with 8 hours average and 6 to 10 hours typical. whether that sleep is all in one block or segmented, getting enough sleep is mostly about creating an environment where you are ready for sleep and have enough time to sleep the amount you need. being ready for sleep means reducing lighting, avoiding physical activity, and relaxing a few hours before bed. if for some reason you can't get enough sleep, a sleep therapist might prescribe you a stimulant to help you stay awake or a sleeping aide to help you get more sleep.

in addition to sleep, nutrition is important to attention, specifically your general nutrition and the balance of chemicals in your body when you try to focus. you need to consume a good balance of fats, proteins, carbohydrates, and micronutrients in order to meet all your body's needs. if you don't you may feel tired, lethargic, or have distracting symptoms that result from malnutrition. and because your brain is the most energy-hungry organ in your body but also one of the least critical to continued life, your brain is one of the first things to suffer when you are malnourished. further, your brain needs readily available energy in the form of sugar, so good focus also means having readily available sugar in your body. if you eat well your body should handle this for you, but if it doesn't you may need to eat fruits or other small sugary foods to help maintain focus until your diet improves.

physical condition also plays a role in attention. healthy physical condition is not so much about being in shape, though, as it is about having a working body. your body is intended to be used for things like running, jumping, bending, and lifting, so if you don't get regular physical activity your body is not operating under normal conditions. this can lead to decreased cognitive function in general, and difficulty focusing in particular.

environmental and metabolic impacts on attention covered, there are the psychological aspects of focus. it's easier to focus on things you care about, even when those things may not be the things you think you should care about. when you lack a reason to care, it's harder to focus and get things done. if you don't much care about something then you either need to start caring or accept you don't actually care and not expect to be able to focus easily. either way is fine, but it underscores the point that caring matters. focusing on things you don't care about is mentally taxing and burns out ability to focus.

but focus is not everything. even if you can focus your attention well on whatever you want whenever you want, you still have to decide what to focus on. and the first part of deciding what to focus on is remembering.

refinding thoughts

better wayfinding depends in-the-moment on improving focus and attention, but wayfinding is also about reading your map, i.e. remembering things. and remembering things is about rethinking old thoughts, i.e. recalling memories: thoughts you previously had that you can rethink quickly without having to regenerate them from scratch. to outperform natural wayfinding, we're going to need to investigate the limits of memory and find ways to work with those limits to achieve more.

memory is actually made up of at least three parts: sensory memory, working memory, and long-term memories. sensory memory is the very short lived (less than 1 second) memory that sense organs use in relaying information to the brain. working memory, also known as short-term memory, is the set of thoughts you can keep consciously aware of without repeated sensory information from outside your body. long-term memories are a collection of different mechanisms you use to store different types of information in different ways for long periods of time outside conscious thought to be recalled later. for example, it appears that long-term smell memory is separate from long-term fact memory, thus you have multiple long-term memories.

your working memory is very limited. most people can only keep about 5 things in their conscious thoughts at a time. it appears that working memory is made up primarily of an auditory loop. this loop repeats sound information inside your head over and over, but it is of a fixed length, so you can't remember anything more than what you can cram in there. although it appears that some people have had success training longer working-memory through exercises, even then the increases are relatively small (and probably not that useful). so even if you can get yourself more working memory, it still means you can keep little more than a handful of things in your conscious thoughts.

in comparison you store an almost limitless number of thoughts in long term memory, but these memory faculties have their own constraints. semantic memory is how you remember independent facts like 'tomatoes are red' and 'the mendoza line is .200'. this is apart from procedural memories that store sensory information and emotional content triggered by other memories. excepting cases of dementia or brain injury, it appears semantic memories can stay around indefinitely once consolidated, but may become almost impossible to retrieve without regular recall to maintain consolidation. in this way semantic memory is like working memory in that you have to keep repeating thoughts to hold on to them, but the process operates over months and years instead of seconds.

episodic memory lets you remember past experiences and other generic information that isn't otherwise stored in another long-term memory. unfortunately episodic memory is highly unreliable: very little detail is stored in episodic memory so most of the details have to be reconstructed on-the-fly when recalling a thought. this means environmental factors can heavily influence the details of episodic memories if not their gist. however episodic memories seem to be more robust than semantic memories, able to be recalled after many years of disuse, so they provide a potential key to very long-term storage of arbitrary information.

given this amazing yet limited memory, we now need to look for ways to improve upon it. techniques fall into roughly two categories: optimizations and augmentations. optimizations are those methods that make better use of your existing memory abilities to maximize their potential. augmentations are processes that put memory outside your head so you can reliably retrieve thoughts later. we'll consider both in turn.

refinding optimizations

let's first consider optimizations for refinding thoughts. there are many well known mechanisms for improving memory, so i'm going to focus on three here that are relatively easy wins: simple to learn and high value. They are chunking, spaced repetition, and the method of loci.

the simplest is chunking, and it applies to both working and semantic memory. rather than remembering, for example, three letters as 'b, r, and t', you chunk them together into 'brt'. when combining just a few things this chunk ends up as a single thought in your memory that you can recall all at once and then split apart when necessary. this is famously used in telephone numbers to make them easier to remember, but you can find examples of chunking everywhere, especially in how you remember tradenames and common phrases.

a naive model of how this works is that thoughts can each hold some maximum amount of information. you store thoughts in your memory rather than raw information, so chunking allows you to better pack information into thoughts and make more efficient use of the space. this is especially useful for working memory since you can only keep about 5 thoughts in your head at once.

but packing your thoughts tightly does no good if you can't remember them. semantic memory relies on consolidation, yet you have little direct control over what gets consolidated. sometimes things you would like to remember fail to get consolidated while other things you perceive as less useful do. to ensure the thoughts you most care about get and stay consolidated, you can use the method of spaced repetition.

the idea of spaced repetition is simple: on some regular schedule you review thoughts to give them repeated opportunities for consolidation. often people optimize and automate this process using technology as simple as flashcards or as complex as anki, but it works even without tooling. for example, when you meet new people most of the time you need to see them a few times spread out over days or weeks in order for them to become fixed in your mind. without reinforcement at regularly intervals you're likely to forget them. the trick of spaced repetition is to take this natural method you have for filtering what information to remember and deliberately using it to choose what you will remember.

the most advanced technique i'll mention, and the one that takes the most effort to get the most out of, has been used for thousands of years and is known as the method of loci or the memory palace. the idea is to use episodic memory to construct narratives where you imagine a place in your mind that you can explore and move through to find information. this is how people could recite epic poems like the homeric epics from memory. before the modern era of cheap writing reproduction, most educated people learned the method of loci so they could reliably store information that they would have no other way to lookup later.

as this suggests, the method of loci is less popular now because we instead use memory augmentation systems, but it is still useful to some people, especially those who may not have the luxury of accessing the world's knowledge from a tiny device in their pocket or need faster recall than such devices allow. anyone who must react quickly and think on their feet without looking things up could stand to build a memory palace.

so now you have techniques to better use working, semantic, and episodic memory, and until a few hundred years ago this would have been the limit of what you could have done to improve your memory. but writing and the technologies dependent upon it allow us to do something more: we can augment our memories by storing thoughts outside our heads for later remembering by reinput.

refinding augmentations

beyond techniques to better utilize memory, the skill of refinding thought can be improved with memory argumentation systems. they are not so complex as the name implies, though: simply put these are ways of storing thoughts outside your head so you can easily access them later. the complexity in memory augmentation lies in the vast web of technologies needed to make these systems practical. that's why even as recently as a few decades ago these sorts of systems were in less use than they are today, and will likely be in greater use in the future.

working memory is short, but the way it functions shows us how to improve upon it. working memory appears to be primarily an auditory loop and visual sketchpad of you repeating things to yourself. it's an internal mechanism for resending information as if you were sensing it for the first time. we can take advantage of the same technique, but using your eyes and putting the 'loop' outside your head.

you can augment your memory by making things visible (as in literally writing them down where you can see them). when you can quickly look around to see things you don't have to put any effort into remembering them other than recalling where to look when you need something. this is a far easier task than trying to keep lots of information in your head, because if nothing else you can quickly rescan everything in your visual field to find what you want.

but for this to work the information has to be actually visible to you. this works well if, for example, you work on a computer and have a large enough screen that you can multiplex it to see many views at once: web pages, documents you're editing, chat logs, whatever. or, if you're working on paper or with a chalk board, writing all the stuff you need to recall where you can easily glance over to see it. all these work by taking the burden of remembering something in working memory and offloading it to a system so you don't have to keep it in your head, freeing up your working memory to have more space to work with new concepts as you think them up but haven't written them down.

if we take this same technique but make it about storing information rather than looking at it right away, you get the great collection of memory augmentation systems we know as notes, diaries, papers, web pages, and books. we take these technologies for granted today, but they have radically changed how much information a single person can use over the course of their life. five hundred years ago the story was very different.

until the invention of the printing press people had to personally write down all information they wanted to store. at best a person could hire scribes to help them, but this was still a very expensive proposition. by using movable-type printing presses people could write things for you, and all you had to do was buy a copy. this began a trend of accelerating growth in knowledge stores that today gives you orders of magnitude more information than you could ever use in a lifetime.

so now that we've covered memory and focus, there remains one last part to your natural wayfinding abilities to be improved: prioritization. this is the last piece of the puzzle of how you can better find your way.

finding priority

when we began our investigation of natural wayfinding we broke it down into three parts: focus, memory, and prioritization. the former two make up the prerequisite skills of wayfinding, but they only take you as far as map reading. the last, the ability to prioritize, is what is needed to find your way.

luckily you are already pretty good at prioritizing! you prioritize all the time without even realizing it. that's because this is the faculty by which you do things: you must always choose what things to do because you cannot do everything at once. do you breathe now? eat? drink? dance? read? the choices you make are the result of a complex process of prioritizing all possible actions.

to appreciate just how hard it is to choose, consider the problem of building an artificial intelligence. early on computer scientists thought it would be easy: just put a bunch of info in the computer, set up some simple, self-improving algorithms to work with it, and you'll get an ai. but it turns out that doesn't work: we today have computers with far more focus and memory than any human could ever have, but we still don't consider them to have intelligence on par with insects let alone humans. why aren't computers already intelligent? because their prioritization mechanisms are too simple.

what humans and most animals can do that computers cannot do yet is rapidly make nuanced decisions among infinitely many possible choices that steer the world towards a particular state. computers instead rely on humans to narrow down their choices enough that they can decide among a finite set of possible actions (this is, in essence, what an algorithm and, by extension, a computer program is). we take prioritization for granted, but trying to understand it well enough to build it has made it very clear that it's amazingly difficult and requires tremendous complexity.

so choosing is hard, but luckily we are like magic prioritization black boxes that somehow know how to do it without really understanding how we do it. we regularly pick from among unlimited possible actions, mostly without any awareness of how it happens. yet there is a small part of the prioritization process we notice because we have access to it, and that part is conscious choice. but conscious choice is expensive and limited. in fact, it's limited by the same 5 thought limit we have on our memory.

because it's hard to think more than 5 thoughts at the same time, it's hard to consciously prioritize among more than 5 potential actions at once. there's not much we can do to optimize prioritization beyond improving focus and memory, but we can augment our ability to choose similar to how we augment memory. so if you need to consciously prioritize among more than 5 things, and you are sure you can't simplify your life so you don't have to choose among more than 5 things, then you need a prioritization system.

all good prioritization systems are built around:

the most popular modern manifestation of a system with these three-cs is getting things done. it's a great choice to get started, and gtd learning materials will teach you how to work a prioritization system. but gtd is only one way: once you master the basics you will likely go on to build your own system that fits your life and the tools available to you.

but i have to end on a warning. prioritization systems can be intoxicating to some people who would give themselves up to the system so they don't have to choose. it doesn't work like that! you are the engine the drives the prioritization system: the system has no power to drive you. if you abdicate decisions to the system no decisions will be made and the system will just drain your time. you are the agent, and the prioritization system is a tool to help you, not the other way around.

0 comments

Comments sorted by top scores.