> I’m totally convinced that a new idea or a new plan or a new technique is never really understood when you just explain it.
> People will often think they understand, and they’ll say they understand, but then their actions show that it just ain’t so.
Isn't this the very reason that homework assignments exist? When I was in college, it took me two years to realize that I could have easily gotten As and Bs (instead of Bs and Cs) in my various math classes, physics, chemistry, etc. had I simply bothered to do my homework properly--or, to put it differently, had I only properly applied the knowledge which I passively acquired by reading the associated textbook sections. Notably, I was always convinced that I "understood" everything I had read, only to find out otherwise when I tried to solve any of the problems. It was only after I actively applied the knowledge I had picked up, that my understanding transitioned from superficial (as in, understanding the underlying logic itself) to concrete (being able to apply the logic toward other problems).
If you apply this to abstract and esoteric technical concepts, it is easy to see why someone might say they understand--and even believe that they do--while in reality only having a superficial understanding at best. The problem then becomes getting the other person to spend the effort to properly internalize the concepts being conveyed, before they have built up the requisite interest in the idea to be sufficiently motivated to carry out said effort.
It's probably also true, however, that this may only apply to sufficiently esoteric and complex ideas in the first place.
> When I was in college, it took me two years to realize that I could have easily gotten As and Bs (instead of Bs and Cs) in my various math classes, physics, chemistry, etc. had I simply bothered to do my homework properly--or, to put it differently, had I only properly applied the knowledge which I passively acquired by reading the associated textbook sections.
I took this realisation to the extreme and decided to completely deprioritise classes in favour of homework and doing the reading in my own time. I figured the cost-benefit, at least for me, was much higher if I spent an hour doing as opposed to an hour listening.
This actually worked really well for me but that might also be because I often struggle with large classroom learning: the pace is either too slow and I get distracted or too fast and I can't keep up. But even when the learning is 'one-to-one' I feel like there's always the tendency for people to zone out and not raise an issue when they are either bored or did not keep up/understand.
I think you're right in that some of that might be the brain pretending that it understood when in fact it did not. Could it also be a social thing? Maybe because the other person expects you to understand and this causes your brain to try its best to believe that it understood when it did not.
Classroom experience is so rarely useful, the ONLY time you benefit is when either a) you already studied the subject and the professor is good and mentions something tangential/extra credit, or b) the classroom size is small, so when you don't understand something you can say so without fear of class disruption and the teacher can spend more time tailoring their curriculum to their class.
Otherwise yes, you are often better off with the book and homework and actually completing both.
And this is where I usually go off on my rant how college is useless for education - you can 'learn' peer interaction but even this way nothing forces you to make friends or connections, so the inherent use is close to nill.
This is a super interesting strategy. Also going to classes can lead one into a false sense of mastery. It almost makes more sense to do homework before class, and attend lecture as a review. It would just become pre-work then and would essentially be considered mandatory. This would also really fix the problem of lectures not being useful, since the lecturer could assume you did the work and then spend more time on things that are a better use of their expertise.
> I often struggle with large classroom learning: the pace is either too slow and I get distracted or too fast and I can't keep up.
This is a problem for everyone. People may not experience the struggle directly, but it's there. It's the inherent problem with any group class.
> I feel like there's always the tendency for people to zone out and not raise an issue when they are either bored or did not keep up/understand.
Tight loops of interaction will make this nearly impossible. What you'll have to watch out for instead is overwhelming students to the point of distress (speaking from experience).
> completely deprioritise classes in favour of homework and doing the reading in my own time
I had two mathematics teachers in high school that took this approach to class time. They would lecture for 10 minutes then have the class do work in groups on homework assignments for the remaining 45 minutes. Exact times varied based on the complexity of the material, of course. And I do not believe their format was sanctioned by the administration. But it sure helped me retain more of the information.
As an extra bonus, it really helped with procrastination. It was a lot easier to get to work on an already-started homework assignment compared to staring at a blank page and a daunting list of problems.
> I took this realisation to the extreme and decided to completely deprioritise classes in favour of homework and doing the reading in my own time. I figured the cost-benefit, at least for me, was much higher if I spent an hour doing as opposed to an hour listening.
That depends very much on the lecturer. If he or she just follows one book for the whole semester and doesn't add any additional insight, explanation of tradeoffs, or enlightening anecdotes, then by any means skip the lecture. But if you have a lecturer giving you all from above and the possibility to ask questions, lectures can be invaluable. Luckily I experienced many of the latter kind, although some of the first.
I think it's to do with the nature of tests. Exam questions in the mathy subjects are sort of two varieties in my experience. One is simply book proof: here's an equation, here's another one, plug one into the other and rearrange and there's this useful result. Come exam time, you just need to remember the steps and you get points.
You'll think you get it.
But there are questions that are about a deeper understanding. For these there's some point to the question that isn't obvious from just reading. I remember the first few question sheets I got in uni, there would be questions that appeared to have nothing to do with what was presented at all. Only by asking around did I discover what the cryptic connections were. If you don't do the question sheets, you won't see this.
From my experience both as a student and as a teacher (formal and informal settings; the formal ones in academia): Different people respond differently to different modes of teaching / exposure to information.
* Some only "get it" when it's explained to them in a classroom, preferably by an engaging teacher; while others have to read it in the book.
* Many, perhaps most, won't really get it unless they do the homework; but others can do all of the exercises you give them, and still fail to get the bigger picture, and will be lost if faced with something slightly different than what they exercised.
I did this when I accidentally enrolled in the astronomy 101 class for astronomers when I meant to enroll in the “fun” astronomy class for non-majors. The professor was extremely uninterested in teaching and I decided to just do the assignments and stop going to class. Still got an A.
Yes. Well, examples. Good homework assignments give you some examples to sit in your head, not all homework assignments are good though.
The way I like to tell people this is, you know that thing where you want to take your kids or your students and make sure that they don't have to go through the same pain that you did to learn what you had learned? That feeling?
If you have ever had that feeling you must understand that it was horribly mistaken. The problem is precisely that learning and pain go hand-in-hand. We learn abstractions precisely because they relieve a sort of confusion, a sort of difficulty, they organize the pain that we have experienced and make it tidy and less painful. Even as kids abstractions like “this is what it means for the stove to be on” work this way, not that you have to get burned, maybe you just need to be yelled at, but it organizes that pain of being yelled at and tells you when are you getting yelled at and why.
Not all learning is this way, for example memorization and repetition... but to a first approximation you have to get lost in the forest before the landmarks on the map are recognizable.
Corollary: you CAN tell people things, but it might be more involved than just a casual conversation. Either you have to establish a shared context, tap into the pain/confusion that they already have... Or you have to get them interested enough that they will follow you down a rabbit hole of confusion and difficulty so that you can finally explain the thing.
> The way I like to tell people this is, you know that thing where you want to take your kids or your students and make sure that they don't have to go through the same pain that you did to learn what you had learned? That feeling?
I am not yet convinced. Let me take the following example because I remember it well: I tried to learn about geometric optics and lenses. To do so, I downloaded multiple lecture PDFs from the internet and read them.
All of them started to explain lenses by describing how large the image of a real object is, defining focal length and such. Literally not a single one of them even defined what an image is. The whole talk about image size and focal length and magnification and what not was totally worthless because it was all based on the same fundamental word that had no meaning. In the end, I tried to make up my own definition that was consistent with all those PDFs, but it left me unsure if I got it right.
The "pain" was there, but there was no relief from it, and I am still convinced that that simple definition could have avoided it. If I were to teach optics, I would give such a definition, and I am still convinced that it would help with the pain.
The other example was one of those PDFs that showed a real object that was "wide" along the distance axis from the lens, and so by my understanding should have an image whose magnification changes along the distance axis, but the explanatory picture in the PDF showed equal magnification everywhere. Today I am convinced that this "explanatory picture" was simply wrong.
Again, there was "pain", no relief (because I could not be sure), and I am still convinced that a learner can be saved a lot of pain when you just exlucde factually wrong content from the learning material.
I strongly agree with your main point that sometimes valuable learning is painful. I've found a lot of mathematics is painful beyond "this is frustrating and difficult," but in a more profound way.
I would say though, about abstractions, I'm not sure hot stove is an abstraction. I think hot is an abstraction that allows you to apply the lesson you learned on the stove to the toaster and the coffee maker and all the other hot appliances you encounter.
> Isn't this the very reason that homework assignments exist?
Absolutely. It is also why I will never understand why it is considered a virtue (especially in the maths and physics community) to set assignments without exemplary solution. I'm not asking to give this exemplary solution to students right away, but it should definitely exist and be given when students have shown honest effort and especially to weaker students that are struggling.
Instead the "left as an exercise to the reader" mentality is celebrated when in reality - in my opinion - it is almost always just a convenient excuse for the laziness of the professor at the expense of their students.
I have seen plenty of miss-use of solutions of the following though:
1. Struggle with exercise.
2. Check proposed solution.
3. Thinking that your understanding of the proposed solution constitutes an ability to solve it yourself.
This is analogous to the original problem, where you assume the ability to follow reasoning translates to being able to reproduce it.
The main use of exercises without solutions is to try and force students to go to their teachers and interact, so that someone can try to pinpoint what exactly the missunderstanding is.
This can probably be better done with assigned homework. However grading can be time consuming, so trying to filter so that only those that have problem with specific exercises come to see you is one way to try and get some time economy going.
Also yeah. This is better in course specific work sheets than in books, since books should be self contained enough to be usable without the guidence of TAs or professors.
IMO mathy homework should always come with the correct end result to the questions. It allows the student to get instant feedback on whether they’re right or wrong. They’ll still have to work out the derivation of the answer.
I had a boss years ago who would encourage me, whenever I was interested in e.g. a new library or framework, to go and reinvent the wheel in my spare time i.e. build it myself from scratch first to understand what the problems it solves actually are.
I've never built my own anything quite as low level as the build-your-own.org projects, but doing things like implementing a PHP web server with a class loader without any frameworks, or a JS templating system which stores the state of the UI in a big ol' object and updates the DOM automatically, has given me the deeper understanding of these things that I've later needed to debug weird issues and find creative fixes.
A friend of mine was going to graduate, and had one required class that she had to take. She had dropped it twice before because she said it was too difficult for her. If she didn't pass, she couldn't take the class again, or maybe for a year, not sure exactly.
Well, at the start of the semester at the first mid-term, she failed the first one. At that point, I decided to step in. I made her read every single chapter, completely, and understand most of it, before going to class, where the teacher went over everything and she could ask the one or two very incisive and important questions that she honestly didn't understand in the book.
I also "made" her do an extra credit paper (meaning I really just put the screws to her to do it, no mercy. For her own good - ie not graduating and failing out with one semester to go.)
As you might expect, she got perfect scores on her next two mid-term tests, and her report she also had an A+.
Reading to understand and working hard at it, you will understand. And reading it before each class, in order to ask the one or two or four honest questions that you have to fill in the honest few things you just can't understand.
And actually, with the internet, for almost all things, you can look up multiple sources on the topic and read them all - because each author has a different perspective and when you see the concept from many different perspectives, you really do understand deeply.
I got pretty much all C's, but then started doing this and then got all A's. Plus I leared memory mnemonics - Tony Buzon - for perfect memory. Learning mnemonics required a lot of work, but so worth it I would make it a required topic to teach starting in 1st grade, and every grade after that. Just like everyone should be required to take probability and statistics classes starting in 7th grade to 12th grade. Just those two.
At my undergrad you had to do the homework and attend all the lectures to get a C+/B-. As/Bs were the realm of studying the class beforehand or, preferably, having taken it once before the time.
This article really frustrated me the last time it was posted, but now I think I have a clear enough idea why to put it into words.
You absolutely can tell people things! But to tell someone something, you need to do three things:
- Establish enough shared context that they can understand you
- Speak about something they actually care about (or convince them to care)
- Use a format that works for what you're trying to communicate
I've told people about technical topics in conversation or presentation plenty of times. I don't really understand how you can be an engineer and not tell people things.
The article is frustrating because it puts the blame onto the listener. But that's not how communication works! You don't have the right to just "tell people things" - you have to put in the work to be understood and to show people it's worth their time to pay attention. Communicating well is part of the job.
The author does this even in this piece. He says: "For example, with just a few magic HTML tags we could stick avatars on a web page — pretty much any web page. For months Randy kept getting up at management meetings and saying, “We’ll be able to put avatars on web pages. Start thinking about what you might do with that.”
Well yeah - if someone stood up in a management meeting and told me to think about what I might do with an avatar, my first thought would be "What's an avatar" and my second thought would be "I'm already busy with other things". He hasn't said what an avatar is or why anyone should care about them. So people pay no attention.
Then the demo shows what an avatar is and lets people see immediately how it might be useful, in a format that works regardless of the skill of the communicator. And so of course now they understand!
And maybe the thing you're trying to communicate is so novel that you can't establish context or convince people to care without showing them a demo. But at least take the time to understand why you can't tell people about it.
Whoever wrote this piece is a bad communicator who thinks he's a good communicator, and then goes through life getting confused about why no one understands him.
For example, in the story about the Japanese, he assumes some context from the reader: "What are they building? What is Fujitsu and what is Habitat and why does Japan need their own special version of it?" It's not even clear they're building anything technical, so I wondered, "Why do they need a client and server? And also, didn't you ever think to check their internal technical details before? What did you think would be the result?"
> “who’s going to pay to make all those links?” or “why would anyone want to put documents online?”
These are good questions and you better have a well-worded answer! In fact, it is easy to answer these if you have prepared for them. Did you just assume that the potential customer would already have familiar with the technicals of your product? If that was the case, they would have bought it from someone else already. Your target demo is the uninformed.
> Whoever wrote this piece is a bad communicator who thinks he's a good communicator, and then goes through life getting confused about why no one understands him.
As much as I appreciate the author writing this piece, I have to agree with your comment. I was half wondering if the entire article was an exercise in “this is how you don’t communicate, here’s the final para which explains everything I wrote!”. Reading the comments on the article and here on HN helped.
Habitat [1] was a massively multi-player online role-playing game, or perhaps one of the first ones, by LucasArts.
It was first released in 1986, and awarded at Game Developers Choice Awards in 2001. So, pretty old-school stuff. The site is about that game, and stories around it, so perhaps one can assume that readers know at least that much context.
> For example, in the story about the Japanese, he assumes some context from the reader: "What are they building? What is Fujitsu and what is Habitat and why does Japan need their own special version of it?" It's not even clear they're building anything technical, so I wondered, "Why do they need a client and server? And also, didn't you ever think to check their internal technical details before? What did you think would be the result?"
None of those details are important to the story. Leaving out extraneous detail is good communication.
Yeah it was a very confusing read, and when I finally understood the point of the writer I wondered if he wrote a confusing article on purpose to create a "Aha!" moment when you finally understand what he's saying, to illustrate his point through this article.
It's also ... if not absolutely necessary, then tremendously helpful to the process ... to observe the subjects' expressed understanding and see if that matches the lesson or message you're trying to convey.
This distinction to me is what distinguishes various informational approaches, in which delivery is one-way, from instructional approaches, in which the teacher closely observes and monitors students to see what understanding they're forming, and to bring them back on course if they're straying from it.
This is challenging at scale, or at distance, or over time (e.g., in writing a book that's used passively in instruction). It's a chief reason I suspect that various methods of scaling instruction perform poorly. It's why even generative-AI approaches to machine-guided instruction are likely to perform poorly --- such tools can explain material or respond to prompts, but it seems don't of themselves address the monitoring-and-guidance approach.
That said, in technical contexts whether in school-based teaching, professional training, or marketing support / vendor-based instruction, efficacy can be hugely improved by adding this step.
"...you have to put in the work to be understood and to show people it's worth their time to pay attention. Communicating well is part of the job."
That's fine and I don't disagree but the problem is that there are very few truly good communicators out there (there is just not enough to go around).
Communication is difficult and really effective communication is very difficult. I'll use myself as an example, When I post to HN I usually know what I want to say and I generally make myself clear—in fact I go considerable lengths ensure I'm not misinterpreted or misconstrued.
To ensure this does not happen I'll often restate what I've said in a different way and or provide examples. This makes my comments long, prolix and boring so few bother to read them.
I don't have a Shakespeare-like talent to be short, pithy and simultaneously convey what I'm saying both succinctly and accurately, and I believe that not many people do.
OK, so when people have some talent, skill or knowledge that needs to be conveyed to the world and they can't get their point across effectively then what are they to do?
> OK, so when people have some talent, skill or knowledge that needs to be conveyed to the world and they can't get their point across effectively then what are they to do?
Try, and try again.
Put yourself in the shoes of your audience. What do they know already? What will they think when you say this or that? Is that what you want them to think?
Re windy comments, I sympathize, but I've learned it's ok to be terse. You will never get across to 100% of your readers. Make your prose worth reading for those whom you reach.
So all this article is really trying to communicate is that you can't get people on board with your product/software without them seeing/using it first. It's the general idea behind the MVP strategy that ycombinator teaches.
>The article is frustrating because it puts the blame onto the listener
That wasn't the impression I got from the article - I think the title was a tongue in cheek way of saying quite the opposite: telling people isn't enough, it's your job as the explainer to guide people through the full path to understanding.
That's more or less what you described in the rest of your post, so I sense you're in concordance with the author
If it's a management meeting at a company developing a massively online multiplayer game as one of their main projects, perhaps you would assume people to know what an avatar is, in that context.
That still doesn't make it sound or feel like a good idea, to me. :)
The whole point is that to create the demo in the first place someone has to approve it, so if they cannot understand the thing then it doesn't get approved.
> Well yeah - if someone stood up in a management meeting and told me to think about what I might do with an avatar, my first thought would be "What's an avatar" and my second thought would be "I'm already busy with other things". He hasn't said what an avatar is or why anyone should care about them. So people pay no attention
> And maybe the thing you're trying to communicate is so novel that you can't establish context or convince people to care without showing them a demo. But at least take the time to understand why you can't tell people about it
Forget novel ideas, even relatively simple ideas within your domain are hard to communicate to other people especially if they are outside your domain.
Add people being busy and you having limited time to explain things on top of that and it becomes extremely difficult to get anything done unless someone trusts you and does not require a full explanation.
Stack ambiguity on top of that (E.g. "we think there's something here but can't pinpoint it) and there's a 0% chance of it happening.
I'm surprised this guy's management actually let him build the thing given the way he recounts how things went down.
> When people ask me about my life’s ambitions, I often joke that my goal is to become independently wealthy so that I can afford to get some work done. Mainly that’s about being able to do things without having to explain them first, so that the finished product can be the explanation. I think this will be a major labor saving improvement.
Agreed. The worst part is when they nod and pretend they got it and then that really important thing that you thought you had communicated everyone that was akin to the sky being on fire is not being prioritized/funded/worked on.
When I first started pair programming, I found the most important part was even if I had a solution that would work, if I couldn't explain it to my partner, it was not a very good solution.
Being able to convey an idea often forces it to be closer to the simplest and most comprehensible version of that idea.
> if I couldn't explain it to my partner, it was not a very good solution
Absolutely. I used to work in an aviation company writing software. When we were discusing architectural decisions (big or small) we often played around with the following thought experiment: Imagine that there was an accident with an airplane which was using our software. Maybe it was our fault, maybe not. Nobody knows. You are sitting in a cramped room with an irate NTSB investigator who points at this part of the code and asks “why on earth did you think this was a good idea?” If you don’t have a good and easy to understand answer to this question maybe we shouldn’t do it that way.
> ...if I couldn't explain it to my partner, it was not a very good solution.
I've found that in situations like that, I just write out a prototype and that clears up the confusion.
Usually, what I'm trying to explain is good, but there's some context that I can't get out of my head and into theirs that's critical to understanding the thing I'm trying to explain.
On the flip side, if you can't get people to get excited about the problem and solution now, what are the chances you're totally on the mark - and even if you are, whether you can get them to care after?
What do you think is the ratio of "they didn't understand my genius until I built it" to "I built it and nobody cared or used it"?
One of the reasons I've intentionally decided not become independently wealthy is that I want to have to explain to other people why I'm doing things. Part of my work is "charity-ish", and by not being able to do things on my own, I'm forced me to improve my communication skills and involve other people in these charity activities. I think that ultimately improves the final outcome, even if the process is immensely more frustrating.
That presumes that communication is just about people understanding the thing you are trying to say, rather than the transfer of thoughts, ideas, and opinions.
Put another way: the only thing that you can verify when communicating that the sky is on fire is that the other party knows you think the sky is on fire. Whether or not they think the sky is on fire, or whether they think you're blowing hot air, is pretty inscrutable. If they don't agree with you, they're still likely to not show it because you (to them) are a neurotic hothead who they don't want to upset.
The only truth in what you manage to communicate to people is the effect it has on how they act. In other words - they can't tell you anything! They can only show you.
I don't know what they had in mind, but I'm guessing something like "we need to refactor/restructure our code, it's a mess right now" might fall into that bucket.
> At Communities.com we developed a system called Passport (I’ll save the astonishing trademark story for a later posting) that let us do some pretty amazing things with web browsers. For example, with just a few magic HTML tags we could stick avatars on a web page — pretty much any web page. For months Randy kept getting up at management meetings and saying, “We’ll be able to put avatars on web pages. Start thinking about what you might do with that.” Mostly, nobody reacted much. After a couple of months of this we had things working, and so he got up and presented a demo of avatars walking around on top of our company home page. People were amazed, joyful, and enthusiastic. But they also pretty much all said the same thing: “why didn’t you tell us that we could put avatars on web pages?” You can’t tell people anything.
I guess they've just proven their point. I was a web developer in 2004 (I don't miss that!), and I still have no idea what this means and why people want it. You can't tell people anything, it seems...
Gravatar, "Login with X", Oauth, OpenID, W3c Solid, Metamask. The idea started with having an identity source-of-truth across websites but it's used to pass personal data around webapps, for e-commerce too.
The article axiom is an extension of the famous John Watson[1] quote :
"Be Kind; Everyone You Meet is Fighting a Hard Battle"
Which naturally extends to "You don't know", and I believe, the equivalent of this 2004 article. Now, as the article suggests, this is best shown - instead of explained.
A Neuroscientist was interviewed in a podcast [2] and illustrated this point with a story. He tells of a time when he dated long distance. Whenever he called his gf, she would ask him to call back and do facetime instead. At first, he thought it was cute, but eventually this quirk started eating at him, making him irritated. Finally, he built the nerve to ask. Why do you always want to do facetime ?
She explained that whenever he calls, she literally couldn't picture him in her head. Apparently she had a rare and poorly research condition[3], where the affected person cannot mentally draw a picture of the interlocutor (caller). So for the gf, resorting to facetime was the only thing that she could do, in order to picture of him.
You can't tell people anything = You just don't know. One of the same. Egotistic vs Humble. Take your pick.
Someone posted a Jan Morris interview here some time ago, and in it she said "Everything good in the world is kindness."
Anyway, I've been humbled a few times where I thought someone wasn't understanding me, but really I didn't understand them or some piece of the problem outside my experience. On HN I learned about Chesterton's Fence, between that and the thought on kindness, I try to just be nice and assume I'm missing something when reasonable people disagree with me and if I'm not the one missing something, my questions will only help me help them understand my perspective.
This puts to clear words a nagging sensation I've had my whole life, that I (also) tried to tell other people but couldn't. Talking just doesn't work as well as people think it does. It seems so obvious once you start noticing, but almost nobody notices this. People are either just repeating memes each other basically already know, or else are talking past each other.
When I overhear people having a conversation, almost always, I can see that the two people aren't really understanding each other, and neither of them realize it. Unless the content is basically emotion only, with no real new information. If I try to be aware of this dynamic when listening to people, it frustrates them... without me jumping to (probably wrong) conclusions and pretending to understand, it seems like I have some kind of communications disability and/or obnoxious personality, where I am drawing out the explanation way more than they expected to be necessary. The social norm is to think or at least act like you understand immediately when you really don't.
I often get the impression, when trying to explain totally new ideas to people, that they just assign it to the nearest known trope/meme, and assume it's that, and are unable to see how it's different. Even when it's really really different!
As an academic scientist this is a HUGE problem. All of my really new ideas, I cannot communicate to anyone, and get funding for them. Only the obvious/stupid things, the things they expect, get understood and funded. I then do my real work, the stuff I later get tons of praise for, "in my spare time." Like the author, I too wish I was independently wealthy, so I could actually do my job!
I like this perspective because it puts the burden back on me, and gives me something to act on. How can I put them into this experience, so they really get it?
> I am drawing out the explanation way more than they expected to be necessary. The social norm is to think or at least act like you understand immediately when you really don't.
I have wrestled with this since I was a child. For a long time I felt I was frequently missing information everyone else was picking up on (which was surely true in some cases, but not all)-- it even led to a brief delusion where I became paranoid about a large subpopulation of telepaths living alongside us, with access to their own hidden world containing significantly richer detail and depth, who pitied disadvantaged individuals like me.
Turns out a whole lot of people have simply felt forced into faking deeper understanding due to pressures of social competition, and gotten really good at it. Like an arms race of bullshitting. A facade of wisdom and knowledge.
Like you, I've often been accused of 'over-explaining' - which many do in fact find annoying - but we only do it because we've found it's necessary to avoid miscommunication.
So, thanks for writing your comment, I feel slightly better about the situation now.
I don't know if it's just stupidity, ignorance, or a lack of effort. Maybe a combination. But this type of deficiency is so prevalent, it makes me want to see fully autonomous AI. Even though I know it's an incredibly bad idea and would never actively work towards it until it becomes an irreversible trend. People are stupid. They will do it soon enough. Hopefully it doesn't immediately turn into a disaster.
But GPT 4 definitely listens and understands. Usually.
You draw a picture of a house. Someone else also draws a picture of a house (but it looks different from yours, as they had a different idea in their head as to what a house should look like).
"That's not right, it's supposed to be on a hill"
You draw a picture of house on a hill.
"Still not quite what I had in mind. It should have 3 front windows, with a chimney on the left side, and a tree between the door and the driveway."
You redraw the picture again.
"Not that kind of tree... it should be pointier"
"You mean like a conifer?"
"What's a conifer?"
Words simply aren't very good at conveying a lot of information, so it can take many words to get a clear message across to someone. Additionally, there's often multiple ways to interpret words, which humor typically plays on. And then, with people not all sharing the same knowledge/understanding of things, it often becomes very difficult to tell someone something, especially complicated things. I mean just imagine how hard it would be to try telling tribal people who live away from civilization, who don't even know what a computer is, about ChatGPT... that is, if they even understand the language you speak.
>When I overhear people having a conversation, almost always, I can see that the two people aren't really understanding each other, and neither of them realize it.
I have noticed exactly the same. But meanwhile, you and me are sitting there understanding what each of the two means, and how each misunderstands the other. So clearly it's possible for someone to understand what is said??
This is the anarchist theory of the unity of means and ends: you can't tell people anything, including what a better society looks like (nor can you impose it on anyone). You have to try to build and live that society in the present in order to even "get it" yourself.
I have some opinions about therapy being the art of asking people questions in order to trick them into telling themselves things that nobody else can tell them.
I've had more than a couple coworkers over the years who pull me into random decisions because they think I ask good questions. Part of me would rather be known for having good answers, but it's better than being discounted, and practically it may end up being more effective.
That’s exactly what therapy is - you’ll never get a “diagnosis” out of a good therapist. Especially in fields like CBT. The point being to make you think (cognitive) about your response/actions (behavioral).
In my perspective this derives from the fact that "knowledge" itself is not some abstract, immaterial thing in the brain, but rather thinking and knowing works through the body and the practical application. Hence the saying "hands-on". Understanding something requires experiencing it, as experience (sensation) is inseparable from consciousness. (We call this phenomenological Emboiment, feel free to check it out!)
You cannot build an ideal society in theory as it will not be realisitic in practice, and you cannot explain reason and feel with words alone, but others need to walk the same experiential path as you did. But since you already figured it out you can help them walk it faster. Thats why we now learn in primary school what originally took a long time until some greek matematician worked it out.
It also sounds rooted in Epictetus, the Greek Stoic. Interestingly, Christian monastics used the Enchiridion of Epictetus to guide their intentional communities.
Of course, these anarchist projects all fail (like Occupy) or have murder rates significantly above baseline society (like CHAZ) so this shows… something.
Like what is this comment? Why do people with an extremely narrow view of some political topic or politcal movement like anarchism/communism/immigration policies/etc. start posting something that dismissive. Like as if they where actually talking about the same thing, without realizing that they dismisse a delusional idea of something which has no to little to do with what the other person was talking about.
Now realize that this applies to everything, not just computers. Unless you've walked a mile in their shoes, you don't really understand the struggles of LGBTQ folks, the homeless, CEOs, assembly line workers... They will try to tell you, but you can't tell people anything, unless you're telling someone with good imagination and empathy.
I deeply appreciate your point that lack of understanding goes both downward (homeless) and upward (CEOs.) I think that's true.
But - unrelated to your point - I think there's more empathy (or understanding, at least) - unexpectedly - downwards. A CEO knows how to become homeless (quit job, give away house and money) if they chose to for some reason. A homeless person has no path to CEO.
Readit News
Isn't this the very reason that homework assignments exist? When I was in college, it took me two years to realize that I could have easily gotten As and Bs (instead of Bs and Cs) in my various math classes, physics, chemistry, etc. had I simply bothered to do my homework properly--or, to put it differently, had I only properly applied the knowledge which I passively acquired by reading the associated textbook sections. Notably, I was always convinced that I "understood" everything I had read, only to find out otherwise when I tried to solve any of the problems. It was only after I actively applied the knowledge I had picked up, that my understanding transitioned from superficial (as in, understanding the underlying logic itself) to concrete (being able to apply the logic toward other problems).
If you apply this to abstract and esoteric technical concepts, it is easy to see why someone might say they understand--and even believe that they do--while in reality only having a superficial understanding at best. The problem then becomes getting the other person to spend the effort to properly internalize the concepts being conveyed, before they have built up the requisite interest in the idea to be sufficiently motivated to carry out said effort.
It's probably also true, however, that this may only apply to sufficiently esoteric and complex ideas in the first place.
I took this realisation to the extreme and decided to completely deprioritise classes in favour of homework and doing the reading in my own time. I figured the cost-benefit, at least for me, was much higher if I spent an hour doing as opposed to an hour listening.
This actually worked really well for me but that might also be because I often struggle with large classroom learning: the pace is either too slow and I get distracted or too fast and I can't keep up. But even when the learning is 'one-to-one' I feel like there's always the tendency for people to zone out and not raise an issue when they are either bored or did not keep up/understand.
I think you're right in that some of that might be the brain pretending that it understood when in fact it did not. Could it also be a social thing? Maybe because the other person expects you to understand and this causes your brain to try its best to believe that it understood when it did not.
Otherwise yes, you are often better off with the book and homework and actually completing both.
And this is where I usually go off on my rant how college is useless for education - you can 'learn' peer interaction but even this way nothing forces you to make friends or connections, so the inherent use is close to nill.
This is a problem for everyone. People may not experience the struggle directly, but it's there. It's the inherent problem with any group class.
> I feel like there's always the tendency for people to zone out and not raise an issue when they are either bored or did not keep up/understand.
Tight loops of interaction will make this nearly impossible. What you'll have to watch out for instead is overwhelming students to the point of distress (speaking from experience).
I had two mathematics teachers in high school that took this approach to class time. They would lecture for 10 minutes then have the class do work in groups on homework assignments for the remaining 45 minutes. Exact times varied based on the complexity of the material, of course. And I do not believe their format was sanctioned by the administration. But it sure helped me retain more of the information.
As an extra bonus, it really helped with procrastination. It was a lot easier to get to work on an already-started homework assignment compared to staring at a blank page and a daunting list of problems.
That depends very much on the lecturer. If he or she just follows one book for the whole semester and doesn't add any additional insight, explanation of tradeoffs, or enlightening anecdotes, then by any means skip the lecture. But if you have a lecturer giving you all from above and the possibility to ask questions, lectures can be invaluable. Luckily I experienced many of the latter kind, although some of the first.
Deleted Comment
You'll think you get it.
But there are questions that are about a deeper understanding. For these there's some point to the question that isn't obvious from just reading. I remember the first few question sheets I got in uni, there would be questions that appeared to have nothing to do with what was presented at all. Only by asking around did I discover what the cryptic connections were. If you don't do the question sheets, you won't see this.
Yes, this is my biggest gripe with learning as well. It's so individual that you almost need one teacher per student.
* Some only "get it" when it's explained to them in a classroom, preferably by an engaging teacher; while others have to read it in the book.
* Many, perhaps most, won't really get it unless they do the homework; but others can do all of the exercises you give them, and still fail to get the bigger picture, and will be lost if faced with something slightly different than what they exercised.
The way I like to tell people this is, you know that thing where you want to take your kids or your students and make sure that they don't have to go through the same pain that you did to learn what you had learned? That feeling?
If you have ever had that feeling you must understand that it was horribly mistaken. The problem is precisely that learning and pain go hand-in-hand. We learn abstractions precisely because they relieve a sort of confusion, a sort of difficulty, they organize the pain that we have experienced and make it tidy and less painful. Even as kids abstractions like “this is what it means for the stove to be on” work this way, not that you have to get burned, maybe you just need to be yelled at, but it organizes that pain of being yelled at and tells you when are you getting yelled at and why.
Not all learning is this way, for example memorization and repetition... but to a first approximation you have to get lost in the forest before the landmarks on the map are recognizable.
Corollary: you CAN tell people things, but it might be more involved than just a casual conversation. Either you have to establish a shared context, tap into the pain/confusion that they already have... Or you have to get them interested enough that they will follow you down a rabbit hole of confusion and difficulty so that you can finally explain the thing.
I am not yet convinced. Let me take the following example because I remember it well: I tried to learn about geometric optics and lenses. To do so, I downloaded multiple lecture PDFs from the internet and read them.
All of them started to explain lenses by describing how large the image of a real object is, defining focal length and such. Literally not a single one of them even defined what an image is. The whole talk about image size and focal length and magnification and what not was totally worthless because it was all based on the same fundamental word that had no meaning. In the end, I tried to make up my own definition that was consistent with all those PDFs, but it left me unsure if I got it right.
The "pain" was there, but there was no relief from it, and I am still convinced that that simple definition could have avoided it. If I were to teach optics, I would give such a definition, and I am still convinced that it would help with the pain.
The other example was one of those PDFs that showed a real object that was "wide" along the distance axis from the lens, and so by my understanding should have an image whose magnification changes along the distance axis, but the explanatory picture in the PDF showed equal magnification everywhere. Today I am convinced that this "explanatory picture" was simply wrong.
Again, there was "pain", no relief (because I could not be sure), and I am still convinced that a learner can be saved a lot of pain when you just exlucde factually wrong content from the learning material.
I would say though, about abstractions, I'm not sure hot stove is an abstraction. I think hot is an abstraction that allows you to apply the lesson you learned on the stove to the toaster and the coffee maker and all the other hot appliances you encounter.
Absolutely. It is also why I will never understand why it is considered a virtue (especially in the maths and physics community) to set assignments without exemplary solution. I'm not asking to give this exemplary solution to students right away, but it should definitely exist and be given when students have shown honest effort and especially to weaker students that are struggling.
Instead the "left as an exercise to the reader" mentality is celebrated when in reality - in my opinion - it is almost always just a convenient excuse for the laziness of the professor at the expense of their students.
1. Struggle with exercise.
2. Check proposed solution.
3. Thinking that your understanding of the proposed solution constitutes an ability to solve it yourself.
This is analogous to the original problem, where you assume the ability to follow reasoning translates to being able to reproduce it.
The main use of exercises without solutions is to try and force students to go to their teachers and interact, so that someone can try to pinpoint what exactly the missunderstanding is.
This can probably be better done with assigned homework. However grading can be time consuming, so trying to filter so that only those that have problem with specific exercises come to see you is one way to try and get some time economy going.
Also yeah. This is better in course specific work sheets than in books, since books should be self contained enough to be usable without the guidence of TAs or professors.
I had a boss years ago who would encourage me, whenever I was interested in e.g. a new library or framework, to go and reinvent the wheel in my spare time i.e. build it myself from scratch first to understand what the problems it solves actually are.
I've never built my own anything quite as low level as the build-your-own.org projects, but doing things like implementing a PHP web server with a class loader without any frameworks, or a JS templating system which stores the state of the UI in a big ol' object and updates the DOM automatically, has given me the deeper understanding of these things that I've later needed to debug weird issues and find creative fixes.
A friend of mine was going to graduate, and had one required class that she had to take. She had dropped it twice before because she said it was too difficult for her. If she didn't pass, she couldn't take the class again, or maybe for a year, not sure exactly.
Well, at the start of the semester at the first mid-term, she failed the first one. At that point, I decided to step in. I made her read every single chapter, completely, and understand most of it, before going to class, where the teacher went over everything and she could ask the one or two very incisive and important questions that she honestly didn't understand in the book.
I also "made" her do an extra credit paper (meaning I really just put the screws to her to do it, no mercy. For her own good - ie not graduating and failing out with one semester to go.)
As you might expect, she got perfect scores on her next two mid-term tests, and her report she also had an A+.
Reading to understand and working hard at it, you will understand. And reading it before each class, in order to ask the one or two or four honest questions that you have to fill in the honest few things you just can't understand.
And actually, with the internet, for almost all things, you can look up multiple sources on the topic and read them all - because each author has a different perspective and when you see the concept from many different perspectives, you really do understand deeply.
I got pretty much all C's, but then started doing this and then got all A's. Plus I leared memory mnemonics - Tony Buzon - for perfect memory. Learning mnemonics required a lot of work, but so worth it I would make it a required topic to teach starting in 1st grade, and every grade after that. Just like everyone should be required to take probability and statistics classes starting in 7th grade to 12th grade. Just those two.
As per article above in which the hubris is its own comeuppance.
Deleted Comment
Deleted Comment
You absolutely can tell people things! But to tell someone something, you need to do three things:
I've told people about technical topics in conversation or presentation plenty of times. I don't really understand how you can be an engineer and not tell people things.The article is frustrating because it puts the blame onto the listener. But that's not how communication works! You don't have the right to just "tell people things" - you have to put in the work to be understood and to show people it's worth their time to pay attention. Communicating well is part of the job.
The author does this even in this piece. He says: "For example, with just a few magic HTML tags we could stick avatars on a web page — pretty much any web page. For months Randy kept getting up at management meetings and saying, “We’ll be able to put avatars on web pages. Start thinking about what you might do with that.”
Well yeah - if someone stood up in a management meeting and told me to think about what I might do with an avatar, my first thought would be "What's an avatar" and my second thought would be "I'm already busy with other things". He hasn't said what an avatar is or why anyone should care about them. So people pay no attention.
Then the demo shows what an avatar is and lets people see immediately how it might be useful, in a format that works regardless of the skill of the communicator. And so of course now they understand!
And maybe the thing you're trying to communicate is so novel that you can't establish context or convince people to care without showing them a demo. But at least take the time to understand why you can't tell people about it.
For example, in the story about the Japanese, he assumes some context from the reader: "What are they building? What is Fujitsu and what is Habitat and why does Japan need their own special version of it?" It's not even clear they're building anything technical, so I wondered, "Why do they need a client and server? And also, didn't you ever think to check their internal technical details before? What did you think would be the result?"
> “who’s going to pay to make all those links?” or “why would anyone want to put documents online?”
These are good questions and you better have a well-worded answer! In fact, it is easy to answer these if you have prepared for them. Did you just assume that the potential customer would already have familiar with the technicals of your product? If that was the case, they would have bought it from someone else already. Your target demo is the uninformed.
As much as I appreciate the author writing this piece, I have to agree with your comment. I was half wondering if the entire article was an exercise in “this is how you don’t communicate, here’s the final para which explains everything I wrote!”. Reading the comments on the article and here on HN helped.
It was first released in 1986, and awarded at Game Developers Choice Awards in 2001. So, pretty old-school stuff. The site is about that game, and stories around it, so perhaps one can assume that readers know at least that much context.
[1]: https://en.wikipedia.org/wiki/Habitat_(video_game)
None of those details are important to the story. Leaving out extraneous detail is good communication.
This distinction to me is what distinguishes various informational approaches, in which delivery is one-way, from instructional approaches, in which the teacher closely observes and monitors students to see what understanding they're forming, and to bring them back on course if they're straying from it.
This is challenging at scale, or at distance, or over time (e.g., in writing a book that's used passively in instruction). It's a chief reason I suspect that various methods of scaling instruction perform poorly. It's why even generative-AI approaches to machine-guided instruction are likely to perform poorly --- such tools can explain material or respond to prompts, but it seems don't of themselves address the monitoring-and-guidance approach.
That said, in technical contexts whether in school-based teaching, professional training, or marketing support / vendor-based instruction, efficacy can be hugely improved by adding this step.
That's fine and I don't disagree but the problem is that there are very few truly good communicators out there (there is just not enough to go around).
Communication is difficult and really effective communication is very difficult. I'll use myself as an example, When I post to HN I usually know what I want to say and I generally make myself clear—in fact I go considerable lengths ensure I'm not misinterpreted or misconstrued.
To ensure this does not happen I'll often restate what I've said in a different way and or provide examples. This makes my comments long, prolix and boring so few bother to read them.
I don't have a Shakespeare-like talent to be short, pithy and simultaneously convey what I'm saying both succinctly and accurately, and I believe that not many people do.
OK, so when people have some talent, skill or knowledge that needs to be conveyed to the world and they can't get their point across effectively then what are they to do?
Try, and try again.
Put yourself in the shoes of your audience. What do they know already? What will they think when you say this or that? Is that what you want them to think?
How do you know?
That wasn't the impression I got from the article - I think the title was a tongue in cheek way of saying quite the opposite: telling people isn't enough, it's your job as the explainer to guide people through the full path to understanding.
That's more or less what you described in the rest of your post, so I sense you're in concordance with the author
That still doesn't make it sound or feel like a good idea, to me. :)
> Well yeah - if someone stood up in a management meeting and told me to think about what I might do with an avatar, my first thought would be "What's an avatar" and my second thought would be "I'm already busy with other things". He hasn't said what an avatar is or why anyone should care about them. So people pay no attention
> And maybe the thing you're trying to communicate is so novel that you can't establish context or convince people to care without showing them a demo. But at least take the time to understand why you can't tell people about it
Forget novel ideas, even relatively simple ideas within your domain are hard to communicate to other people especially if they are outside your domain.
Add people being busy and you having limited time to explain things on top of that and it becomes extremely difficult to get anything done unless someone trusts you and does not require a full explanation.
Stack ambiguity on top of that (E.g. "we think there's something here but can't pinpoint it) and there's a 0% chance of it happening.
I'm surprised this guy's management actually let him build the thing given the way he recounts how things went down.
Agreed. The worst part is when they nod and pretend they got it and then that really important thing that you thought you had communicated everyone that was akin to the sky being on fire is not being prioritized/funded/worked on.
Being able to convey an idea often forces it to be closer to the simplest and most comprehensible version of that idea.
Absolutely. I used to work in an aviation company writing software. When we were discusing architectural decisions (big or small) we often played around with the following thought experiment: Imagine that there was an accident with an airplane which was using our software. Maybe it was our fault, maybe not. Nobody knows. You are sitting in a cramped room with an irate NTSB investigator who points at this part of the code and asks “why on earth did you think this was a good idea?” If you don’t have a good and easy to understand answer to this question maybe we shouldn’t do it that way.
I've found that in situations like that, I just write out a prototype and that clears up the confusion.
Usually, what I'm trying to explain is good, but there's some context that I can't get out of my head and into theirs that's critical to understanding the thing I'm trying to explain.
What do you think is the ratio of "they didn't understand my genius until I built it" to "I built it and nobody cared or used it"?
Put another way: the only thing that you can verify when communicating that the sky is on fire is that the other party knows you think the sky is on fire. Whether or not they think the sky is on fire, or whether they think you're blowing hot air, is pretty inscrutable. If they don't agree with you, they're still likely to not show it because you (to them) are a neurotic hothead who they don't want to upset.
The only truth in what you manage to communicate to people is the effect it has on how they act. In other words - they can't tell you anything! They can only show you.
Life will slap you with this lesson time and time again.
> At Communities.com we developed a system called Passport (I’ll save the astonishing trademark story for a later posting) that let us do some pretty amazing things with web browsers. For example, with just a few magic HTML tags we could stick avatars on a web page — pretty much any web page. For months Randy kept getting up at management meetings and saying, “We’ll be able to put avatars on web pages. Start thinking about what you might do with that.” Mostly, nobody reacted much. After a couple of months of this we had things working, and so he got up and presented a demo of avatars walking around on top of our company home page. People were amazed, joyful, and enthusiastic. But they also pretty much all said the same thing: “why didn’t you tell us that we could put avatars on web pages?” You can’t tell people anything.
I guess they've just proven their point. I was a web developer in 2004 (I don't miss that!), and I still have no idea what this means and why people want it. You can't tell people anything, it seems...
Which naturally extends to "You don't know", and I believe, the equivalent of this 2004 article. Now, as the article suggests, this is best shown - instead of explained.
A Neuroscientist was interviewed in a podcast [2] and illustrated this point with a story. He tells of a time when he dated long distance. Whenever he called his gf, she would ask him to call back and do facetime instead. At first, he thought it was cute, but eventually this quirk started eating at him, making him irritated. Finally, he built the nerve to ask. Why do you always want to do facetime ?
She explained that whenever he calls, she literally couldn't picture him in her head. Apparently she had a rare and poorly research condition[3], where the affected person cannot mentally draw a picture of the interlocutor (caller). So for the gf, resorting to facetime was the only thing that she could do, in order to picture of him.
You can't tell people anything = You just don't know. One of the same. Egotistic vs Humble. Take your pick.
[1] https://checkyourfact.com/2019/08/15/fact-check-plato-hard-b... [2] econtalk.org [3] https://en.wikipedia.org/wiki/Aphantasia
Anyway, I've been humbled a few times where I thought someone wasn't understanding me, but really I didn't understand them or some piece of the problem outside my experience. On HN I learned about Chesterton's Fence, between that and the thought on kindness, I try to just be nice and assume I'm missing something when reasonable people disagree with me and if I'm not the one missing something, my questions will only help me help them understand my perspective.
Dead Comment
When I overhear people having a conversation, almost always, I can see that the two people aren't really understanding each other, and neither of them realize it. Unless the content is basically emotion only, with no real new information. If I try to be aware of this dynamic when listening to people, it frustrates them... without me jumping to (probably wrong) conclusions and pretending to understand, it seems like I have some kind of communications disability and/or obnoxious personality, where I am drawing out the explanation way more than they expected to be necessary. The social norm is to think or at least act like you understand immediately when you really don't.
I often get the impression, when trying to explain totally new ideas to people, that they just assign it to the nearest known trope/meme, and assume it's that, and are unable to see how it's different. Even when it's really really different!
As an academic scientist this is a HUGE problem. All of my really new ideas, I cannot communicate to anyone, and get funding for them. Only the obvious/stupid things, the things they expect, get understood and funded. I then do my real work, the stuff I later get tons of praise for, "in my spare time." Like the author, I too wish I was independently wealthy, so I could actually do my job!
I like this perspective because it puts the burden back on me, and gives me something to act on. How can I put them into this experience, so they really get it?
I have wrestled with this since I was a child. For a long time I felt I was frequently missing information everyone else was picking up on (which was surely true in some cases, but not all)-- it even led to a brief delusion where I became paranoid about a large subpopulation of telepaths living alongside us, with access to their own hidden world containing significantly richer detail and depth, who pitied disadvantaged individuals like me.
Turns out a whole lot of people have simply felt forced into faking deeper understanding due to pressures of social competition, and gotten really good at it. Like an arms race of bullshitting. A facade of wisdom and knowledge.
Like you, I've often been accused of 'over-explaining' - which many do in fact find annoying - but we only do it because we've found it's necessary to avoid miscommunication.
So, thanks for writing your comment, I feel slightly better about the situation now.
Human-powered GPT.
But GPT 4 definitely listens and understands. Usually.
"Draw me a picture of a house"
You draw a picture of a house. Someone else also draws a picture of a house (but it looks different from yours, as they had a different idea in their head as to what a house should look like).
"That's not right, it's supposed to be on a hill"
You draw a picture of house on a hill.
"Still not quite what I had in mind. It should have 3 front windows, with a chimney on the left side, and a tree between the door and the driveway."
You redraw the picture again.
"Not that kind of tree... it should be pointier"
"You mean like a conifer?"
"What's a conifer?"
Words simply aren't very good at conveying a lot of information, so it can take many words to get a clear message across to someone. Additionally, there's often multiple ways to interpret words, which humor typically plays on. And then, with people not all sharing the same knowledge/understanding of things, it often becomes very difficult to tell someone something, especially complicated things. I mean just imagine how hard it would be to try telling tribal people who live away from civilization, who don't even know what a computer is, about ChatGPT... that is, if they even understand the language you speak.
I have noticed exactly the same. But meanwhile, you and me are sitting there understanding what each of the two means, and how each misunderstands the other. So clearly it's possible for someone to understand what is said??
https://theanarchistlibrary.org/library/anarchopac-means-and...
I've had more than a couple coworkers over the years who pull me into random decisions because they think I ask good questions. Part of me would rather be known for having good answers, but it's better than being discounted, and practically it may end up being more effective.
Deleted Comment
You cannot build an ideal society in theory as it will not be realisitic in practice, and you cannot explain reason and feel with words alone, but others need to walk the same experiential path as you did. But since you already figured it out you can help them walk it faster. Thats why we now learn in primary school what originally took a long time until some greek matematician worked it out.
But - unrelated to your point - I think there's more empathy (or understanding, at least) - unexpectedly - downwards. A CEO knows how to become homeless (quit job, give away house and money) if they chose to for some reason. A homeless person has no path to CEO.
Is not how people become homeless. There's no path to empathy there.