Andy Matuschak — The reason most learning tools fail

Domain knowledge matters for teaching inquiry-oriented classes (requiring improvisation) but not for rote instruction. Teachers need to understand material well enough to think on their feet and integrate student contributions, but not at expert level.

“if you aren't very familiar or comfortable in math, then you will struggle specifically to do inquiry oriented classes...you have to be able to think on your feet...But I don't think it requires anything like extraordinary ability.”

The education system for the median student is primarily structured around making students do things they don't want to do and achieve goals that aren't theirs, rather than helping them achieve their own goals more effectively.

“For the median student, the education system mostly wants to make the student do things they don't want to do. It's not about helping them achieve their goals more easily or more effectively. For the most part, it's about achieving goals that aren't theirs!”

Video has taken off as a pedagogical medium enormously (Grant Sanderson's 3Blue1Brown videos get millions of views from people who would never read a Springer graduate textbook), but video produces engagement without producing detailed understanding—most people don't retain much from watching explanatory videos.

“One is that I would argue that one medium has taken off in an absolutely enormous way and that's video. People love video. People will watch Grant Sanderson spend an hour going through some explanation of an esoteric math problem, people who would never crack a Springer graduate textbook in mathematics or something like that. The issue is that they will not walk away from that interaction with much understanding but they're much more engaged.”

Virtually everyone in education is focused on the bottom quartile performing students (not the median) because of equity arguments—they want to provide opportunities others had—and because marginal impact is greater for struggling students than for those already doing fine.

“Basically everyone in the educational space, this isn't totally true, but like 90+% of people in the educational space are focused on the bottom quartile, not even the median. And there's a good reason for this. Many people who are in education are motivated by arguments of equity and opportunity. They want everybody to have the opportunities they had.”

The shift of intelligent women away from teaching (historically one of few options) toward other fields may have reduced teaching quality, though it's unclear whether high intelligence is necessary for effective teaching compared to empathy and communication.

“one of the consequences of the very unforeseen circumstance of the mid-20th century, was that one of the very few occupations an intelligent woman could pursue was teaching...now that other options are available...there's other competition for the same very intelligent woman”

Predictive utility theory suggests the probability of retrieving a memory correlates with the brain's model of how useful that memory will be in the situation—frequent retrieval practice increases the brain's prediction of utility.

“One of the theories about how well we remember stuff in what circumstances is actually called predictive utility theory. And it suggests that the probability of retrieval of a particular item in a given situation actually does correspond with basically a model of to what extent the brain predicts it will be useful.”

Recasting Patreon from 'please support my work' (tip jar framing) to 'become a member' (membership framing with specific benefits) increased Matuschak's subscriber conversion and revenue, suggesting framing and positioning matter more than the underlying offering.

“I had much more success with my Patreon when I recast it as, 'Oh, please subscribe to support my work', like the thing you were describing, to an offering to become a member. 'When you become a member, these things will happen.' These things need not be terribly substantial necessarily. There's a difference between a tip jar and a membership in people's minds.”

Memory champions and multilingual people have memorized orders of magnitude more information than typical people, showing that we're far from human memory's actual limits.

“Some of these memory champions have memorized maybe two orders of magnitude more things than I have practiced. Certainly people who are multi-lingual have really, really absurd numbers of things memorized.”

Spreading activation theory explains forgetting as a functional feature—if you didn't forget, you'd have a cacophony of noise in memory retrieval; forgetting helps you focus on what's salient and relevant to the current situation by tuning weights on memory connections.

“We start from some point that is like a stimulus, and speaking very informally, we kind of expand outwards from there and there are effectively like weights on those connections. By tuning those weights effectively, we route the packets on the network effectively. Memory is encoded in these weights, at least partially. So if you didn't forget things, then you might just have this weird cacophony on the network and in particular, what's salient? What to do next?”

Scaffolding in education—providing simpler versions of problems first, partially worked solutions, or worked examples—is effective for building capability, but should be faded (gradually removed) as learners become more capable, because continuing scaffolding becomes ineffective as learners progress.

“We give you simpler versions of questions first, that's a kind of scaffolding. We partially work the answer first, that's a kind of scaffolding. We give you worked examples first, where we might ask you to predict the next step of the work example. That's also a kind of scaffolding. Where the metaphor breaks down is that once you become more capable, we try to remove the scaffolding. It's called fading. The idea is that once you have solved a lot of calculus problems, you don't need half of it worked out and you're just filling in one of the blanks anymore. And in fact, doing that would not be as effective a learning experience.”

Memories are not actually lost completely—decades later, properly cued, people can retrieve things they thought were forgotten, though you need to be careful about false memories created by suggestive cueing.

“Another weird thing about memory is that as far as we can determine, memories aren't lost exactly, at least not completely. There's a series of interesting experiments that people have used to demonstrate that decades later, things are still there. If you can cue them right, people can bring things back, even things that they feel are lost.”

When cognitive demand is high, metacognition becomes difficult—it's hard to monitor your own understanding, plan your next steps, and reflect on learning while also thinking hard about the material itself.

“One enduring result in education psychology is that when you're doing a lot of cognition, metacognition is difficult. What I mean by that is when you're thinking really hard about the stuff on the page, it's very difficult for you to plan, to regulate yourself, to figure out what the best next action to do is, to reflect and evaluate on whether you're understanding things.”

Memory champions and people using explicit memory systems don't lose other normal human knowledge—there's no resource management constraint; they simply know more. This is supported by experimental literature and the fact that multilingual people have absurdly high numbers of memorized things.

“One of the things that I think is kind of weird about this memory system stuff, or like memory champions, or something like that is 'Oh, if you do these things, will you start to forget other normal human stuff?' And what's weird is, no. I've been doing this memory system stuff for years and I just know more stuff now... Some of these memory champions have memorized maybe two orders of magnitude more things than I have practiced. Certainly people who are multi-lingual have really, really absurd numbers of things memorized. So there isn't a resource management argument.”

Predictive utility theory suggests the probability of retrieving a memory depends on how much the brain predicts it will be useful—repeated practice increases the predicted utility, and practicing in varied situations increases utility prediction across broader distributions.

“One of the theories about how well we remember stuff in what circumstances is actually called predictive utility theory. And it suggests that the probability of retrieval of a particular item in a given situation actually does correspond with basically a model of to what extent the brain predicts it will be useful.”

Creative insight often depends on memory: noticing surprising connections between disparate ideas requires those ideas to be in memory and salient; if you must consciously retrieve an idea, it won't seem salient when you encounter related information, preventing the creative leap.

“there's a bunch of studies trying to catalog case studies of how it is that people have flashes of insight. It's a little hard to talk about that but one of the things that's a pretty consistent source of insight for people is noticing a surprising connection or a surprising contradiction... Now that's only possible if the other thing is in your memory. If you have to think to look up the other thing, the newspaper wouldn't seem so salient to you.”

'Cool shit' in innovation is likely to come from disaffected and bored high-school students (20th percentile performers) who are not engaged by traditional education, suggesting that by not helping these students learn, society loses potential innovations.

“Part of the trouble here is that the cool shit is very likely to come from students who are performing at the 20th percentile in school, because they're disaffected and bored and none of this stuff matters to them, right? Part of the trouble here is that by opting out of helping these people learn, there are all kinds of interesting inventions that could probably occur that aren't occurring.”

Choose-your-own-adventure books and hypertext novels force authors to write the lowest common denominator because any page must work as the destination of multiple entry points. This constraint destroys narrative coherence.

“There were all these dreams of hypertext novels, for instance, and some people wrote them. And one of the problems that a hypertext novel has, and it actually can be seen in a choose your own adventure book that existed before there was digital hypertext, that the author is forced to write something like a lowest common denominator story, the page that is the destination of a hyperlink, it has to work as the endpoint of all of its reference. And so it can't establish any kind of coherent or consistent arc, unless there's a kind of sameness to all of the reference. And the more that there's sameness to the reference, the less useful hypertext is.”

Matuschak systematically avoids marketing despite having prominent work because he's worried about the corrosive effects of marketing and audience concerns on honest inquiry—marketing pushes you to present work in the best light and downplay limitations, which is 'death for discovery.'

“I systematically avoid marketing it. That's a self-protection thing. I am really worried about the corrosive influence of audience and marketing on honest inquiry. It is very easy to distort my work.”

Computing's appeal historically was personal empowerment and possibility. Many who got into computers were affected by formative experiences of self-directed learning and world-opening through computing, making any narrative about expanding human capacity compelling.

“Here's a more optimistic and generous response. Many of us got into computing because computers portray a sense of personal empowerment and possibility. We remember growing up, being locked in our bedrooms at midnight, fooling around. We have this very powerful tool at our disposal and it's opening up these worlds for us. For many people here, that was a formative part of their personal development. Anybody pointing to that and saying we can do more stuff like that is going to be pretty compelling.”

There may be pedagogical value in forgetting: the brain forgets things to manage attention and salience, and by only retaining gist rather than details, people may generalize insights better than if every detail remained equally salient.

“Do you think there's pedagogical value in forgetting? Some anecdotal or unrelated evidence is in neural networks where sometimes you can improve performance by pruning some of the weights. Obviously, we forget things and we don't remember everything... Is it possible that by not getting the details and only getting the gist, that actually helps us better generalize the insights we're getting from text?”

ACT-R theory by John Anderson explains learning through 'knowledge compilation'—the process of converting individual facts into higher-level patterns that can be generalized and applied in new contexts—and Anderson's theory suggests this compilation requires repeated exposure, active use, and variety of contexts, not just reading.

“Very practically speaking, it's funny that he uses the word compile, because one of the prominent theories of cognition, that is how we come to know and learn things, is this theory called ACT-R by John Anderson. A key part of it is this process that he calls knowledge compilation. This is the process by which we take individual facts and turn them into higher level patterns that we can generalize and apply in more contexts.”

Photoshop's layer abstraction is a powerful primitive that fundamentally changed what photo editing could be. Layers introduced a new way to think about the task, enabling non-destructive, composable manipulation.

“In Photoshop, for instance, you can manipulate a photo by means of a construct called a layer...The layer is like a new primitive that is introduced into the activity of photo editing. It utterly changed what you could do in photo editing.”

Hey Siri required a dedicated low-power coprocessor rather than using the main CPU for always-on voice detection because the main CPU couldn't run audio processing continuously without draining the battery; technical constraints drive architectural decisions.

“To be able to activate a voice command at any time without interacting with the device is great... That requires having the main CPU on all the time, processing audio buffers. You simply can't do that – it drains the battery. That attempt led to eventually having this dedicated co-processor that runs at a lower power that's very limited and restricted.”

The 'ceiling' of educational achievement (what the very best learners can achieve) has not been meaningfully raised historically; great minds from centuries past (Thomas Jefferson, von Neumann) demonstrate that top-tier intellectual achievement may not have improved.

“Has the ceiling been raised significantly? Depends on what we mean by the ceiling... Because you can go back like hundreds of years and the most learned people around... It's just incredible you look back on how many books Thomas Jefferson read... And certainly it's not at all clear that they're now producing more von Neumanns or something like that.”

Schoolhouse (pandemic education pod company) was innovative: hiring elementary teachers for small groups (5-6 kids) cost about 1/5 of private school tuition, making quality tutoring tractable for middle-income families.

“During the pandemic, I was struck by a company called Schoolhouse, which is now defunct, started by Brian Toball...Buying this one teacher's time split five or six ways was actually really very tractable...say you want to pay the person $50 an hour, maybe that seems reasonable for a teacher, this is not that hard to do, and actually costs substantially less than a private school.”

For some domains and some information (rare diagnoses, rare events that happen once every couple of years), naturalistic reinforcement is insufficient; out-of-band mechanisms like explicit spaced repetition are necessary because natural practice frequency is too low.

“For example, the linking that you brought up, one issue for doctors is rare diagnosis. So if it's only going to be once every couple of years that you see a patient that's going to present with these symptoms, that's not going to be frequent enough to naturally reinforce your memory of that. You're going to need some out of band mechanism.”

Demanding, active reading (taking seriously the requirement to understand each part) is more exhausting than passive reading, yet the exhaustion may be a signal of engagement rather than a sign that the approach is unsustainable.

“Whereas just the attention and the intensity to go through sentence by sentence and constantly being paying attention seems to be way more exhausting.”

From Nand to Tetris exemplifies active learning in a different form: instead of questions, the course structure itself is naturally active—students build a virtual computer from basics, with doing and making woven throughout rather than as separate practice.

“I don't know if you've seen From Nand to Tetris. This is a very interesting project that's kind of along these lines. And what characterizes it, like games, is doing. It's active. So when I was asking all those questions of the book, that was active learning, active reading, Nand to Tetris is naturally active... You start with basically nothing. You start with memory and you build a virtual computer and build Tetris... The whole thing's active.”

The feeling of 'almost working' during programming creates a compulsive debugging loop: small progress increments ('almost done', 'almost running end-to-end') trigger continued effort, generating sustained engagement without external motivation.

“There's this feeling that's very common for programmers, and it's laden with doom. The feeling is it's like 9 p.m., and you've been working on a thing all day, and it's almost working. It's almost working. And you know, if you just debug this one thing, then your project will be done... Before you know it, it's 2 a.m. You keep going because it feels so good. You feel the sense of forward progress.”

The question of whether all fields can be transformed into experiences with visible, rapid feedback cycles that create game-like engagement is open; Matuschak thinks we can get much closer in most domains than we currently are.

“Can every field be transformed into something where you can feel the sense of forward progress? You can get this rapid feedback cycle. I think that's really hard. It's not clear to me that some fields can be transformed in that way. I think we can get a lot closer in most cases than we're at right now.”

There's a large difference between helping an 80th percentile student be 'fine' and helping them be 'supercharged'—the marginal impact argument underestimates the potential value of improving excellent performance.

“It's also true that there is a big marginal difference between fine and supercharged.”

Matuschak chooses not to publish minimum viable papers—the strategy of publishing every marginal insight—and instead waits to publish substantial work, which has cost him financially.

“One thing that I've chosen not to do, and it's a choice that's definitely cost me financially, is to publish what academics would call minimum viable units of paper. They have a pithier phrase than that. Minimum viable papers. It's very common to take any new marginal insight that is above a particular bar and publish that. I just haven't done that. I've written informal letters to my patrons, 'Hey, I figured this thing out this month.' If I were an academic, I would have published that as a paper. If I were a marketing-oriented crowdfunded researcher, I would have done some glossy thing and promoted it 'Look at this thing I figured out.' But actually, I just don't think it's that big a deal and I'd rather get on to the next thing.”

There's a fundamental tension in unschooling philosophy: while letting children pursue their interests is valuable, if a child's passion (e.g., ocean geology) is not represented in their immediate environment, they'll miss exposure to opportunities they might not have self-selected but could inspire them.

“There's a fundamental tension in unschooling, for instance. Just let the kids pursue what they're interested in. That's cool. There's a lot of good things about that. But say that a kid's true passion turns out to be ocean geology or something and they're in a landlocked country and there's just no one around them that talks about ocean geology, then they're missing out on some great opportunity.”

Early in his time at Khan Academy, Matuschak learned education market details via memory systems, which enabled him to contribute strategic insights in executive meetings by being able to recall specific statistics about instructional spending, market growth, and student demographics on demand.

“Early on in my time in Khan Academy I learned a whole lot of details about the education market in a very thorough way using memory systems. This let me be in high level executive kinds of conversations where we're trying to figure out strategy stuff and somebody would propose a particular direction and I could say things like, "Well the total budget spending for instructional materials is this and that market is growing by this percent per year and 10% of students in the US are in this place" and so on and so forth. Basically I could evaluate ideas on the fly in a way that others couldn't.”

Matuschak spent extensive time thinking about and managing his energy levels, structuring his day around them—suggesting that energy management, not fixed stamina, is the key variable in sustained work.

“I've actually written extensively about all my struggles with energy and ways of managing energy. I spent a lot of time thinking about it, managing the energy levels and structuring my day around it.”

In the quantum mechanics study session, the social context (studying with Dwarkesh and performing for him) provided substantial energizing effect; studying alone would have required more willpower and been less sustaining.

“So if I were alone and studying that book for three hours, and I wasn't effectively trying to perform for you Dwarkesh, it wouldn't have been nearly as energizing for me. And I definitely would have taken breaks. I still would have been able to go for three hours, I think. Part of the reason for that is that it's simply way less hard than things I normally do.”

Despite many pedagogical innovations and new media (video, games, fiction-based learning), most have failed to achieve mass adoption or dramatically improve learning outcomes, suggesting either an inherent limitation in non-text/lecture media or a failure in content design and approach.

“There are ideas that people have come up with for different pedagogical tools, which are mediums that give closer connection to the reader. One is, you have some sort of fiction account, where a concept is introduced and reinforced, or you have a video game with characters you care about. As far as I know, there isn't something that has really taken off using these sorts of new mediums.”

The Witness by Jonathan Blow is an extraordinary game that teaches complex systems through environmental design and interaction rather than text; players feel in conversation with the designer despite no explicit communication.

“I think The Witness is an absolutely extraordinary work of art. It's a game that has no text, at least no text that's relevant to the game elements... the game proceeds to explain to you, without using words, but just by shaping your environment, a series of extremely complex mechanics of a system that exists in this world... It's like a medium of action... it feels like you are in conversation with the game's designers.”

Educational games (explicitly about arithmetic, math, etc.) have largely failed to match The Witness's success because games are fundamentally in an aesthetic form serving an experience, not an explicit subject—games and educational texts serve different purposes.

“Some people have tried to make educational games, games that are explicitly about arithmetic or something, Jonathan Blow's game is not about that... The issue seems to be that games are ultimately in aesthetic form. The purpose of the game is to have an experience that feels a particular way. And so they're sort of serving a different purpose than Grant's videos or a text.”

In fields like design and musical composition, the standard approach to teaching is studio classes with intensive hands-on feedback, making it difficult to scale to many students; programming has more scalable resources because it's more legible and amenable to textbook/video instruction.

“In particular, design, but lots of other domains that are like that, like musical composition, architecture, something like this. Nope. It's normally done in studio classes. Lots and lots of hands-on feedback. The feedback is highly contingent. It's highly contextual. We just haven't figured out how to communicate this.”

Matuschak advocates for more designer and domain-specialist streamers to make tacit knowledge in fields like design more accessible at scale, paralleling what programmer streamers have enabled.

“I really want to see the streamers in these other domains.”

Design, unlike programming, often feels like exploring a combinatorial search space without clear progress indicators—you try many things, learn which don't work, and eventually stumble on something good, but lack the sense of 'getting closer.'

“What's hard about designing a user interface is that often there's this feeling of exploring a combinatorial search space... Design often feels less like that... Often it just feels like, "Nope, I just need to try 300 things." ... You don't have this feeling, or at least I don't, and I think many designers don't have this feeling of progress. You're just kind of exploring parts of the search space, and you're learning that parts of the search space don't work.”

People who become obsessed with design work typically have skill (pattern recognition, intuition for good solutions) and tackle highly constrained, tractable problems where search space exploration yields quick progress—not unbounded design challenges.

“So to some extent, I think they are skillful. The people that I know who are like this, it's a combination that they're often skillful and the nature of the problems that they're solving are highly tractable.”

Matuschak uses Beeminder (a tool that charges him if he fails commitments) to enforce his spaced repetition practice, despite viewing it as at odds with his belief that learning should be joyful—an example of tension he doesn't fully resolve.

“I use a tool called Beeminder, which charges me if I don't do certain things. This sounds kind of military, but it's certainly more authoritarian than this kind of freewheeling butterflies kind of gesture I was making a moment ago. And I use it to make sure that I do my memory practice. Shouldn't my memory practice be so joyful? It's at the center of my research, right? It should be the most interesting, exciting part of my day, but often it's not. And so I use this to do it anyway. So there's some tension here.”

Khan Academy experimented with dynamic media (interactive representations and visualizations) to enable learners to explore concepts through play, like manipulating Cuisenaire rods to discover patterns.

“We did a lot of experiments with dynamic media, representations of things. The idea being that, you've probably seen these like plastic blocks or things that people can play with when they're kids to get an idea of numbers and number systems. Kids will play with these things unprompted because they're fun.”

Much of the most interesting innovation and creation comes from students performing at the 20th percentile in traditional schooling because they're disaffected and bored by conventional education, so investing in their engagement could unlock creativity.

“Part of the trouble here is that the cool shit is very likely to come from students who are performing at the 20th percentile in school, because they're disaffected and bored and none of this stuff matters to them, right? Part of the trouble here is that by opting out of helping these people learn, there are all kinds of interesting inventions that could probably occur that aren't occurring.”

Memorizing taxonomies and foundational facts using modern memory systems is now trivial compared to pre-system memorization, and Matuschak endorses this traditional educational priority precisely because the cost is now so low.

“I think I do want to say the reason why I'm willing to endorse this headmaster's view about the taxonomy has to do with the price. I did a bunch of memorization in high school, and it was very inefficient, and it was very uncertain. It was emotionally difficult because I wouldn't even feel confident that I'd learned the stuff. I didn't know what it was to learn something reliably, to be confident that I'd be able to recall it. And it was hugely time consuming because I didn't have techniques or tools. And now, part of why I respond so favorably to just learning the taxonomy is that, for me, it's just trivial. Like, yeah, sure, whatever, throw it into the deck. It'll consume a total of 15 minutes over the next few weeks and then I'll know it. It just doesn't cost anything.”

Swanson-linking (named after researcher Don Swanson) is the ability to notice connections between disparate bodies of knowledge by reading widely across domains and recognizing that findings in one field link to findings in another, enabling serendipitous discovery.

“Swanson was just somebody who read the medical literature and he was just familiar with a lot of esoteric results. Different things would come up and he would be able to figure out what different things are connected. For example, he noticed in one case that headaches are linked to some other symptom and that other symptom is linked to magnesium deficiency. Apparently a whole bunch of people's headaches were solved once they were given magnesium supplements.”

Programming already has game-like properties (immediate feedback, sense of progress, ability to create things) that make it intrinsically engaging, unlike more abstract subjects.

“programming has an element where it already feels like a video game. I have a friend who has a sort of intense sort of manic energy, and he used to be addicted to video games when he was a teenager, and now he just stays up all night in these coding binges. It's just the same part of the brain.”

A common mistake is trying to follow a curriculum exhaustively and in the 'right way,' but losing motivation when the material doesn't connect to intrinsic interest. Learners are better off skipping around according to their interests and continuing, rather than abandoning the domain.

“I think a very common mistake that people make is to feel that they need to do the thing the right way and that is exhaustive and completionist. If they fail because they find themselves bored or unmotivated because the material doesn't actually seem to relate to what they want to know, but they're just going on faith that, "Well, if I follow what the author says, everything will be good." Anyway, they find themselves having trouble for that reason, and then they just stop. This is bad. They would be better off just skipping around according to their interest and continuing.”

Matuschak endorses the traditional view that taxonomy memorization is important (following a 1900s headmaster's view), but only because modern memory systems make it trivial in cost—the value proposition changes fundamentally when memorization takes 15 minutes total rather than months.

“I think I do want to say the reason why I'm willing to endorse this headmaster's view about the taxonomy has to do with the price. I did a bunch of memorization in high school, and it was very inefficient, and it was very uncertain... And now, part of why I respond so favorably to just learning the taxonomy is that, for me, it's just trivial. Like, yeah, sure, whatever, throw it into the deck. It'll consume a total of 15 minutes over the next few weeks”

At Apple, Matuschak was much better at engineering work than he is now at research work, because engineering drew on deeply familiar knowledge in a well-understood domain requiring flow and execution, while research requires constantly answering 'what should be done' questions that demand synthesizing unfamiliar literature.

“I was much better at what I was doing then than I am at what I'm doing now. That's pretty funny. It was just totally different. Let's talk about this a little bit. This feels very, very juicy for me. Most of what I was doing was engineering. Some of it was very difficult engineering, but mostly engineering, and mostly on things that were fairly well understood.”

Streaming (like George Hotz streaming code) can transmit tacit knowledge—things not in books that engineers learn through long experience dealing with unusual situations—and people can absorb pattern recognition and heuristics by watching, but streaming lacks feedback and contains lots of chaff.

“One of the things that you have to learn to do as an engineer is to learn to deal with 100,000 different weird situations where something is not behaving the right way. Eventually you learn pattern recognition, you learn ways of dealing with this. Much of this is not described in any book. It's not explicitly taught. You just learn it by doing it over a long period of time. By watching George do it, I think that people do absorb stuff.”

Quality at a certain level (like Gwern's research) probably doesn't need traditional marketing because high-quality work spreads naturally, but Matuschak's experience is that many people don't know he has a Patreon despite being familiar with his work—suggesting some marketing of membership offerings (different from promoting content) is efficient.

“Gwern is an interesting example because there's a simpler failure mode. I still routinely run into people who will tell me, 'Oh, I've really liked your work for a long time. I didn't know you had a Patreon.'... The fact that they aren't converting into patrons is mostly a matter of the way that he talks about it and the way that he presents it.”

Dwarkesh's interviewing skill has improved over time not because he remembers details from past conversations, but because he's developed automatic patterns for generating good questions about material—through repeated practice reading essays and blog posts, these question-generation patterns have become automatic through knowledge compilation.

“I expect the main way in which you're getting better, is actually not really about any of the details of those materials. I think it's about your practices as an interviewer, the way that you generate questions... in the course of repeatedly doing that, you've made parts of it automatic, so that you don't need to do it consciously, you can focus more on the material, you can probably take on more difficult material, or actually understand material at a higher level than you could have before, because less of yourself is engaged in this question of how do I make the questions from the material.”

Content infovores like Tyler Cowen integrate large amounts of information without formal note-taking because their writing obligations (daily blog posts, weekly columns, frequent books) function as distributed note-taking—the act of writing down their thinking, often in first-draft form, is how they process and integrate material.

“One of the other things that's so interesting about Tyler is his writing obligations. This is a man who's blogged every day since 2007 or something and has a weekly Bloomberg column, something like 1500 words, and also has published something like a book a year for a decade or more... That is notes.”

Matuschak used memory systems to learn education market details in depth (budget spending, market growth rates, student demographics), which allowed him to evaluate strategic ideas in real-time during high-level executive conversations—demonstrating how explicit memory practice enables decisive action in complex domains.

“Early on in my time in Khan Academy I learned a whole lot of details about the education market in a very thorough way using memory systems. This let me be in high level executive kinds of conversations where we're trying to figure out strategy stuff and somebody would propose a particular direction and I could say things like, 'Well the total budget spending for instructional materials is this and that market is growing by this percent per year and 10% of students in the US are in this place' and so on and so forth.”

A mass medium combining active learning (like Nand to Tetris) with structured material (like games or courses) that could be applied to many domains hasn't been created yet—we have non-mass apprenticeship, which can't scale like books, and mass education without the action-feedback loop.

“I think some kind of mass medium that is like that is waiting to be created, but that can be applied in many, many circumstances. We have the non-mass medium version of it already and it's apprenticeship. If you want to be a good yoga teacher, you go hang out in yoga studios... This isn't a mass medium. We can't print billion copies of it like we can with a book.”

Most attempts at educational games fail because games are primarily aesthetic forms pursuing particular experience goals, while education requires transmission of specific knowledge—Kerbal Space Program and Zachtronics games might teach tangentially (orbital mechanics, assembly language) but aren't designed for that purpose.

“Some people have tried to make educational games, games that are explicitly about arithmetic or something, Jonathan Blow's game is not about that. It's the mechanics that you learn are they're about the environment. I don't think anybody has yet really succeeded in doing this about explicit subjects... The issue seems to be that games are ultimately in aesthetic form. The purpose of the game is to have an experience that feels a particular way.”

Design work often feels like exploring a combinatorial search space where you need to try hundreds of variations without feeling like you're getting closer—the 'Figma approach' of duplicating artboards and trying things—which is psychologically very different from programming's sense of progress.

“What's hard about designing a user interface is that often there's this feeling of exploring a combinatorial search space. Programming often feels like a search problem too... Design often feels less like that... Often it just feels like, 'Nope, I just need to try 300 things.' The core action of Figma is to duplicate. You have an artboard, you tried something, and that didn't work so you select it, and you press Command D.”

Early in his research (influenced by 'fail fast' startup advice), Matuschak didn't think deeply about problems, instead rapidly prototyping ideas. This led him to miss opportunities for developing strong theoretical understanding before building.

“When I was getting into this kind of work, what that advice led me to do is to not think all that deeply about the ideas I was exploring...There wasn't this sense of building a theory of what the problem is and what it would mean to solve it.”

The most powerful design comes from observing something true about the world (e.g., writing is nonlinear but tools linearize it), then building interfaces that express that observation. Strong ideas precede strong designs.

“The most powerful design work has ideas in it. What makes a nonlinear text editor, the text editors that we all know and love, so powerful is this observation that writing is a nonlinear process, but writing with a pen linearizes it.”

Some necessary learning (like memorizing 200 chemistry names in organic chemistry) involves unavoidable unpleasantness even with optimal tools, but can be made less painful by embedding it in interesting contexts like Fermi problems rather than rote flashcards.

“In the case of something like organic chemistry, where you truly do just need to learn 200 names or something. One answer is that it can be done very cheaply using modern memory systems... But even with modern memory systems, you're probably going to spend a total of 100 or so minutes across some weeks, studying all of these formulae. That still is unpleasant so can that be resolved? And I think the answer is yes, actually.”

After three hours of intensive studying quantum mechanics with Dwarkesh present, Matuschak was able to go directly to a piano lesson without needing recovery time, which he attributes partly to social motivation (performing for someone) reducing demand on willpower, and partly because studying a textbook is less cognitively demanding than his usual research work.

“First off, social. So if I were alone and studying that book for three hours, and I wasn't effectively trying to perform for you Dwarkesh, it wouldn't have been nearly as energizing for me. And I definitely would have taken breaks.”

With Orbit (his mnemonic medium platform), Matuschak learned that highly general platforms might not be the right way to explore an idea; working directly with specific authors on one-off implementations would have been more effective for understanding where the format succeeds and fails.

“I would have been better building more one-offs rather than a self-serve thing that anyone could use. And this comes down to the difference in aim. I'm not trying to build some kind of scalable thing for the world at this moment. I'm trying to build the idea. The prototype is an expression of the idea.”

Video has become an enormously popular educational medium and people will watch lengthy explanations on video (like Grant Sanderson's math videos) that they'd never read in a textbook, but the issue is that viewers don't retain much detailed understanding—engagement has increased without comprehension gains.

“One is that I would argue that one medium has taken off in an absolutely enormous way and that's video. People love video. People will watch Grant Sanderson spend an hour going through some explanation of an esoteric math problem, people who would never crack a Springer graduate textbook in mathematics or something like that. The issue is that they will not walk away from that interaction with much understanding but they're much more engaged.”

For material that's genuinely just rote memorization (like learning 200 chemistry nomenclature terms), the task can be made much cheaper and faster with spaced repetition, reducing necessary time from months of suffering to 100 minutes spread over weeks.

“In the case of something like organic chemistry, where you truly do just need to learn 200 names or something. One answer is that it can be done very cheaply using modern memory systems. Organic chemistry students suffer through this and they don't need to. But even with modern memory systems, you're probably going to spend a total of 100 or so minutes across some weeks, studying all of these formulae. That still is unpleasant”

Having to decide whether old journal entries are important enough to promote to evergreen notes presents a trade-off: over-curating feels like excessive preparation, but under-curating means missing durable insights.

“You could view that as a failure of this note writing system. In some ideal sense, I shouldn't be looking at these journal entries, because if something's important, and it's going to be something I refer to a year later, it should be in some durable evergreen note. I don't know. You don't always want to do that. It feels like prepping... There's an amount of prepping that feels like a reasonable amount to do and there's an amount that feels kind of dutiful and unpleasant.”

The startup advice to 'fail fast' and 'get to users immediately with MVP' can reduce the depth of thinking about problems and ideas; this can lead to less thoughtful product development than building a theory of what the problem is before building solutions.

“When I was getting into this kind of work, what that advice led me to do is to not think all that deeply about the ideas I was exploring. An idea would come up and I'd think, "Oh, I can try that." I would try that, then I'd learn something, and then I'd repeat. There wasn't this sense of building a theory of what the problem is and what it would mean to solve it. Instead, it was just a theory of action.”

Quantum Country embedded review questions every 1500 words to help people remember what they read, but this had an important secondary benefit: it created opportunities for readers to notice they didn't absorb what they read, not that they forgot it, but that they never fully apprehended the material in the first place.

“When you're asked these questions after reading 1500 words it is an opportunity for you to notice that you did not in fact absorb what was in that thing. Not that you don't remember but that there's a word in the question that is apparently important that you simply didn't even notice.”

When self-learning, there is a tension between using a syllabus or expert's curation to guide learning (outsourcing what's necessary to know) and using your own judgment to skip material that doesn't seem relevant to your goal.

“One problem you might have as a self learner is you have some goal, a reason for learning, and then you start thinking, "Well, do I really need this chapter? Do I really need this content?" At this point, you're doing the metacognition that you were trying to use a syllabus to avoid.”

Tyler Cowen doesn't use a note-taking system in the traditional sense; instead, his writing obligations (daily blog posts, weekly columns, regular books) function as implicit note-taking where he externalizes and refines his thinking.

“Tyler's a good example... One of the other things that's so interesting about Tyler is his writing obligations. This is a man who's blogged every day since 2007 or something and has a weekly Bloomberg column, something like 1500 words, and also has published something like a book a year for a decade or more... That is notes.”

Stamina is not fixed but subjective; Matuschak often perceives himself as having low stamina, yet others perceive him as having remarkable stamina, suggesting a mismatch between self-perception and actual performance.

“One of the things that I think is funny about stamina is first off, there's some kind of weird grass is always greener kind of situation where, I often feel struck by other people's stamina and feel like I have very little of it. I struggle with energy.”

LLMs depend critically on humans' ability to externalize and make legible—to articulate ideas and information in language—so that they can be fed into systems; this dependency on externalization limits what LLMs can help with.

“LLMs depend on our ability to externalize things and to make them legible.”

The most expensive tutors might not be the best, and research on tutoring shows that even having a 'warm body' provides large effect on tutoring efficacy, with improvements as tutors gain expertise, though Matuschak suspects some of tutoring's value is communicating a worldview rather than pure instruction.

“It may be the case that some tutors have been priced out of the market, but it's not clear to me that the most expensive tutors actually would be the best... And I also have a kind of healthy skepticism of these studies. I think part of the role of having Aristotle as a tutor is communicating a worldview. It's not something that would show up on a test or something that these studies would be measuring.”

Matuschak uses Beeminder (which charges him financially for not meeting goals) to maintain his spaced repetition practice despite it being theoretically central to his research, suggesting a tension between his philosophy that things should be joyful and the reality that he needs external enforcement.

“Just as an example of something totally contrary to everything I was saying, I use a tool called Beeminder, which charges me if I don't do certain things. This sounds kind of military, but it's certainly more authoritarian than this kind of freewheeling butterflies kind of gesture I was making a moment ago. And I use it to make sure that I do my memory practice. Shouldn't my memory practice be so joyful? It's at the center of my research, right? It should be the most interesting, exciting part of my day, but often it's not.”

Matuschak is broadly opposed to David Goggins-style 'embrace suffering' attitudes, viewing most reported learning misery as stemming from sources that can be addressed (lack of care, not understanding the value, negative self-judgment about struggle).

“Let me say first off that I am, broadly speaking, very opposed to what I understand to be David Gogginsesque attitude towards almost anything.”

Quantum Country seemed like the obvious next project after success (generalize the approach to a platform), but it was the wrong project because the idea wasn't strong enough and building a general platform wasn't the right way to test it—more targeted one-offs would have been better.

“Michael Nielsen and I made Quantum Country and when I was trying to think about what to do next, the most obvious or natural idea was 'What if we just tried that with lots of other things?'. That idea occurred to me and the pandemic had just struck, so I was feeling a little timid, creatively or emotionally. I wanted something that felt safe and I knew I could do that. I can build a platform that generalizes this thing that we did for this textbook. So I did. And I did it relatively quickly. And that wasn't the right thing to do.”

The distinction between building a prototype as an expression of an idea and building an MVP as a deployable product is important—an MVP is infrastructure-heavy while a prototype is idea-heavy, and they serve different purposes.

“It was an MVP in the sense that it has very few features and it's very simple. But it was highly general. It's a deployed thing that has infrastructure and has accounts, it has all this stuff that you do when you're building a real thing. That's very different from, 'Let me work with this one author and see if I can make it work with this one other book that's very different from Quantum Country to form a specific question or specific theory about, it worked for this text, what's the next kind of text that would be good to test with?'.”

If Matuschak had hypothetical children, he would likely explore homeschooling or schoolhouse pod models with hired teachers, unbundling school's social, behavioral, and instructional functions rather than using traditional schools.

“This is very difficult. The school, of course, has many purposes other than instructional, right? It has a social purpose, it has a societal purpose, it has a behavioral purpose, and it also has like a pragmatic purpose of basically babysitting. Those things can be unbundled. I think it's pretty interesting to consider that. If I actually did have a kid, I would probably consider that project pretty thoroughly. I think it's pretty likely that some kind of homeschooling situation would occur. It probably wouldn't be [unclear] the teacher, but it would probably be the people I would hire.”

Byrne Hobart's newsletters have limited context and often assume reader familiarity with topics, relying on old blogosphere norms of hyperlinks for exposition; he uses links even to non-explanatory sources (like linking 'utilitarian bets' to Crime and Punishment on Amazon) because narration would lose the sarcasm and jokes embedded in hypertext.

“One example of this is his write up about SBF and his collapse. It has a bunch of links - if you want to learn more about margin calls, read this. And he goes, and if you want to learn more about the psychology of utilitarian bets, read this, and it's just a link to the Amazon page of Crime and Punishment.”

Most people interested in tools for thought are engineers (95%+) but building interesting tools for thought is fundamentally a design problem, not an engineering problem, and many engineers' design ideas are not very good even when their tools solve problems for themselves.

“I meet tons and tons of people who tell me that they're interested in tools for thought and 95 plus percent of them are engineers. The problem with this is that building an interesting tool for thought is basically entirely a design problem. Their design ideas are usually not very good, or troubled in a variety of ways.”

At Apple, most engineering work applies well-understood principles to well-specified problems. The thermals team sets boundaries others must work within; executives set top-level strategic choices like 'no home button, edge-to-edge screen.' Individual engineers optimize within these constraints rather than discovering new principles.

“It is iterative. We'll run into stuff where there's a thing we really want to do, but we can't pull it off, because it drains too much power. So Hey Siri is an interesting example. To be able to activate a voice command at any time without interacting with the device is great. People prototype that just like having a thing listening in the background and watching for it. But that requires having the main CPU on all the time, processing audio buffers. You simply can't do that – it drains the battery. That attempt led to eventually having this dedicated co-processor that runs at a lower power that's very limited and restricted and it can be on when the main CPU is not on, and it can listen for that sound.”

Spaced repetition appears 'accidental' in many professional knowledge practices: writing background sections, mentoring conversations, medical rounds. The practices achieve spaced repetition effects naturally without explicit intent.

“For a researcher, when you need to write papers regularly and you're writing those background sections, you're repeatedly explaining the history of a particular line of research...That is a kind of spaced repetition...All this is kind of accidental.”

The Witness by Jonathan Blow is an extraordinary work of art demonstrating how a game can teach complex mechanics without text by shaping the environment and creating a conversation-like interaction between player and designer, suggesting that medium and form matter for learning.

“I think The Witness is an absolutely extraordinary work of art. It's a game that has no text, at least no text that's relevant to the game elements. In kind of classic Myst style, you wake up on an island, and figure out what's going on. And the game proceeds to explain to you, without using words, but just by shaping your environment, a series of extremely complex mechanics of a system that exists in this world.”

The 'doom-laden' feeling of programming at night (9 p.m., almost working, stay up to fix one thing, then another issue appears, keep going until 2 a.m.) is powerful because each fix creates visible forward progress, making it hard to stop despite exhaustion—each success triggers continuation rather than quitting.

“There's this feeling that's very common for programmers, and it's laden with doom. The feeling is it's like 9 p.m., and you've been working on a thing all day, and it's almost working. It's almost working. And you know, if you just debug this one thing, then your project will be done... And then you start debugging it, and you get it, and you solve it, and that feels great then immediately you run into one more thing, like, 'Oh, it's almost running all the way through, it's almost going end to end,' and you're like, 'Well, I'll just stay up a little bit longer.'”

Design work on constrained, finite problems (like designing a poster) is more intrinsically engaging because the search space is tractable, you can explore it rapidly, and then polish—it feels good even though it's not 'progress' in the sense of solving a problem.

“An example of a kind of thing that designers will often rabbit hole into is designing a poster. It actually often used to be kind of a cliche that at Facebook, there were all these posters up on the wall of the office... Why did somebody put all this effort into it? Well, it feels really good, because a poster is really constrained, it's finite, it's ephemeral. You can start it and yeah, there's a search space, but you can find a decent part of the search space pretty rapidly. And once you're there, there's this beautiful and very enticing feeling of turning the crank and like making it better and polishing it”

At Khan Academy, Matuschak focused on the 25th percentile learner (not even the median) and explored inquiry learning and dynamic media to transform education into something related to student goals—asking questions they authentically want to answer in ways that feel natural.

“At Khan Academy, we were mostly thinking about not just the median learner, but like maybe the 25th percentile learner. One of the angles that felt most relevant, maybe not from an efficacy perspective, but for me, from like a breaking out of this, getting them to follow goals that aren't their own perspective, was to focus on inquiry learning and to focus on transforming the learning experience into something that actually is related to their goals.”

Byrne Hobart's newsletter heavily relies on old blogosphere norms around hypertext and links for jokes, sarcasm, and references, which makes it difficult to adapt to other formats like narration or podcasting—the medium shapes what's possible.

“I don't think it would work out as well for mine, because I heavily rely on the old blogosphere's norms around hypertext, where you can add jokes and sarcasm... One example of this is his write up about SBF and his collapse. It has a bunch of links - if you want to learn more about margin calls, read this. And he goes, and if you want to learn more about the psychology of utilitarian bets, read this, and it's just a link to the Amazon page of Crime and Punishment.”

Matuschak crowdfunds his research through Patreon after Michael Nielsen encouraged him to start early; it now generates between a grad student and junior faculty member level salary, which works for him but wouldn't work for most researchers who need to support teams or institutions.

“If I'm going to be honest, it's because it has worked. The history of the crowdfunding of this project, like many things in my life, is the result of goading from Michael Nielsen... It turned out he was right. The process of building a subscriber base and crowdfunding a project takes a couple of years, at least in my experience. Starting earlier was better.”

Crowdfunding probably works best after initial research stages have been completed—you need something to show to convince people to support you, but very early-stage research can't explain what it's doing because the direction hasn't been determined yet.

“Unfortunately it's probably very difficult to use crowdfunding in the very early days of a research project. I've already chosen a research agenda or direction, and I can express it. Crowdfunding probably applies after the first few stages of research have been completed.”

Tools for thought are nerd-sniped in Silicon Valley because: (1) they promise expanded thinking capacity (appealing to intellectuals), (2) they're actionable for engineers (even though design, not engineering, is primary), (3) they're meta (appealing to people who tinker endlessly with their tools).

“It contains this message for technologists that they can potentially be very powerful...It also feels very actionable...95 plus percent of them are engineers. The problem with this is that building an interesting tool for thought is basically entirely a design problem.”

Matuschak's website note system (the graph visualization) was built in a weekend and has become influential in the design community, though he finds it has limitations—he hasn't spent more time improving it because he doesn't think it's important enough to prioritize.

“That was a weekend... I feel bad about it because it's made its way into tons of commercial projects now... I spent like a couple days on it... Nope... I have design intuitions that led me in a particular direction. But there's lots of things I don't like about it. I just haven't allowed myself to spend any more time on it because I just don't think it's important enough.”

When programming hits a brick wall (stuck on a problem with no visible progress), the engagement disappears and the experience becomes unpleasant, showing that the 'game-like' feeling depends on visible forward progress feedback.

“For the programming problems where you are at a brick wall, it doesn't feel like this. It feels bad.”

Poster design is emotionally satisfying and often done as a 'snack' or 'treat' by designers because it provides the sense of progress and control that larger, unbounded design work lacks.

“It's a snack. It's a treat. It's also something they get to control whereas ordinarily, they don't.”

When tools for learning are well-designed, the question remains: does the learning process need to feel fun, or is some level of sustained difficulty and discipline necessary?

“Once all the tools are built, when you're at the end of your career, is the learning process supposed to feel fun? Or does it have to feel fun? Is there an element of even when all the tools are there, that there's just like a level of David Goggins, this is going to be miserable, but I've decided to learn this in this way and I just had to go through it.”

One theory is that mass education may have actually worsened conditions for producing top-tier talent relative to aristocratic tutoring: top tutors may be priced out of the market because they can earn more in Silicon Valley or Wall Street than in tutoring.

“Although the reason that might not be the case is the opportunity costs for people who might become teachers or aristocratic tutors is much higher now, whereas the kind of person who would be, your tutor can now directly be making lots of money on Silicon Valley or Wall Street.”

Matuschak's personal journal entries are written in complete sentences and paragraphs, often with links, because 'past me is another person' and he frequently refers back to entries from a year ago.

“Usually my journals are complete sentences, complete paragraphs. Sometimes bullets, sometimes veering and breaking and changing to new subjects suddenly. They tend to be filled with links to the things that I'm talking about, in part because I'm trying to accumulate context in those things... past me is another person, it's kind of a cliche. I am routinely looking at journal entries from a year ago.”

A 'theory of action' focuses on what you can do and build from your current position, while a 'theory of change' works backward from a desired goal state—the latter requires more uncertainty and is harder but more directional than iterative experimentation.

“A theory of action as opposed to a theory of change is, imagine you're in your current position, and eventually want to get to some goal state, a theory of action is you look around you and you say, "Well, what can I do? What can I build? What do I see as possible?". A theory of change is to look at the endpoint to try to work backwards.”

Paul Graham's claim that reading and experience train your model of the world, and this effect persists even if you forget the specific content (like a compiled program where you've lost the source code), is true to some extent, but the sufficiency of this depends on your specific needs.

“What he's saying is true, to some extent, whether or not that extent is sufficient is going to depend a great deal on the situation and on what you need. If your aspiration actually depends on having a deep, detailed understanding of the material, then the imprint on your worldview or on your automatic responses made by the book may not be sufficient.”

Andy Matuschak's memory systems work has improved his capacity as a podcast interviewer by automating the pattern-recognition and question-generation aspects of his practice, allowing him to focus more cognitive capacity on the material and less on technique.

“in the course of repeatedly doing that, you've made parts of it automatic, so that you don't need to do it consciously, you can focus more on the material, you can probably take on more difficult material, or actually understand material at a higher level than you could have before, because less of yourself is engaged in this question of how do I make the questions from the material?”

Matuschak knows many postdocs who work as tutors for $60/hour, a low rate showing both the supply of expert tutors and a sad undervaluation of their expertise.

“I'll also say that I know lots of people who have postdoc tutors right now. These people, as graduate students, they're very pleased often to have a $60 an hour tutoring kind of commission. And that's a little sad.”

Being a good teacher requires empathy, communication, and care—personal and intimate skills—more than high subject mastery, so the correlation between high intelligence and teaching ability is weaker than assumed.

“Being a good teacher is often about empathy and effective communication and care. It's very personal. It's very intimate. You need to understand the subject but to teach a 15 year old or something, you actually don't need to understand it at a postgraduate level necessarily.”

Byrne Hobart's daily newsletter often re-explains the same concepts (like interest rate risk) repeatedly but differently each day, inflected by current events—an argument for ephemerality and recreating explanations rather than accumulating permanent notes.

“One thing I find interesting about Levine's newsletter is that when he's talking about a topic repeatedly, like the recent bank collapse or something, he will have to explain some concept like interest rate risk over and over again for days. Every day he has to explain it, but every day he explains it anew, and every day the explanation is colored a little bit by that day.”

Some topics don't require durable notes because you can recreate explanations effectively each time, making permanent notes unnecessary—a critique of over-systematization of note-taking.

“He's leaning more on his past explanations, which is interesting, because he can't update them. That format of writing a newsletter and then linking it to past newsletters, or as you say, the former blogosphere thing to do, you have a series of six words and each word is linked to a previous post on the topic... Maybe for many topics, you don't need that ability.”

When studying a subject in which you have deep expertise, you need less external structure (syllabi, scaffolding) because the cognitive demand is lower due to familiarity, and you can deploy more effective personal planning because you can model the path between your starting point and destination.

“If I'm studying something in computer science, which is a domain that I know really well, I don't need those syllabi, not in the same way for most subjects, and I think that's mostly just because the amount of cognitive demand that's placed on me by the subject is just much lower than it is for other subjects. So much of it is familiar already that I can deploy my own planning more effectively as I go. But it's also the case that because I know so many things about the subject, I can do a better job from the get go of making a plan.”

Reading an introductory textbook in a subject is valuable not primarily for its instrumental utility to a specific problem, but because it provides context and conceptual scaffolding that enables more effective learning in the domain later.

“if you do want to learn about a subject, it might just be helpful to just do an Intro to X subject course or textbook, not necessarily because it is instrumentally valuable to whatever problem you're interested in but because it'll give you the context by which to proceed on, the actual learning.”

Memorization has multiple meanings and purposes: rote memorization of trivia (like nucleotide volumes), and deeper memory work that serves as a dictionary or index enabling complex reasoning and creative insight.

“Maybe this is a good time to talk about what memorization is or what it's for. We could use that word to refer to the practice of learning more trivia... But I just want to make the case that so much of what you do and experience day to day is memory bound, or is memory influenced in important ways.”

Creative insight often comes from noticing surprising connections or contradictions between ideas; this is only possible if both ideas are available in memory—if you must look up one idea, the connection won't be salient and you'll miss the creative opportunity.

“in creative work, there's a bunch of studies trying to catalog case studies of how it is that people have flashes of insight. It's a little hard to talk about that but one of the things that's a pretty consistent source of insight for people is noticing a surprising connection or a surprising contradiction... Now that's only possible if the other thing is in your memory. If you have to think to look up the other thing, the newspaper wouldn't seem so salient to you.”

At Apple, Matuschak was much better at his engineering job then than he is at research now, because engineering worked with well-understood problems and familiar domain knowledge, requiring less metacognitive load and discovery work.

“I was much better at what I was doing then than I am at what I'm doing now. That's pretty funny. It was just totally different.”

Paul Graham's claim that reading trains your model of the world and those effects persist even if you forget the specific content is true to some extent, but whether that extent is sufficient depends heavily on what you need—if your aspiration requires deep detailed understanding, then just imprinting on your worldview may not be enough.

“What he's saying is true, to some extent, whether or not that extent is sufficient is going to depend a great deal on the situation and on what you need. If your aspiration actually depends on having a deep, detailed understanding of the material, then the imprint on your worldview or on your automatic responses made by the book may not be sufficient.”

The opportunity cost for smart people who would have been tutors in the late 19th century is now higher because they can make money in Silicon Valley or Wall Street, which might explain why finding good tutors is harder now than in the past.

“Although the reason that might not be the case is the opportunity costs for people who might become teachers or aristocratic tutors is much higher now, whereas the kind of person who would be, your tutor can now directly be making lots of money on Silicon Valley or Wall Street.”

It's difficult to develop strong ideas through rapid MVP iteration. While testing ideas is good, it requires first developing ideas worth testing, which requires stillness, confusion, and contemplation.

“it's difficult to have ideas by means of building an MVP very rapidly...Part of why I'm confused in my response here is that it's good advice once you have something worth testing. For me, adopting that mindset...it makes me not sit in stillness and in confusion and in contemplation with the ideas long enough for them to be good.”

After studying quantum mechanics intensely for three hours, Matuschak had stamina to go directly to a piano lesson without break, which suggests that what drained him wasn't cognitive load per se but metacognitive and volitional load—the uncertainty and decision-making.

“I was actually surprised that you went the entire three hours without interruption. Afterwards, I was packing up and you're like, "Hey, I'm about to actually go to my piano lesson." I was so confused at how you had the stamina to keep going.”

Matuschak uses hypertext in his personal notes (not for reading) as a navigational aid to build up ideas over time. His notes on the mnemonic medium accumulate from interviews and papers, so when designing the next system, he has a place to synthesize relevant knowledge.

“I'll say that I do find hypertext very useful in my own notes, not really for reading. I actually don't think it makes for a very good reading experience for others. Today, I was reading this very old cognitive psychology paper on the topic of adjunct questions, which we discussed earlier, the effects of asking questions while you read, not on remembering the information covered in the questions, but on the general effect it has on stuff that isn't touched by the questions. I have some notes on the design decisions of the mnemonic medium, this Quantum Country thing that I was talking about earlier, interleaving the questions into the text. Those notes are kind of partial. They evolve over time. What was the impact of doing this? My notes about that, they've come from interviews with readers. They expand when I read a paper that's relevant to them. It means that when I go to design the next system, and I'm thinking about the role of questions in text, I'll have a place to look.”

At Apple, product design and strategy involves compartmentalized domains with clear boundaries—each leader has a small 5% zone of direct control, a surrounding ring of things they monitor and advise on, and delegated domains they only hear about if there are problems.

“There's a really interesting Harvard Business Review piece from a few years back about Apple's management structure and how they have different concentric rings of responsibility for any given leader... say there will be 5% of things that you're responsible for – that you have your hands on at all times and you are directly manipulating, controlling. There's a ring outside of that, that's a little bit bigger. Those are the things that you're keeping an eye on.”

The best sight-readers Matuschak knew played church regularly—they had naturalistic repeated exposure where they had to play new hymns every Sunday, providing immersive practice that built expertise, but they had a foundation of explicit practice before they could take that church pianist job.

“All of the best pianists at sight reading that I knew in university played with churches. They were so good at sight reading because they had to show up every Sunday and they're playing a different thing. New hymn every Sunday. So this is immersion also. Over time, they're learning all these cadences and these things that are really common and whatever. But you can't show up and be the church pianist every Sunday in the first place if you don't already have some decent foundation.”

Nand to Tetris is an interesting project that combines the active learning and doing of games with explicit learning goals—you build a complete computer from basic logic gates, which is naturally active learning where you're constantly seeing the system grow, requiring less external regulation than reading.

“We have the non-mass medium version of it already and it's apprenticeship... I don't know if you've seen From Nand to Tetris. This is a very interesting project that's kind of along these lines. And what characterizes it, like games, is doing. It's active. So when I was asking all those questions of the book, that was active learning, active reading, Nand to Tetris is naturally active.”

The median student in education is not the focus of Matuschak's work; he focuses on more motivated or advanced learners; the tools he builds are 'especially geared towards very intelligent and very motivated students.'

“it seems like the tools and the ideas you're developing seem especially geared towards very intelligent and very motivated students. If they would be different, what would the tools that you would develop for a median student in the education system look like?”

A significant portion of Matuschak's notes (at least a third, possibly more than half) remain in the state they were after first major revision, not undergoing subsequent updates.

“I wonder now what fraction of my notes are in the state they were when I did my first major revision of them. It's probably at least a third, it might be more than half.”

Crowdfunded research only works if the research is understandable and interesting to others, produces visible progress, and applies to domains (like tech/learning) that have audiences with disposable income—it wouldn't work for marine geology or research on disadvantaged communities.

“I realize that this crowdfunding project only even slightly works, because it's understandable and interesting to others. It's already in a place where there's some results that look promising. It's very easy to imagine other projects that are not broadly applicable. If I were doing marine geology stuff, I probably wouldn't have a big crowd of internet people, not nearly as large or excited.”

Matuschak's Patreon works because of early momentum from Michael Nielsen's suggestion to start it early (years before having concrete results), giving time to build subscriber base; high-net-worth individuals rarely offer to fund him despite his prominence.

“It's because it has worked. The history of the crowdfunding of this project, like many things in my life, is the result of goading from Michael Nielsen. Early on when we were working on this Quantum Country project, he suggested we set this up and I hemmed and hawed and said, 'Yeah, it's going to be a distraction. We don't really need this right now. Let's deal with it later when we have something to show.' He's like, 'No, no, no, let's just get it started. It's going to take a long time to get enough subscribers.' It turned out he was right.”

Gwern (a prominent researcher) makes a tiny fraction of what Matuschak does on Patreon despite being more impactful and having a larger audience, suggesting the difference is in how they talk about the membership offering rather than in audience size or impact.

“Gwern is an interesting example because there's a simpler failure mode. I still routinely run into people who will tell me, 'Oh, I've really liked your work for a while. I didn't know you had a Patreon.' That's a simple failure of a certain kind of marketing on my part. Gwern actually has this even worse. I adore Gwern. I have learned so much from him. You can go to his Patreon page and he actually makes public his revenue. He makes a tiny fraction of what I do on Patreon. I think this is inappropriate. Gwern is a much more impactful researcher than I am and he has a much bigger audience than I do.”

Not every field can be transformed to have rapid feedback cycles and a sense of forward progress—it's very hard and unclear whether some domains can be redesigned that way.

“Can every field be transformed into something where you can feel the sense of forward progress? You can get this rapid feedback cycle. I think that's really hard. It's not clear to me that some fields can be transformed in that way. I think we can get a lot closer in most cases than we're at right now.”

Spaced repetition adoption is probably mostly efficient—it's widely used among medical students, language learners, and other highly motivated groups with incentives to learn, which are the domains where it's most valuable.

“It's probably mostly efficient. In places where spaced repetition, as it stands without substantial novel cultural knowledge that's difficult to transmit and isolate, is valuable, we see a lot of space repetition usage. Among medical students who are highly motivated and have lots of reason to study, the material is shaped in a way that's highly amenable to spaced repetition usage, there's tons of spaced repetition usage.”

Learning quantum physics with spaced repetition is faster than without, but doesn't solve the fundamental problem—it still requires additional scaffolding and isn't like learning anatomy where studying a deck essentially solves the learning problem.

“As to why spaced repetition isn't widely used, for instance, to learn quantum physics, it's basically correctly priced. I can use spaced repetition to learn quantum physics a bit faster. It doesn't make it a fait accompli or anything like that. It's not like learning anatomy, where basically if you study the deck, you'll be done.”

Matuschak finds himself doing a mix of ephemeral journaling (daily writing about today) and durable evergreen notes (intended to stand outside of time as useful references), and hasn't developed a clear model of when each is better—both feel useful.

“I find myself doing a mix these days. I have a journal that's about today, and I'll do a bunch of writing. Often I'm recapitulating stuff I've written before, and I have other things that are trying to be more durable, and be a useful reference that can stand outside of time. The combination feels useful. I don't yet have a clearer model of when one is better than the other.”