Software Engineer
Raghu Veer S
Karma: 36
Borrowing from the post, how about we categorise between a lone genius: someone who comes up with an idea using a combination of different means available to them such as collaboration, intense study, iterated experiments etc; and a solitary genius: one who’s completely withdrawn from the world and has chosen to rely solely on their cognitive prowess to come up with new ideas, independent of any interactions with the world or the processes and entities in it that are usually known to assist in the generation of new knowledge.
(Differentiating based on the other meanings of the word, that is, lone can mean solitary, but it can also mean single, whereas I feel solitary necessarily denotes isolated)
Maybe this distinction reduces the need for mythifying knowledge creation, as the latter is next to impossible and former is pretty much how knowledge creation works, no?
I have a feeling that the reason we are still having a conversation about ideas like the lone genius and the great founder is because different people are talking about different things when they talk about the lone genius. Some implicit assume the interactivistic elements, whether it be with fellow humans or processes; while others see it as an attempt at undermining the credibility of an interconnected structure(because they think the idea of lone genius signifies an individual in seclusion without any interactive components). We first need to get the denominator right so that we know we are not talking past each other.
Not necessarily. Globalization has had many negative second-order effects. For example: As much as air connectivity has helped us travel across the world, it has also increased the risk of infections travelling longer distances quicker than if we had a localism-based model. If we are epistemically humble enough, it is not difficult to see how many COVID-like events might have happened in the past in various isolated parts of the world, that we do not know of, but never ravaged the entire world.
Globalization has benefits, not saying that it is not useful, but describing progress as a function of globalization is what I take issue with. Progress is a multiscale phenomenon. You need a strong localism-based core for innovation and you also need decentralization to accentuate the process, and then you can use globalization to scale. And then there is also the part where you need a lot of wisdom to know what should be scaled and what shouldn’t be.
Neat article. Few things I myself have been struggling with and would like some insight into are:
Were you able to address the issue of ankifying the algorithms? If yes, how?
How do you handle the Anki burnout(or for that matter Anki inertia)? I am guessing with multiple subjects at a time, you at least have to review 100-200 odd cards per day(maybe more), and based on the non-atomic nature of your example cards, it seems to me that getting yourself to review that many cards every day is not a sustainable activity?
How do you preserve your sanity between learning new things and practicing new cards? It seems to me that it can be taxing to spend time revising on your own while also learning new things.
How do you handle the inertia that comes with creating new cards or editing cards for so many concepts and subjects?
Does the SM-2 algorithm suffice your needs in terms of frequency, priority, scheduling, etc?
Finally, have you found ways to avoid things like wrist issues, back issues, cognitive exhaustion, etc? Like maybe good computer etiquettes or something?
Sorry for the long list of questions. I am curious to know more because I have been struggling with getting my way into auto-didacting for more than 2 years now due to issues like being unable to differentiate between wanting-to-want and really want, burnout, haphazardness of it all etc. Any pointers on it would be really appreciated.
One reason I can think of is the intersection of the English speaking populations of the world. India, Pakistan, Indonesia, Philipines, and Bangladesh constitute over 25% of the world population. And these are also some of the largest English-speaking nations. It seems to me that the only way to cater to their needs given that only a fraction of the population can speak fluently would be through reducing the number of redirections. This is to say that it would be surprising if the incentives of one of the largest dictionary manufacturers in the world were not affected by a customer base that large.
If I am understanding this correctly, I feel the map is multilayered.
First, you have the territory and the map within language itself, a few of the aspects of “linguistic freedom” in this context I feel can be addressed right away using existing linguistic tools like Polysemy, Metonymy, Metaphor, etc. They clearly tell you that the freedom that you are enjoying is due to the categorical representations(the map), which holds true even for non-existent and newly formed words with tools like aureation, retronym, portmanteau, etc. The only place I feel where the existing tools fail is in addressing the emergent aspects of these phenomena like the potentialities of a word, possible use cases, etc, but if you leave the emergent aspects out, it seems to me that it is less about freedom or the awareness of freedom, and more about the mapping between the two, that is, people realize they are free-running, they also realize that they have this freedom as in they are walking on a map, but what they never realize is the weight on the edge between the two nodes. This is to say that their memory of why they chose to exercise this freedom in the place always eludes them. I think it is the cataloguing of that weight that we are missing and not the awareness itself.
Second, you have the entire domain of linguistics inside the map, and then there is the “structure of reality” as you mention elsewhere, which could be thought of as territory. And I feel this fails to present itself to anything beyond a heuristic. You can guide them linguistically but the realization is largely contingent on their umwelt. So at least, in my opinion, the issue is still with the weight on the edge and less on the freedom or the awareness of it.
I am sorry if I am missing something here or misinterpreting your point. I am just trying to understand the core theme.
Mental model that works for me is that of Self-Organization (from cybernetics and complexity theory). Use interactions to enforce redundancy of potential command(See McCulloch). This way you still retain the agency as the command is initiated by you, but the emergence of order(i.e., the execution of that task in our case) can be orchestrated by changing the nature of interaction. One example would be to decentralise the decision making process when you sense the beginning of dissolution of willpower: commitment contracts(eg. beeminder), substitute activity(eg. using stairs to compensate for lack of exercise), automating tasks that can be automated, fractal attention(eg. picking up a different book when you are bored with the book, instead of stopping reading altogether) etc.
I can understand the sentiment of being subtle, articulate, and precise as long as it remains within the bounds of “nitpickers gonna nitpick” or “dislikers gonna dislike” stage, but if you are treading in the realm where you have (or can possibly cultivate) haters, I don’t think focusing on presentation works. Also, IMO, if the emotions have already gotten to the stage of being hateful, you might as well stand your ground and say what you have to say instead of selling nuance and subtlety as a point. The reason I say this is because it is a great heuristic to follow when you are already aware of the theme driving your writings; and more often than not the untrue cases(5%) that you mention are not going to change, while the ones who are already well accustomed to the topics and your style are probably only going to be more interested if you stand your ground. The classic example of this would be someone like NNTaleb, whose cult has only grown since he started being more unabashed about his ideas. Not to mention the number of death threats he gets on mainstream twitter alone and who knows how many via other sources. I think the best solution is to follow the heuristic(I mean haters gonna hate heuristic) and be courageous in your writing style, helps offset a lot of other issues such as stress, anxiety, and what not.
I feel that framing matters here. For eg. Look at how the words tradition and celebration complement one another in some of these situations. In the case of Olympic fireworks or Times Square gathering, how much effect does the instinct to preserve a long-standing tradition come into play? but when you look at scientific accomplishments it is too disparate and of varying significance to even be equated with the likes of Olympics or a New Year’s Eve. I have a strong feeling that to be celebrated, an event must either form a part of an existing (celebratory) tradition or create a new one, which is also why I feel Nobel(Old) and Twitter trends(New) are an equally interesting celebratory methods unless one wishes to restrict the definition of celebrations to rallies and fireworks. In this case the former being traditional/ceremonial, and the latter unconventional/momentous like the railroad project or light bulb of modern times, which is seen through the occasional burst of congratulatory enthusiasm in the form of a hash tag or likes.
Have you used Syntorial, the synth-learning/tutoring software? I think it makes great use of adaptive interactivity(learning), which I feel tools like brilliant or explorable explanations, although great in terms of UX, lack severely. In fact, I have also found Syntorial to be very effective in terms of memory-related things like remembering patches etc. I think it has that neat quality of helping with both learning/doing and remembering what you learn. Maybe you could look into that too for some inspiration.
Neat. I too have been logging for quite sometime now, the biggest cost initially for me was with Storage, but I’ve found some neat ways to handle the storage issues: 1) If you are on your desktop, create a private stream on YouTube using OBS—obviously helps you save diskspace while also helping you maintain privacy and providing unlimited storage as these are basically just youtube videos that are kept private. 2) If you own a tripod, you can record your sleep patterns and do the same by mounting your smartphone on it and hooking it up to your PC and streaming it to YouTube privately. 3) Obviously Go Pro is the best all-purpose recording device as Mati says, but I am not much of an underwater person, so I use my smartphone for everything, which I feel is the second best as you can use the YouTube’s go live feature to record yourself with private mode on.
Apart from that I would say, always carry a small pocket notebook and a pen incase your phone’s battery dies on you or something like that. Also write everyday, blog! (Not Twitter), because reconstructing the thought process from an elaborate medium is much much easier than Twitter. If it is you who is going to revisit it later, you don’t just want to sympathize, you want to empathize; and pithy sentences(280 chars) are bad for that. I’ve tried. I realized it after reading Montaigne’s Essays and a nice little book called the Diary of a Nobody by guy named George Grossmith. There is so much that can be extracted from that as opposed to Meditations where you can feel for Marcus but forget it so easily that you never get to ponder about what would Marcus be doing right now, which can be done with Montaigne or the other guy.
Scale and Complexity? — Thinking in terms of variables and states works fine to some extent, but I think that it is always in the interactions between those variables that the states get screwed up(to a point of complete unrecognizability). And when you scale the number of variables, the interactions become too non-linear for effective management/monitoring of states. Cliched, but I feel disintegration of large is a more plausible event than coalescence of small, for with scale comes uncertainty and complexity.
Learnt this from my twitter/email exchanges with some productive people I know and it has been serving me quite well so far:
-
Don’t read the referenced articles before you’re done with the original/current one. (Maintain a document of references if needed)
-
If you already have more than 5 tabs opened on a same topic, switch to a new window instead of a new tab with just the current article. Helps with the distraction.
-
Read in multiple-passes instead of a focused single-pass.
3.1. Pass 1: Skim through the structure of the post and the comment section for some context.
3.2. Pass 2: If you find it worthwhile, read through the post carefully and collate all the references and links in a document once you are through the post.
3.3. Pass 3: Skim through the references and embedded links to see if there is anything of value beyond what the original article had to provide. Look for the themes that inspired the original post.
3.4. Pass 4: Read through the ones that you found to be valuable i.e., the ones that had more(different) to offer than the original post.
3.5. Pass 5: Repeat 1 through 4.
(Note: If you are reading a paper or a book, during the 1st pass read through the Table of Contents, Bold faces, Headings, Abstract as part of skimming; Note all the jargons and key concepts in the second pass; and map these concepts and jargons to the known concepts during the 3rd pass; Now go through bibliography like in pass 4 and 5).
-
Be well rested before any cognitively demanding task, it doesn’t matter if it is a blogpost or a paper. And I seriously can’t emphasise the endurance benefits of being well rested when it comes to cognitive challenges even as trivial as cursory reading.
-
When reading try to read from the perspective of a specialist as opposed to a generalist, it helps with pruning the articles and tabs that you think are important but don’t serve beyond the session. You can still be a generalist, but the idea is to not be all over the place. This helps with the cognitive exhaustion and the feeling of being overwhelmed. A related idea would be to ask if I would still read it if I were to never talk about it ever to anyone. (Note: Also helps discern between curiosity and the laziness involved with evading responsibility in the name of curiosity).
-
You can use tools like Zotero, Evernote, etc for quickly jotting down interesting points in case of a single pass read, but I would say instead of doing that go for multipass and when you are done may be try to write a small commentary post debating the key ideas to solidify your understanding and post it here. Or conversely engage right away by commenting in the comment section.
-
Always timebox. Mine is usually 6pm-7pm everyday. If you are an infovore, you can make it everyday 7-9 or something like that. I do that because I have a pretty low capacity for consumption, but I can adhere to my own schedule. If you cannot try to time it with your office commute time etc.
-
I feel “self-fulfilling prophecy” does the same thing, but I like your term better — catchier, pithier, and more functional. :)
Exactly this. Studying behaviours and averaging it has reduced us into easily categorisable beings. The complexity just goes out of the window when the question itself has a design constraint that the answer is expected to meet. My idea is that even if there is an irreducible unit to which you can be reduced to—which I don’t think there is—the temporally emergent aspect of interactions with a larger whole such as the society that are combinatorially so large as to be intractable just do not allow for a siloed theory/inquiry to explain it all.
Culture-specific habits like eating with hands in India or with chopsticks in other South Asian countries. Could be considered a social norm but I don’t think they are, as these are mostly a matter of preference that can be changed without violating any social agreement.
It definitely has taught me some epistemic humility, and especially after reading contents by people like Eliezer, Gwern, and Scott I realized the amount of introspection that I had to do to be able to come to terms with the knowledge deficit I had. I always had an emotional alignment with their content, but the fact that these guys could think the way they do, and all using the same set of tools that I have has made me less envious and more curious in general.
My view: Although I think it is a neat thought experiment, my intuition is it is a false dichotomy to separate between compute and algorithm, and I think so because: narrowing the path dependence of a domain that consists of multiple requirements for it to evolve optimally to an “either/or” situation usually leads to deadlocks that can be paradoxical(not all deadlocks have to remain paradoxical, pre-emption/non-blocking synchronization is a way out) like the one above.
My answer: Not much difference, because twenty-year timescale doesn’t seem very significant to me; and also because neither has there been any fundamental revolution in the semiconductor/compute-manufacturing industry that has benefitted us in ways other than cost, and nor has there been any revolutionary algorithms found that couldn’t be run with old hardware scaled to today’s standards. (But in complex systems (which ML is) interactions matters more than anything else, so I might be way off here)
I have/had all the problems that are mentioned in the post, but as of late I am observing that as I read more books on a single topic, it allows me to maintain my natural rythm without having to strain myself to be hyper-attentive. And the added benefit I see is even if I miss some quirky details in the first book due to lack of attention, it somehow starts to come together on its own by the nth book(for me it has been 2nd or 3rd). On the other hand, I don’t think my attention span has improved drastically, but I would say that it has definitely improved by some margin due to meditation and reading more, at least to a point of being able to realize that it has.
As for note-taking, I would also love to know how do people take notes, to me it feels like a flow-breaking activity to a point that I’ve come to detest doing it. Also as you say, sometimes the divided attention b/w I have to mark/note important things vs I have to maintain my focus makes it a tiresome activity to read, and sometimes it just feels like almost everything is equally important. So if someone could answer that I’ll be grateful. To be precise my interest is in converting offline margin notes to online notes, and knowing how people decide what is important in the first read.
And among the books you’ve listed, I have read the “The Art of Reading”, but I felt it was rather underwhelming, meaning, it never says anything about how to read, in fact, the entire book is weaved around the idea of kindling your interest in reading by explaining snippets of various prose written by various accomplished authors on how good reading requires good writing. Not good if you are already interested and don’t want to blame the author for not writing well or making it too dense; and are only looking to up the attention/retention game, note-taking etc.
When you say that the reversal tester loses the argument, do you mean that one could easily refute the rhetoric of the question posed, as in, homeostasis being the optimal condition, or one can counter that with a flaw in the rhetoric, as in, there is an implicit assumption there in the question of some sort that defeats the main intention of the question itself. If it is the second one, I am genuinely curious as to how that can be countered.
Remarkable achievements indeed, but I would also note that there is always more to it than meets the eye. He lives a solitary life now, but he was an active mathematician up until 2004-05 at Steklov. We don’t know how his stay at Courant, Stony Brook, or UCB shaped his ideas. Additionally, his work builds upon Hamilton’s Ricci flow, which he had been working on, more or less, since 1992, which happens to coincide with his time at Courant and Stony brook. We don’t know the people with whom he was in regular correspondence. Also, the analogous results were proved by Smale and Freedman for dimensions higher than 3, now what did he borrow from their work, we don’t know. Also we know nothing about him other than that he is a recluse, who has withdrawn himself from the world, or that he espouses a ethical and moral philosophy that stands contra to what the global institutional mathematics espouses in terms of fame and recognition.
Please note that my point is not to refute the notion of a genius, but to show that the language we use undermines the shared understanding of the process of knowledge creation. I never said that you can’t work in complete seclusion, but that it misses the “middle”(think excluded middle) that we don’t usually convey. This can often lead to people talking past each other. Even in the case of Perelman, let’s strip him of his math education, his exposure to modern mathematical results, his correspondence with people involved in modern mathematics, could he have built everything by himself in order to prove Poincare conjecture? Now you might intuitively understand that this is not what we usually mean by a lone/solitary genius, but think about the numerous times you’ve spoken to people about a lone genius and people have found it difficult to understand. It’s because they feel you are undermining the fundamental nature of knowledge creation, that is, it is an iterative process that involves knowledge acquisition, conjecturing, formulation, correspondence, criticism, error correction, etc, which needs other people and processes. And when we speak in simplistic terms like lone genius and great founder, it is inevitable that we will end up getting entangled in the meta war and lose the sight of the object of communication.
My understanding is that the object of communication in our case is what leads to some people contributing significantly while others only do incrementally. Better questions would be how do they build on other’s work? What were the incremental components of their work that went unnoticed? How did the correspondence they had with other people in the field affect their trajectory? etc. Now see that all of these allows for a solitary life and work but not a ground up creation of new knowledge without any interaction whatsoever with other people or the processes in it.