What are you learning?

This is a thread to connect rationalists who are learning the same thing, so they can cooperate.

The "learning" doesn't necessarily mean "I am reading a textbook / learning an online course right now". It can be something you are interested in long-term, and still want to learn more.

 

Rules:

Top-level comments contain only the topic to learn. (Plus one comment for "meta" debate.) Only one topic per comment, for easier search. Try to find a reasonable level of specificity: too narrow topic means less people; too wide topic means more people who actually are interested in something different than you are.

Use the second-level comments if you are learning that topic. (Or if you are going to learn it now, not merely in the far future.) Technically, "me too" is okay in this thread, but providing more info is probably more useful. For example: What are you focusing on? What learning materials you use? What is your goal?

Third- and deeper-level comments, that's debate as usual.

Comments

sorted by
magical algorithm
Highlighting new comments since Today at 2:49 AM
Select new highlight date
Rendering 50/128 comments  show more

I'm working through a group theory / analysis syllabus to remedy my gormlessness with proofs. If anyone else is gormless with proofs, we could form a Remedial Proofs Club, where we all fruitlessly push variables round a page for three quarters of an hour before giving up. It'd be like a secret handshake.

I've also been sitting on the Daniel Solow and Polya books with minimal motivation to work through them. Remedial Proofs Book Club, maybe?

I was about to begin trying to work through Elias Zakon's Basic Concepts of Mathematics, available free online at here. My motivations are similar to yours, and the idea of a Remedial Proofs Book Club is attractive to me. I chose this particular text only because it was free online and seemed to be written specifically for people at my level (got a 5 on the Calculus AB AP exam after high school, thought I was hot shit, took a rigorous Calculus course my first of year of college, couldn't do a single problem, cried real tears, traumatized). If you had a strong preference for different source material, I would switch to it if we could work through it together.

I want a deep understanding of elliptic curve cryptography. This led me to study algebraic geometry, which led me to study category theory. I think I'm ready to go back to algebraic geometry now.

I am not sure how comfortable you are with mathematics (and algebra in particular), but my courses in algebraic geometry were among the hardest I have taken so far (and the study of elliptic curves is a specialisation of the study of algebraic geometry). After studying algebra for 4 years I can now finally understand most of the introductory chapter of my book on elliptic curves, though not all of it. Since you want to learn about a specific algorithm rather than all of elliptic curves I think it shouldn't take you the 4/5 years that it is taking me, but be warned that acquiring a deep understanding might prove to be very hard.

I'm studying very basic Lie group theory by working through John Stillwell's Naive Lie Theory. The end-goal in this direction is to acquire the basics of modern differential geometry. If I make it to the end of this book, I've got Janich's Vector Analysis (differential manifolds, differential forms, Stokes' theorem in the modern setting, de Rham cohomology) and Loomis & Sternberg Advanced Calculus (all this and more, starting from basic linear algebra and multivariable calculus in a principed way). Not decided yet which of them I'll try to work through or both.

Independently of this, I would like to refresh probability and acquire statistics in a mathematically rigorous way. I tried Wasserman's All of Statistics that is sometimes recommended, but it's too dry and unmotivating for me. I like the look of David Williams's Weighing the Odds, which seems to be both suitably rigorous and full of illuminating explanations, but I haven't really tried reading it yet.

I haven't read Weighing the Odds, but for what (very little) it's worth I attended one course lectured by Williams years ago and I thought he was an outstandingly clear lecturer.

I'm trying to learn Linear Algebra and some automata/computability stuff for courses, and I have basic set theory and logic on the backburner.

I'm writing a novel about metafiction: Some of the characters are aware that they are fictional, or rather that they live within a simulation where the laws of physics seem to follow a narrative. Unlike other metafiction stories, however, this isn't a comedy and the ontological and practical implications are treated seriously. Also, the main character is basically following timeless decision theory, but since it operates on very different timescales than humans, this has quite strange implications.

I find working on the background, the setting, characters and plot quite easy and captivating, but I hit a writer's block whenever I want to transform my notes into complete chapters. This has reached the point where I have far more notes than actual story.

Outlining is fun, but you just have to sit down and write at a certain point. Accept that the ideas won't be expressed perfectly the first few times. It's just like baking brownies; the best recipe in the world won't make you a tray of chocolaty goodness.

I know, but writing is hard :-( Also, I have made it way too hard for myself. It's easy to write notes about the personality of a completely non-human character, as long as you can intellectually understand its reasoning. But once I am forced to actually write its dialog, my head just hits a brick wall. The being is very intelligent and I want this to be rationalist fiction, so I have to think for a very long time just to find out in what exact way it would phrase its requests to maximize the probability of compliance. Writing the voices of the narrators/the administrator AIs of the simulation as they are slowly going insane is not easy, either.

Maybe I'm too perfectionist here. Do you think it's better to write something trashy first and rewrite it later, or is it more efficient to do it right the first time?

You can't write your characters more intelligent than you in fact are

Why? Your characters can have better memory than you, since you are writing things down; this also applies to keeping in memory several steps of something that you personally can't keep in memory. Your characters can take less time to make decisions than you do (since your fiction is not written in realtime). Your characters can notice things that you would miss if you saw them (because since you have defined their world, you already know what things are important to notice without having to notice them yourself). Your characters can have more skills than you have. How does this not ultimately add up to "your characters can be more intelligent than you"?

Of course, you could also cheat and have your character deduce something that he couldn't possibly have really deduced, but that doesn't mean all intelligent characters are examples of such cheating.

I know, and that is part of what makes this so hard. Thankfully, I have several ways too cheat:

-I can take days thinking of the perfect path of action for what takes seconds in the story.

-The character is a humanoid avatar of a very smart and powerful entity. While it was created with much specialized knowledge, it is still human-like at its core.

But most importantly:

-It's a story about stories and there is an actual narrator-like entity changing the laws of nature. Sometimes, 'because this would make for a better story' is a perfectly valid criterion for choosing actions. The super-human characters are all aware of this and exploit it heavily.

I'm making an effort to read more fiction, read about writing fiction, critique others' amateur fiction, and write my own.

I'm studying and trying to learn about Alzheimers disease: the causes and potential therapies. Please get in touch if you are interested in this as well, or aging more generally.

Last night, for the first time in my 31 years of life, I had a second date. Thank you, OKCupid matching algorithm.

On the subject of OKCupid...

Conferring with other LWers at the last London meetup, I seemed to enjoy a disproportionately large amount of success with OKCupid given my modest looks and overall questionable value as a human being. Over the ~18 non-continuous months I was actively using OKCupid in London, I went on dates with at least 30 different women, many of which had positive outcomes of one form or another. Others present seemed to think this was quite a lot.

I don't think I have any particularly great expertise on the subject, and it seems likely I just found whatever worked for me, but I would like to offer myself up as a resource for anyone who wants to mine my experience for useful information.

I am considering using OKCupid "seriously" for the first time; I've had an account for years but mostly for entertainment, with no profile.

I would definitely be interested in trying to understand how you were so successful.

Some observations of my OKC experiences:

A big factor was being in London. I've lived in various other UK towns and cities, and London is the only place OKCupid "worked", in the sense of being a semi-reliable place to obtain a date, with a high rate of turnover in the pool of available matches. By way of comparison, in about 18 months of online dating in Birmingham, (the UK's second-largest city), I went on maybe half a dozen dates.

Age was probably also a salient feature. I was 29-30 at the time, and had a sliding window of 26-32 on ages of prospective matches. I imagine the numbers were probably in my favour as far as site demographics were concerned.

Speaking to other people, I seemed to have enjoyed a large number of 99% matches, even for London. I'd expect to log into my account and see around 15-20 99% matches, which I gather is also unusual. I almost exclusively dated high 90% matches. I didn't engage in any clever strategy for answering questions, though I did answer a lot (> 1000). I do wonder if there are some particularly discriminate questions that most men answer "incorrectly", and I happened to fall on the right side of them.

(My take on the OKCupid matching algorithm is that it's sensitive but not very selective. People who you get on well with will probably be high matches, but people who are high matches won't necessarily be people you get on well with. A disproportionate number of 99% matches were tied to groups in my existing social network.)

I'm pretty sure my comparative advantage on the dating market is a combination of eloquence and dirty-mindedness. There seems to be a large subset of women who I match highly with who really appreciate the ability to subtly encode filth in language. This probably carries well over text-based communications and may account for some of my relative success.

My subjective experience of dating on OKCupid seems to be similar to everyone else, in the "seriously, fuck OKCupid" sense. I would regularly compose thoughtful messages to interesting-sounding women only to get no response, which was disappointing and downheartening. (I do have quite a bit of sympathy for the women on OKCupid in this regard, but that's a whole other essay). This seems to be a fixed experience of being a dude on OKC. I have no idea how much effort I put in compared to other people, or even how to go about quantifying it, but this might be a factor.

Patterns of actually going on dates were very much Feast or Famine. Sometimes I'd go for months without any responses. Sometimes I'd have an elaborate scheduling nightmare. On a couple of occasions I got to second-date territory with two women simultaneously, which was a novel experience for someone who spent his formative years pretending to be mythical creatures and developing strong opinions on which starships were the best. There was a particularly gruelling stretch in early 2012 where I'd just come back from a date and didn't have another one in the calendar, and it felt like I'd gotten out of some sort of debt.

The most sensible approach seemed to be treating the whole process as a way to meet new friends, who happened to be single women who hadn't ruled out sleeping with me. In this regard OKCupid was pretty successful. A little under half of the women I met I maintain some sort of social contact with, even if it's just the occasional bit of banter on Facebook. Eight or so are people I'll actively hit up for social activity, and a couple I'd consider good friends. Romantic outcomes were mixed, but generally positive: a few brief casual affairs, one ongoing long-term relationship and one ongoing intermittent play partner.

After a recent event where I encountered someone I'd been on an OKCupid date with way back in 2011, but didn't remember where I knew them from, I went to the effort of listing every date I could remember to make sure it didn't happen again. This was surprisingly difficult. The number currently stands at thirty women, but there could easily be a couple I don't remember. Prior to making the list, I somehow had the idea that I'd been on quite a few "bad dates", but looking over them, there was only one I'd describe as bad, and a few I'd describe as so-so. The dates themselves were overwhelmingly positive, but I think the overall process can be quite draining.

I think I'm out of observations for now.

I'd expect to log into my account and see around 15-20 99% matches, which I gather is also unusual. I almost exclusively dated high 90% matches. I didn't engage in any clever strategy for answering questions, though I did answer a lot (> 1000).

The more questions you answer, the more 99% matches you get. If you and another person have given the same answer to many questions, OkCupid tends to overestimate your compatibility.

One idea to avoid this is to first answer a thousand questions, see which ones are marked as Unacceptable most often, and then delete all of your answers and answer only those questions and the ones you consider particularly important.

I'm using The Charisma Myth, Neil Strauss' Rules of the Game, Nonviolent Communication, some other books and quite some LW posts to improve my social skills with good results. Apparently I made quite a good impression on my last job. I'm careful to clearly use only non-dark techniques (I avoid any that involve lies; The Game actually involves lots) and to stay authentic.

I'm using the Ultimate Geography Anki Deck to complete my knowledge of countries.

I was motivated to do so my this LW Comment::

A few months ago I started using the Ultimate Geography Anki deck after performing quite abysmally on some silly geography quiz that was doing the rounds on Facebook. I now know where all the damn countries are, like an informed citizen of the world. This has proven itself very useful in a variety of ways, not least of which is in reading other material with a geographical backdrop.

This I fully second. Knowing some lesser known *istan countries has already been helpful in GJP.