So, rationality fail: I'm currently in the process of buying plane tickets for a flight that's happening in two weeks. I could have purchased these tickets months ago and saved at least a hundred dollars, not to mention probably avoided layovers, but I had some ugh fields around both deciding whether I would go to the event that I'm flying to and around exactly when the flight should be timed. I could've easily resolved all of these issues by talking to a few people earlier but I had some ugh fields around that too. Bad times.
Started translating the Sequences into Slovak language (with Eliezer's permission). Two articles already published (yesterday, today), another eight are already waiting translated on my hard disk (I decided to limit publishing to at most 1 article per weekday, to avoid possible quick start followed by gradually slowing down later).
How much useful is that? Not sure -- this is a meta action, so it is naturally a multiplier, and I don't have a good estimate about what it multiplies: how many potential LW fans are there in my country. If many, this is one of the best ways to find them; someone can recommend them an article, or they can find it by putting some words in google. If there are few, this is a way to find out, and move on to some other plan. Perhaps this is also a way to create them.
What I learned:
Translating from English to Slovak is surprisingly easy for me. Understanding the context helps tremendously. I also have two translating programs open in browser (an English-Slovak dictionary, and Google Translate), so when I find an unfamiliar word, I first try to guess, and then look at the provided translation. If feels almost like using Anki. Seems like I found my comparative advantage. Depending on the results, this could be my most efficient "sanity waterline increasing" activity by far. I wish I had this idea half year ago; but better now than never.
Translating those articles and seeing the paragraphs written in my first language helped me understand the text better. (For example, in Simple Truth, which is a rather long text with some boring parts, I understood the motivation for having Autrey as a person separate from the speaker, as they represent different approaches. I somehow missed that on the first reading.) Generally, reading a text in a foreign language imposes a cognitive tax; with a lot of practice this tax gets smaller, but at least for me it's still far from zero. For the same reason I guess my readers will appreciate the translations even if they understand English.
It seems that you could capture benefits of both having the material online and searchable and retaining interested readers with regular updates by publishing everything at once, and then regularly posting your analytical readings of Eliezer's material. If you have the audience already, those could naturally grow into discussion posts.
Thanks for the idea! However, I started this project at an exceptionally un-busy time (makes sense in hindsight: that is when I have time and energy to think about new things to do), which seems to be over now; and these days I am not able to translate more than 1 article per day. Oh, the planning fallacy; happens to me all the time!
I'm reading Feeling Good by David Burns, and noticed that my expected value of many everyday activities is seriously out of whack with how it actually turns out. I don't have quantified data - yet. I'm starting to log expected value of time usage decisions, and compare with the actual experienced value, in order to calibrate things and hopefully improve my "I really don't want to clean my room even though I feel good about having a clean room and accomplishing stuff".
I suspect that a lot of my miscalibration is a holdover from the adversarial relationship I had with my parents over getting arbitrary tasks done. More specifically, it's an attempt to increase their costs of making me do something (since they don't want to see me unhappy, I can use being whiny or sullen as negotiating leverage). Turns out self-sabotage is an extraordinarily unhelpful habit, but I haven't really had the insight or tools to fix it up until now.
Anyhow, the two things I'm doing are making a daily schedule and comparing plans to actual accomplishments, and comparing predicted
Did my best to get Cat to come to Vienna. Applied for C-FAR minicamp. Publicly precommited to doing 4 papers or wouldn't go out on weekend. Have started using music when doing non-important work (raises happiness, minimal work impact). Started using rewards to make myself do work (after a bout of work watch a short tv series or something).
I just received my first performance evaluation in my current job. It occurred to me, just after reading it, that I should consider, for each statement in it, whether that information was surprising. From that, I obtained two plausible and relevant conclusions about where I am miscalibrated about my own performance.
The next step could be to estimate which criteria seem to be valued most, and optimize for them. Make list of things you could do, estimate their costs, and based on your evaluation try to estimate the benefit. Choose the ones with the best ratio.
Sometimes you are miscalibrated about your own performance. But sometimes the management is not aware of what you do. Part of success is to make all your good work known. Mathematically speaking, your work is w, how it appears to management is a×w, and your reward is r(a×w), not r(w). It is good to increase w, but it is also important to increase a.
(If you don't communicate something to the management, it can be forgotten, or if you happen to have a colleague skilled in Dark Arts, they might present it to the management as their own success.)
I am now doing Anki about 3 times a week now. I am currently creating an Anki deck for my Psychology 101 book and another for "An Idiot's Guide to Game Theory" which I picked up the other day. I am also learning HTML and CSS on an online course, again using Anki and practice tests to reinforce what I have learned.
I picked up these techniques at LessWrong.
May I suggest applying CSS skills to styling Anki cards? A personal anecdote: I have a lot of decks for various languages, and having cards styled differently helps with switching context. It's also nice to have them look pleasant.
I signed up for HabitRPG a couple of weeks ago, and so far have experienced a couple of probably obvious things:
I tried outlining the different parts of my accessible image viewer/editor, and realized that the Ugh Fields keeping me from working on it are specifically around the UI and user-controlled settings; I can toss out code for the actual viewing part in five minutes, but working it into a larger program collapses from the weight of the rest of it. The trouble is that it's not testable without a minimum of lots of user controls.
Other than breaking it up into as many different chunks as possible (bound to be messy, but possibly more likely to get done), I haven't come up with a means of exploiting this to complete the program.
What are you using to program the UI?
I've had my UI-design experiences vary from easy (webapps via servlets and jsp, Hypercard), to easy so long as I was trying to do exactly what it was made for (Java with Swing), to so obnoxious that I gave up (C with GTK UI elements).
I've tried it two different ways, but the one that seems best is Java via Swing (the other approach involved me needing to build an imageIO library from scratch, which would just be silly). Making Swing accessible, though, is rarely worth attempting, so I was hoping to make liberal use of shortcut keys, spoken feedback and component titles.
Really? I'd think that routing things through subclasses of AbstractAction and then hooking them up as Listener to the various access methods would make it comparatively straightforward.
This is the public group instrumental rationality diary for May 16-31.
Thanks to cata for starting the Group Rationality Diary posts, and to commenters for participating!
Next diary: June 1-30
Immediate past diary: May 1-15
Rationality Diaries archive