All of Ratcourse's Comments + Replies

The point in the first paragraph is well made, but in a way that might be interpret as misvaluing the content which is in fact, very good. It shifts the means from "Find the right advice" to "Figure out how to implement the advice you already know is right" which is a very notable change.

Excellent post, OP.

yeah no worries be as late as you like :P

"Hmm. I didn't interpret a hypothetical apostasy as the fiercest critique, but rather the best critique--i.e. weight the arguments not by "badness if true" but by something like badness times plausibility."

See Plausibility depends on your current model/arguments/evidence. If the badness times probability of these being wrong dwarfs the former, you must account for it.


N=1: Time to stop self-identifying with thoughts was less than 5 total hours of meditation practice (scattered across months, but still). This was especially helpful in diminishing neurotic behavior - the thoughts are still there just not engaged with.

Yes, a blank spot and one that makes everything else near-useless. This needs to be figured out.

That automation makes sense, thank you. Trying to think of how to generalize it, and how to to merge it with the first suggestion.

Anki doesn't work for me on this, agreed. The above suggestion seems to dominate this one.

In response to this post:

Robert Wiblin got the following data (treated by a dear friend of mine):

89 Confirmation bias

54 Bandwagon effect

50 Fundamental attribution error

44 Status quo bias

39 Availability heuristic

38 Neglect of probability

37 Bias blind spot

36 Planning fallacy

36 Ingroup bias

35 Hyperbolic discounting

29 Hindsight bias

29 Halo effect

28 Zero-risk bias

28 Illusion of control

28 Clustering illusion

26 Omission bias

25 Outcome bias

25 Neglect of prior base rates ... (read more)

One could try ranking biases by the size of the correlation between susceptibility-to-the-bias and damaging behavior, for example, using the correlations in
This is totally worth a discussion post.
You may not have noticed when you posted this, but the formatting of your post didn't show up like I think you may have wanted, with the result that it's hard to read. (If you're wondering, it takes 2 carriage returns to get a line break out.) If you intended the comment to look like it does, I apologize for bothering you.

How do you correct your mistakes?

For example, I recently found out I did something wrong at a conference. In my bio, in areas of expertise I should have written what I can teach about, and in areas of interest what I want to be taught about. This seems to maximize value for me. How do I keep that mistake from happening in the future? I don't know when the next conference will happen. Do I write it on anki and memorize that as a failure mode?

More generally, when you recognize a failure mode in yourself how do you constrain your future self so that it doesn't repeat this failure mode? How do you proceduralize and install the solution?

Discuss the failure in person, face to face, with a helpful colleague. Admit your failure. Make a conversation out of it. Brainstorm the fixes you've already made to the bio and any others that come up. Let the conversation have an attached emotion, whether it be a feel good problem solving session or a public shaming. Memories with emotions stick around better. I found that even the small act of saying "Yep, I did forgot to update the new widget part number" to a supervisor / team-member helps me remember to in the future.
For a problem like this, remembering for something rare in the indefinite future, the important thing is to remember at that time that you know something. At that point, if you've put it in a reasonable place, you can find it. It seems to me that the key problem is the jump from "have to write a bio" to "how to write a bio," that is, making sure you pause and think about what you know or have written down somewhere. Some people claim success with Anki here, but it doesn't make sense to me. What most people do with bios is that they reuse them, or at least look at the old one whenever needing a new one. As Maia says, if you write an improved bio now, you can find it next time, when you look for the most recent version. But that doesn't necessarily help remember why it was an improvement. If you have a standard place for bios, you can store lots of variants (lengths, types of conferences, resume, etc), along with instructions on what distinguishes them. But I think what most people do is search their email for the last one they submitted. If you can't learn to look in a more organized place, you could send yourself an email with all the bios and the instructions, so that it comes up when you search email.
I'm not sure this applies to your particular situation, but a general solution for proceduralizing behaviors that was discussed at minicamp (and which I'd actually done before) is: Trigger yourself on a particular physical sensation, by visualizing it and thinking very hard about the thing you want yourself to remember. So an example would be if you want to make sure you do the things on your to-do list as soon as you get home, spend a few minutes trying to visualize with as much detail as you can what the front door of your house looks like, and recall what it feels like to be stepping through it, and think about "To do list time!" at the same time. (Or if you have access to your front door at the time you're trying to do this, actually stepping through your front door repeatedly while thinking about this might help too.) And if there's some way to automate it, then of course that's ideal, though you said you don't know when the next conference will happen so that's more difficult. Or another kind of automation: maybe you could save the bio you wrote in a Word document, and write a reminder in it to add the edits you want... or just do them now, and save the bio for future use. Then all you have to remember is that you wrote your bio already. Which is another problem, but conceivably a smaller one: I don't know about your hindbrain, but upon being told it had to write a bio, mine would probably be grasping at ways to avoid doing work, and having it done already is an easy out.
For a while I was in the habit of putting my little life lessons in the form of Anki cards and memorizing them. I would also memorize things like conflict resolution protocols and checklists for depressive thinking. Unfortunately it didn't really work, in the sense that my brain consistently failed to recall the appropriate knowledge in the appropriate context. I tried using an iOS app caled Lift but I found it difficult to use and not motivating. I also tried using an iOS app called Alarmed to ping me throughout the day with little reminders like "Posture" and "Smile" and "Notice" to improve my posture, attitude, and level of mindfulness, respectively. This worked better but I eventually got tired of my phone buzzing so often with distracting, non-critical information and turned off the reminders. My very first post on LessWrong was about proceduralizing rationality lessons, I think it's one of the biggest blank spots in the curriculum.

WRT S.M.A.R.T. goals, Nick Winter says in the motivation hacker:

When you do pick your goals, forget the advice about SMART goals. Use Piers Steel’s slightly improved CSI Approach. Your goals should be Challenging (if they’re not exciting, they won’t provide Value); Specific (abstract goals can leave you vulnerable to Impulsiveness, since it’s not clear what you need to do); Immediate (avoid long-Delayed goals in favor of ones you can start now and finish soon), and Approach-oriented. (As opposed to avoidance goals, where you try not to do something, you should instead reframe it positively as an attempt to do something—it just feels better.)

Nick Winter knows about habit formation

I agree with the challenging bit, but for a different reason. Quoting from Piers Steel, "We are motivational misers who constantly fine-tune our effort levels so that we strive just enough for success." For low complexity goals, up to a point, there is a linear relationship between goal difficulty and goal performance, even when the reward is held constant. That is, more difficult goals require more motivation; provided that the goal is valuable, that motivation is provided. The difficulty with choosing challenging goals is ensuring that you feel motivated, not excited. Excited is thinking of the benefits and feeling positive emotion in anticipation. Motivation is the energy necessary to actually complete all the crap in between, like doing your pushups. I use mental contrasting for this. It works better than nothing, but still leaves much to be desired.

Intimate relationships by Miller/Perlman/Brehm

Found out intimate relationships are a part of my life in which I feel I could do better. Found out it is overlooked in x-rationality groups. Bought The textbook on it. Am learning it.

Sorry if this is a noob question, but what is "the textbook" in this case?

Single data point but: I can alternate between inner monologue (heard [in somebody else's voice not mine(!)]) and no monologue (mainly social activity - say stuff then catch myself saying it and keep going) - stuff just happens. When inner monologue is present it seems I'm in real time constructing what I imagine the future to be and then adapt to that. I can feel as if my body moved without moving it, but don't use it for thinking (mainly kinesthethic imagination or whatever). I can force myself to see images, and, at the fringe, close to sleep, can make up symphonies in my mind, but don't use them to think.

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).

Is this really want went trough your mind or is it a rationalization?

I can't understand. LW is obviously important to you. You know this is a touchy topic. Why not provide sources (who's burden is it if not the author's?) and turn this into an amazing post? If you have sources for everything that you claim this is an amazing work. If not, it's worse than useless: it imprints wrong thoughts that will hang around for a while.

I don't understand.

It is an after the fact reconctruction, as I mentioned. There is no "what went through my mind", there is only us, making sense of ourselves. I didn't provide sources for specific things because these are not easily accessible in my mind, I'm not a computer, with folders separating everything and eidetic memory. They are the product of a mix of information from those papers, books, etc... that later on I cited, with a lot of experience talking about sex with people. And please don't ignore the most important reason: Suppose you want to give advice to a friend who just say a woman and felt interested at a bar. You say: "Go to her, try to talk to her, see if you really like her, and at some point try to kiss her" This is very, very simple advice. You would not be able to pull off citations for it, even though it is obviously true that 1) Approaching 2)Talking 3)Seeing if you like her and 4) At some point trying to kiss her... will help your friend. Advice and citations don't go together as neatly as would be desirable. Suggestions, even when based on rocket science, include some element (value? desire? intention?) that doesn't fit the dry writing style of academic papers. Some like to call this the fact/value differentiation, debate etc... Which is why the PUA community has more knowledge about human courtship than most ethologists (even according to the ethologists), they test things differently, with some intention, by trying specific things, not just trying to take a picture of reality and reverse-engineering the minds of all who belong in the picture with Omega level intelligence and Oracle AI level of knowledge.

Agreed. Downvoted original post because of lack of sources.

Edited. Oh, when editing I forgot Sperm Wars, The Emotion Machine (2007) The Company of Strangers (2005)

Maybe this should be in the Open Thread?

Nonetheless, I feel that if you can't explain it without using jargon, that gives some evidence for you not understanding it in the first place (whatever it is).

I feel that if you can't explain it without using jargon, that gives some evidence for you not understanding it in the first place (whatever it is).

Eh. Maybe if you're using it in a guessing-the-teacher's-password kind of way, but sometimes you need jargon because you need to say something very precise (e.g. in mathematics).

What is your goal? Why would get a PhD help you meet it? Is your goal to have a PhD?

As did shimnux, I started a blog to much of the same effects.

I also got into habitrpg, this has been, easily, the best thing I ever did to install new habit; highly recommended.

Language note: Your first paragraph sounds like it means, "Under the name 'shminux', I started a blog [...]". It would be better to say, "Like shminux, I started a blog", or "As shminux did, I started a blog". (Assuming I take your meaning correctly.)

I'd like to know why this is getting downvoted. I think you are making a serious point in a short, funny manner. I actually preferred this to your longer posts.

I've seen insane-er stuff argued seriously before.


Thank you, this is my experience - it feels stressful. Updated prob.

I did try it once. It was less bad than german but still bad: I memorized, but only after actually reading up on them I knew what I meant - leading me to think that I ended up not saving time in the long run, for if I had written them myself my memorization would probably be better.

I will try it again. For science.

Thank you all for your responses. I've updated my estimate that this is just a me-problem.

Still, if you are trying to learn material from a field that requires a general framework for understanding, then I'd advise you not to learn directly from an Anki deck, but to first study the original source from which the deck was created. Perhaps you may want to try the List of Cognitive Biases and Fallacies deck, which many people (myself included) have had good experiences with, to determine if your "atrocious experience" was the result of bad luck or bad choice of decks on your part, or if it instead reflects something else about how you are using Anki.

benthamite, have you had success using decks you have not built yourself?

I once tried with the list of cognitive biases, and again with german and it was an atrocious experience. I thought then that I was violating rule 2:

"Learn before you memorize Before you proceed with memorizing individual facts and rules, you need to build an overall picture of the learned knowledge. Only when individual pieces fit to build a single coherent structure, will you be able to dramatically reduce the learning time. This is closely related to the problem comprehension ... (read more)

I've tried it and run into what feels like the same problem. I've only ever kept one deck built by someone else, and I very frequently edit and delete cards from it. It's still my only Anki deck that I don't really enjoy spending time on - it feels more stressful. And for me, the process of creating the cards is definitely one of the most valuable components of the procedure. That said, I still haven't got rid of that deck, and I'll post a few of my decks that I think could be somewhat generally applicable in this thread in case anyone else wants to use them.

People who learn about spaced repetition are always tempted to look for existing decks to memorize, but I think this is a false efficiency. In order to memorize a deck of material, you are going to be dedicating a non-trivial portion of your life to memorizing its content. The time spent actually creating the cards is substantially less than the time spent quizzing for almost any material. I'm not completely against pre-existing decks, but I think people just need to keep perspective.

One of the biggest lessons SRS will teach you is humility: the amount you will ever be able to learn is extremely finite. Use it wisely.

Thank you all for your responses. I've updated my estimate that this is just a me-problem.
Funny you should mention German -- I started learning German with someone else's German vocab deck (one of the ones that's on Ankiweb, possibly even the same one you used). Those first 200 words were not terribly well-chosen, in retrospect, but I digested them in 12 days, 20 words a day plus 2 review-only days at the end. Since then I've been adding to it, and have created other decks for grammar and irregular verbs and things like that. It was the first Anki deck I've ever used, and I've never used Anki for anything but German, so I don't have a very broad perspective. But it worked for me.
Yes, both to learn new material and to retain material I had previously learnt. From the list above, examples of the first type include Get Motivated, List of Cognitive Biases and Fallacies, 100 Greatest Paintings of All Time, and The Major Mnemonic Memory System, while examples of the second type include Richard Wiseman's 59 Seconds, How to Formulate Knowledge and Marshall Rosenberg's Nonviolent Communication.
I used a deck I didn't build myself to learn GRE vocab.

Upvoted, but it felt more like a lament than a stab at a problem. Not that there is anything wrong with that.

I expected a pointing at a solution. Nothing came. I can imagine some books have higher priority than others. I can also imagine that some insights have higher priority than others (in my case, more than 20 insights a day get forgotten, so I have to put an upper bound on how much I learn per day).

Looking at what people you already consider bright recommend most highly might be a way.

Rationalists should win, I feel your pain, but sketch me a solution.

Before you can sketch a solution, you first have to characterize the problem. Sometimes the solution isn't apparent immediately. And sometimes someone else has it.

Please pm me your email and I will forward it, no idea why it didn't go trough.

Please pm me your email and I will forward it, no idea why it didn't go trough.

Maybe a mod can help with that, I confess to my ignorance: I don't know how to make that happen

You should make a meetup post, not a discussion post (use the button "Add new meetup"), and specify the address explicitly and in enough detail so that at least the city is correctly recognized by the inline google map (is "Cafe im schottenstift" sufficient information for the locals? Google Maps says it's at "Schottengasse 2, 1010 Innere Stadt (1.Bez)").

Yes, I forgot that one, quite good indeed

Coursera is fine, just fine. I took intro to model thinking and game theory and they've put some thought into it. Udacity is better, cs101 is the best online course out there afaik. intro to ai, is so-so

Udacity stats 101, is also very good.

what do you mean rarely have time for importing? going to anki import file it takes maybe 30 secs?

highly discourage this method. "reviewing" starts taking too long and one starts procrastinating.

i copy and paste stuff i want to learn to a text file, and when i have down time go into it and turn it into question/answer/tag.

once everything is done i import it into anki

My method is better than processing each card you decide to add, which breaks your reading workflow and discourages you from adding more cards. This is a good workflow. Definitely better than mine. However, I don't think it will work well for cloze cards.
I do this too, but I find I rarely have time for importing. Edit to add: for vocabulary words I've come up with a set of scripts and vim macros that put the definition (from wiktionary) and some tags into a csv file for importing into Anki. But even with this amount of automation it's more work than I normally want to deal with.