This is the public group instrumental rationality diary for October 1-15.

It's a place to record and chat about it if you have done, or are actively doing, things like:

  • Established a useful new habit
  • Obtained new evidence that made you change your mind about some belief
  • Decided to behave in a different way in some set of situations
  • Optimized some part of a common routine or cached behavior
  • Consciously changed your emotions or affect with respect to something
  • Consciously pursued new valuable information about something that could make a big difference in your life
  • Learned something new about your beliefs, behavior, or life that surprised you
  • Tried doing any of the above and failed

Or anything else interesting which you want to share, so that other people can think about it, and perhaps be inspired to take action themselves.  Try to include enough details so that everyone can use each other's experiences to learn about what tends to work out, and what doesn't tend to work out.

New this month:  If you have preferences about the future frequency of rationality diary posts, please express them in the poll below!

Thanks to cata for starting the Group Rationality Diary posts, and to commenters for participating.

Immediate past diary:  September 16-30

Rationality diaries archive

New Comment
39 comments, sorted by Click to highlight new comments since: Today at 1:56 AM

So my plan yesterday was to take a day with no computer, tablet, Kindle, or physical books in order to get some metacognition done about why I was having trouble getting started on my MIT talk.

Result: Spent almost the entire day lying in bed in a pseudo-REM state, then fell asleep a couple of hours early.

Next day, lo and behold, writing the talk doesn't seem so daunting. I think I was just exhausted, and having access to a computer, maybe even a Kindle or book, was sufficient to prevent previous rest days from being restful.

Is there a way for this to inform your sleep/rest schedule moving forward?


Yep, this happens to me too when I forget the other half of the secret: close the computer and go outside.

[This comment is no longer endorsed by its author]Reply

I had a problem I was trying to work out. I thrashed about on it for a long time without making apparent progress. I decided I could ask a forum of people more experienced in this.

So I began I composing the question. I didn't want to seem like a total noob, so I set about forestalling the obvious suggestions by explaining what hadn't worked. I sort of set up an imaginary interlocutor and had it suggest things. Eventually, it suggested something and instead of my showing why it didn't apply to my situation, I showed that it did. So I copied the solution out of the forum post box and closed the browser window.

Now, the process of composing this question wasn't exactly quick or easy, but it did take a lot less time than I'd been using before.

I wonder if I can replicate this on purpose. I imagine writing to (or start writing to) someone more knowledgeable than me whom I don't want to seem stupid to, and explain the problem. It seems a lot like just writing a summary of the question for myself, but by giving it that additional purpose, it became more focused and self-critical.

Or maybe there's nothing to it, and I got lucky that I happened to be sitting on everything I needed and was one inference away from the answer.

Similar experience here. When I was doing my grad research and hit a snag, my first reaction was to ask my supervisor if I was unable to make any progress for a few days. I did it a few times, and sometimes the feedback was unexpected and based on the experience, skills and knowledge I simply didn't have at the time. But quite often, to my embarrassment, the useful reply was well within my capabilities. That made me uncomfortable and the look I got from my supervisor didn't make me feel any better. So eventually I started imagining asking the question and getting a reply, as per my model of the interlocutor. That cut down on stupid questions significantly at the actual meetings, saved me time waiting for them, and (eventually) improved my supervisor's opinion of me.

I rarely face similarly challenging issues while writing software, so simply clearly writing down the issue I am facing tends to be enough to see possible solutions.

Not quite. Conversing with an imaginary intelligent interlocutor is more advanced than simple rubber ducking.

You're supposed to have the rubber duck ask these questions too.

The difference here was that I was actually writing a letter / forum post that other people could have seen. I was doing this with the urgency of 'If there's an obvious solution to this, I'm going to look like a moron'. In contrast, a rubber duck, well... it's just me.

I think the distinction you drew in the other branch of this conversation is more important. Getting a little skin in the game of 'get this right now (or satisfy yourself that it's genuinely hard)'.

I got lucky that I happened to be sitting on everything I needed and was one inference away from the answer.

Even so, finding a way to make the missing inference is not necessarily trivial.

I just resisted bystander apathy. But probably all I did was call the cops on some cops.

I was driving home from work, and I saw some people getting pulled over, but the people pulling them over didn't look like police. It didn't look any different than what I would see if some plain-clothes police officers had seen someone speeding, but I had a really bad feeling about it for some reason, so it didn't seem like a good idea to do nothing. My phone wasn't charged, though, and my destination didn't have phone reception, so I let my phone charge for a few minutes when I got to my destination (which was nearby). I also asked a friend what they thought I should do (it seemed like a good idea to a basic "am I being stupid?" check), looked up the local non-emergency police number, and wrote down everything I could remember about what I'd seen. After charging the phone, I had to drive back to somewhere where I could get reception, so I didn't make the call until half an hour after the fact.

I think the main thing that would've helped me handle this better is having thought out in advance what to do in common emergency-like situations. Something really general, like "if there are IDs of any sort, consider writing them down", could have been useful here. Also, "borrowing things from other people is possible" (like charged phones), and "consider safety before entering a situation" (having nowhere to pull over was the first reason I decided not to - there's some chance I could have acted too quickly if there were a convenient driveway). Putting phone numbers for police into my phone would also be a good thing to do in the future.

I dropped my home internet service about a month ago. I activated my cell phone's hotspot capability, so I still effectively have internet, just much slower and I'm much more conscious about using it.


  • Saving $45/mo.
  • I spend more time doing productive things.


  • Streaming multimedia is no longer a viable option for entertainment.
  • The internet isn't as fun anymore.

Interesting artifacts:

  • I'm getting to sleep earlier. Now that the internet is significantly less fun, it's much easier to shut it off.
  • Since text loads more easily than video and graphics, I'm reading more stories.
  • I'm also working through my backlog of things to do, rather than acquiring yet more things to do.

I started microdosing melatonin based off of Eliezer's success. I have the same sleep disorder. So far it is playing havok and giving me different results every night, but I am going to stick with it for a month to see if it evens out. Taking the melatonin at 11 and going to sleep actually feeling sleepy at 2am is still a huge improvement over the status quo before.

Isn't Eliezer's solution to take the melatonin 5-7 hours before you want to go to sleep?

MetaMed's solution, to non-24 sleep disorder in particular.

I'm pretty sure I have something similar-ish with regards to drifting sleep cycles, but it's a whole lot less resistant to sleep treatment - forget 5-6 hours, 1mg melatonin an hour before bed is fine.

Yes, the interval has fallen out organically. I take it at 11 and go to sleep whenever I get sleepy. This has averaged 2am so far.

Can you describe, or link to a description of this method?

I have a notetaker device that was out of date when I got it in 2006, but since it's primary functions are as a word processor, calculator, and occasional calendar/alarm/stopwatch, that isn't too problematic. The problem is that the battery stopped holding a charge last year, and I haven't been able to replace it. My productivity has declined noticeably since then. It also seems to lend itself to working with external distractions: when I was attending college, I could get quite a lot done during classtime without it negatively impacting academic performance, it made going places or watching TV with family both tolerable and not-useless, and is much better for working with music than my laptop.

I decided to try using it this morning, but that did not go as planned. This evening, people left the house with comedy central playing (the Stewart-Dawkins interview replayed), so I decided to try bringing it out again. And this worked pretty much perfectly, even after people returned.

I should fix that battery. The company that sold the device will do replacements for prices several times that of batteries with the same specs on the internet, but I don't know if the battery is too unique for random internet batteries to work (and I can't find a screwdriver). This should not be a difficult problem. That it is says more about me than the device/battery/company.

I am beginning to track my motivation so I can have something of an idea of how often I ought to 'refuel' as well as what sorts of things work for 'refueling.'

I began to schedule a chunk of time to make a to-do list for the next day in my daily to-do lists. They work great when I actually make them, but I remember to make them less than half the time.

What types of things have y'all successfully beeminded?

I tried increasing the number of consecutive push-ups I could do, eventually got sick and failed. Beeminder isn't very forgiving in such situations. I tried forcing myself to write every day and managed to fail that despite putting up an actual pledge and setting a really low bar for what counts as writing. I love the idea of beeminder and it has actually helped me in the sense that it worked for a while in each previous attempt, which is better than not working at all. I've even read Nick Winter's The Motivation Hacker yet I can't seem to find goals in my life that it makes sense to beemind.

So, to return to the question, what things have you successfully beeminded?

If I understand it correctly, much of the power in Beeminder comes from the threat of losing money when you fail. How many times did you fail at writing before giving up? I have not used BM in a while, but I did successfully use it for writing.

(Number of consecutive push-ups doesn't seem like a good thing to Beemind. If your body doesn't have 50 push-ups in it, wanting it really bad isn't going to help much. Tracking number of push-ups done in a given amount of time would probably work better, which would naturally increase consecutive pushups.)

I very much agree with the parenthetical about pushups. I beemind 30 pushups per day -- -- with the idea that I'll gradually ramp that up as my max reps increases. Except I'm failing to ever do that and have been at 30/day forever. If I cared more I'd ramp it up though. Right now I'm just happy to be forced to maintain some semblance of baseline upper-body strength.

The general point: beemind inputs, not outputs. Ie, things you have total control over.

PS: The Beeminder android app has a pushup counter built in, where you put your phone on the floor and touch your nose to it on each pushup and it tallies them for you.

Beeminded successfully:

  • Anki.
  • Exercise, including pushups. I was also successful doing pushups with the Seinfeld method for about a year before Beeminder.

Beeminded with middling success or unsuccessfully (not totally sure why, but I think these are just harder goals for me):

  • Hours worked (that I get paid for).
  • GTD/Remember the Milk tasks.
  • Hours spent on personal programming projects.

As far as pushups are concerned, they've been successful in the sense that I've been doing pushups regularly for two years and am still going, but -- due to my genes and my current choice not to expend a huge amount of mana points on this goal -- I've been plateaued at 12 to 16 consecutive pushups for a year or more. I second Daniel's recommendation [edit: Daniel being dreeves; also, jetm made the same point] to set a goal that's under the control of you, not of your body's ability to increase its strength.

I have lost money to Beeminder a few times. I think the money helps motivation, though maintaining the graph and "not breaking the chain" are powerful on their own.

Short answer:

Long answer: Anything I want to repeat, with reminders, that is measurable (the enforcement is nice, but actually much less important than the reminders, tracking and ease of entering data for me).

My beeminder results are generally bad. The only three successful goals were: making pictures for my textbook, doing sit-ups, and running.

I guess the difference was mostly that my successful attempts were short-termed. I can motivate myself to do something for 30 days. For a longer period, after some time I start hating doing the thing. It is a bug I should fix somehow, but beeminder wasn't any help in fixing that bug.

I've been using a balance board with my standing desk. It subjectively feels like it increases my peek productivity when I'm really into my work, although I haven't tested this. At the very least, it's given me more exercise.

I'd prefer for rationality diaries to be posted:


If you have further specifications to your preference (e.g. biweekly on Tuesday) please mention them in comments.


Is biweekly twice a week in this poll?

It would be every other week. So, slightly more often than the current schedule, but always on the same day of the week.

I probably should have said "twice monthly" on the third poll option.

I would like the rationality diary, open thread, and stupid questions to be combined into one weekly thread.

I would dislike this relative to the status quo. In my case, I have comparatively little interest in the "stupid questions" thread, and appreciate that the current setup allows me to read the other two threads without having to sort them out.

I agree with Dorikka, and for much the same reasons. Having these separate allows me to read what kind of content I want to read whenever, and if reading is made easier and posting is not really made harder, there is no reason not to keep it as it is.

I have particular interest in the "stupid questions" thread and want it to stay separate. I'm neutral on the separation of rationality diary and open thread.

What do you find especially interesting about the "stupid questions" thread?

  • I'm definitely too shy to make a top-level comment in the open thread, in part because I fear my thoughts are not clever enough to belong on this forum. It seems good to have a designated place where this is less of a concern.
  • I am more likely to have a useful or interesting response to a "stupid question" than to a typical open thread comment.

I can't subscribe to the RSS feed for this page. Other LW pages work fine. Anyone else have the same problem?

New to LessWrong?