I was reading the NY Times article on Decision Fatigue, when I came upon a hypothesis I would like everyone's feedback on.

I take as a premise that there seems to be a high prevalence of akrasia in the lesswrong community.

I also take as a premise that the sequences give us a more-than-usual detailed model of the world, one that presents us with more possible trade-offs we could be making in every day life.

So the conjecture that by trying to reduce bias and perform a lot of cognitive calculation, we effectively spend large parts of our days in a decision fatigued state, leading to akrasia problems.

Does this sound (un)reasonable? Why? How would you go about turning this into a testable proposition?

UPDATE: Anna Salamon has put up a detailed poll here that may shed some light on the situation. Please take some time to fill it in.

New Comment
50 comments, sorted by Click to highlight new comments since:

there seems to be a high prevalence of akrasia in the lesswrong community.

There were plans to check this but people never got around to it.

I just made a short poll here, to better estimate both how often LW-ers have akrasia, and what it correlates with (e.g., decision fatigue); please take it if you have a few minutes.

Edited to add: poll results now here.

For one of the questions, it will tend to help to know that Less Wrong began about 30 months ago.

I'm interested in analyzing a Google spreadsheet like this one is linked to. I use a form like this for a daily self-survey. Are there any tricks you know of to analyze the data on these things bit more effectively than just reading through them? The charts feature on Google Docs seems a little weak.

BTW, filled out the survey. It's nice but the last 'blank field' type question is a duplicate.

I'll just be importing the data into Matlab, which has nice data analysis tools.

This is a great question. A quick guide (or link to one) on how to do easy journaling/analysis using Google Docs or similar simple tools would be great. I just started a food/biojournal myself. There's probably one or more out there in the self-analysis literature, but I haven't seen it much discussed here.

I have the data from the LW-ers, and have done most of the analysis, but it would aid interpretation to have a "control group".

Just posted to Reddit; should have data with control group analysis writtten up in a couple days, if Redditors in fact fill it out.

Any update on this?

Argh, pressed "return" in one of the form boxes and accidentally submitted the form early.

If I am currently a U.K. undergraduate, what was my college G.P.A.?

Good question; please leave questions blank if they don't apply to you, such as that one.

Oh, ok. (For some reason I thought the answer to that question involved time travel.)

Also, "What's your present income, in dollars per year?" is repeated twice, once with the GPA subheading.

Thanks; fixed.

I had virtually the same thought while showering this morning. I was running late for an extremely important exam and found that I was sphexishly staring at nothing.

I said to my brain, "We're late. Finish showering."

My brain replied, "Hold on, I'm optimizing something unrelated to this."

"This is kind of urgent," I suggested.

"It's not very rational of you to ask me to stop thinking about this and thus risk allowing a potential cognitive error to pass unchecked," my brain said haughtily.

"I think it is instrumentally rational for us to not be late to the exam."

"Perhaps, let us consider the value of sacrificing global correctness for local instrumental rationality - "

"Shut up," I said out loud, and finished showering. But of course I continued thinking and realized how often I allow meta-optimization to take priority over actually doing things, and how it only really becomes apparent when I am frantic about a deadline. This also explains why deadlines actually do motivate me despite the fact that I hate them.

Many LW-ers seem also to have had unusual amounts of akrasia before finding LW, suggesting that it isn't the sequences that are the cause of the trouble. (For me, the sequences helped with productivity.)

But even so, it could be that folks here tend to use system 2 processing more than folks elsewhere (and tended to do so also as kids), and that this leads to decision fatigue. It sounds worth testing, if we can find a way to do that.

Someone without "too much time" on their hands might find it difficult to find this place and then spend enough time to evaluate it to become a regular.

Will you or have you already published more detailed analysis of the survey?

[-]jimmy130

What sounds more likely to me is that people start identifying as someone that wants to get more stuff done, but their motivation level doesn't increase (fast enough?) so what used to be "I just don't want to do X" becomes "I want to do X, but I'm not doing it so it must be akrasia or something"

I'm familiar with the feeling of decision fatigue, but casually using and building a more detailed model of the world doesn't seem to have that effect - only if I'm really struggling on a hard problem.

Academics may be relevant as a similar "thinking community". I've heard many academics say they are severe procrastinators. Possible reasons for this are: 1) Procrastinators are attracted to the job, its independence and long-term deadlines 2) The nature of the work makes people procrastinate; research is hard, plus no boss and long-term deadlines mean immediate punishment for procrastination is rare 3) The job makes people feel they have akrasia even when they don't, perhaps because colleagues and competitors seem smarter and harder-working than in other fields

If nothing else, reading LW makes me feel 3)

2 seems like most of the explanation. You can't procrastinate on an assembly line.

I procrastinated when in academia, but did not feel particularly attracted to the job, so option 1 is not always true. Comparison with people not in academia makes it seem that option 3 is not true for me either.

Poll: How has your level of akrasia changed since you found lesswrong?

About the same.

Decreased significantly.

I voted decreased significantly, but I don't think the cause is LW - that's confounded with going to grad school, which I think is more likely to have been the cause, but those are both confounded with increased maturity due to, well, aging.

Poll: only answer this one if your akrasia has improved significantly since you found LessWrong:

Let t be the time since you found Less Wrong (e.g., "two years", if you found LW two years ago). Did your akrasia decrease more in the t years since you found LW, or in the t years preceding your finding LW?

ETA: The poll I just added (as a Google form, here) is better; please use that one instead, so we can see cross-correlations. (That's why I "retracted" this, but I can't figure out how to delete it.)

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

About the same.

Improved significantly more in the t years before you found LW, than in the t years since you found LW.

Improved significantly more in the t years since you found LW, than in the t years before finding it.

Increased significantly.

I don't necessarily see a larger amount of people on Less Wrong who suffer from akrasia. It seem to me like more people identify their procrastinating as akrasia and see it as problem that can fix if they try harder or find the right tools.. As a student, I hear others/myself complaining about how they didn't have the willpower to complete their homework, or wishing for better time management skills, or that they didn't give in to playing video games/the internet. No-one uses the word 'akrasia' though, and many do search for solutions.

[-]Crux10

I disagree with the conjecture. I don't think to be rational one needs to perform an unusual amount of conscious calculation per unit of time. To be rational, one must simply change one's default settings in a specific way. This could be done bit by bit over the course of many years.

It certainly takes an unusual sum total of introspection and conscious awareness and whatever to discover just how broken and unreliable our default settings are for the modern environment, but once changed it takes no further conscious intervention to keep those settings in their new position.

For example, it might have taken a long time and a lot of conscious intervention for me to identify and stop eating junk food (or what I consider such), but it certainly no longer takes any conscious thought to reject a piece of candy. I don't think through everything bad it could do to me or anything and weigh that against how good it would taste; I simply feel a few painful bodily sensations and think, "Ew! Can't have that!"

So I think this comment by Jonathan_Graehl is spot on. It seems that your conjecture would only apply to people currently making an unusually large amount of changes to their thinking and behavior. Perhaps that's the case with a lot of people on LW though.

it might have taken a long time and a lot of conscious intervention for me to identify and stop eating junk food (or what I consider such), but it certainly no longer takes any conscious thought to reject a piece of candy. I don't think through everything bad it could do to me or anything and weigh that against how good it would taste; I simply feel a few painful bodily sensations and think, "Ew! Can't have that!"

When I began cooking and grocery shopping on my own, I've thought things through and decided that it would probably be better to mostly eat fresh vegetables, beans and, rarely, some meat. I kept at it for maybe two years, but then I moved in with different people and since I didn't have any strong reasons to stay with my old diet I began eating more junk food and sweets. I don't feel or observe any effects, ill or otherwise.

I guess I didn't make the relevant dietary knowledge truly part of myself. It might bite me back later, but right now I don't have ugh-reaction to hamburgers or candy, althrough I do have a mental model of subjective yumminess as a manifestation of adaptation-execution, and I understand that this is far from being fitness-maximizing behavior in today's environment.

Can you think of a simple way to check the territory and discover how broken and unreliable our maps have grown?

I don't try to perform a lot of cognitive calculation except about things that I expect will make a difference. I suppose if I felt like I were working on my rationality-chops Rocky-training-montage style, what you suggest would hold.

Poll: Have you found anti-akrasia tips that have helped you since you found lesswrong?

No/not particularly

I've found anti-akrasia tips that would have helped if I'd actually carried them out.

Might this be called the Akratic Paradox?

I think someone was planning to write a paper with that title.

Uh... oops? Was it someone here on LW? Google isn't turning up much.

(I should probably check to see if my next clever idea is already, you know, a thing before I start assigning Names with Capital Letters.)

Sorry this was already a pretty tired joke by this point. No, as far as I know no one is even procrastinating writing a paper with that name.

I've seen some variation of that joke in an actual academic article on procrastination. I should probably look it up or something.

Ah. Well, in that case, I am hereby stating my intention to write a paper by that name, to be published no later than seventeen years from now. Remember, I called dibs!

I have found anti-akrasia tips that devoured my time while seeming effective. I reckon I may be overdoing it.

How confidant are you of that? It sounds surprising to me, but that could be the typical mind fallacy or just insufficient creativity in imagining what sort of evidence you have.

Anyways, for the purpose of evidence, you should vote no if you have not already.