There's a couple interesting things about the 2011 LessWrong post, "Approving reinforces low-effort behaviors."

The post itself is fascinating to re-read — there's some valuable and profitable ideas in there.

But two sentences also stood out to me on more of a meta-level, and prompted some interesting thought —

>"Unlike the first class of behaviors, we expect to experience akrasia when dealing with this latter sort."

>"This process is especially important to transhumanists."

Reading that post made me remember what LessWrong was like in 2011 — discussions on akrasia ("acting against one's better judgment") and transhumanism were much more common.

Both of these concepts have, to some extent, cycled out of the common discussions on rationality. They still happen, sure, but much less frequently.

On the other hand, certain lines of thought and concepts seem to have become part of the fundamentals of LessWrong and the broader rationalist community. When I mean by "fundamentals" here is more like in sports than in religion — dribbling is a fundamental skill in basketball, and gets practiced by all basketball players.

It's interesting to reflect on why this might happen with a given set of concepts in a community. Akrasia certainly seems like it could become a fundamental area of constant study for people in this space, and yet it did not to the extent that, say, cognitive biases did.

Cognitive biases became a fundamental piece of the rationalist community and has been stably so, whereas it seems like akrasia was more of an intellectual fashion for a time.

I started thinking on why. Here's a few reasons I came up with —

  1. "Solved Problems" are more likely to become fundamentals than areas with disputed or inconsistent solutions. This seems obvious. The Conjunction Fallacy is a cognitive bias that's very common. With some (very basic) statistics and probability theory, it's very clearly proven why one's natural intution in that area is mistaken. After learning it, you make those mistakes less often going forwards. Compare to akrasia which has been around thousands of years — we're using the Ancient Greek word for it, after all — and which hasn't been anywhere close to fully solved.
  2. Ideas that are at the "bottom of tech trees" are more likely to become fundamentals. Game theory has many practical applications for individual decisionmaking, group coordination, predicting the behavior of others, incentives, etc. So game theory becomes a fundamental, similarly to how arithmetic is a fundamental in mathematics. These are necessary to build upon. Whereas ideas at the "top of a tech tree" are more likely to be explored in a cyclical fashion — a topic like, say, designing cities from scratch could conceivably become popular for a while but is rather unlikely to become a core fundamental area of study here, even if there were some "solved problems" in that branch of study.
  3. Current events likely drive some topical interest. It seems politics and communication are both more fashionable here than in the past, which is probably because we're in a somewhat strange and surprising political climate.
  4. We're possibly more likely to entertain certain ideas in optimistic or pessimistic periods, and that drives some fashionability of discussions. Transhumanism is an inherently very optimistic idea. I think it'd be fair to say that much of the world at large is in a slightly more pessimistic environment right now than it was in 2011. That might make it harder to get mainstream traction and positive feedback loops on transhumanism, but might make it easier to get traction and discussion on AI Safety and Existential Risk.
  5. Randomness and variance. A great post goes up on a day that a lot of people are logged into LessWrong for some reason, a lot of people read it and comment and go on to write on that subject too, and it becomes a popular subject for a while. Some randomness might drive ideas into fashionability for a while, or not, just due to unpredictable chance around who is logged in and reads a given post or not, independent of the post's quality. Even if an idea was more tangential and not really critically fundamental, perhaps if it had a few "randomly boosted" cycles of importance, maybe it'd get a lot of thought and interest and become a fundamental topic. This last point is particularly interesting and perhaps even a little scary to behold — how much of what becomes an intellectual lineage is dominated by randomness? Maybe moreso than we'd think. Survivor's bias and all that, too.

That's not a comprehensive list, but I think it might kick off an interesting conversation.

Why do some ideas in a community — like ours, and other communities — merely cycle through fashionablity and unfashionablity, while other ideas become a fundamental and enduring part of the discussion that all new members are encouraged and generally expected to learn?

An interesting topic, no? Your thoughts?


New Comment
15 comments, sorted by Click to highlight new comments since: Today at 6:08 PM

I think it's straight up a mistake to think in terms of akrasia and I'm glad we stopped talking about it. The optimistic story is that people recognized that talking about akrasia was not solving their akrasia so they talked about something else instead.

(It would take a top-level post to explain my thoughts here, but for starters, I think every time you feel tempted to diagnose yourself with "akrasia" you should instead say "I notice that I didn't do this thing even though I claimed I wanted to." I think this is a much more productive starting point.)

Incidentally, I agree with you. You want to 80/20 your top-level thoughts on it in a comment, though? Sounds like a good thought.

Akrasia is a socially convenient narrative: "I, the real me I identify with, wanted to do this thing, but some mysterious thing, which is not the real me and which I don't identify with, prevented me from doing it." One way of describing what's happening is that you're identifying with your System 2 and distancing yourself from your System 1 so you can lay the blame on it.

But your System 1 is where your motivation comes from, especially your deepest motivation; it's where you love from, it's where you defend your loved ones from, etc. It is you, not just a subsystem of you that you have to wrangle. You are the elephant too, not just the rider.

So you don't "have akrasia," you didn't want to do the thing, and there's some social / psychological weirdness around admitting that fact to yourself or others. You can further try to figure out why you didn't want to do it and whether you could want to do it later, but that's secondary to just admitting to yourself that you didn't want to do the thing.

While I 100% agree with your views here, and this is by far the most sane opinion on akrasia that I've seen in a long time, I'm not convinced that so many people on LW really "get it". Although to be sure, the distribution of behavior that signals this has significantly shifted since the move to LW2.0.

So overall I am very uncertain, but I still find it more plausible that the reason why the community as a whole stopped talking about akrasia is more like, people run out of impressive-seeming or fresh-seeming things to say about it? While the minority that could have contributed actual real new insights turned away for better reasons.

Right, that's why I labeled the above "the optimistic story." The pessimistic stories were left as exercises to the reader.

Edit: I rewrote this to use "Alice" and "Bob" instead of "you" and "me" as characters to clarify that it's a thought experiment and not a question about Less Wrong user arundelo (though it is inspired by actual events). I also added a paragraph at the end.

Let's say Alice asks Bob why he didn't watch the most recent episode of $TVSHOW and he says, "I didn't feel like it", and she asks for more detail. He might tell her that he doesn't really like $TVSHOW, or that he likes it but wasn't in the mood and maybe will watch it tomorrow.

Now let's say she asks him why he didn't work today and he says, "I didn't feel like it", and she asks for more detail. He might tell her that he decided to take a day off because he's been working a lot lately, or because the weather was nice and he wanted to spend the day hiking.

All these responses seem pretty similar compared to Bob telling Alice, "I don't know why I didn't feel like working. I guess work is hard and I'd rather goof off. Or maybe I have some sort of subconscious fear that if I do work I'll prove that I'm stupid or incompetent. But I've only worked a couple hours so far this week and I get paid by the hour, and I'm afraid I'm gonna be late to pay my rent again, and my landlord told me if I'm late again she's going to file an eviction notice."

The most important thing is solving the problem, which may involve figuring out if Bob does have a subconscious fear of failure or whatever. But when I use words like "akrasia" or "procrastination", I'm using them as shorthand for long descriptions like the one in the previous paragraph.

Is it really worthwhile for Bob to avoid the words "akrasia" and "procrastination"? If so, should his short answer to "Why didn't you work today?" really be "I didn't feel like it"? Or is there something better?

Yes, I think it is really worthwhile for Bob to avoid the words "akrasia" and "procrastination," and that the short answer "I didn't feel like it" is better.

It's an important feature of this scenario that Bob must work in order to survive, which the akrasia / procrastination frame masks; poetically, he is a slave to Moloch, and it's important that he uses language that clearly distinguishes what he wants (which is to not work) from what Moloch wants (which is to continue his enslavement).

(A mantra for Bob: is it akrasia or am I a slave?)

Thanks for the response! (I've seen you say similar stuff about "akrasia" once or twice before and had been meaning to ask you about it. I'll think about this.)

("Meditations on Moloch" link for anyone who didn't understand the reference.)

I think the immediate problem for most people isn't akrasia, but straight up addiction to shiny screens. It's not as harmful as other addictions though, apart from being time-consuming. And it's hard to quit cold turkey because many jobs involve shiny screens as well. The only remedy is finding substitute activities that will make you happier. That depends on your tastes, so it makes sense that general advice has dried up (except "hit the gym" which is a good idea for everyone).

I'm going to think about this (he said, staring at his shiny screen).

I think my own view is something like the a combination of this, and Qiaochu's point. I've never found the tag "Akrasia" that helpful. I do think I often find myself working at cross purposes, but almost all of those times fall into the "shiny screen is addictive" category.

Yep. Uninstalling Facebook messenger is more rational than reading yet another scientific article on hyperbolic discounting.

Agreed with other comments. Said differently - Akrasia is an applause light word. You put a block up to solving the problem any time you call it Akrasia. Mainly because there are several possible causes of Akrasia and resting on the name makes you stop looking.

As for transhumanism - it seems trivially true and musing is not helpful now. We all know that the future is rapidly approaching. We do wonder about it but there are more pressing things.

Or maybe no "The Book"s have been written on these things that people no longer talk about.

In that case, sooner or later there will be The Book, and we'll be talking about it again.

Or maybe biases started separately and got concerted later, akrasia started as a hydra with many heads and then just got propagated into many unnamed things nobody here is going to rigorously refute or support with a study.

In that case, the unnamed things might segregate into more natural categories (or maybe they already have and it is simply false that akrasia doesn't get talked about).

Or maybe new people came who were never told the reason why the earlier discussions should be interesting.

In that case, I expect biases to come out of fashion after a while.

I think in both of those cases there are good reasons again using the words. Tabooing them is helpful. In the case of akrasia there are many different reasons why people don't do what they want to do and mixing them up under the term of Akrasia doesn't help.

In the case of transhumanism, it's a bad idea for different reasons. It invites belief in belief. I think as this community got thinking more seriously about the dangers of AI the blind trust in believing that moves past being a human is unambigiously positive disappeared.

New to LessWrong?