Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Comment author: Qiaochu_Yuan 06 June 2017 06:23:11AM *  3 points [-]

I strongly recommend that no one attempt to use the phrase "expected utility" without understanding, at a reasonable level of detail, the proof of the von Neumann-Morgenstern theorem. For my take on the proof see this blog post. Among other things, understanding the proof teaches you the following important lessons:

  • Utilities can be assigned completely arbitrarily. All the vNM theorem tells you is that a collection of preferences satisfying some axioms ("being vNM rational") is equivalent to a collection of preferences described by maximizing expected utility with respect to some utility function, but it puts no constraints whatsoever on the utility function.
  • The vNM theorem also does not imply that you ought to make decisions by maximizing expected utility, only that if you are vNM rational then your preferences can be described in this way. (Also, humans aren't vNM rational and it's not at all clear that we should try to be, just so we're all clear.)
  • The vNM theorem makes no mention of time or of making multiple decisions; the justification for maximizing expected utility, in this setup, has absolutely nothing to do with long-run averages of repeated decisions, it is in some sense a mathematical trick for expressing certain kinds of preferences and that's it. In the proof of the vNM theorem utility falls out as "that thing which we must be maximizing the expected value of, if we're vNM rational."

The standard way to interpret the relevance of the vNM theorem for an agent acting in the world over time is that your preferences should actually be over world-histories, not world-states, hence if you're vNM rational then your utility function takes as input a world-history and you're maximizing expected utility with respect to probability distributions over world-histories (possibly once, ever: say, when you make a decision at the beginning of time about what you're going to do in all possible futures). Needless to say nobody has ever or will ever do this.

Anyone who's actually interested in formal theories of how to make decisions over time should be learning about reinforcement learning, which is a much richer framework than the vNM theorem and about which there's much more to say.

Comment author: HapiUG 01 June 2017 01:56:55PM 2 points [-]

Anathem. Phenomenal little sci-fi book about theoreticians living in seclusion until called out into the world to deal with a large external threat

Comment author: Qiaochu_Yuan 03 June 2017 01:02:56AM 0 points [-]

I really liked Anathem the first time I read it (sometime in college), but then I read it more recently and completely reversed my position. I don't remember why in very much detail; the whole book just annoyed me a lot more the second time around.

Comment author: sone3d 02 June 2017 07:14:50AM 0 points [-]

Books about human status.

Comment author: Qiaochu_Yuan 03 June 2017 01:00:47AM 2 points [-]

Impro has a nice chapter on status. The rest of it's great too.

Comment author: Qiaochu_Yuan 31 May 2017 12:26:57AM 7 points [-]

I would like everyone posting criticism, especially heated criticism, to keep very firmly in mind that Duncan did not have to write this. Whatever your opinion of him, at least make sure you've factored in the evidence that he wrote this whole, weird thing, complete with references to Ender's Game, Fight Club, etc. instead of writing either 1) nothing or 2) something much more reassuring.

There are critics who think Duncan is incompetent and overconfident, and about this hypothesis I can say at least that it is consistent with Duncan having written this post. Then there are critics who think Duncan is, I dunno, evil or power-hungry or something, and I think those people are mostly failing to see what is in front of them.

Comment author: malcolmocean 28 May 2017 08:30:52PM *  4 points [-]

I want to publicly express my strong support for this experiment/meta-experiment.

I think that my support is particularly noteworthy as I'm presently a core member of a different taking-each-other-seriously co-living experiment that is profoundly different in its philosophy. (Mine is not in Berkeley, nor rationalist.) Therefore some people might assume that I would be opposed to Dragon Army Barracks.

Things in common between the experiment I'm part of and Dragon Army Barracks:

  • is "high-commitment, high-standards, high-investment"
  • is trying to actually make & achieve something together
  • is addressing unanchored abandoned loneliness thing
  • has consciously explicated commitments and assumptions
  • is intended to produce a high-level of consistent excellence and ability to effectively collaborate

Things that are different:

  • We're very far from authoritarian or hierarchical. Although we're also not egalitarian, consensus-based, or even democratic per se... but we have essentially zero of telling-other-people-what-to-do
  • Our basic collective navigating framework is Kegan-5 / fluid mode / post-rational, rather than Kegan-4 / systematic mode / rational (good summary of this distinction)
  • Our focus is almost entirely on the meta-level of building the new cultural platform we're building. We don't have any expectations of each other on the levels of specific object-level projects or explicit behavioral norms (aside from ones necessary for the house's function)

I think that these differences are core to why I am part of this project that I'm part of, and why I consider it to be the most valuable investment I could be making with my time and energy. I am, therefore, non-Berkeley-residence aside, not going to be applying to DA. As I said above though, I strongly support Dragon Army Barracks as an experiment and potentially as an ongoing resource to individual and collective growth.

Reasons why I think that DA is a good idea:

  • Expected value of high amounts of worthwhile object-level output. As Sebastian Marshall says, "the gains made from living more purposefully are forever - the time you've spent well will remains well-spent even if you fall off for a while sometimes. Most people don't even try, which is why most people don't succeed."
  • I expect it will also produce a lot of developmental progress for people involved; that if you were to be able to sort rationalists by amount of growth in a year, the Dragons would all be in the top quartile, and would occupy many of the top 10 slots. This, even if the experiment were to end after 6 months.
  • The DA Barracks is an intervention that is attempting to produce change on a very fundamental level of the system that is a group house. This is a powerful leverage point (see Donella Meadow's article... I would say this is around a 2 or 3, and most group houses have only done mild experiments at the 4-6 level.)
  • I agree with and/or resonate with the six points that Duncan makes in Section 2 of this document.
  • The project-level value of learning here is also very high: this will greatly inform future experiments, whatever their leadership basis.
  • If I had kids, I would absolutely sign them up for any summer camps or classes Duncan was running. I think the amount of power he would have in relation to them would be similar to the amount of power he'll have in this situation.

A final reason is this: I think that we as humanity need to rapidly make progress on being able to effectively coordinate in non-hierarchical ways, which is what the project I'm part of is about. Corollarily, humanity is kind of mediocre at doing this in many contexts. Therefore if non-hierarchical projects aren't emphatically directed towards solving that challenge itself, I expect them to be outperformed by projects that are leveraging existing understanding about how to coordinate effectively in hierarchical ways. i.e. in this case, Dragon Army Barracks.

Comment author: Qiaochu_Yuan 29 May 2017 12:06:18AM 3 points [-]

I really, really wish Kegan levels didn't come in an order, so a claim to be at a higher Kegan level than someone else didn't look so starkly like a claim to superiority. It's turning me off even trying to take them seriously, because everyone who uses them looks like they're just self-aggrandizing to me.

Comment author: Kaj_Sotala 28 May 2017 04:11:22PM 6 points [-]

Hypothesis for why the antibodies exist: People noticed that there were standard methods for creating in-group identification, and these methods were exploited by con artists, advertisers, managers trying to get their employees to work harder, teachers trying to get their students to behave, etc. Antibodies formed in response.

Given that the standard response to "weird" groups that demand cohesion/commitment seems to be "that sounds like a cult", it feels like these antibodies could have developed after the cult scares, which Wikipedia tells me showed up seriously in the 1970s.

Comment author: Qiaochu_Yuan 29 May 2017 12:02:04AM 1 point [-]

Yeah, this is my hypothesis. Vietnam and Watergate probably seriously contributed to a general erosion of trust in authorities as well.

Comment author: Nisan 27 May 2017 08:49:43PM 1 point [-]

I'm not proposing a house policy here. I'm suggesting that a Dragon would do well to have regular followups with someone outside the house, and I'm proposing that some members of the wider community offer to be those someones.

In the past I've had regular video calls with a couple people who were doing long-term experiments with their lifestyle; I think it was helpful. I believe such an arrangement was part of the Leverage polyphasic sleep experiment.

Jacob is right: There's a difference between a friend one can reach out to if one needs to, and a friend one is scheduled to talk to once a week. Personally, I struggle to keep up with friends without scheduled meetings, and it sounds like the Dragon Army will be very busy.

Also, there is a difference between reaching out to a friend when things have gone very wrong and one needs to get out; and bringing up a less drastic problem during a weekly check-in. In the first case, you need a couch to crash on and maybe a lawyer. In the second case, you need someone who will listen to you and bring an outside perspective, and maybe refer you to other resources.

Partially, I'm afraid that if this doesn't go well, our community will lose a cohort of promising people. It would be a shame if that happened because we failed to pay attention to how they were doing.

But also, if the experiment goes very well, this arrangement would be a means by which the wider community can learn from what went right.

Comment author: Qiaochu_Yuan 27 May 2017 10:19:02PM *  0 points [-]

Partially, I'm afraid that if this doesn't go well, our community will lose a cohort of promising people.

I really don't know what you mean by "lose" here (and I'm worried that others will have varying interpretations as well). Do you mean they'll become less promising? Not promising? Leave the community? Go crazy? Die?

Anyway, this seems sensible, but I still want to nudge you and everyone else in the direction of sharing more explicit models of what you think could actually go wrong.

Comment author: jsteinhardt 27 May 2017 05:29:34PM 8 points [-]

I think there's a difference between a friend that one could talk to (if they decide to), and a friend tasked with the specific responsibility of checking in and intervening if things seem to be going badly.

Comment author: Qiaochu_Yuan 27 May 2017 06:56:21PM 0 points [-]

Sure, but what I'd like to know is why Nisan thinks that difference is important in this case.

Comment author: username2 27 May 2017 10:54:10AM 6 points [-]

Romantic entanglements and their fallout are not ruled out by all male environments even if the members do not identify as homosexual. So still important to consider these issues even if there are no women at all.

Comment author: Qiaochu_Yuan 27 May 2017 06:56:07PM 6 points [-]

Can confirm. I was in a fraternity in college with many gay members, some of whom occasionally hooked up and caused manageable levels of drama. This was a relatively recent phenomenon in the history of the fraternity; I think as recently as 10 years before my time nobody was out, and then some people came out after joining.

Comment author: Sinal 27 May 2017 05:24:08AM 2 points [-]

This post makes me miss my days in marching band, or in the Boy Scouts. Honestly it doesn't sound all that authoritarian. Can you not accomplish the same thing using a traditional organization and a meeting place? Why does it have to be a house?

Comment author: Qiaochu_Yuan 27 May 2017 09:00:16AM 5 points [-]

Can you not accomplish the same thing using a traditional organization and a meeting place? Why does it have to be a house?

A couple of reasons occur to me. First, everyone's real goddamn busy. If you already live in a rationalist house and also have a job there's not gonna be a ton of time or attention left in your life for other stuff as big as what Duncan wants Dragon Army to be. Second, Duncan wants people to do things like exercise with each other first thing in the morning before heading off to work, and it seems really annoyingly difficult to coordinate something like this with anyone other than the people you live with.

In general it's just way, way easier to coordinate all sorts of activities with the people you live with than with anybody else. My most direct experience with this was living in a fraternity and seeing the difference between the brothers who did and didn't live in the house; there was a big difference in terms of social accessibility and bonding, and accordingly we strongly encouraged people to live in the house when at all possible.

View more: Next