Open Thread May 2018

by Elo1 min read1st May 201821 comments


Open Threads

If it’s worth saying, but not worth its own post, then it goes here.

Notes for future OT posters:

  1. Check if there is an active Open Thread before posting a new one (use search for Open Thread ).

  2. Monthly open threads seem to get lost and maybe we should switch to fortnightly.

  3. what accomplishments are you celebrating from the last month?

21 comments, sorted by Highlighting new comments since Today at 7:55 PM
New Comment

Quick meta note

[j/k this wasn't quick]

We've had a sort of awkward relationship with Open Threads for the past few months (previously our policy has been "they are fine but shouldn't be on frontpage."). There are some reasons habryka and I are both a bit skeptical about Open Threads as a thing.

tl;dr: We chatted and decided, despite still having those reservations, to try out a change in policy and leave the May Open Thread on the frontpage (without any of the usual rules about comments, i.e. talking about meta/community stuff is fine.)

I'm more hesitant about talking about Mainstream Politics because I have a sense that'd attract Random Internet Denizens rather than rationalists, and a bit wary about explicit calls to action since that might incentivize people turn use the Open Thread as a giant billboard. The line here is still a bit blurry.

...Some off-the-cuff thoughts on my own reservations about Open Threads:

My experience with them on old LW is that they're a weird mishmash of topics that mostly seemed... kinda meh? It didn't feel like a place where fun casual conversation was happening. It felt like a place where kinda-meh-ideas went to die, and contributed to my overall sense of old LW being a withered ghost town.

Previous open threads on LessWrong2.0 seemed like almost all the discussion there was about "how great it would be if there were an OpenThread", as opposed to any actual interesting conversation happening. But, to be fair, Old LW had a bunch of other problems, and on New LW, it's not really a fair test of their value if they're only visible to many people via the Recent Discussion section.

[Edit: though I note, open threads that were more goal oriented "let's have a conversation about X" seemed more often fruitful to me]

Open Thread vs Shortform/CasualLongterm, Open Threads don't make sense to me as the mechanism by which people have off-the-cuff conversations, ask questions, say things that don't quite fit in other threads, etc. They feel like a makeshift solution for a forum that hasn't yet built better tools for any of those things. Solutions that seem more sensible to me might include:

  • A dedicated, integrated IRC platform for LW where people can drop in, hangout, ask questions etc.
  • A dedicated version of the Shortform Feeds that some people have been experimenting with. (These also feel like a makeshift solution for a forum that hasn't yet built a real version of the thing, and have some pluses and minuses compared to Open threads. Although in this case I think the tools that need building to make it "not makeshift" have less to do with the feeds themselves, and more to do with our Recent Discussion section and/or subscription options).

But, those tools won't be built for awhile. My current take is that having one dedicated slot for casual conversation on the frontpage makes sense as a way to handle that need. We were avoiding leaning into that because it wasn't our intended longterm solution. I currently think it was a mistake to have a few month period where that meant there weren't really any good solutions at all.

(I think shortform feeds could end up working in a way that gets a best-of-both-worlds, where there's a view that lets you see everyone's current shortform conversations that feels kind of like an open thread, while still being able to then look at an individual person's feed to see all their off-the-cuff thoughts. The Recent Discussion section is sort of trying to do that but isn't quite pulling it off)

It feels to me that I lack a good way to subscribe to the short-form feeds of other people.

I'd be interested in a dedicated version of Shortform Feeds.

The blogosphere equivalent of this was the main/sideblog setup -- think SSC and Scott's Tumblr. This seemed to work well for a lot of people, myself included: if you have something that isn't quite substantial enough for a 'main' post, a quote from a book that you might want to link to later, or something like that, you just put it on your sideblog.

This might have been what Main vs. Discussion was intended to be on old LW, but it obviously didn't work out like that: insofar as something like the setup existed on LW, it was Discussion vs. Open Thread.

These feeds wouldn't be places for fun casual conversation, but that's what the asteroid belt of largely invite-only chatrooms and satellite forums are for, at least if the Age of Forums is any guide. (These days it'd probably be social media instead of satellite forums.) Communities tend to grow their own places for fun casual conversation -- I don't think this has to be engineered, unless the point is to ensure that at least some of it happens on If they're set up right, I think they'd be a lot like what existing sideblogs are like, which would IMO be a Good Thing.

I'm not a UI guy, so take this with a Dead Sea or two of salt, but my guess is that you'd need sideblog feeds to only be reachable from a link on someone's userpage for this to work. If there's a firehose, people might worry about polluting it; if sideblog posts are directly visible from userpages, people might worry about keeping their userpages looking pretty and respectable and so on. (In the traditional setup, main blogs are easily discoverable from sideblogs but not vice versa. That's trivial if you're on Tumblr or, but it probably wouldn't mesh well with how things work here, and I doubt it's strictly necessary. But some degree of distance from the agora probably is.)

Quick thought: since the Berkeley community currently has about Dunbar's number of people, shouldn't we want this number to remain stable? Of course new rationalists want in on all the fun and all the impact, and currently it looks like Berkeley is their only option. Should we be focusing on a second hub that aims to rival Berkeley in it's size and awesomeness?

What places should we be recommending? I'm thinking London or Berlin. Any other options?

I think London has a lot of potential that I sometimes worry is going untapped because I'm not a great organiser. But our assets include:

  • Meetups that currently draw 15+ people more often than not
  • A group house that sometimes hosts those meetups (and has also hosted two secular solstices)
  • An EA community that seems to be fairly vibrant
  • In terms of serious organisations, I'm not sure if Deep Mind counts but it feels at least adjacent? And for people who want to do politics-related things, it surely doesn't hurt that we're the seat of the UK government.

I feel the same about Sydney. I have been meaning to strategise for more attendance and a bigger meetup.

As far as I understand the Berkeley community has more than Dunbar's number of members and I don't think that's a real problem. Not everybody has to know each other.

Even in a city like Berlin not everybody in the LW/EA sphere knows each other.

With 263 people our LW Berlin mailing list is also over Dunbar's number and our meetup group has also has 339. Not everybody who subscribes to either lives in Berlin but it shows how you always have many people who are somehow connected even if they don't come to the regular meetup.

Villages and cities has synergy that gets created even with much more than Dunbar's number of people.

That said, I think Berlin is great and I would be happy if you and other people move here (I'm currently organizing our main monthly meetup).

I think Berkeley can afford to have up to 3 Dunbar worth of Rationalists without spiraling out of control.

Ideally, there would be three separate social "domains" that people could compete within, with some crossover spaces to facilitate cross-pollination.

And right now we have that. If we actually directly split the community into X-Risk, EA, and Community / Self Improvement, I think most people wouldn't feel too much of a shakeup in their tribal configurations.

Maybe that’s the Berkeley optimum, but is it the optimum from an international perspective? My intuition is that 3 dunbar size communities in different cities is still better

From within the Bay's vantage point (and I think also for basically every city that's stuck at the "small meetup stage"), a related thing I'm worried about is what to do about the smaller group structure. i.e. meetup-esque things tend to scale to around 12-20 people before they get awkward.

I have more thoughts on this, still mulling it over. Some of it is relevant to the REACH post and I may write the rest of the thoughts there.

The last meetup that I ran that had that people before they get awkward.

Running a meetup with 15 people is different than running a meetup with 30 people but I see no reason why 30 is awkward.

The last meetup that I ran that had that attendance of around 30 people went well. I see no reason why that format wouldn't scale to 40 people as well.

Sure, but if people ever want the intimacy effects of a 15 person meetup, you either can’t do it, or weird social dynamics come into play about who gets to be in those 15 people.

I also often haven’t had access to spaces that fit more than 20 people, or getting them is at least more work.

Should we be focusing on a second hub that aims to rival Berkeley in it's size and awesomeness?

I take Amazon's HQ2 as evidence that this general class of structure is workable.

I'd like to see one on the East Coast, but people here tend to end up in Berkeley.

I think creating a hub requires there to be a critical mass of Serious Organizations™ that give an outlet for people who have ramped up the agency-ladder to "community leader", who are then looking to move into "do something that seems more directly impactful than community leadership".

NYC used to have MetaMed, Givewell and Leverage. Boston has FLI, although I'm not sure what the current state of that is. The problem is networking effects compound (i.e. network effects dragged Givewell and Leverage over to the Bay)

I tentatively think Boston makes most sense as East Coast Hub because the existence of Harvard and MIT creates an intellectual locus that's easier to build around. (This necessarily means that the sort of organizations you can expect to stick around in Boston, or wherever you're building, needs to be oriented in such a way that the local network is more useful/relevant than other city's networks)

To be a little cynical, Berkeley has the community-hub advantage of imposing a strong selection effect: it's far from everything that isn't the West Coast, it's hideously expensive, and as a city it isn't all that great -- I know New Yorkers who tried to move out there and came back with a litany of horror stories to rival those of my friends in Baltimore. So only the hardcore (or people competent enough to land a SF tech industry job) move out there.

The East Coast, on the other hand, has a lot of nice cities and, for most Americans, isn't so far away. I moved to Boston from a location that's pretty far away in east-of-the-Appalachian-range terms, but I could still take a day trip (by plane) to visit my parents, which I couldn't do if I lived in Berkeley.

As much as I like Boston, I think there's an important advantage DC has over it: where Boston has students, DC has people who opted to take safe, cushy government jobs and now have a lot of intellectual energy and no channel for it. (It also has the advantage of being Where The Government Is, which might be important at some point, for some purposes or other.) And, while it's a worse city than Boston in many respects, there's more to do -- if I lived in DC, I'd try to put together a group for going to free concerts (of which there are many in DC) and so on, but I don't know of anything like that here.

I picked Boston because there was at least one existing rationalist org there (although I don’t know if there’s actually much opportunity for new community members getting involved)

so far from everything that isn’t the East Coast.

This seems transparently... failing to notice the entire west coast exists?

This seems transparently... failing to notice the entire west coast exists?

It exists, but it's less populous.

The Northeast and South together (by Census Bureau definitions) contain 55% of the population of the US. The West contains 24%, and the Midwest contains 21%.

But the West extends as far east as Colorado, and also contains Alaska and Hawaii, which should be excluded here; and the Midwest contains states like Ohio and Michigan, which aren't all that far out.

Unfortunately, Wolfram Alpha can't tell me how many people live within whatever distance of Berkeley or DC, so I have to ballpark. A thousand miles seems like a reasonable number -- it's about a thousand miles from Boston to Atlanta, and that's about a three-hour flight. While you'd have to sleep on your parents' couch overnight, it wouldn't be a _huge_ excursion the way a cross-country flight is.

For Berkeley: the total population of California, Oregon, Washington, Nevada, Idaho, Utah, and Arizona is about 65 million people. (I can't even use Census _divisions_ here -- Alaska and Hawaii are in the same division as the West Coast.)

And for DC: the total population of the Northeast Census region, the South Atlantic and East South Central divisions of the South Census region, and the East North Central division of the Midwest census region is about 186 million people.

The list of states could be quibbled with -- maybe Colorado should count for the West, maybe Missouri should for the East -- but I doubt it'd make much difference. The underlying factor here is that population density drops off sharply a few meridians before 100° west and doesn't pick up again until you hit the Pacific.

Fair, I suppose. But... doesn't really feel that compelling to me as a reason. I rarely left NYC when I live there. (Also, my overall experience living in the Bay so far has been comparably expensive and nice to NYC). Distance between cities just doesn't feel like a major motivation to me.

Maybe my experience is unique, but I think the much more obvious explanation is network effects, first from Silicon Valley existing (I think this is the main reason that Givewell and Leverage moved), and then from the community growing up around it.

Does DC currently have a LW community? If so, how large is it?