Pre-commitment and meta at the Cambridge UK meetup

by rebellionkid1 min read29th Apr 201214 comments


Personal Blog

At today's Cambridge UK meetup I made an observation: It seems that LW meetups are very good at having meta-discussions. But they are not so good at acting effectively on them.

The point of meta discussion is to make object level discussion better. The meta questions aren't themselves interesting and dont automatically produce win. Object level discussions are themselves interesting and do produce win (if not then stop talking about boring things). So if one has a meta discussion it should be such that the improvements made to the object level discussions outweigh the cost of the meta discussion.

I notice we have meta discussions which (like a lot of discussions in LW groups) dont resolve themselves into actions. This means that improvements to the meetups aren't in fact implemented. This is a double fail: first because the object level discussion isn't improved, and second because the unresolved meta took resources away from the body of the meeting.

We could cheaply improve this with the internet. Doodle polls solve the problem of when to schedule a meeting far more efficiently than verbal discussion. Likewise the time-consuming question of "what shall we talk about" can be thought about outside the meeting where there are far fewer constraints on time. Both these problems should be outsourced to the google group and not mentioned in the meeting itself.

A point that was raised is that it is very easy for the group to decide that such and such a thing must be done, that does not automatically translate into the actions of specific people. Someone mentioned the parable of the rabbi raising funds, and we started the following pre-commitment game. 

The Napkin

We got out a napkin and Douglas drew a table of "who, what, by when" on it. He was the first to write down a commitment so as to overcome everyone's reluctance to be the first to act. We then went round and asked for commitments that would be made public in front of the group. I'm now posting those commitments online.

  • Douglas: "Post a meeting format to discussion" Wednesday midnight
  • Paul: "Kahneman AD/BC* example on LW wiki" Thursday midnight
  • Paul: "David Styles" Monday midnight
  • Adam: "Post this list, post on meta/object interaction" Wednesday midnight
  • Jonathan: "Directions to JCR" Tuesday midnight
  • Ben: "Keep diary for 1 week, identify biases" next Sunday
*(did I read that right? edit: no I didn't)
 The case of Ben is worth commenting on. At the end of the meeting as everyone was walking out the room I made to fold the list away and shouted "last chance to commit to anything if you want to", Ben took this opportunity to sign up when he had not done so before in the meeting. This may be a generally useful technique.
Those at the meeting: have you completed your task? 

What are easy ways to overcome the reluctance of people to be the first to act?

How can we have meta-discussions that are targeted at concrete actions?


14 comments, sorted by Highlighting new comments since Today at 9:01 AM
New Comment

I've fulfilled my first promise: AD/BC problem. I have punted on the second; I've supplied contact details, but someone who is actually going to be at the meetup will have to invite Styles to speak.

Upvoted, on the grounds that rewarding people fulfilling promises made at LessWrong meetings seems a reasonable usage of LessWrong karma.

Have updated the post, sorry for misreading your writing.

I am amazed you were able to make as much sense of it as you were! Thank you for posting this!

What are easy ways to overcome the reluctance of people to be the first to act?

Bribe people with Jaffa cakes? :-)

Sounds like an (in)joke: but I'm always shocked by the effectiveness in terms of incentivising behaviour etc. of tiny bribes that are meaningless in terms of monetary value. Like cake.

Yes, small bribes can actually be more effective. If you offered £5 for the first person who acted first, I think that would work as a disincentive - people would be embarassed, not want others to think they needed it (or maybe that's just us being Brits?). Whereas a Jaffa cake shows esteem, and lets someone pretend they're just doing it to get the cake.

People enjoy play and small ritual. Making light of something lowers the barriers, doesn't ring the cthulhu-horror-of-being-manipulated bell because it is openly ironic.

{I know "ironic" isn't the right word, but I'll stick with it, because I think people will guess from context what is meant.}

Douglas: "Post a meeting format to discussion" Wednesday midnight
Those at the meeting: have you completed your task?


The post was interesting, but please don't post a separate discussion post again. The clutter from one discussion post per meetup event is justifiable, but more than one is excessive. In short, I respectfully request that you post the content as a comment to a previous discussion post unless the content is of general interest to the online LW community.

FWIW I disagree; I feel we don't hear from meetups enough, and these are two very separate posts.

Better directions to the JCR (with images) are here.

ETA: Also fixed the list of meetups to link there.

5 and 6 link to the same photo.

How can we have meta-discussions that are targeted at concrete actions?

Decide beforehand what you want the meta-discussion to achieve.

So instead of "Discuss how we hold meetings." make it "Come up with three concrete changes to how we hold meetings that we will test out within 1 month from today." or "We spend too long at the start of meetings deciding what to discuss and just saying 'take it to the list' isn't working because lots of people at meetings don't participate on the list. Discuss possible ways to improve on this specific situation, pick one to test out, and measure if it actually reduces initial dither time."