You… know I don’t optimize dinner parties as focus groups, right? The people who showed up that night were people who like chili (I had to swap in backup guests for some people who don’t) and who hadn’t been over too recently. A couple of the attendees from that party barely even post on LW.
You… know I don’t optimize dinner parties as focus groups, right?
It is perhaps more importantly dinner parties are optimised for status and social comfort. Actually giving honest feedback rather than guessing passwords would be a gross faux pas.
Getting feedback at dinner parties is a good way to optimise the social experience of getting feedback and translate one’s own status into the agreement of others.
If I were to guess, I’d guess that the main filter criteria for your dinner parties is geographical; when you have a dinner party in the Bay area, you invite people who can be reasonably expected to be in the Bay area. This is not entirely independant of viewpoint—memes which are more common local to the Bay area will be magnified in such a group—but the effect of that filter on moderation viewpoints is probably pretty random (similarly, the effect of the filter of ‘people who like chili’ on moderation viewpoints is probably also pretty random).
So the dinner party filter exists, but it less likely to pertain to the issue at hand than the online self-selection filter.
The problem with the dinner party filter is not that it is too strong, but that it is too weak: it will for example let through people who aren’t even regular users of the site.
You… know I don’t optimize dinner parties as focus groups, right? The people who showed up that night were people who like chili (I had to swap in backup guests for some people who don’t) and who hadn’t been over too recently. A couple of the attendees from that party barely even post on LW.
It is perhaps more importantly dinner parties are optimised for status and social comfort. Actually giving honest feedback rather than guessing passwords would be a gross faux pas.
Getting feedback at dinner parties is a good way to optimise the social experience of getting feedback and translate one’s own status into the agreement of others.
FWIW, I eat chili but I don’t think the strongest of the proposed anti-troll measures are a good idea.
If I were to guess, I’d guess that the main filter criteria for your dinner parties is geographical; when you have a dinner party in the Bay area, you invite people who can be reasonably expected to be in the Bay area. This is not entirely independant of viewpoint—memes which are more common local to the Bay area will be magnified in such a group—but the effect of that filter on moderation viewpoints is probably pretty random (similarly, the effect of the filter of ‘people who like chili’ on moderation viewpoints is probably also pretty random).
So the dinner party filter exists, but it less likely to pertain to the issue at hand than the online self-selection filter.
The problem with the dinner party filter is not that it is too strong, but that it is too weak: it will for example let through people who aren’t even regular users of the site.
That’s kinda the point.