Prevention over removal. Old LW required a certain amount of karma in order to create posts, and we correspondingly didn’t have a post spam problem that I remember. I strongly believe that this requirement should be re-introduced (with or without a moderator approval option for users without sufficient karma).
Agreed. Also, in the slightly longer term, there must be automated spam detection services that could be incorporated or hired to reduce the moderators’ spam-filtering work load? (If not, it seems like a business opportunity for someone.)
FYI, we’ve added both of these since this post, although they don’t affect greaterwrong yet since they use different views than we do. (An upcoming patch will change that)
Does that mean spam will still show up if I use the “All” view in GreaterWrong? I use that all the time to catch people’s personal posts… Although if you’ve also added automated spam filtering then perhaps that’s not a big deal anymore.
Yes, that means spam just shows up regularly in the All view. (In general, changes made to LW views do not affect greaterwrong, although views based on manual-curation still work because we don’t put spam on frontpage or curated)
The initial version of the “must be approved by admin” filter just applied it to the LessWrong frontpage views (among other things, because we wanted new users to at least be able to find their posts on their personal page).
The upcoming patch moves the filter into the default view (so it’ll affect GreaterWrong), and then instead manually removes it from the user profile view.
The spam filter in general seems to be working okay-but-not great (part of why we went ahead and implemented the “must be approved by admin” requirement)
Admins receive notifications for all new users so going forward I don’t expect spam posts to be too much of a problem.
We haven’t yet implemented the “admin approval” requirement for comments, largely because there isn’t currently a wave of comment spam so we just haven’t gotten around to it, but if comment spam becomes a problem we’ll prioritize it.
Prevention over removal. Old LW required a certain amount of karma in order to create posts, and we correspondingly didn’t have a post spam problem that I remember. I strongly believe that this requirement should be re-introduced (with or without a moderator approval option for users without sufficient karma).
Agreed. Also, in the slightly longer term, there must be automated spam detection services that could be incorporated or hired to reduce the moderators’ spam-filtering work load? (If not, it seems like a business opportunity for someone.)
FYI, we’ve added both of these since this post, although they don’t affect greaterwrong yet since they use different views than we do. (An upcoming patch will change that)
(Instead of minimum karma people just can’t have their posts appear on the home page until they’re approved by an admin)
Does that mean spam will still show up if I use the “All” view in GreaterWrong? I use that all the time to catch people’s personal posts… Although if you’ve also added automated spam filtering then perhaps that’s not a big deal anymore.
Yes, that means spam just shows up regularly in the All view. (In general, changes made to LW views do not affect greaterwrong, although views based on manual-curation still work because we don’t put spam on frontpage or curated)
The initial version of the “must be approved by admin” filter just applied it to the LessWrong frontpage views (among other things, because we wanted new users to at least be able to find their posts on their personal page).
The upcoming patch moves the filter into the default view (so it’ll affect GreaterWrong), and then instead manually removes it from the user profile view.
The spam filter in general seems to be working okay-but-not great (part of why we went ahead and implemented the “must be approved by admin” requirement)
Admins receive notifications for all new users so going forward I don’t expect spam posts to be too much of a problem.
We haven’t yet implemented the “admin approval” requirement for comments, largely because there isn’t currently a wave of comment spam so we just haven’t gotten around to it, but if comment spam becomes a problem we’ll prioritize it.