The standard policy for blogs and online forums is for everyone to be free to add comments unless they repeatedly violate rules against swearing or personal abuse. In the past I have taken this approach on my personal blog and Facebook profile and so only blocked a handful of people over many years. This policy ensures that all comments, even those judged negatively by the original author, can be found somewhere in the resulting thread. But it has some major downsides, and I now wonder if it was a mistake.
if I t ain't broke don't fix it!
Coming back to this discussion, I am struck by) Robert Wiblin's smug confidence that he is capable of judging the quality of posts (without demonstrating any basis for so supposing) and 2) the lack of almost any recognition that what is called "quality" on a site like this has ideological agreement as its principle component..
The complaints about "quality" come exclusively from political rightists, and they are implicitly directed at liberal or socialist Commenters. There are some reasons for this that I can't completely disparage. "Dissident" Commenters tend to comment more; it's easier to contribute to a discussion when you disagree. (Also, Comments will generally tend to be a little "meaner" across ideological lines.) Aversion to this political trend among Commenters is the sole reason for Robert Wiblin's post. He hides this reality under a mask of intellectual overconfidence and disparagement of opponents (as slackers, not "busy people.")
LW is organized to exclude those with different ideologies. I can't say that's necessarily a bad idea, but it should be done honestly (not LW's way, either), not by pompously claiming the mantle of expert when you're really a greenhorn.
Now, take this post. To the point, it is one that Wiblin would like to exclude. I think it makes new points that needed to be stated. Would it be better to exclude its abrasiveness? A supportable contention, but it wasn't really examined because Robert Wiblin was too concerned with claiming the mantle of quality and signaling that he's a busy person.
Great suggestion. This is also suggested in Kevin Kelly's "Out of Control," in several areas, discussing feedback. I highly recommend that optimizing different kinds of emergence be the goal of all posting, and taking advantage of higher and higher levels of emergent order, and being able to ascend and descend the hierarchy based on threshold-based filters. Several filters should be able to be applied. (Positive, Negative total rank above and below a threshold, but also total "disagree with politics rank," or neutral of judgment categorized from dropdown of various philosophies, "objectivist-categorized," "collectivist-categorized," "self-political-categorization" "self-political-categorization of all comments mismatches community assessment of remark's philosophy by over 10 points" etc.)
To some extent, LW already does this (allows one to hide comments below a threshold, which assumes that longtime commenters will be smarter, better judges than raw emergence, etc.), but having several of my highly critical, but high-utility and rationally prioritized posts downgraded and hidden by people who simply put in a lot of time at LW has repelled me. It's a garden that has a lot of beautiful flowers planted by the gardener (Eliezer) but also a lot of people stomping on things they think are "weeds" (the symbiotes of those flowers).
It's enough for me that interested artilects scan the comments at some future point. I don't need to impress a bunch of stupid humans who can't even prioritize keeping themselves out of sociopath deathcamps (and are too intellectually cowardly to even discuss the subject with reference to reality). Seems to me, some of Marvin Minsky's "negative expertise" is called for on LW. My guess is that one only gets significant "negative expertise" by communicating with a lot of smart people, in both meatspace and cyberspace and different networks of both. Before one does that, the problem can't be clear, because of evolutionary social programming.
I find Caledonian/melendwyr more interesting than the average commenter. He can be overly brief/cryptic sometimes though.
For spectacularly bad comment threads see "Pharyngula". The problem there is not just the quality of the comments but the quantity, people get into lengthy tiffs that end up taking over the threads.
You could have a system where everyone has a limited amount of comments, only a few trusted commentors who have a good track record would have more than this limit. The people who enjoy making poor comments tend to want to reply obsessively to everything; this would keep them away.
Actually, both systems could coexist on a single website.
Imagine a system where users can upvote and downvote, and they can select whether they prefer to calculate the total karma from both upvotes and downvotes, or only from upvotes (or anything between).
I think this cartoon sums it up.
Any ideas what would work better?
It is not for every discussion, but some debates would I think benefit from a tool which could dynamically represent the logical/evidential structure and state of an evolving multiparty debate. I believe that many complex debates get lost in the weeds in part because the available tools (essentially text threads or trees) only very crudely support the logical structure of the arguments.
If anyone is aware of such a tool, I'd love a pointer.
I rarely delete comments at my blog. But I see the troll farms that some other places wind up with, and imagine getting a lot of use out of Hellbanning. Even if they do catch on and get enraged, it's way more work for them to shift computer / IP if enraged than for me to add one to the ban file.
That's too easy to notice and circumvent and would probably enrage people even more than downvotes.
The best strategy of all is to figure out how those people that you want to be banned ended up posting on your blog/forum in the first place and what exactly is it that bothers you about them.
Real "trolls" are extremely rare and I don't think I've seen even one on e.g. LessWrong.
I am always amused how oblivious e.g. people associated with MIRI (formerly known as the Singularity Institute) are to the fact that those people who they perceive to be trolls seem to be solely focused on them rather than e.g. the Future of Humanity Institute. The forgone conclusion that they must be trolls prevents them from arriving at any other strategy than trying to gag them.
Another obvious insight should be that those people falsely labeled as trolls almost never show up at posts which are technical and precise or simply well-argued. The conclusion being that reducing vagueness should make it much harder for anyone other than a real troll to say something that you don't like.
Unfortunately there are too many intelligent and insightful bloggers whose attitude to comments is far less than healthy. There is a trade off to be made that is not at all pleasant to deal with.
Talking to your readers about this was probably a mistake.
I like the "featured comments" idea. All the smart people would soon learn not to read non-featured comments, and the Caledonians of the world could just drop into the void. Or an "author hide" function which is easy for people to override in the preferences, but by default gives people the experience the author wants them to have. Don't trust the author not to hide intelligent comments? You shouldn't be reading an author like that in the first place! (Don't trust *me* with that much power? Stop reading my stuff. Seriously, go away.)
This comment has been posted here because it would get downvoted on Less Wrong.
Hah I think another commenter is enjoying the irony too! :) I think I've managed to restore it.
Now, this is a bit humorous/ironic. My earlier comment has just now (at least, at the moment I am typing this) been replaced with "This comment was flagged for review." Now, I doubt this particular flagging was your personal action, Mr. Wiblin. Um... Right? :-)
"If you find their writing worthwhile ... you can probably also trust them curate comments effectively..."
Aren't there any psychological reasons why people capable of intelligently producing new ideas might nevertheless not be the best at rationally evaluating criticisms of their own ideas? Perhaps some reasons we should be looking at Overcoming?
I think the reason why we shouldn't look too much at the "quality" of comments is a) confirmation bias (no matter how smart he/she is, the author can quickly fall into the pitfall of branding comments they don't agree with as being "bad" or "uninformed") and b) even if you could filter for quality you might end up with so few commenters that the blog runs dry.
If you're going to implement a vote system, don't have a downvote button, only an upvote button (which this blog does right). Also, have indentation for replies (which this site has but only for two replies, maybe that should be five).