Sure do. On stuff I know a little about, what gets upvoted is “LW folk wisdom” or perhaps “EY’s weird opinions” rather than anything particularly good. That isn’t surprising. Karma, being a numerical aggregate of the crowd, is just spitting back a view of the crowd on a topic. That is what karma does—nothing to do with quality.
Every crowd thinks they are such a place where it’s actually true.
Some of the extreme sceptics do not believe they are much closer to the truth than anyone else.
Outside view: they are wrong.
There does not exist a group such that consensus of the group is highly correlated with truth? That’s quite an extraordinary claim you’re making; do you have the appropriate evidence?
I think Ilya is not claiming that no such group exists but that it is well nigh impossible to know that your group is one such. At least where the claim is being made very broadly, as it seems to be upthread. I don’t think it’s unreasonable for experimental physicists to think that their consensus on questions of experimental physics is strongly correlated with truth, for instance, and I bet Ilya doesn’t either.
More specifically, I think the following claim is quite plausible: When a group of people coalesces around some set of controversial ideas (be they political, religious, technological, or whatever), the correlation between group consensus and truth in the area of those controversial ideas may be positive or negative or zero, and members of the group are typically ill-equipped to tell which of these cases they’re in.
LW has the best epistemic hygiene of all the communities I’ve encountered and/or participated in.
In so far as epistemic hygiene is positively correlated with truth, I expect LW consensus to be more positively correlated with truth than most (not all) other internet communities.
Talking about LW, specifically. Presumably, groups exist that truth-track, for example experts on their area of expertise. LW isn’t an expert group.
The prior on LW is the same as on any other place on the internet, it’s just a place for folks to gab. If LW were extraordinary, truth-wise, they would be sitting on an enormous pile of utility.
The prior on LW is the same as on any other place on the internet.
I disagree. Epistemic hygiene is genuinely better on LW, and insofar as Epistemic hygiene is positively correlated with truth, I expect LW consensus to be more positively correlated with truth than most (not all) other internet communities.
Presumably, groups exist that truth-track, for example experts on their area of expertise.
A group of experts will not necessarily truth-track—there are a lot of counterexamples from gender studies to nutrition.
I would probably say that a group which implements its ideas in practice and is exposed to the consequences is likely to truth-track. That’s not LW, but that’s not most of the academia either.
Sure do. On stuff I know a little about, what gets upvoted is “LW folk wisdom” or perhaps “EY’s weird opinions” rather than anything particularly good. That isn’t surprising. Karma, being a numerical aggregate of the crowd, is just spitting back a view of the crowd on a topic. That is what karma does—nothing to do with quality.
What if the view of the crowd is correlated with quality.
Every crowd thinks so.
I think Lesswrong might be (or at the very least was once) such a place where this is actually true.
Every crowd thinks they are such a place where it’s actually true. Outside view: they are wrong.
Some of the extreme sceptics do not believe they are much closer to the truth than anyone else.
There does not exist a group such that consensus of the group is highly correlated with truth? That’s quite an extraordinary claim you’re making; do you have the appropriate evidence?
I think Ilya is not claiming that no such group exists but that it is well nigh impossible to know that your group is one such. At least where the claim is being made very broadly, as it seems to be upthread. I don’t think it’s unreasonable for experimental physicists to think that their consensus on questions of experimental physics is strongly correlated with truth, for instance, and I bet Ilya doesn’t either.
More specifically, I think the following claim is quite plausible: When a group of people coalesces around some set of controversial ideas (be they political, religious, technological, or whatever), the correlation between group consensus and truth in the area of those controversial ideas may be positive or negative or zero, and members of the group are typically ill-equipped to tell which of these cases they’re in.
LW has the best epistemic hygiene of all the communities I’ve encountered and/or participated in.
In so far as epistemic hygiene is positively correlated with truth, I expect LW consensus to be more positively correlated with truth than most (not all) other internet communities.
Doesn’t LW loudly claim to be special in this respect?
And if it actually is not, doesn’t this represent a massive failure of the entire project?
Talking about LW, specifically. Presumably, groups exist that truth-track, for example experts on their area of expertise. LW isn’t an expert group.
The prior on LW is the same as on any other place on the internet, it’s just a place for folks to gab. If LW were extraordinary, truth-wise, they would be sitting on an enormous pile of utility.
I disagree. Epistemic hygiene is genuinely better on LW, and insofar as Epistemic hygiene is positively correlated with truth, I expect LW consensus to be more positively correlated with truth than most (not all) other internet communities.
A group of experts will not necessarily truth-track—there are a lot of counterexamples from gender studies to nutrition.
I would probably say that a group which implements its ideas in practice and is exposed to the consequences is likely to truth-track. That’s not LW, but that’s not most of the academia either.
I don’t think LW is perfect; I think LW has the best epistemic hygiene of all communities I’ve encountered and/or participated in.
I think epistemic hygiene is positively correlated with truth.