I heard a rumor that not that many people are writing reviews for the LessWrong 2019 Review. I know I’m not, haha. It feels like a lot of work and I have other things to do. Lame, I know. Anyhow, I’m struck by how academia’s solution to this problem is bad, but still better than ours!
--In academia, the journal editor reaches out to someone personally to beg them to review a specific piece. This is psychologically much more effective than just posting a general announcement calling for volunteers.
--In academia, reviews are anonymous, so you can half-ass them and be super critical without fear of repercussions, which makes you more inclined to do it. (And more inclined to be honest too!)
Here are some ideas for things we could do:
--Model our process after Academia’s process, except try to improve on it as well. Maybe we actually pay people to write reviews. Maybe we give the LessWrong team a magic Karma Wand, and they take all the karma that the anonymous reviews got and bestow it (plus or minus some random noise) to the actual authors. Maybe we have some sort of series of Review Parties where people gather together, chat and drink tasty beverages, and crank out reviews for a few hours.
In general I approve of the impulse to copy social technology from functional parts of society, but I really don’t think contemporary academia should be copied by default. Frankly I think this site has a much healthier epistemic environment than you see in most academic communities that study similar subjects. For example, a random LW post with >75 points is *much* less likely to have an embarrassingly obvious crippling flaw in its core argument, compared to a random study in a peer-reviewed psychology journal.
Anonymous reviews in particular strike me as a terrible idea. Bureaucratic “peer review” in its current form is relatively recent for academia, and some of academia’s most productive periods were eras where critiques came with names attached, e.g. the physicists of the early 20th century, or the Republic of Letters. I don’t think the era of Elsevier journals with anonymous reviewers is an improvement—too much unaccountable bureaucracy, too much room for hidden politicking, not enough of the purifying fire of public argument.
If someone is worried about repercussions, which I doubt happens very often, then I think a better solution is to use a new pseudonym. (This isn’t the reason I posted my critique of an FHI paper under the “David Hornbein” pseudonym rather than under my real name, but it remains a proof of possibility.)
Some of these ideas seem worth adopting on their merits, maybe with minor tweaks, but I don’t think we should adopt anything *because* it’s what academics do.
Yeah, several those ideas are “obviously good”, and the reason we haven’t done them yet is mostly because the first half of December was full of competing priorities (marketing the 2018 books, running Solstice). But I expect us to be much more active/agenty about this starting this upcoming Monday.
Maybe we have some sort of series of Review Parties where people gather together, chat and drink tasty beverages, and crank out reviews for a few hours.
Maybe that should be an event that happens in the garden?
Maybe we give the LessWrong team a magic Karma Wand, and they take all the karma that the anonymous reviews got and bestow it (plus or minus some random noise) to the actual authors.
Wouldn’t this achieve the opposite of what we want, disincentivize reviews? Unless coupled with paying people to write reviews, this would remove the remaining incentive.
I’d prefer going into the opposite direction, making reviews more visible (giving them a more prominent spot on the front page/on allPosts, so that more people vote on them/interact with them). At the moment, they still feel a bit disconnected from the rest of the site.
I heard a rumor that not that many people are writing reviews for the LessWrong 2019 Review. I know I’m not, haha. It feels like a lot of work and I have other things to do. Lame, I know. Anyhow, I’m struck by how academia’s solution to this problem is bad, but still better than ours!
--In academia, the journal editor reaches out to someone personally to beg them to review a specific piece. This is psychologically much more effective than just posting a general announcement calling for volunteers.
--In academia, reviews are anonymous, so you can half-ass them and be super critical without fear of repercussions, which makes you more inclined to do it. (And more inclined to be honest too!)
Here are some ideas for things we could do:
--Model our process after Academia’s process, except try to improve on it as well. Maybe we actually pay people to write reviews. Maybe we give the LessWrong team a magic Karma Wand, and they take all the karma that the anonymous reviews got and bestow it (plus or minus some random noise) to the actual authors. Maybe we have some sort of series of Review Parties where people gather together, chat and drink tasty beverages, and crank out reviews for a few hours.
In general I approve of the impulse to copy social technology from functional parts of society, but I really don’t think contemporary academia should be copied by default. Frankly I think this site has a much healthier epistemic environment than you see in most academic communities that study similar subjects. For example, a random LW post with >75 points is *much* less likely to have an embarrassingly obvious crippling flaw in its core argument, compared to a random study in a peer-reviewed psychology journal.
Anonymous reviews in particular strike me as a terrible idea. Bureaucratic “peer review” in its current form is relatively recent for academia, and some of academia’s most productive periods were eras where critiques came with names attached, e.g. the physicists of the early 20th century, or the Republic of Letters. I don’t think the era of Elsevier journals with anonymous reviewers is an improvement—too much unaccountable bureaucracy, too much room for hidden politicking, not enough of the purifying fire of public argument.
If someone is worried about repercussions, which I doubt happens very often, then I think a better solution is to use a new pseudonym. (This isn’t the reason I posted my critique of an FHI paper under the “David Hornbein” pseudonym rather than under my real name, but it remains a proof of possibility.)
Some of these ideas seem worth adopting on their merits, maybe with minor tweaks, but I don’t think we should adopt anything *because* it’s what academics do.
Yeah, several those ideas are “obviously good”, and the reason we haven’t done them yet is mostly because the first half of December was full of competing priorities (marketing the 2018 books, running Solstice). But I expect us to be much more active/agenty about this starting this upcoming Monday.
Maybe that should be an event that happens in the garden?
Wouldn’t this achieve the opposite of what we want, disincentivize reviews? Unless coupled with paying people to write reviews, this would remove the remaining incentive.
I’d prefer going into the opposite direction, making reviews more visible (giving them a more prominent spot on the front page/on allPosts, so that more people vote on them/interact with them). At the moment, they still feel a bit disconnected from the rest of the site.