01:37 < Athrelon> http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/ my prescription is way way more elitism
01:38 < mporter> and finally, a secretly famous tsarist reactionary jurist http://en.wikipedia.org/wiki/Konstantin_Pobedonostsev
01:38 < Athrelon> first-rate science needs first-rate minds at the helm, not "slightly cleverer institutional rules"
01:39 < mporter> athrelon interesting, i don't think i've seen anyone from the neoreo cluster develop a specific alternative philosophy of scientific organization
01:39 < Athrelon> mporter moldbug did it first
01:40 < mporter> antiversity?
01:40 < Athrelon> no, more object-level
01:40  * Athrelon finds link
01:40 < Athrelon> https://news.ycombinator.com/item?id=3819346
01:42 -!- JenRM|afk [~jrodrigu@108-74-165-2.lightspeed.sntcca.sbcglobal.net] has quit [Ping timeout: 252 seconds]
01:42 < Athrelon> brb
01:55  * mporter tries to grasp what moldbug's model is
01:56 < mporter> his hypothetical example is a billionaire with cancer who uses his resources to fund multiple teams who generate and test hypotheses rapidly
01:58 < mporter> this scenario has a number of features, any of which might be selected as salient
02:00 < mporter> is the idea that science would work better as an autocratic command economy, rather than as byzantine bureaucracy dispensing state patronage?
02:03 < mporter> i find athrelon's own maxim clearer [18:38] <Athrelon> first-rate science needs first-rate minds at the helm, not "slightly cleverer institutional rules"
02:06 < mporter> directly applies a reactionary principle (good government needs good governors) to the scientific domain, and contrasts it with an alternative philosophy (that appropriate institutional design is the answer; compare the idea that "rule of law" is the key to good government)
02:07 < mporter> if the dying billionaire in moldbug's scenario is meant to be an example of a first-rate mind in charge...
02:07 < um> "laws rather than men" i think is the relevant fragment
02:08 < mporter> ...it seems that we are talking about excellence as an executive or general-purpose decider, rather than excellence as a scientist
02:10 < mporter> i suppose one implication of neoreo governance is that science shouldn't be left to itself
02:11 < mporter> so even if science as such is organized autocratically, with the best scientists in charge, the best governors will still ultimately be in charge of them, because the best governors should be in charge of everything
02:13 < um> i want to say something about aligned incentives but i'm not sure what would be correct
02:16 < Athrelon> back
02:17 < Athrelon> mporter excellent questions and the meta "how to get from here to there, what's the meta-structure of 'there'" has yet to be worked out
02:17 < Athrelon> re: Moldbug: I take him to be making several points
02:18 < Athrelon> 1. that medical research ethics is a hilarious sham that's costing us a lot of progress (agreed; I wrote a term paper on a related topic that received the only A in the class)
02:18 < Athrelon> 2. that Steve-style applied science >> the current way basic med research is done (plausible)
02:20 < Athrelon> 3. that a decentralized billionaire-pet-project driven hack-y research effort with sane medical ethics is more likely to get us to (2) than current lapping-at-the-research-trough model
02:22 < Athrelon> in general the observations that drive my "philosophy of science institutions" are:
02:23 < Athrelon> 1. people, even very smart people, vary hugely in agency, ability to see the bigger picture of science, and produce significant advances.  A scientist is not a scientist is not a scientist.
02:24 < Athrelon> 2. current institutions assume fungible scientists and their rules seem to be driving a lot of dysfunction
02:24 < Athrelon> 2a. the funding-trough model creates a snake eating its own tail; smart self-interested people forced to work on very short time horizons on hot topics and play nice with politics - all anathema to geniuses
02:26 < Athrelon> 2b. educational model requires long apprenticeships with limited autonomy, also emphasizes relatively short (publication-cycle) time horizons, scientists aren't taught how to lead and manage large projects or how to raise money outside the funding trough (limited flexibility)
02:26 < Athrelon> 2b(1). James Watson was educated as a scientist at the best schools but his story was more like "lousy college student, read "What is Life?" and got excited, went to Indiana to work on it" than "competing for his spot at the top."
02:27 < Athrelon> 3. as is becoming a cliche, our institutional safeguards are inevitably circumvented
02:29 < um> there's probably something about scientists being a pressure group that wants reliable-ish jobs in their sector, but i'm not sure how their power works politically
02:29 < um> and what imprint pressure of that form leaves on the rules for who gets funding
02:30 < Athrelon> 3a. What is alternative?  I think you have to trade off "firmly enforce obvious but easily bent rules" and "accept that there will be, upon retrospect, obvious quacks."  There will be quacks regardless; the question is whether you want them to be obvious-upon-retrospect (N-rays) or just ubiquitous and unobtrusive (Ionnadis's scoops)
02:30 < Athrelon> um right!  In retrospect science becoming high status and "throw money at scientists" becoming applause light probably doomed the institution
02:31 < Athrelon> Golden age of science seems to have been Royal Society through Victorian gentleman-science up to early Apollo, with some ups and downs in that period
02:32 < Athrelon> some assorted related points:
02:34 < Athrelon> Some in the Thiel cluster have made noises about scientists being able to be much more effective by learning to pitch rich people on funding real science; the limiting factor seems partly "market-making" and partly that we don't train scientists to fundraise or lead large teams.
02:34 < Athrelon> SarahC has estimated that a scholar "working on actually-useful things rather than hill-climbing" in her field could produce ~10x social good pretty reliably
02:34 < um> re "accept", a lot of the time the norms gain power from people not willing to take on the risk of having lent credibility to a quack personally, whether or not they would accept it in aggregate
02:35 < um> although it's a thing with equilibria, maybe if the norms hadn't had so much power people would be more willing to take on that risk
02:36 < Athrelon> The mega-science projects (Manhattan, Apollo, Bell Labs) are a really interesting case study of a successful model besides gentleman-science; I need to read more about it but they seemed to have much more functional cultures and leadership than the joke that is contemporary mega-science.
02:36 < Athrelon> It could be that once they set the precedent that "this shit works, throw money at us" the incentives that made them good to begin with collapsed
02:37 < Athrelon> um you personally don't have to accept it; heck if you're a good scientist who can debunk a quack, status points for you!
02:38 < um> and when the norms do have power, then people are less willing to be associated with rules that would clearly-in-prospect allow clean causal attribution and blame-propagation from obvious quacks to the people who supported the rules that made them possible
02:39 < Athrelon> um yes, of course; that's one of the things keeping the current medical ethics system in place
02:42 < um> i don't have a view of the big picture though i was just commenting on the aggregate-personal-level feasibility of one of your implied engineering subtasks
02:43 < um> anna tells a story of how four-year-olds (five? eight?) could only just barely keep the rules of checkers in their head well enough to not constantly attempt illegal moves
02:46 < um> i feel like an important part of the current written and unwritten rules of science, the part that engages emotions and keeps the whole thing socially running, was dumbed down enough to engage Wason Selection Task type instincts, and a lot of the participants are not far above the level of ability to keep science from flying apart at the ligations which corresponds to that level of checkers playing
02:47 < um> ("ligation" i mean like in "religion" or "obligation", or "seams" except the metaphor is more like a flywheel and you don't make those out of cloth and stuffing)
02:49 -!- JenRM|afk [~jrodrigu@108-74-165-2.lightspeed.sntcca.sbcglobal.net] has joined #goodintentionspavingco
02:52 < Athrelon> welcome to the slide room of the academic
02:53 -!- JenRM|afk [~jrodrigu@108-74-165-2.lightspeed.sntcca.sbcglobal.net] has quit [Ping timeout: 252 seconds]
02:53 < Athrelon> um what does flying apart at the seams look like?
02:54 < Athrelon> it seems like cargo cult science could hum along nicely, becoming more cargo culty, without any really obvious signs
02:55 < Athrelon> until it's 30 years later and it turns out nothing interesting has happened technology-wise
02:56 < Athrelon> and even then you could say "but we've made advances in constructing these [now necessarily high status] edifices in basic science for its own sake, which is pretty awesome"