Spending two months improving a paper so that each of 200 other researchers spends half a day less struggling with it is a net win.
Indeed, but notice that Aaronson himself says that the people who will actually end up reading the paper can be counted on one hand. And even that’s unusually good—except for the tiny percentage of blockbusters and citation classics, most published research papers and theses are never read by anyone except for the authors and reviewers/committee members (if even the latter). Unless you’re in the ultra-elite in your field, and perhaps even then, there is no way that thousands of man-hours will be invested in reading your work.
Moreover, academic writing is usually not at all aimed at improving the paper, in the sense of optimizing it for easy conveyance of accurate information. It is optimized for jumping through the bureaucratic hoops of the review and editorial process (and perhaps also the subsequent citation impact). This process should theoretically be highly correlated with actual improvements, but in reality, this correlation is very low—or even negative, since the “publish or perish” pressures often force one to employ every possible spin short of outright data falsification to make the work look better. This is further exacerbated by the ossified bureaucratic rules from the days when printed journals and proceedings were crucial for dissemination of results, which make no good sense in the age of the internet.
All this is even without getting into the issue of how much sense the work itself makes when evaluated in an honest, no-nonsense way, and how much it is fundamentally pointless, with zero or even negative contribution to human knowledge, and aimed only at padding one’s resume, essentially a peculiar species of bureaucratic makework. (This is usually not a problem for people working in healthy fields and at sufficiently elite levels, but it certainly is a problem for a very considerable percentage of people doing academic research and scholarship.) All things considered, I am not at all surprised to see rampant akrasia, nihilism, and impostor syndrome.
And even that’s unusually good—except for the tiny percentage of blockbusters and citation classics, most published research papers and theses are never read by anyone except for the authors and reviewers/committee members (if even the latter).
Indeed, but notice that Aaronson himself says that the people who will actually end up reading the paper can be counted on one hand. And even that’s unusually good—except for the tiny percentage of blockbusters and citation classics, most published research papers and theses are never read by anyone except for the authors and reviewers/committee members (if even the latter). Unless you’re in the ultra-elite in your field, and perhaps even then, there is no way that thousands of man-hours will be invested in reading your work.
Moreover, academic writing is usually not at all aimed at improving the paper, in the sense of optimizing it for easy conveyance of accurate information. It is optimized for jumping through the bureaucratic hoops of the review and editorial process (and perhaps also the subsequent citation impact). This process should theoretically be highly correlated with actual improvements, but in reality, this correlation is very low—or even negative, since the “publish or perish” pressures often force one to employ every possible spin short of outright data falsification to make the work look better. This is further exacerbated by the ossified bureaucratic rules from the days when printed journals and proceedings were crucial for dissemination of results, which make no good sense in the age of the internet.
All this is even without getting into the issue of how much sense the work itself makes when evaluated in an honest, no-nonsense way, and how much it is fundamentally pointless, with zero or even negative contribution to human knowledge, and aimed only at padding one’s resume, essentially a peculiar species of bureaucratic makework. (This is usually not a problem for people working in healthy fields and at sufficiently elite levels, but it certainly is a problem for a very considerable percentage of people doing academic research and scholarship.) All things considered, I am not at all surprised to see rampant akrasia, nihilism, and impostor syndrome.
I have made a stab at collecting some citations on citations in one of my footnotes: http://www.gwern.net/Culture%20is%20not%20about%20esthetics#fn48 If anyone wants to suggest more...