3 ms·

>I don't see this changing without fundamentally realigning incentives in academia. I think the corruption in science is a manifestation of Goodharts law. Peop

by forcry 5y ago

>I don't see this changing without fundamentally realigning incentives in academia. I think the corruption in science is a manifestation of Goodharts law. People started to believe everything with the science label blindly. Like if something that starts with "research has shown", then it must be true (now it is "overwhelming evidence"). And whenever that happens, what ever the label is, tends to get bought. (This is why we have toothpaste commercials with people in lab coats). So the long term solution is to actually decrease the trust of people in "science", the label. Then it will come to a point that just the label of science is not enough to sway the people. Then it has a chance to progress, un-disturbed by these corrupting forces. https://en.wikipedia.org/wiki/Goodhart%27s_law https://en.wikipedia.org/wiki/Goodhart%27s_law

murgindrag 5y ago
That's part of the solution. Another part is to somehow change incentive and organizational structures. MIT, Harvard, Stanford, and similar schools made their reputations decades if not over a century ago, and the current people and structures aren't the same as the people who built those institutions. These are big institutions, with big money and big egos. The people at the top are politician types good at climbing corporate structures. It's hard to overstate the distorting impact billions of dollars and those levels of brand recognition have. It's also hard to overstate the degree to which Goodhart's Law is exaggerated when you have 1000:1 competition for faculty slots, among otherwise qualified applicants. It's Goodhart's law on steroids, turbo, and overdrive. You CANNOT get a faculty position at some labs without cheating, at least a little bit. I don't see this changing unless federal and philanthropic funds are radically redistributed.

by forcry 5y ago

murgindrag 5y ago

That's part of the solution. Another part is to somehow change incentive and organizational structures. MIT, Harvard, Stanford, and similar schools made their reputations decades if not over a century ago, and the current people and structures aren't the same as the people who built those institutions. These are big institutions, with big money and big egos. The people at the top are politician types good at climbing corporate structures. It's hard to overstate the distorting impact billions of dollars and those levels of brand recognition have. It's also hard to overstate the degree to which Goodhart's Law is exaggerated when you have 1000:1 competition for faculty slots, among otherwise qualified applicants. It's Goodhart's law on steroids, turbo, and overdrive. You CANNOT get a faculty position at some labs without cheating, at least a little bit. I don't see this changing unless federal and philanthropic funds are radically redistributed.