I recently decided to try reading this blog to see what the fuss was and this leapt out at me:
“At that time, there were no Conspiracies, no secret truths; as soon as Eld scientists solved a major problem, they published the solution to the world and each other. Truly scary and confusing open problems would have been in extremely rare supply, and used up the moment they were solved.”
It occurs to me that Mr Yudkowsky is proposing that having science (or scientific fields) incorporate something like a “hidden secret” of the sort mystery religions use would actually be beneficial for science or the world. It’s not the first time I’ve heard the idea connected to him.
Also, I’ve heard interpretations of Noam Chomsky’s early publishing tactics in linguistics described in roughly this way (not publishing enough to replicate his work, letting special people in on the secret who then publish papers based on it, immunizing his theories from disconfirming argument by explaining that critical papers aren’t criticizing the full true theory (the one he was working on but hadn’t yet published to people not specially selected and sworn to secrecy)). I’m not sure if this is is true or not. It’s academic gossip mostly.
But it might explain why linguistics is in such a “pre science” state even now, with many competing paradigms co-existing in the community so that linguists spend much time re-arguing fundamentals and relatively less solving puzzles about language. At the same time, Chomsky’s citation tree is breathtakingly large. Tenuous conclusion on the mystery cult tactic from this example: good for Chomsksy, bad for linguistics?
Other than modeling experiments, it’s hard to even test the theory because the object of study would be scientific communities and it would be difficult and (ahem!) ethically dubious to experiment on them… but the thought is worrisome when bearing in mind that the payoffs of the dynamic are (on first glance) structured like an N-person prisoner’s dilemma with no obvious regulatory agent.
EDIT in 2023: To augment the link, with a residual broken one, and a better link to a hopefully more stable archive that helps maintain the reliable infrastructure that supports Bacon’s Project.
I recently decided to try reading this blog to see what the fuss was and this leapt out at me:
“At that time, there were no Conspiracies, no secret truths; as soon as Eld scientists solved a major problem, they published the solution to the world and each other. Truly scary and confusing open problems would have been in extremely rare supply, and used up the moment they were solved.”
It occurs to me that Mr Yudkowsky is proposing that having science (or scientific fields) incorporate something like a “hidden secret” of the sort mystery religions use would actually be beneficial for science or the world. It’s not the first time I’ve heard the idea connected to him.
Also, I’ve heard interpretations of Noam Chomsky’s early publishing tactics in linguistics described in roughly this way (not publishing enough to replicate his work, letting special people in on the secret who then publish papers based on it, immunizing his theories from disconfirming argument by explaining that critical papers aren’t criticizing the full true theory (the one he was working on but hadn’t yet published to people not specially selected and sworn to secrecy)). I’m not sure if this is is true or not. It’s academic gossip mostly.
But it might explain why linguistics is in such a “pre science” state even now, with many competing paradigms co-existing in the community so that linguists spend much time re-arguing fundamentals and relatively less solving puzzles about language. At the same time, Chomsky’s citation tree is breathtakingly large. Tenuous conclusion on the mystery cult tactic from this example: good for Chomsksy, bad for linguistics?
Other than modeling experiments, it’s hard to even test the theory because the object of study would be scientific communities and it would be difficult and (ahem!) ethically dubious to experiment on them… but the thought is worrisome when bearing in mind that the payoffs of the dynamic are (on first glance) structured like an N-person prisoner’s dilemma with no obvious regulatory agent.
If this is a failure mode of scientific disciplines, it would tend to occur where someone unilaterally broke with the cultural norms of academic science.
EDIT in 2023: To augment the link, with a residual broken one, and a better link to a hopefully more stable archive that helps maintain the reliable infrastructure that supports Bacon’s Project.