Entry tags:
outsourcing intellectual work in science; semantic proofs
Computer gamers solve problem in AIDS research that puzzled scientists for years
Yay for gamification! This is very exciting!
Where else could we outsource intellectual labor to computer gamers? What other bits of science can be formalized and modularized away so as to not require the context that only comes with years of experience? Is this the same set of problems where you might use AI search algorithms?
For anyone interested in the logical structure of scientific theories and inter-theory relations, I recommend Theo Kuipers's "Structures in Science".
---
Tangentially, I've been working on a homework problem: if X and Y are independent geometric random variables, show that min(X,Y) is geometric. It is easy to write a semantic proof of this statement: just interpret X as "first time period in which a man arrives" and Y as "first time period in which a woman arrives", then min(X,Y) is "the first time period in which a person arrives". X and Y can be thought of as arising from memoryless arrival processes; and since they are independent, min(X,Y) arises from the combination of the two processes, which is clearly also memoryless.
Despite what some might say, this argument is 100% rigorous; though it is probably a distraction from the intent of the exercise, which is to play with equations... which is understandable, since that object-level axiomatic system is more well-known by mathematical bureaucrats who can rubber-stamp proofs as "valid".
To use a term from diagrammatic reasoning, semantic proofs give us a "free ride".
Yay for gamification! This is very exciting!
Where else could we outsource intellectual labor to computer gamers? What other bits of science can be formalized and modularized away so as to not require the context that only comes with years of experience? Is this the same set of problems where you might use AI search algorithms?
For anyone interested in the logical structure of scientific theories and inter-theory relations, I recommend Theo Kuipers's "Structures in Science".
---
Tangentially, I've been working on a homework problem: if X and Y are independent geometric random variables, show that min(X,Y) is geometric. It is easy to write a semantic proof of this statement: just interpret X as "first time period in which a man arrives" and Y as "first time period in which a woman arrives", then min(X,Y) is "the first time period in which a person arrives". X and Y can be thought of as arising from memoryless arrival processes; and since they are independent, min(X,Y) arises from the combination of the two processes, which is clearly also memoryless.
Despite what some might say, this argument is 100% rigorous; though it is probably a distraction from the intent of the exercise, which is to play with equations... which is understandable, since that object-level axiomatic system is more well-known by mathematical bureaucrats who can rubber-stamp proofs as "valid".
To use a term from diagrammatic reasoning, semantic proofs give us a "free ride".