The idea behind Laplace's Computer is that in principle one can predict the future exactly and accurately. Only a couple of small things are needed. First, the universe must be deterministic. Second, one should know all the forces and laws of nature and be able to model them precisely. Third, one should know all the positions and velocities as well as other properties of each and every item that the universe is composed of. Fourth, one should have a computer with enough computing power to run a model of the universe. Now, given the current state of the universe and the laws of nature, one could compute the future states of the universe.
First I thought I had a big problem with Laplace's Computer. The problem is that when a computation is being executed on a computer, that computation changes the computer's states. And because computer's states are physical things (swarms of electrons), changing computer's states alters the universe. So while computing the future states of the universe, the computer should also take into account how its modeling of the universe will affect the universe. And it should also take into account how "taking into account how its modeling of the universe will affect the universe" affects the universe. And so forth.
So Laplace's Computer cannot work, on the pain of infinite regress, unless--and this is the beautiful part--the computer is not in this universe. Is that a problem to you?
Subscribe to:
Post Comments (Atom)
7 comments:
I'd say that rather a having the computer in other universe, the computer itself will be a universe.
Is anyone around with infinite energy to bootstrap my new universe? These big bangs explosions are skyrocketting my electricity bill!
Yep... if you think that the universe is vast and complex, just think about the computer modeling that universe! :-)
It is written down beautifully this post. It even sounds as a good introduction chapter for a science fiction book (with or without sense).
Wondering if you have contact with the life form that has access to such type of computer in other universe, would you like to know the results of such type of computation?
Playing the devil's advocate, wouldn't Laplace's computer, using the knowledge of the state of all entities in the universe, all forces and rules acting on them, and possessing the computational power to handle all of that information be also able to extrapolate the universe's affect on itself?
I find it funny that your gripe is with the idea that the computer would not be able to determine it's own affect on the universe, and not with determining the state of everything in the universe, or with the computer having full knowledge of all forces acting in the universe, their rules, and how to apply them correctly.
Members of the jury: things that defense (Devil's advocate) mentions -- like determining the state of everything in the universe, having full knowledge of all forces, etc. -- are so trivially problematic [sic] that it's boring to get a gripe with them.
It's much more interesting to note that Laplace's computer is logically impossible because of the infinite regress we described. The defense's case relies on universe's effect on LC, but misses the LC's effect on the universe.
Prosecutor rests.
Even more interestingly -- perhaps one could argue that the universe is a computer which computes the laws of nature. (cf. Lloyd & Ng's work (1, 2)
(in the back of the courtroom, a man dressed quite identically, but with a different accent than the defense speaks up) The possibility and impossibility of Laplace's Computer, then, would rely solely on its affect on the universe and retroactively the universe's affect on it being an infinite regress. Could it not converge?
Post a Comment