4 ms·
So this article does the Maxwell Demon some injustice. There is a key point about the demon that when he is 'sorting' the particles into two bulbs or rooms, th
by lubesGordi 10y ago
So this article does the Maxwell Demon some injustice. There is a key point about the demon that when he is 'sorting' the particles into two bulbs or rooms, that the gate he is working on is frictionless. In this way, you can can see that there is no energy entering the system, yet the entropy is decreasing.
Now, this is where things get interesting to me (please correct me if I'm wrong here). What the demon is adding is information to the system. It decreases entropy. Natural selection is a sort of maxwell demon in it's selection process. I have a theory though that it all evens out. The more complex/evolved the organism, the more entropy the organisms generate themselves.
- trentlott 10y agoI think that was Schrodinger's argument in "What is Life": nature seeks to maximise entropy, and life does a very good job of it.
- eli_gottlieb 10y agoThat theory has gotten a bit more elaborated these days: http://rsif.royalsocietypublishing.org/content/10/86/20130475 http://rsif.royalsocietypublishing.org/content/10/86/2013047...
- swiley 10y agoI thought the idea with Maxwell's daemon was that detecting the particles heated it up which made the particle detector less accurate.
- taneq 10y ago> There is a key point about the demon that when he is 'sorting' the particles into two bulbs or rooms, that the gate he is working on is frictionless. Exactly. Maxwell's Demon is a magical construct. In reality, any active device that sorts molecules into high and low energy bins would take power to run, and would generate more heat (or other entropy) than it removed by doing the sorting.
- lubesGordi 10y agoWhile what you said is true, it sort of misses the point. The thought experiment demonstrates that there is some quantity being added to a system each time the demon opens or closes the gate, and that quantity is information. Now we have a relationship between information and entropy. Edit: I think you might be speaking to my point that in the end it evens out, and that natural selection in turn does create entropy even as its generating information.