TheoMorphic writes:
quote:
Can someone (or a group) write a series of physical laws into a computer program, enter in some matter to begin with, and see what happens to that matter over time?
Already done.
The
Golem Project at Brandeis will have your computer evolve an organism that can walk. In fact, if you allow it, it will contact other computers with Golem installed to see how well your walker will work in the physical landscape of other simulations.
And best of all, the results correspond to actual physical processes so you can actually make a robot from the plans generated by the computer (and they've done that...a robot made a robot that walks from a computer-generated design.)
Alas, it only has versions for Windows.
But, there's no need to do this through computer simulation. You can do it in a test tube. Self-replicating, auto-catalyzing, homochiral molecules are not hard to produce.
Your original statement of "so we always see these absurdly large improbability numbers for the spontaneous generation of self replicating molecules," while true, suffers from the problem of thinking it is an accurate assessment of the question. That is, we do see these absurdly large improbability numbers for the spontaneous generation of self-replicating molecules...
...and they are all bogus. The probability of generating a self-replicating molecule depends upon the reagents you start with. Start with the right ones, and the probability is precisely 1. You cannot help but get self-replicating molecules. And you can even go one step further and make them auto-catalysing which means that the reaction starts to speed up one it gets going. And you can even go yet another step further and make them homochiral which means they all have the same three-dimensional structure.
------------------
Rrhain
WWJD? JWRTFM!
™ Version 4.2