I haven't actually read the book yet, but I did think about playing around with some very complex cellular automata recently.
I didn't get very far after I realised that every new rule I added (more than one data bit per cell for example) made the automata more and more like the conventional method of simple numerical equation solving - i.e. grid the problem out then change the value of some numbers at each grid point according to a differential equation. Any sufficiently complex system should be able to model most of physics, so is this really anything new?
That said, it's always handy to have different maths methods around, and this may be handy for some systems...
Slashdot Mirror
I didn't get very far after I realised that every new rule I added (more than one data bit per cell for example) made the automata more and more like the conventional method of simple numerical equation solving - i.e. grid the problem out then change the value of some numbers at each grid point according to a differential equation. Any sufficiently complex system should be able to model most of physics, so is this really anything new?
That said, it's always handy to have different maths methods around, and this may be handy for some systems...