The important question for a "new kind of science" is whether it can make predictions in a better way than the old science without resorting to using that old kind of science, i.e. without using the good old mathematical formulas in the algorithms.
Cellular automata are supposed to do something like that, but probably we'll still have to use at least number theory, some kind of mathematical logic, and some kind of good old geometry to interpret the answers in space and time -- I do not expect geographical maps to be replaced by algorithmic instructions, because that would be very inefficient for the massively parallel human senses.
At first glance, cellular automata look like an attempt to give up differential equations and any analytical mathematical stuff that describes motion and change, but I may be wrong here.
I have read lots of claims and justifications, but still I'll have to see a particular example of using this new kind of science to reproduce results of the old kind of science -- such as the results of gravity between two physical bodies, without using anything like Newton's law, but only using the states of cellular automata and their neighbours'.
Of course, in that case, the problem of developing new physical theories would become a problem of recognizing kinds and states of cellular automata, or even assigning states to cellular automata. Do I ask too much?
---- This message was posted by rhinoceros to the Virus 2002 board on Church of Virus BBS. <http://virus.lucifer.com/bbs/index.php?board=51;action=display;threadid=25625>
This archive was generated by hypermail 2b30 : Sun Sep 22 2002 - 05:06:15 MDT