🚀 go-pugleaf

Dave Griffiths writes:

>There is no such thing as infinity. Physicists are going to have to bite
>the bullet and develop a discrete TOE at some time or other. To do so
>may require going right back to basics to develop an information based
>theory. But with so much momentum in current theories, so much resistance
>to the radically new, so little money to be made from such a reformulation,
>I wonder if it will ever happen.

This is what computer scientists often say.  We eagerly await the day
when one of them actually does some physics.  :-)  Do physicists go
around saying that present-day computers are inadequate for physics
because they are digital and not analog?  Maybe some do, but
that would be equally misguided.  

The main point is that "there is no such thing as infinity" is an
extremely vague and misleading statement.  Is there such a thing as a
universal Turing machine?  (I've never seen one.)  Is there such a thing
as the square root of minus 1?  All these things are mathematical
constructs which can play an important role in modeling reality - all
three can also be misused.  

Secretly, my theory about computer scientists' love of discrete physics
is that it's a still poorly-understood side-effect of staring at pixels
all day.  I have long had a humorous battle with Nate Osgood over this
issue, replete with such slogans as "Wreck the rectangle!" (which I
stole from Zippy the Pinhead), "Crush the continuum!" and "Death to the
discrete!"  

By the way, I have nothing against discrete physics as long as it's not
claimed as a panacea.  I have, in fact, long been working on a paper on
mathematical methods for discrete mechanics.  It seems as I may soon
have a grad student, and I may inflict this topic on him, since there is
a lot of nice stuff to do with it.  (In case James Dolan is reading this
- no, I don't mean you.)

Work in discrete physics *is* proceeding nicely, *not* because some
folks hope the universe is just like a computer, but because it's nice
to understand the sort of quasi-physical processes that a digital
computer can simulate exactly.