Mike Hollinshead <[EMAIL PROTECTED]> repeated some comments he made
earlier about chaotic versus non-chaotic models, saying

> One major problem you are going to run into is the inability of non-chaotic
> systems models to generate surprise (like innovations) as such models are
> completely defined.  Innovation is one of the major factors from the human
> side and Mother Nature has tricks up her sleeve too.  Any model which
> cannot explain it endogenously is fatally flawed.  ...

A simulation is a tool, that's all, just a tool.  All a tool needs to 
be is useful for some purposes.  I am not undertaking to create a 
crystal ball to magically foretell the future, just to do the best we
can with existing technology.  

> So the conventional algorithmic approach ultimately won't help, ditto 
> non-chaotic simulation.  You need something like biologist Stuart Kauffmans
> bootstrap model which is a chaotic simulation model, but it is only a 
> baby step towards what you will need.

I don't think I've ever said anything that restricts the simulation to
being "non-chaotic".  I'll look around for Stuart Kauffman's model, 
but I don't expect it to include anything beyond the capabilities of 
what I have described.

I am somewhat out of sympathy with Mike Hollinshead's views, which seem
to set up a non-chaotic strawman for him to knock over. If you think
what I am proposing is "fatally flawed", Mike, you'll have to spell
it out in more detail.

> In this I am agreeing with Ed and Eva, i.e.  You have to have some theory
> of how things work, and you have to be able to conceive of and simulate
> alternative systems to the one we have now (entirely different social and
> technical systems).  Otherwise you are simply projecting the status quo,
> which might be interesting in the short run, but is of no earthly use in
> the long run when we want to be able to imagine things being different than
> they are.

Of course I have theories of my own, but I don't want to build them 
into the software, which should not depend on anybody's theories.  
Once we have the software running and have some data to use with it,
then I'd like to try out some theories, but I'd like to make the
software "theory-proof", if possible.  "Projecting the status quo" 
doesn't sound to bad to me, and if we could reliably do that I think
it would help a lot.  But yes, I do want to be able to imagine things
being different than they are.  Until I see evidence to the contrary
I will continue to think that the ability to run simulations will
help out our imagination a lot.

      dpw

Douglas P. Wilson     [EMAIL PROTECTED]
http://www.island.net/~dpwilson/index.html

Reply via email to