Robert writes: > Models such as Schellings segregation and Axtel and Epsteins > artificial societies typically take place on some bounded > checker board through which nothing flows. By the defintion > below are these therefore not complex systems?
Yeah, I've been wondering for a few years how to reconcile dissipative structures with the classic computational models in complexity (eg flocking, schelling, ant foraging, CAs etc). In particular, what does it mean for a model to be far-from-equilibrium. The only thing I have to offer at this point is to look at any particular model and try to figure out what is flowing through the model and what constitutes a gradient. In physics, a gradient can be an asymmetric distribution of a conserved quantity (mass, energy, charge, momentum, etc). The more asymmetric the distribution is (ie far from equilibrium, low entropy), the more work can potentially be extracted. - In *most* toy ABMs, I think flow is in the updating of the agents and is probably closely equivalent to energy flow. - Gradients in ABM, I suspect, are mappable to asymmetries in the model description/setup. This can be in the initial configurations of agents or resources, asymmetries in behavioral rules or asymmetries in agent-agent communication networks. If you remove asymmetries in your model setup, you will not get emergent structure in your run. For example: In ant foraging, asymmetries are found in the initial placement of the ants, the location of the nests/food and the forward bias of ant movement. In Schelling's segregation model, for spatial patterns to exist, agents must asymmetrically prefer to live near like agents (ie an asymmetry in a behavioral rule). To get emergent patterns in the beer game, one needs an asymmetric penalty where it is more expensive to be out of stock than to hold inventory (asymmetry in behavioral rule). In flocking, boids need an asymmetric forward-biased cone of vision and the asymmetric forward-bias in turning (ie no 180 degree turns) in order to break symmetry in the system's linear momentum. In the case of 1D CAs, I'm guessing it's the case that all type-IV CAs have asymmetric rule tables. Anyone know for sure? -Steve > -----Original Message----- > From: Robert Holmes [mailto:[EMAIL PROTECTED] > Sent: Sunday, July 23, 2006 3:07 PM > To: The Friday Morning Applied Complexity Coffee Group > Subject: Re: [FRIAM] formalization of Complexity (was > Dynamics of ComplexSystems by Yaneer Bar-Yam) > > Models such as Schellings segregation and Axtel and Epsteins > artificial societies typically take place on some bounded > checker board through which nothing flows. By the defintion > below are these therefore not complex systems? > > Robert > > > On 7/21/06, Stephen Guerin <[EMAIL PROTECTED]> wrote: > > > > Yet when I ask for a formal treatment, I get no answer. > > I very much like Hubler's deceptively simple definition > of complexity: > "A complex systems is a system with large > throughput of Energy, > Information, Force, .... through a well designed boundary." > > His notes from the SFI CSSS school with this definition > are here: > http://www.how-why.com/ucs2002/tutorial/ > > > As a restatement of the same ideas that formalizes what > "large" means, I would > offer: > "complexity emerges when a gradient acting on a > system exceeds the > capacity of the internal degrees of freedom of the > system to dissipate the > gradient". > > > Is that formal enough? or, does the statement need to > be mathematized? > > -Steve > > ________________________________________ > [EMAIL PROTECTED] > www.Redfish.com > 624 Agua Fria Street, Santa Fe, NM 87501 > mobile: (505)577-5828 > office: Santa Fe, NM (505)995-0206 / London, UK +44 (0) > 20 7993 4769 > > > ============================================================ > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at > http://www.friam.org <http://www.friam.org> > > > > ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
