Evolutionary Dynamics of Knowledge

1
Evolutionary Dynamics of Knowledge

• Understanding Complex Systems
• UIUC
• May 19, 2005

Carlos M. Parra ltTokyo Institute of Technologygt
2
Contents

• Importance of knowledge in economics
• Question being addressed
• Our approach and fields involved
• EDGE Model
• Artificial implementation of model
• Genetic Algorithms
• Artificial Interpretative Devices
• Problematic issues
• Ways of handling variation and justifications for
  variety
• Real-life example (Evolutionary Institutions)
• Conclusions

3
Knowledge and Economics
NEOCLASSIC ASSUMPTIONS RELATED FIELDS -Perfect
Information - Informational Economics
-Perfect Competition - Evolutionary
Economics -Individual Rationality -
Behavioral and Experimental
Economics
Complexity Economics Agent Based Modeling
(ABM) Agent-based Computational Economics (ACE)
4
Question

• How can ACE increase the predictive power of its
  models by incorporating the salient features of
  actual human decision-making?

5
Approach
FIELDS INVOLVED DESCRIPTION -
Neurophenomenology - Cognitive scrutiny of
human experience. - Piercian Semiotics -
INTERPRETANT as a modification of
consciousness that is itself an experience.
- Embodied experiences (Distinctions). -
Capabilities Approach - Liberty instead of
utility. - Embodied distinctions
(Choices).
Experiences ? Distinctions ? Choices
- EVOLUTION - Solves MENOS Dilemma
(Animates Knowledge). -
Variation/Selection/Retention.
6
Embodying experiences EDGE Model of inner world
construction
Background embexp1,, embexpi where i is
the total number of experiences an individual has
embodied.   Interpretant embexp1,, embexpj
Any subset of embodied experiences belonging to
Background set.
There are many other interpretants influencing
the constitutions of interpretants
7
SoWhy should the nature of knowledge be
evolutionary?

• How does evolution resolve Meno's dilemma?
• Because during knowledge acquisition the
  individual is getting increasingly familiarized
  with an interpretant that lived in a pool of
  potentiality and that was finally selected for to
  constitute an alternative subset, and provide the
  individual with a new (more adequate)
  perspective!
• Knowledge emerges out of a neuro-phenomenological
  evolutionary process that is evolutionary at
  every single stage, which is why it may be
  simulated using Genetic Algorithms.

8
Computational ImplementationGenetic Algorithms
(GA)
Artificial Interpretative Devices (AID)
GA MISINTERPRETING DISRUPTING AID
INCORPORATING CHANCE EEFECTS
9
Computational ImplementationIssues of applying
model

• Detailed specifications of each AID (i.e. changes
  in initial conditions)
• Interaction mechanisms between AID in an agent
• Interaction mechanisms between agents
• Groupings of agents, and the interaction
  mechanisms between groupings

• Very difficult to find the detailed
  specifications of the decisionmaking process!
• If found, access would be limited

10
AndWhere do we go from here?

• Change focus from examining regularities to
  scrutinizing exceptions (because these may lead
  to innovative changes in the system)
• Manage unpredictability by
• Unrestricting the possibilities!
• Real politiks of modeling.

11
Key question in social-computational terms

• How should variety generating mechanisms, or the
  process of generating new variety (variation) be
  managed or handled?

12
Revisiting suggestions made here at UCS 2005

• Medicine Timothy Buchman
• Interventions that follow regular patterns (i.e.
  provision of oxygen) inhibit patient recovery.
• Mathematics Bruce West
• Fractal correlations between heart rates.
• Psychology Daniel Miller
• Crises management by moving away from
  homeostasis.

13
Suggestion from socio-economic systems modeling
seems to be the similar

• When variation alludes to the process of
  generating new and alternative variety (i.e. in
  terms of possible states of a system).
• And variety refers to a measurement of complexity
  that directly relates to the number of possible
  system states.
• Variation may be managed by providing additional
  variety.

14
Real Life ApplicationEvolutionary Institutions
Knowledge
DYNAMIC EQUILIBRIUM OF KNOWLEDGE
DEMOCRACY MARKETS
15
Additional socio-economic justifications for
variety

• Ethics Heinz von Foerster
• In-principle un-decidable questions
• Capabilities Amartya Sen
• Human Development as augmenting peoples choices
• Organizational Theory of the Firm Malerba
• Increasing experience generates more
  variability, which can act as source of
  creativity and innovation.

16
Conclusions

• The field of economics is changing for the
  better.
• Inner world reconstruction and the emergence of
  knowledge may be modeled through an evolutionary
  neurophenomenological process.
• When agent-based modelling incorporates the
  salient features of human decision-making,
  detailed access to the computational process is
  hindered.
• Institutions can advance the evolutionary
  dynamics of experiences, distinctions and
  choices.
• Variation ought to be handled by increasing the
  available variety.

17
FUTURE WORK

• Find work (JOB).
• Perform computational implementations.
• Develop experimental applications and find
  evidence at the organizational level (i.e.
  Implications for Knowledge Management in firms).

18
Selected references

• Arthur, W. B., Durlauf, S. N. and Lane, D. A.
  (1997) The Economy as an Evolving Complex System
  II. Boulder, CO Westview Press.
• Fogel, D. B. (1997) Evolutionary Computation A
  New Transactions. IEEE Transactions on
  Evolutionary Computation, 1 (1) 1-2.
• Hodgson, G. M. (2002) Darwinism in Economics
  From Analogy to Ontology. Journal of Evolutionary
  Economics, (12) 259-281.
• Lakoff, G. and Johnson, M. (1999) Philosophy in
  the Flesh The Embodied Mind and its Challenge
  to Western Thought. New York Basic Books.
• Parra, C. M. and Yano, M. (2002) Tridimensional
  Recursive Learning Model. Cybernetics Human
  Knowing, 9 (34) 79-99.
• Parra, C. M. and Yano, M. (2004) Distinctions as
  Embodied Experiences. Semiotica, 151 (1-4)
  75-96.
• Peirce, C. S. (1931-35) Collected Papers of
  Charles Sanders Peirce. Vols I-VI. Hartshorne and
  Weiss (eds.) Cambridge, MA Harvard University
  Press.
• Sen, A. and Nussbaum, M. (1993) The Quality of
  Life. Oxford Clarendon Press.
• Tesfatsion, L. (2003) Agent-based Computational
  Economics. ISU Economics Working Paper No.1, Iowa
  State University. Downloaded January 2004 from
  http//www.econ.iastate.edu/tesfatsi/acewp1.pdf
• Varela, F. (1995) Resonant Cell Assemblies A new
  approach to cognitive functioning and Neural
  Synchrony. Biological Research (28) 81-95.