Play the slideshow

AIW02 (December 2010 @ CRI)

Collaboration and rule systems

or why which hands you shake make all the difference

by Fabien Benetou for AIW, December 2010 @ CRI

Problematic

Shift from physical and spatial boundaries to an interconnected set of communities thus leading to appearance of "virtual" boundaries.

cf AIWProposal presented the previous month.

Why is it still web related?

  • system of rules to promote collaboration and supporting exclusion for institutional success without physical boundaries also exist outside of the web
    • but explicit mechanisms are harder to study (rarely formalized)
    • they inherit limitations from older technologies (e.g. paper)
    • are not available to everybody on one media, "1-click away" from each other (thus competing for attention)

But... why really?!

  • the web is a network of networks, a network of machines executing algorithms
  • but, and maybe even more importantly, a network of people behaving in specific way and constantly deciding to collaborate or not
  • those decisions are
    • always guided or at least influenced
      • by an explicit or implicit rule system
    • recorded
      • by the software stack allowing participant to collaborate

Hypothesis

Knowledge communities can more efficiently fulfill their role by adapting their rule system from our immune system and from simulations.

but rather

Knowledge communities can more efficiently fulfill their role by adapting their rule system from simulations e.g. based on the human immune system.

cf AIWProposal presented the previous month.

Exploring the initial bibliography

cf AIW02#ResourcesRulesSystem

First ideas from this bibliography

  • existing basis for rule systems
    • compatibility of licenses
    • Martin Nowak (Harvard PED) 5 rules
    • Elinor Ostrom (Indiana) 8 rules
  • ideas
    • cognition is offloaded through the rule system
      • e.g. behavior and visual codification
    • automation of rule system enforcement (as software) lower transaction costs (easier collaboration)
  • possible datasets
    • licenses of specific collaborating projects
      • not limited to software
    • MIT CCI handbook

2 papers focused on

  1. Five Rules for the Evolution of Cooperation by Martin A. Nowak, Science 2006
    • kin selection, direct reciprocity, indirect reciprocity, network reciprocity (ref 31-39), and group selection
      • network reciprocity is especially important regarding the problematic since it "should not" be there thus is inherited by habits and/or supports by explicit rules
  2. Finding Optimal Policies for Online Communities with CoSiMo by Felix Schwagereit, Sergej Sizov and Steffen Staab, Proceedings of the WebSci10
    • impossible to find the actual tool and thus to try it

+ bonus An-arrgh-chy: The Law and Economics of Pirate Organization by Peter T. Leeson, Journal of Political Economy 2007