Artificial Societies and Virtual Worlds

1
Artificial Societies and Virtual Worlds

• A Presentation by
• Andrew Krausnick

2
What is an artificial society?

• A collection of computer simulated agents
  designed to use emergent behavior to grow
  meaningful social structures using communities of
  agents.
• As Dr. Joshua Epstein and Dr. Robert Axtell wrote
  a book Growing Artificial Societies about their
  project, Sugarscape.
• In which, they wrote, We apply agent-based
  computer modeling techniques to the study of
  human social phenomena, including
• Trade
• Migration
• Group Formation
• Combat
• Interaction with an Environment
• Transmission of Culture
• Propagation of Disease
• and Population Dynamics

3
Example Sugarscape

• Three ingredients of an artificial society
  agents, environment, and rules.
• Agents are the people of constructed artificial
  societies, they have any number of traits, from
  sex to vision to wealth.
• The environment exists in a physical capacity,
  to be operated on and interacted with by the
  agents.
• The rules are everything else, modes of
  behavior for agents or physical rules of the
  environment itself.

4
Picture!
5
What is a virtual world?

• Dr. Richard Bartle, co-creator of what is
  heralded as the first true virtual world - MUD1
  outlines what he considers to be most important
  features of virtual worlds, these are
• underlying rules (or physics)
• user representation
• real time interaction
• a shared world
• world persistence
• Actors in a virtual world are known as avatars.

6
Examples

• Multi-User Dungeons (MUDs), usually text based.
• Everquest
• World of Warcraft
• Second Life
• Traits Mostly commercial, usually games. Always
  social spaces.

7
Pictures!
8
Why study them? (Part 1)

• Nicholas Yee, currently a Ph. D. student at
  Stanford University. (http//www.nickyee.com/daeda
  lus/)
• traditional personality assessment techniques
  are typically transparent and reactive.
• actions in massively multi-user online
  environments can be tracked unobtrusively by the
  server
• users are personally invested in their avatars
  and the environment, every decision they make is
  personally revealing

9
Why study them? (Part 2)

• Dr. Edward Castronova of the Department of
  Economics at Cal State Fullerton.
• Economists believe that it is the practical
  actions of people, and not abstract arguments,
  that determine the social value of things.
• One does not study the labor market because work
  is holy and ethical one does it because the
  conditions of work mean a great deal to a large
  number of ordinary people.

10
What virtual worlds and artificial societies have
in common

• Agents, Environment, Rules.
• Of academic interest to social scientists.
• Emergence from large numbers of independent
  actors.

11
Why combine them?

• Possible to compare artificial rules to natural
  ones.
• Pass ideas, diseases or goods between real and
  virtual actors.
• Examine virtual world behavior by attempting to
  grow a virtual society to mimic the behavior.
• Create an emergent society with meaning for human
  interaction.

12
How to combine them

• Actions and goals need to be synonymous for
  agents and avatars.
• Limiting avatar abilities to bring them on par
  with agents.
• Providing a meaningful interface for human users.
• Rules -gt underlying rules and human action.
• Environment -gt shared world.
• Agents -gt user representation and human action.

13
Example AlphaWolf

• Bill Tomlinson and Bruce Blumberg at MIT Media
  Lab
• 3D graphical simulation of a Grey Wolf pack.
• The demonstration allowed for user interaction,
  but in a limited capacity, the user could by
  using a microphone to howl, growl, or bark cause
  his or her pup to howl, dominate, submit or play
  based on context.
• In addition, these actions would then affect how
  each of the social participants would view the
  actor in the future, giving the action importance
  over time.
• Relationships are conveyed to the user by way of
  graphical interface buttons, giving some
  indication to the user in which way their
  orders to the pup will be enacted with regards
  to any other wolf.

14
Picture!
15
My Work

• Resources, Needs, Skills, and Actions are
  abstract.
• Communication such as trade is enacted as an
  Action, therefore it is usable as a method for
  cross agent/avatar communication.
• Rules determine agent behavior as well as how
  interaction works (ie. they remember trading
  locations).

16
Example Action Types

• AskAboutTrades
• Construction
• Goto
• Trade
• UseSkill

17
Pictures!