TokenTrade 1.0

Java applet that simulates a psychological economy.

Instructions for using this applet are here.
Click here for an example experiment.

Note: you must highlight a group of cells in order to Edit them.


If you have trouble getting this applet to work in your browser, please send email to turknett@amoda.org describing the problem, and be sure to include your operating system and browser version.


Graduate Thesis

TokenTrade was developed as part of Rob Turknett's master's thesis, An Artificial Society of Token Traders: Computer Simulation of Life, Happiness, and Complexity in Trade, completed at The University of Texas at Austin. The TokenTrade simulation is based on ideas contained in the forthcoming book, A General Theory of Value, by Michael Benedikt. Professors Michael Benedikt and Greg Lavender are advising this project. For more information, please contact the author at turknett@amoda.org

The thesis can be downloaded in Adobe PDF format here.

TokenTrade Simulation

TokenTrade is a simulation of a psychological economy. In this economy, individual agents have needs which are fulfilled by tokens.

Needs and Cell Satisfaction

Each individual has three needs:

    1. Confidence/Freedom
    2. Legitimacy/Approval
    3. Survival/Security

Likewise, there are three types of tokens, one corresponding to each need (we will refer to them as x, y, and z tokens). x tokens correspond to Survival/Security, y to Legitimacy/Approval, and z to Confidence/Freedom. Each agent, or "cell" is initially endowed with a random number of x, y, and z tokens.

A cell's total satisfaction depends upon the fulfillment of its needs. The lower, more basic needs dominate the higher needs, so the lower needs must be attended to first. For example, confidence/freedom (z) tokens are worthless to a cell if its need for survival/security is not being met. In economic terms, tokens that fulfill higher needs are dependent goods; their value is dependent upon an individual's possession of other goods, namely, tokens that fulfill the more basic needs.

The model of agent satisfaction is based on Abraham Maslow's heirarchy of basic human needs. The terminology employed here (survival, security, legitimacy, approval, confidence, and freedom) follows Benedikt's revision of Maslow.

The interpretations ascribed to tokens are based on the formal relationships between the three needs. What is being modeled here is the hierarchy, rather than the emotional or behavioral aspects of the needs. More generally, the simulated hierarchy of needs would be described like this:

    1. Ability of the agent to derive "satisfaction" from z tokens
    2. Ability of the agent to derive "satisfaction" from y tokens
    3. Ability of the agent to derive "satisfaction" from x tokens

Satisfaction is a function of the number of x, y, and z tokens that an agent has:

0 <= S = a(1+b(1+c)) <= 3

where

a=x/(1+x)
b=y/(1+y)
c=z/(1+z)

The function for S models the hierarchy, in that y tokens are worthless if the agent doesn't already have some x tokens. The calculation of a,b, and c models diminishing returns as more tokens are accumulated. In other words, there is a limit to how much satisfaction an agent can get from accumulating any one type of token.

A cell's happiness is simply the change in satisfaction that occurs after a round of trading. Increases in satisfaction produce happiness, decreases in satisfaction produce unhappiness (sadness).

Trading

In order to fulfill their needs (and thereby increase their satisfaction), cells trade tokens with each other. Tokens are traded in a barter-style economy. In a single round of trading, each cell is given an opportunity to trade. Trading proceeds as follows:

    1. A cell is chosen from the trading pool at random. At the beginning of a round, the pool contains all the cells in the universe.
    2. The cell negotiates a possible trade with each of its neighbors.
    3. Of these possible trades, the trade that results in the greatest combined happiness (subject to a fairness test) is executed and the cell and its trading partner are removed from the trading pool.

OR

If there is no trade that results in an increase in happiness, the cell (and none of its neighbors) is removed from the trading pool.

This process repeats until the trading pool is empty. The display is updated, and a new round of trading begins.

Additionally, some cells produce tokens, while others consume them. This is controlled by a cell's dissipation/production rate, which is described in the "Cell Parameters" section below. The dissipation/production rate controls how much a cell produces or consumes per round.

Experimentation

This simulation is intended to provide a laboratory for experimentation. The user is encouraged to experiment with various configurations and parameter settings, in order to find configurations that result in interesting overall behavior. Some configurations will be more stable than others, and some configurations will yield more predictable behavior than others.