☰
⌕

Page

Table Of Contents

Simulations and Games Overview

Simulations are experiments to test how different rules and settings influence decision-making and outcomes. In simulations we can observe the emergent behaviors of individuals and groups under different scenarios and constraints.

We're writing both in browser games (like the one below) and more complex simulations that run on servers and use AI agents to make decisions. See our Simulations Basics page to learn more about how we build simulations.

What we'll explore.

  1. Will AI fall trap to the tradjedy of the commons?:

    • Hypothesis: AI agents will not fall into the trap of the tragedy of the commons because they are not motivated by self-interest.
    • Example: Compare outcomes when agents are motivated by self-interest versus when they are not.

  2. Ability to corrupt a decision:

    • Hypothesis: Corrupting agents (e.g., bribing, threatening, or lying) can shift decisions away from the group's best interests.
    • Example: Add an agent who spreads false information to influence others.

  3. Delegation to knowledgeable agents:

    • Hypothesis: Delegating decisions to agents with higher knowledge improves outcomes for the group.
    • Example: Compare outcomes when the most knowledgeable agents decide versus when the least knowledgeable decide.

  4. Allowing remote decisions

    • Hypothesis: Allowing agents to vote remotely increases participation and improves outcomess for those who cannot attend in person.
    • Example: Test whether remote voting improves outcomes for agents who cannot attend in person.