Why are only human beings possessing many intellectual abilities among many creatures? The social brain hypothesis shows that human beings have increased their brains while reading the intentions of others. A relationship where each other reads the intention of others, cooperates or betrays it creates intelligence arms race competition. In this study, we simulated this kind of arms race by using evolutionary computation on finite automaton using games with three types of gain tables, and as a result confirmed whether the complexity of the agent is increased. As a result of the analysis, we found that the complexity of the agent ‘s strategy tree which got a high score increased with the gain condition requiring delay to compensation.
The findings of this research can be applied to interaction with artifacts that do not get bored.
- Hirotaka Osawa. 2014. Intelligence Arms Race: Delayed Reward Increases Complexity of Agent Strategies. International Conference on Autonomous Agents and Multiagent Systems: 789–796.
- Hirotaka Osawa and Michita Imai. 2013. Evolution of Mutual Trust Protocol in Human-based Multi-Agent Simulation. In 12th European Conference on Artificial Life, 692–697.
- Hirotaka Osawa. 2014. Sustainable Relationship with Product by Implementing Intentional Interaction. In Autonomous Agents and Multi-Agent Systems, 1677–1678.