I-1-3: Mind-reading communication under the conflict of interests
| Number | I-1-3 |
| Title | Mind-reading communication under the conflict of interests |
| Authors |
Akira Ito (Gifu University) Kazunori Terada (Gifu University) |
| Abstract | In designing a human-agent interaction, most authors implicitly suppose that an agent should and would behave for the interests of a human. We should investigate what kind of problem exists in HAI when the conflict of interests exists between an interacting agent and a human. For the above goal in mind, we investigated the emergence of human-human communication under a partially conflicting situation. We designed an artificial game where cooperation is necessary for good performance, but still there exists a conflict of interests. The players can communicate through very restricted media, and there is no pre-defined meaning for signals. The partner may behave dishonestly, or betray you. Hence the only means for communication is through mind-reading. We report in this paper how the conflict of interests modifies the communicative behaviour of the players. Next we discuss how HAI should be designed under conflictive situations. |
| [Link] |
