The PlayMate should be able to plan which information processing actions to take to achieve particular goals. Rather than planning is specific domains this process should involve determining when various internal actions should in enlisted. This will require the PlayMate to have representations of its internal processes and be able to connect them in various ways.