Give computers a goal so they know when they are not achieving it.
Two parts to this:
Related to ``rational choice theory'' (economics).
We're not trying to make machines intelligent (whatever that means anyway)... we're trying to get them to be more rational.
