6: Internal modeling (mechanism): All cas anticipate. Internal models are the mechanism for anticipation. They are also called “schema”, a term which generally applied (only) to the study of genetic algorhythms. >A basic maneuver for creating models is aggregation. We eliminate details so that selected patterns are emphasized. When the models are interior to the agent, the agent must select patterns from the torrent of input it receives and then must convert these patterns into changes in its internal structure. >Finally, the changes in structure (the model) must enable the agent to anticipate the consequences that follow when the pattern, or one like it, is encountered again. >There are two types of internal models, tacit, and overt, which are sometimes difficult to distinguish from each other. >A tacit internal model simply prescribes a current action under an implicit prediction of some desired future state, such as the bacterium swimming to sugar. >An overt internal model is used as a basis for explicit, but internal, exploration of alternatives, a process called lookahead. An example of lookahead is the metal exploration of possible moves in chess prior to moving a piece. >A model allows us to infer something about the thing being modeled. A given structure in an agent is an internal model if we can infer something of the agent’s environment merely by inspecting that structure. We can infer a great deal about the environment of an organism by studying its morphology and biochemistry. Therefore, those pieces constitute an internal model. >The structure from which we infer the agent’s environment should also actively determine an agent’s behavior. >If the resulting actions anticipate useful future consequences, the agent has an effective internal model and will be favored more than an ineffective one. >In all cases of cas, the organism’s chance of survival is enhanced by the predictions implicit or explicit, that the model entails. Thus variants of the model are subject to selection and progressive adaptation.