BF5: One of the main conditions, which must be fulfilled by any model, is to become stable (harmonic or logic). That is, any simulation on a model must reconfirm the model in the same shape. If a disharmony or logical contradiction is detected, the model must regain its stability by IR (from external reality or from other models). Moreover, as any model is already integrated in a structure of models, any other model must accept any result of any simulation on any model. This condition ensures the general stability of a structure of models. A real brain does not easily meet this condition. BF6: Any ZMs are able to activate, in time sharing, many ZAM models to do many activities. However, there is a single ZAM which can act on the external reality at a given moment of time. The reason for it is that, before activation, any model needs to initialize using data taken from external reality. When a ZAM is deactivated, it needs to store data for future reactivation. This method is fast, but if external reality is changed too much, such data is no more valid. In this case, the activated ZAM has to find the new conditions of initialization, based on ZMs. The brain uses both methods. By description of the process, we see that it is not easy to do many activities at the same time (in time-sharing), and it is easy to make mistakes. BF7: The facility of any model to gain information from any other model of the brain. However, due to the technological implementation, it is possible that some models have a better communications with some models, and not as easy communication with other models. BF8: Any model has the facility to develop any of its elements as models. Thus, it is possible to have a nested structure of models. The "depth" of this structure has only technological limitations. These facilities generate the knowledge and the consciousness, based on a structure of stable (harmonic or logic) models. Such structure is able to self-develop in an unknown external reality. For a given brain, in interaction with an external reality, there are a number of features which will be described now. That is, as a brain has many modes of interaction with external reality, a particular brain could use mostly only some of them, as follows: SF1: If there is a difference between reality (prediction) and the external reality (IR), a brain has some possibilities: SF1.1: to correct the model based on IR (knowledge) SF1.2: to modify the external reality (creativity) SF1.3: to store IR in a story-type model SF1.4: to ignore or to forget that IR SF2: When a model is "correct", but it cannot be integrated in the structure of models, there are some possibilities: