This is my free interpretation of the knowledge accumulated at the moment, both directly from neurophysiology and related sciences, and in the design of dynamic neural networks.The brain has a distant resemblance to virtualization machines. For example, in that he is able to simulate the behavior of surrounding dynamic entities in himself, starting with his owner. Its elements are ideally suited for such a simulation, as are
FPGA elements. And this simulation or reflection itself takes place continuously and automatically, as the “new facets” of the object are discovered and in some way anticipating or predicting them.
For example, self-reflection implies the perception of a stimulus not in relation to the objective self, but to its simulation. And also in dreams - we fully immerse ourselves in these simulated worlds with simulated heroes.
Thus, a significant part of the perceived information is generated by the brain itself, therefore, the thin line between objective and subjective perception, reality and illusion is so. And this is one of the reasons for deviations - personality disorders and perceptions or the vision of something that does not objectively exist. It is a kind of “materialization” of the simulated world. Something close to
augmented reality .
In terms of "objectivity" of perception, there is a
list of cognitive distortions . It is also interesting that, in order to exchange exclusively objective information, D. Bom tried to create a language “reomod”, which was never created.
')
The other side of the simulation is that there is no objective time for it. Even taking into account the delay of perception and the speed of propagation of nerve impulses (3 - 120 m / s). But one should also take into account the total number of connections, the distribution of axon lengths, the effect of diffuse neurotransmission, feedback, and much more. All of this works together at speeds much higher than the speed of perception. As multitasking, in which one program "thinks" that it gives all the CPU time and it is continuous, while a huge number of parallel programs are running. Neural networks can be assembled into blocks of dynamic scenes after the fact, both by building them into a ready-made chain by the time they are converted, and dynamically update them as the “reading” of memories (and it is possible that there is no sequence as such - there is a flexible “redirection” to different chains both from current settings and general background / state).
An important role in the formation of the basic neural network configuration of the brain and the whole organism is assigned to genetics. A set of chromosomes is a kind of 22-pair prototype hash function with an additional pair of gender, which additionally determines species compatibility. And this is a self-replicable / recursive hash function, which allows you to fully restore a full-fledged working prototype. In this sense, A. Turing in his car was quite close to reproducing the work of molecular computers. This is similar to bootstrapping. By the way, in the final phase of the “preparation” of the brain for the beginning of work, targeted migration of neurons, differentiation of the functional under the local influence of hormones and directed axon growth occur.
In addition, briefly touch the power. It is provided by an exogenous system (digestive and respiratory systems), as well as a filtration and recycling system that promote reuse. A self-oscillation pump with innervation from four types of nerves - slowing down, accelerating, enhancing and weakening, as well as an extensive network of "delivery" are responsible for transporting nutrients.
So, in the end, we have a full-fledged working "suit" on self-sufficiency with a universal processor of external / internal information. During the period of operation - use them carefully, intelligently and consciously.
And do not forget to at least sometimes recheck whether you really turn to reality or to cash simulations. After all, the “matrix” is not outside, it is inside.