In 1972,  Gerald Edelman (born 1929) received the Nobel Prize in Physiology or Medicine for his discovery of somatic selection in the immune system of mammals. It was his answer to the question of how our body can produce as many different antibodies, each directed against a specific invader.

Previously it was thought that the drawings of all the antibodies were encoded somewhere and was activated during an infection. But the number of all possible pathogens that our species is encountered in the past and may yet meet in the future are so overwhelming that strained credulity hypothesis. In addition, different people at very different antibodies in response to the same attacker.

Gerald Edelman has shown that the immune system operating on the principle of evolution. While all other cells in the body carries the same genes, certain immune cells is an exception to the rule. Their genetic composition can be varied. When a new infectious agent is encountered, guns the engine of the immune system in a crazy, busy trying different combinations of genes in immune cells, until an adjustment is made.

This architecture allows a quick response to any attack that may never be offended. In a few days, moving in what may have been decades of rational design to accomplish.

But Edelman did not rest on our laurels. He suggested that the brain also works on the principle of evolution. This was the birth of neural Darwinism paradigm in neuroscience.

Evolution occurs in the brain in several ways. First, insofar as its structure is encoded in the genes, the brain is a product of evolution, natural selection.

Another in a growing organization that competes neurons to make connections between them. Again, we see how developments are better than the rational design. Instead of pre-programming of a specific structure rigid, neural development suggests competition for self-optimization of the method of connection. This choice of development ensures that even identical twins or clones would never have the same brain. However, the random is not allowed to run wild, General, are the high-level structure in the brain that has preserved a sort of combination “free market” and control, honed to perfection over the centuries evolution.

Third, a functioning brain, neurons compete for a chance to fire it is to send signals to other neurons. There are two types of neurons in the brain: excitatory and inhibitory. When a neuron sends an excitatory signal to the other, it encourages under fire again, while an inhibitory neuron trying to silence its target (or neither succeeds depends on the current situation and a series of thresholds).

If we had only excitatory neurons, they quickly synchronized, all neurons in the brain firing in perfect harmony, like the clocks to be influenced the same floor of the other through mutual feedback spontaneously synchronize their oscillations in a process called clocks ticking along entrainment.Their start. But perfect harmony is a very simple structure and it does not support the complexity. neuronal oversynchronization is what really happens during a seizure (grand mal), one would expect that the person is unconscious, while it lasts.

inhibitory neurons creates a complexity, by providing competition. When an excitatory neuron fires for a second he wakes up from its inhibitory allies who seek to silence other neurons that want to send similar signals. Winner takes all. In addition, the winner is rewarded far: synapses firing neurons neurons get stronger, so they are more likely to earn in the future. Synapses become weaker if they do not fire for some time. This process is called brain plasticity, the brain is constantly changing to become more intelligent, more sensitive to new situations. Thus, for example, we can learn tasks to a level of perfection that we can do on autopilot learning new sap, reconnect.

In the brain, neuronal ensembles massively parallel and competing for the best results by comparing their predictions with the reactions of external actions to make corrections. The impression that our brains are “single-threaded” is an illusion, because we only consider the results of massively parallel computing, as many teams working on the same task. And if you think that our memory is weak, just because we can juggle with only a handful of objects in our mind at the same time, think about how much information is involved in a single object, taking into account all sensory data, to say nothing of the interaction with the object, so that the number of decisions of forking paths taken in the light properties of the object may grow exponentially. Daniel Dennett, a promoter of neural Darwinism, wrote that if you take a skeptical a doubtful case, and in a second or two of balancing, stripes, tasting rings, and watch the reflection of sunlight on the surface, consuming the skeptic more information bits of a Cray supercomputer to organize in a year.

Neurons are natural oscillations of electrical potential across the cell membrane. When the fire together, they can synchronize, producing what we call oscillations of the brain, or brain waves.

Random variations that are necessary to operate the neuronal selection, follow the distribution of pink noise as 1 / f, the amplitude (strength) of the oscillations is inversely proportional to their frequency. Noise after a more general 1/fadistribution called fractal noise, where the number of its boards dimension.When fractal chaos is stronger than his and when a> 1, is so strong that the chaos. But when a = 1, this area of highest complexity, where complexity is measured by the number of states of the system to distinguish from each other. In other words, a = 1, where the butterfly effect is more visible. Of course, the number of states of the system can distinguish between the greater is the amount of information in the system can hold. This type of noise is the most dense low noise information in the universe.

Once again, the process of evolution spontaneously established a perfect balance between order and chaos. On the one hand, fluctuations in the brain to synchronize, because this is precisely how the inputs from different sources are combined to produce a cognitive event. But on the other hand, brings oversynchronization epilepsy, a state of mind when large groups of neurons are activated in order to maintain a too simple structure of consciousness.

In a normal brain awake, synchronization be transient. The alert (or dreaming) mind is always in a mode of phase change, like a ball on top of a hill in an unstable equilibrium, which selects and fall maximum complexity mode, driven by the constant driving and The Butterfly Effect.

I suggest the idea that we do not need to know how the brain to simulate the functionality that is currently prevalent in AI research is flawed. Instead we should learn from the brain.

Until we deterministic chaos harnesses, we can never create a true artificial intelligence. Call this the principle neuromorphic.

But how the brain develops? And why? Neurons are very hungry, energetic animal cells, but the brain continues to grow in size from one species to another. What is the evolutionary advantage of consciousness?

One of the fathers of modern neuroscience, Rodolfo Llinas suggested that the brain has evolved into an active, conscious movement organizations to predict the outcome of the motion. Plants do not move voluntarily, so they do not need and therefore did not have a brain. Exactly, ascidians spend the first short phase of their lives as larvae actively moving animals with a small brain. But once they find a good place to live, they grow into plants, digest their own brain.

We tend to underestimate the complexity of our movement. If you take into account the number of muscle groups in one hand, and the number of motor neurons activated in each tenth of a second in different sequences, so that the number of degrees of freedom to move just the hand is so huge is a computer-based CPU should have a truly astronomical CPU frequency to cope, and 100% CPU, no. But our brain performs the task without difficulty, with only a small fraction of its neurons, which leaves a lot of processing power to think things like the others.

The computing power of the brain is overwhelming. It may not be to add numbers very quickly, but as a processor of movement and decision, he finds a computer at any time. Robots can be programmed to perform well, with repeatable accuracy in predictable environments. In contrast, the brain never repeats itself exactly, thanks to its architecture of motor development. But for the same reason it is able to respond quickly to any situation in different environments, the members of the species may be in addition to several million years.

patterns of action saps-fixed and certain emotions are necessary “optimization” of the prediction engine of the brain. Awareness is essential for survival in an unpredictable world, taking over from the autopilot when something unexpected happens. Thus, neither emotions or awareness is limited to humans. Many animals have them just to be functional. I guess our children first AI will be more driven by emotion than we are because the feelings come first, even before the cause.

To predict the brain constructs an internal model of the world. During the action, the results obtained from the forecast and the model is spontaneously changed plasticity predict better next time.

It is important to understand that this model is internally generated. sensory impressions of the outside world changes, but not completely define the model that can operate on the basis of internal contributions (as it does in dreams, or say in the planning for the future), even in the absence of sensory stimulation from outside. The brain is a virtual reality machine.

How can we understand each other? Why not have internal models of different brains so different than being mutually incomprehensible? Yes, they are unintelligible different species. But in a case are the basis for the model with the history of evolution. The model, after all, must adequately reflect the shared world to survive.

With the loop action-reaction-action is universal in the embedded world in the course of evolution, the structure of the brain. Llinas offers a metaphor for a gelatinous cube of electrically conductive material with electrical contacts on its surface. condensed gelatin in endless if the power is often between the contacts, but relaxes back to the amorphous state, where no current flows in a while.

In this regard, you acknowledge brain plasticity already in the workplace.

If power is based on sensory information from, for example, playing football and finally our cube of gelatin would be to develop a structure which, in a sense, coding rules for playing football, but it would be very different from familiar game with a ball, a team of players and a referee.

Similarly, the codes of the brain of our experiences in a different format. It is useless to ask where exactly in the brain images that we see and our thoughts are, because they are products of the process, encoded in our brains by the interaction with the loop in the world of action-reaction -action.

Although many degrees removed, our thinking is ultimately an internalization of our movement.