The Scientific Way To Turn Computer Into Brain

2021-02-03   |   by CusiGO

The background question is: do they look like computers and brains? Can we replicate a human brain? Can we make machines think like people? If we succeed, will we have a consciousness and a thought? The human brain is a piece of material weighing 1.5 kg, but it is the most complex and perhaps the most fascinating object in the universe. In fact, the only one trying to understand the universe. Can he understand himself?

The first concept that allows us to compare computers with brains is that both entities can store and process information, for example, they can do mathematical calculations, even though computers make them much faster than humans. In general, computers perform faster operations that can be broken down into a series of simple steps (algorithms). But the brain is much higher in more complex functions: creativity, emotional development, and ultimately, everything that makes us human.

We can also find some fuzzy similarities in its design: if a computer works with a transistor circuit through which current flows, we will find a very complex neuron circuit in our skull (1011 in each brain, just like stars in the galaxy, And 1015 synapses), in which the cyclic speed of the electrochemical signal is much slower: besides the electric pulse, chemical neurotransmitters are also used.

Synapses are more complex than electronic logic gates. The collection of neural connections is called connectome: Salk’s Institute calculates the storage capacity of the brain, in order of Pb, according to the number of connections (a Pb is 1 billion MB, like 6.7 million MP3 music CDs). In terms of processing speed, the working frequency of a neuron is kilohertz, which is one million times slower than the processor of a smart phone. The processor of a smart phone can work at kilohertz. That’s why silicon processors do logic math faster. So transistors and neurons are completely different: the brain is not digital, it doesn’t work with 1 and 0, it’s analog.

“The brain is a machine that can perform very complex operations efficiently: it consumes only 50 watts of power, less than a light bulb on a bedside table,” explains Francisco clarska. Professor of anatomy and human embryology at the Autonomous University of Madrid (UAM), who studies axiomatic networks of the brain related to attention, consciousness and intentional movement. “To achieve brain like functions, supercomputers like mare nostrum in Barcelona need a lot of energy,” the professor added. “It’s amazing that a brain does so much with so few things.”

Some projects try to simulate the brain. This is the case with the blue brain project launched by IBM and the Lausanne Institute of technology, or the European Union’s flagship project, the human brain project (HBP), whose third phase will end in 2023, bringing together scientists from multiple disciplines to better understand, learn and even simulate the brain. To represent a large brain or a part of it, you need to have a connection graph (set of connections), know its dynamics (mathematical form), and have a very powerful computer.

Simulating the brain, even partially, can help us understand the basis of its function, understand some diseases or develop drugs. Computational neuroscience is a subject that attempts to simulate the neural network of our brain and its interaction through computers and mathematical models. Neuromorphological computing (silicon brain) attempts to simulate neural connections, not on a computer, but physically, using tangible circuits. On the one hand, it helps to understand the function of the brain, on the other hand, it helps to improve technology.

“Any brain is a model in which you can draw inspiration and learn from what you’ve done over millions of years of evolution, from the most capable, resilient and effective solutions,” Kraska said. In addition, psychologist Gary Marcus and other researchers define the brain as a Kluge, an abbreviation for the English word “clumsy, lame, ugly but pretty good”: it is defective, full of patches and gadgets, because it is not designed, but the fruit of evolution. It is precisely because it is the result of natural selection, which is a progress for millions of years. Instead, computers are designed entirely by humans to perform their functions as efficiently as possible.

One of the most unique and surprising abilities of the brain is fast learning. Machines may be easier to handle multitasking, but they are hard to learn on their own. Artificial neural network is a part of artificial intelligence. It tries to simulate these abilities by developing machine learning. Javier de Felipe, a neurobiologist at the Ramon and Cajal Institute (CSIC) and director of the blue brain Institute in Spain, explains: “neurons have several dendrites to capture information, and then an axon to send a signal.”. Artificial neural networks try to simulate these systems with multi-layer neurons. (an important difference between the brain and the computer is that a transistor connects to two or three other neurons, and a neuron in the cerebral cortex can connect to hundreds or thousands of other neurons.) When there are a large number of levels, information processing becomes more complex. We are talking about deep learning. These technologies have made progress in speech, image or facial emotion recognition and computer vision. For machines, the function of the human brain is very complex.

Trying to reduce a brain to a computer is called reductionism. You can simply understand how it works, but it’s very incomplete. “The key difference is complexity: in the human brain, there are about one million synaptic connections, and the level of complexity varies,” explains Louis pastor, a professor at the University of Juan Carlos, king of Madrid. These levels of complexity range from molecules to neurons, networks or brain regions: it’s a very complex organ, no matter where it’s seen, from a detailed point of view. The shepherd explained that in order to simulate the brain on a large scale, we need computers with more computing and storage capacity than existing computers. “The amount of data that the analysis will get will also be a problem,” he stressed.

The concept of wetware (a bit like wet software) attempts to approach the brain computationally. It’s not software, it’s not hardware, it’s the third thing: wet stuff. This “humidity” refers to the brain’s plasticity and constantly changing neural connections, rather than the rigidity of the computer, as well as the above-mentioned ability to adapt and learn. “Although the brain processes signals like a computer, it doesn’t use silicon chips, it uses neurons, which are connected in a network of dynamic interactions,” Kraska said. The brain is always changing, and perhaps this neural plasticity is the main feature that distinguishes it from computers. For example, when a brain is damaged by any part of it, it can learn to work in another way. Over the years, we’ve lost a lot of neurons, but the system has withstood wear and tear.

In the field of philosophy, computational thinking theory proposed by Hillary Putnam and Jerry Fodor has been developed. According to this theory, the brain is physically based on brain activity and functionally equivalent to a computer, that is, a symbol processing machine that follows the rules in order.

For Fodor, the brain is modular, with different parts dedicated to music, mathematics or language. “These capabilities,” Fodor wrote, “operate through abstract algorithms, just like computers.” These abstract computers will be inspired by the universal machine of computer science pioneer Alan Turing, who predicted the future of computers. A machine that operates symbols according to certain rules and algorithms, regardless of the physical mechanism of symbols. Turing himself wondered whether the machine would think.

In these areas of philosophy, the question is: if we simulate a brain, will a thought appear? Do you have a conscience? It’s a mental problem, it’s about whether two entities are two different things or the same thing, and how they relate to each other. He has been challenging philosophers for centuries. Is brain a new brain property, just like ant colony’s collective thinking, an apparent phenomenon of neuron activity? Then the whole will be more than the sum of the parts. If consciousness, thought, self, is a side effect of brain activity, a bit like an unexpected mistake, it explains the absurd and meaningless feelings we experience when we are alive. Will this phenomenon produce machine consciousness, just like some science fiction movies? Everything is a mystery.

So far, what we know is that the computer is completely understandable, it is not the product of human thinking. The brain is the most complex object in the universe. “So far, we’re not sure we can understand all this,” Phillip concluded.