Posted 31 марта 2021,, 10:14

Published 31 марта 2021,, 10:14

Modified 24 декабря 2022,, 22:36

Updated 24 декабря 2022,, 22:36

There will be no artificial intelligence because the brain is not a computer

There will be no artificial intelligence because the brain is not a computer

31 марта 2021, 10:14
Experts say that it will take more than one century to understand how the human brain works.

The famous American psychologist, analyst at the American Institute for Behavioral Research and Technology in California, former editor-in-chief of Psychology Today magazine Robert Epstein published an extremely interesting material on the website of the scientific publication Aeon that the mystery of the human brain will not be solved yet soon, so the creators of artificial intelligence need not worry. Here, in particular, what he writes:

“No matter how hard brain researchers and cognitive psychologists try, they will never find in the brain a copy of Beethoven's 5th Symphony, or copies of words, pictures, grammar rules or any other impressions of the environment. Of course, the human brain is not really empty. But it does not contain most of the things people think about, or even things as simple as "memories".

Our understanding of the brain has deep historical roots, but the invention of computers in the 1940s completely baffled us. For more than half a century, psychologists, linguists, neuroscientists and other experts on human behavior have argued that the human brain works like a computer..."

That is, our brain does not process information, does not extract knowledge and does not store memories. He is not a computer or an information system.

“...Think how complex the problem is. To understand even the basics of how the brain supports human intelligence, we may need to know not only the current state of all 86 billion neurons and their 100 trillion interconnections, not only the different strengths with which they are connected, and not only their states of more than 1000 proteins that exist at every junction point, but how the current brain activity contributes to the integrity of the system.

Add to this the uniqueness of each brain, due in part to the uniqueness of each person's life story, and Candela's prediction starts to sound overly optimistic.

(In a recent article in The New York Times, neuroscientist Kenneth Miller suggested that it would take "centuries" to figure out basic neural connections).

In the meantime, huge sums of money are being raised for brain research, based in some cases on erroneous ideas and promises that cannot be kept.

The most egregious example of a neuroscience failure documented recently in a Scientific American report concerns the $ 1.3 billion Human Brain Project launched by the European Union in 2013.

The charismatic Henry Markram was convinced that he could create a simulation of the entire human brain on a supercomputer by 2023 and that such a model would revolutionize the treatment of Alzheimer's and other diseases, EU officials funded his project with virtually no restrictions.

Less than two years later, the project turned into a "brain wreck" and Markrama was asked to resign.

We are organisms, not computers. Accept this. Let's continue to work, trying to understand ourselves, but without burdening ourselves with unnecessary intellectual baggage.

The metaphor of the brain as an information system has existed for half a century, and little has been done on this path. It's time to press the DELETE key..."


However, network analyst Fyodor Kuznetsov believes that to create Artificial Intelligence it is not at all necessary to thoroughly know how the human brain works:

“And for utilitarian purposes, you don't have to try to copy the whole brain. The brain is an extremely redundant system due to the imperfection of the electrochemical nature of signal processing. Elementary stimulation activates beams of at least 700-1000 neurons in order to ensure the stability of the passage of the electrochemical signal, as well as to smoothly adjust the output signal. For an artificial brain, the first is not a problem at all, and the second is solved by choosing a suitable activation function.

Thus, an artificial brain can be obtained by reducing the present by two orders of magnitude - and if we discard all crap from emotions and control of the vital activity of the body and the brain itself, then by three orders. Moreover, since the brain is imperfect, the goal of copying the brain is completely harmful and artificially limits us in striving to surpass nature in the ability to process information and make decisions. The most important thing: it should be remembered that the mind itself appears only in the complex of brain-sensors.

The shovel will not dig itself, the shovel only has the ability to dig. You need a digger to dig. The brain itself will not work productively without sensors - all the more, if all sensors and their controllers are completely turned off, consciousness will completely disappear. The most important question: how many sensors are needed for the birth of consciousness? One, two, three or more? I am inclined to believe that no less than two. One - for the actual reaction to the outside world. The other is for analyzing the output of the first sensor. The more sensors, the more levels of complexity at which consciousness is able to work..."