Transhumanism And The Future Of Humanity. Can human thoughts be encoded, decoded and manipulated to achieve symbiosis of the brain and the machine

This article discusses the current state of neurointerface technologies, not limited to deep electrode approaches. There are new heuristic ideas for creating a fast and broadband channel from the brain to artificial intelligence. One of the ideas is not to decipher the natural codes of nerve cells, but to create conditions for the development of a new language for communication between the human brain and artificial intelligence tools. Theoretically, this is possible if the brain “feels” that by changing the activity of nerve cells that communicate with the computer, it is possible to “achieve” the necessary actions for the body in the external environment, for example, to take a cup of coffee or turn on your favorite music. At the same time, an artificial neural network that analyzes the flow of nerve impulses must also be directed at the brain, trying to guess the body’s needs at the moment with a minimum number of movements.

The most important obstacle to further progress is the problem of biocompatibility, which has not yet been resolved. This is even more important than the number of electrodes and the power of the processors on the chip. When you insert a foreign object into your brain, it tries to isolate itself from it. This is a multidisciplinary topic not only for doctors and physiologists, but also for engineers, programmers, mathematicians. Of course, the problem is complex and it will be possible to overcome it only with joint efforts.

The global goal of projects like Neuralink (and quite a few research teams around the world are involved in creating brain-computer interfaces) is human improvement. This technology allows you to expand the capabilities of the brain by connecting external devices to it, such as the same computer. Intermediaries in the form of not only a keyboard, but also many different programs are now needed to interact with a computer. If Neuralink succeeds, the brain will communicate directly with the computer. What is it for? Today, to make a decision, one must take into account a huge number of factors that are difficult to analyze and sort through the shelves in his head. And here, in theory, implanted chips come into play and allow you to store in memory and easily process large databases. Which makes smart decisions more reasonable and reasoned. But this is a matter of the very distant future.

Today, brain-computer interfaces solve more mundane tasks, in particular they help completely paralyzed people (patients with strokes, serious traumatic brain injuries, cerebral palsy, ALS — Stephen Hawking’s disease, etc.) to communicate with the outside world. There are millions of such people in the world. In Russia alone, about half a million people suffer a stroke each year. One of the interfaces developed in Russia as part of the Neurochat project allows such people to enter messages on a computer screen with the effort of thought. Unlike Elon Musk’s Neuralink, this technology is not invasive, meaning there is no need to drill a hole in the skull. Patients wear headphones that pick up electrical signals from the brain. While the writing speed is low — 12–15 characters per minute.

The development of neurocomputer interfaces (NKI) began in the 1970s. NCIs are devices that connect the brain and the computer.

There are two types of NKI:

· Invasive interface (electrodes connected directly to the brain)

· Non-invasive interface (external devices that read information from the brain).

An example of a non-invasive interface is electroencephalography (EEG): a signal is transmitted through electrodes and the activity of the cerebral cortex is recorded. With the help of EEG you can control a speech machine or a wheelchair.

As early as the 1990s, Stanford University learned to split nerves, reconnect them, and then (using a chip) transmit information to a computer. Research has begun to replace biological neurons with artificial ones. The most advanced development in this industry has long been the artificial hippocampus for the rat.

For the past ten years, scientists have been trying to make an artificial neuron that can be transplanted. In 2019, a group of researchers from the University of Bath succeeded. They built a neuron on silicon microchips.

The difficulty was that depending on the excitation or relaxation of the brain area, as well as the direction of the signal, neurons react in different ways. A slightly stronger or weaker impulse that goes beyond the optimal range at this point will prevent “real” neurons from reading information. To create an artificial neuron, it was necessary to study the mechanisms of electrical conductivity of real cells.

Some companies are trying to recreate the brain in machine space. This requires a detailed map of the brain — the content. This is a complete description of the structure of nerve connections in the body. Drawing such a map is extremely difficult. In 1993, scientists modeled a content for a primitive, 1 mm long roundworm.

In January 2020, a team from the Howard Hughes Medical Institute completed a brain connection of the Drosophila fly, more precisely one third of it. The study took 12 years, cost $ 40 million and 250 scientists worked on the map, aided by neural networks from Google. The decoded part of the Drosophila brain is 20,000 neurons and 20 million connections between them. The human brain is made up of 100 billion neurons, each of which forms 15,000 connections.

In 2006, the Technical Institute of Lausanne modeled a column of rat cerebral cortex containing 10,000 neurons. The Blue Gene supercomputer with 8,000 processors was used to create the model. Simulating a brain column in a Blue Gene computer generates hundreds of gigabytes of information per second.

A rather sinister transfer idea has been proposed by Nectome. Starting promises to save and then charge your mind. It is assumed that he will perform high-tech embalming of the brain, which will remain unchanged for hundreds of years, and in the future it will be scanned and turned into a computer model. In order for the brain to be in good condition, the substances must be applied to a person who is still alive.

The creators claim that the procedure is legal because it is considered voluntary euthanasia. Nectome is negotiating with terminally ill patients and invites those interested to join the project. The startup has already performed cryopreservation of the brain of a deceased woman.

Earlier, the company received a grant from the US Federal Institute of Mental Health for research on objects with a whole brain.

Neuroscientists compare a dead brain to a switched off computer that still stores information. But most of them doubt the possibility of a “resurrection.” Eventually, the neurons are constantly renewed, and when they freeze, we will lose short-term information about the brain, a sense of self-perception, and continuity of consciousness. In addition, there is no guarantee that a person’s thought or memory can be obtained even from a complete connection.

Let’s assume that the transfer becomes a reality. Then humanity will have problems with electricity. The rat brain experiment took 10 years. Advanced supercomputers consume about 20 MW of power per hour (as a small town). If we have to maintain these computers (and maybe hundreds of times more powerful) to create massive human simulations, we will have to find alternative ways to generate electricity.

Semiconductor quantum dots have been used successfully to stimulate brain cells. With Alzheimer’s disease, epilepsy and a number of other diseases, non-surgical methods are needed to stimulate brain cells.Quantum dots are one of the best solutions to this problem.In the not too distant future, they will be used to reactivate damaged retinal cells to restore vision to the blind.

Many brain problems are associated with unbalanced neuronal activity, manipulation of certain neurons can restore normal levels of activity.

Existing methods for stimulating brain cells have several disadvantages. The incorrect signals that the neurons of a patient with parkinsonism emit, causing tremors in the arms and head, can be suppressed, but the necessary electrodes require implantation in the brain, which is performed surgically.

The method of magnetic stimulation of cells does not require surgery to open the skull. It is true that when used, too many cells, including normally functioning ones, are exposed to stimulation.

Other researchers have been able to stimulate genetically modified gray matter cells with a stream of light, but how many patients would agree to genetically modify their brains?

As you can see, a significant part of the roadmap for these projects is not as transparent as we would like.

In the next 10 years, we are unlikely to have a comprehensive interaction with the whole brain … Neuralink is not something like neural lace … Nonehow to put 100 million [threads or electrodes] in the brain … There are practical limitations in terms of tissue destruction, heat dissipation and computing power … I share this vision [of radical interfaces with the whole brain], but we will learn to do this [brain interface development] in parts, with many different applications and many brain areas, in the foreseeable future … So one is still far from realizing the idea of ​​merging with AI.

Our thoughts will remain with us, their decoding is impossible. We encounter certain physiological barriers. Our consciousness spreads to 86 billion nerve cells in the brain. Connecting to any neuron is unrealistic. It is not even about the cells, but about the contacts between them, where the issue of information transfer is resolved. We don’t even know exactly where our memories are stored. Every fact immersed in memory has an emotional, auditory, visual component. Therefore, to read a thought, it is necessary to register this activity in different parts of the brain, to know the codes for decoding signals and to be able to combine information from different areas into one whole. Therefore, so far, even in the future, there is no possibility of connecting with memory, emotions and the thought process.

You can read my full article on the subject here

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Nikolay Raychev

Nikolay Raychev

An expert in software process improvement and software engineering technologies with two decades experience as a software engineer, a software architect, a CIO,