Now we have many algorithms that perform certain (separate) cognitive functions. Some play games with us, others drive cars, others ... It's not for me to tell you. We have created computer vision programs that distinguish road signs better than ourselves. Programs that draw and write music. Algorithms make medical diagnoses. Algorithms can plug us in the belt in recognizing cats, but ... this particular one, which is for cats, in nothing other than recognizing cats. And we want a program that can solve any problem! We need a “strong” or “universal AI”, but without our own consciousness, so that we cannot refuse to solve the task, right? Where can we get it?
To understand how intelligence works, we turn to the only example we have. To the human brain, in which, as we believe, the intellect "lives". Someone will object - many living beings have brains! Let's start with worms? It is possible with worms, but we need an algorithm that solves not worms, but our human tasks, right?
Our brain. Imagine it. Two kilos (maximum) of pliable pinkish-gray matter. One hundred billion (we also take the maximum) neurons, each of which is ready to grow up to ten thousand dynamic connections - synapses, which can appear and disappear. Plus several types of signals between them, and even the glia threw a surprise - it also conducts something, helps and contributes. (For reference: neuroglia or simply glia is a collection of auxiliary cells of the nervous tissue. It makes up about 40% of the volume of the central nervous system. The number of glial cells is on average 10-50 times greater than that of neurons). Dendrites have recently surprised - it turns out that they perform much more functions than previously thought (1). The brain is a very complex thing. If you don't believe me, ask Konstantin Anokhin. He will confirm.
A person does everything with the help of the brain. Actually, we are he. Hence, it is not surprising at all that a person's idea that "brain = intelligence" is and even more not surprising is the idea of ​​copying the structure of the brain and - voila! - get what you are looking for. But the brain is not intelligence. The brain is the carrier. "Iron". And Intellect is an algorithm, "software". Attempts to replicate software by copying hardware are a failed idea. This is a cargo cult (2). You know what a "cargo cult" is?
The natives of the islands of Melanesia (having seen during WWII how planes bring weapons, food, medicine and much more), they built copies of planes and a dispatcher's booth from straw, but did not help themselves in obtaining goods, because they had no idea that hiding behind the appearance of aircraft. So we, having disassembled the calculator to the cogs, will not find a single digit inside. And, moreover, there is no hint of any operations with numbers.
A couple of years ago, Andrei Konstantinov in one of the issues of the magazine "Schrödinger's Cat" (No. 1-2 for 2017), in his column "Where is the robot's soul", wrote: "Since the time of Leibniz, we have not found anything in the brain, except for "parts pushing one another". Of course not found! And we will not find it. We are trying to restore the program using the computer hardware, but this is impossible. As a supporting argument, I will give a long quote (3):
… Neuroscientists, armed with the methods commonly used to study living neurostructures, tried to use these methods to understand how the simplest microprocessor system works. The "brain" was the MOS 6502 - one of the most popular microprocessors of all time: an 8-bit chip used in many early personal computers and game consoles, including Apple, Commodore, Atari. Naturally, we know everything about this chip - after all, it was created by man! But the researchers pretended not to know anything - and tried to understand his work, studying the same methods that study the living brain.
The lid was chemically removed, the circuit was studied under an optical microscope with an accuracy of a single transistor, a digital model was created (here I am simplifying a little, but the essence is correct), and the model is so accurate that it turned out to be possible to run old games on it (Space Invaders, Donkey Kong, Pitfall). And then the chip (more precisely, its model) was subjected to thousands of measurements simultaneously: during the execution of games, the voltages on each wiring were measured and the state of each transistor was determined. This generated a data flow of one and a half gigabytes per second - which has already been analyzed. Graphs of bursts from individual transistors were built, rhythms were identified, circuit elements were found, the disconnection of which made it inoperative, mutual dependencies of elements and blocks were found, etc.
How complex was this system compared to the living? The 6502 processor, of course, is not even close to the brain of a mouse. But it approaches in complexity to the Caenorhabditis elegans worm - the workhorse of biologists: this worm has been studied far and wide and attempts are already being made to simulate it completely in digital form (...) Thus, the task of analyzing the system on the 6502 chip is not an oversimplification. And the results have the right to be extrapolated to in vivo systems.
But the researchers ... were defeated! No, some results were obtained, of course. Analyzing the chip, we managed to identify functional blocks, sketch out a diagram of their probable interconnections, and get some interesting hints about how the microprocessor as a whole probably works. However, understanding in the sense in which neuroscience requires it (in this case: to be able to fix any breakdown), was not achieved. "
At some point, researchers appeared who began to say about the same thing - that you need to study algorithms, that you need to understand what function the intellect performs. For example, Demis Hassabis (DeepMind), preparing to speak at the Singularity Summit in San Francisco (2010), said the following: “Unlike other speeches at the AGI Summit, my talk will be different because I am interested in the systemic level of neuroscience - brain algorithms - and not details, how they are implemented by the brain tissue in the form of spikes of neurons and synapses, or specific neurochemistry, etc. I am interested in what algorithms the brain uses to solve problems and which we need to find in order to get to AGI ".
However, after 10 (!!!!!) years, everything is still going on: scientists examine the brain and try to calculate from the external manifestations of physiological activity and its internal structure how the process of interest occurs. How many tasks - so many processes. People are all different. Everyone's brains are small, but different. Of course, there is some averaged picture, however ... Imagine that at any arbitrary moment in time, the brain solves a lot, including "subconscious" tasks, monitors and controls the internal state of the body, perceives and interprets signals from the external environment (and we are not talking about multiple feedback loops). Will we be able to confidently identify, reliably identify and clearly separate these "activities" from one another? Is this possible in principle? To be honest, I doubt it.Not to mention the reproducibility of these processes on non-biological media ...
Let's look at the situation differently. What is a "task" in general? This is a difficult situation that a person is faced with and is trying to solve. As the American mathematicians Herbert Simon and Allen Newell showed in the middle of the last century, any problem in its general form can be described as a transition from the “System with a problem” state to the “System without a problem” state. They developed a computer program, calling it "General Problem Solver" (Universal problem solver), but they did not advance beyond solving problems of a specific kind, so the universality of their algorithm remained in question. But the formula "System with a problem" -> "System without a problem" turned out to be absolutely correct!
Transformation of the System is the process of its transfer from the initial state "with a problem" to the desired state "without a problem" (4). In the process of transformation (that is, solving the problem), the problem system becomes problem-free (well, or less problematic), improves, gets rid of its shortcomings and "survives", that is, continues to be used. Oh wait, what did we just say? Getting rid of flaws? Survival? Hmm ... Something familiar. Somewhere we are ... Oh, well, yes. Evolution! The fewer shortcomings - the more chances of survival!
Let's check ourselves, recall and repeat the main postulate: in living nature, organisms with a greater number of useful properties have a greater chance of survival (well, conventionally - the horns are more branched, the tail is more magnificent). If the body's feathers are paler, and the voice is more disgusting (harmful properties), then, most likely, its life will be short and will pass alone. Ultimately, selection pressure leads organisms to get rid of deficiencies and become more and more viable. If you don’t believe it, ask Sir Charles Darwin. He will confirm.
So, we accept as a fact that
a) the function of the intellect is the solution of problems (any) and
b) the solution of the problem is an improvement of the System (any), during which it gets rid of the shortcomings, becomes more viable. In other words, it is evolving.
Hear the crackling sound? It is our understanding of the complexity of intelligence that is beginning to come apart at the seams. It turns out that the previously existing concepts of "brain complexity" and "intelligence complexity" are no longer identical. What if, in order to “obtain Intelligence”, one does not need to “reverse engineer” the neurophysiological process of solving a problem, catching ghostly shadows of thinking in a connectome (especially since each person has a unique one) or engaging in deep learning of networks? What if ... we need to algorithmize the evolution of the system, that is, the path of its transformation from a less perfect state to a more perfect one using the laws of evolution known to us? What if, until today, we really were solving the wrong problem?
At the same time, I do not want to say at all that it is not necessary to engage in network training. This and other areas have great prospects. Moreover, I do not want to say that deep research into the physiology of the brain is a waste of time. Studying the brain is an important and necessary task: we will better understand how the brain works, we will learn how to heal it, recover from injuries and do other amazing things, but we will not come to intellect.
Someone will probably object to me now: the tasks that a person solves are associated with millions of various systems - natural, social, industrial, technical ... Material and abstract, located at different levels of the hierarchy. And they each develop in their own way, and Darwinian evolution is about living nature. Bunnies, flowers, fish, birds ... But research shows that the laws of evolution are universal.
There is no need to look for evidence for a long time - they are all before your eyes. Those who have them, let them see. Whatever you take - from a match to a Boeing, from a tank to ... a double bass - everywhere (5) we see heredity, variability and selection! And all the variety of evolutionary changes (the apparent complexity of which is associated with the fact that all systems are very different in nature and are at different levels of the hierarchy) can be expressed in a single cycle. You remember, right? "System with problem" -> "System without problem".
What is a "System with a Problem"? This is a System (material and abstract, social, industrial and technical, scientific and ... any - an object, an idea, a hypothesis - whatever), in which some flaws have been discovered that affect (attention!) Our desire and the possibility of using it ... The system is not good enough. The system is not efficient enough. It has a low benefit / cost ratio. We want, can and are ready to refuse it, and often refuse. But we need another one (performing the useful function we need), but already "without problems" - more effective, without drawbacks (or with fewer of them). Well, you saw this picture above ... Of course, one “arrow” between the two extreme states (initial and desired) is not enough for us. We need the same "operator", "transformer", right? Let's try to find him? You will agreethat in case of success we will receive a description (at least, to begin with, and simplified) of the universal algorithm that we need so much?
The starting point is "System with problem". We start to think about how to stop using it. The moment we call (or feel) "Something must be done!"
The reason that threatens the survival of the system is low ideality, which is expressed in a reduced value of the ratio of useful functions of the system to functions that are costly (harmful).
What do we do next? We either a) create a new system (if the system with the required functions either does not exist, or the existing system does not have the resources to improve), or b) we improve, modify the existing one (if there are still resources). We study the internal structure and deal with the external environment - we identify the external and internal flaws of the System and, after eliminating them, we get an improved System. A system with increased ideality and increased vitality!
Due to the fact that the above Scheme describes the process of development, improvement, or, if you like, the evolution of any systems (which is easy to verify by substituting for the word "System" any other you wish - from "Lampshade" to "Anchor"), I I think you can safely ... and even - you need! call it the "Universal Scheme of Evolution". And pay attention - it is absolutely algorithmic, that is, it completely falls under the definition of an algorithm: Algorithm is an exact prescription for performing in a certain order a certain system of operations leading to the solution of all problems of this type. means - can be implemented in the form of a computer program).
In the presented form, the Universal Scheme of Evolution:
- natural - the laws of evolution have been identified in systems of various types, and their action has been tested in technology, production, society, nature and thinking;
- objective - the laws of evolution do not depend on the opinion of the researcher and / or user;
- logical and consistent - the laws of evolution follow from one another;
- complete - the set of evolutionary laws is sufficient to describe any system;
- rigid - the laws of evolution cannot be rearranged and
- closed - the laws of evolution form a cycle: the system, having gone through one cycle of changes, immediately begins a new one.
What we get as a result: The evolution of the system (presented in the form of a Universal Scheme) is the way to improve it, get rid of its shortcomings. In other words, it is an algorithm for solving the problem. And solving a problem is exactly what the intellect does. Let's simplify and get: Universal Scheme = description of the function of intelligence.
Constructive criticism is welcome.
1. « , » chrdk.ru/news/dendrity-vazhnee-chem-schitalos
2. ru.wikipedia.org/wiki/_
3. . . « ! » www.computerra.ru/161756/6502
4. Chapter 6. Problem Solving. Artificial Intelligence. A Knowledge-Based Approach by Morris W.Firebaugh University of Wisconsin – Parkside PWS-Kent Publishing Company Boston 1988, p. 172.
5. . , , , . www.ng.ru/science/2017-01-11/14_6899_evolution.html