What Is Intelligence?
General intelligence can’t be achieved based on the current golden standard artificial intelligence algorithms because it’s absolutely different from them by its very nature. Here I want to light up the foundation of the true and invariant general intelligence present in humans and other animals.
The Picture of the World
The world view of Plato, Aristotle, Avicenna, Albert Einstein, Max Plank, Erwin Schredinger, Ivan Pavlov, Claude Levi-Strauss, Norbert Wiener, John von Neumann, Edward Tolman, Alan Turing, Endel Tulving, Andrey Kolmogorov and many more outstanding scientists of the past and of today rests on their passionate curiosity, their ability to wonder, their love of resolving the mysteries of nature and the acceptance of the fact that any knowledge is incomplete unlike imagination that encircles the entire world.
What is humanness? How did human species emerge? How do our brains work? What is intelligence? How humanness and intelligence interrelate? Is intelligence contingent to humanness or vice versa? Many thinkers and scientists tried before and keep trying now to find answers to the above questions. Following the links in this post you will find an overview of their thoughts.
Shrinking Brains
From that very moment when 20 thousands years ago humankind stepped on the way of the industrial revolution the evolutionary pressure on humans to make autonomous decisions in the face of uncertainty began to evaporate. Artificial selection began to favour obedient but stupid over independent and smart.
Human brains shrank by some ten percent (the size of a golf ball) over that period of time. That brain shrinkage doesn’t impose any immediate danger to a person. Scientifically documented cases show that humans can survive and act normally under stable conditions while retaining only 10% of their brain tissue.
The existential risk for humanity emerges from the condition that the huge and complex body of the modern civilization (all our infrastructure, cities, industry and technology) is being managed by a brain that is getting smaller and smaller.
The brain of our civilization — the collective intelligence of all people is represented currently by a tiny layer of decision makers, people whose ability to take independent decisions is also diminishing with each new generation as well as with ageing of the older ones.
We need to embrace the art of understanding to revert this negative trend and to open future horizons to humanity in us and in the Universe..
The Genesis
There’s only one way how people can create and update their picture of the world. It’s invariant and it has nothing to do with reinforcement, the currently uncontestable cornerstone of psychology, neuroscience and artificial intelligence.
Paradoxically, the most accurate description of this method was given not by a psychologist, neither by a neuroscientist but by Max Plank, a physicist who was granted a Nobel prize for the discovery of the elementary quantum of action and the entire field of quantum physics.
The evolution of the scientific concept of the art of understanding (science-of-concrete) went through the following phases:
Ivan Pavlov discovered that animals can establish cause-effect relationships between external objects without any interaction (conditioning or response reinforcement) with them.
Edward Tolman proposed that such relationships are sketched on a sort of a map in the minds of animals.
Jean Piaget suggested that Tolman’s maps are schemata that humans use for thinking.
Claude Levi-Srauss discovered that the thinking process based on schemata of relationships between external objects is the basis of thinking of primitive tribes. He underlined the importance of categorisation of everything in the process that he named ‘science of concrete’.
O’Keefe, Nadel and Mosers discovered place and grid neurons as a substrate physiologically underlying cognitive maps in the brain.
George Lakoff proposed conceptual metaphors as the basis of categorisation of everything by humans. The core of a category is a prototype, a kind of schema, according to him. Schemata is a metaphor of the external world.
Vassili Nalimov extended Bayesian inference to deal with external priors (meanings packed on a continuum of semantic vacuum) for the establishment of internal priors (meanings unpacked into ‘texts’).
Michael Rabinovich proposed that stable heteroclinic channels established by neurons are the mathematics underlying schemata of thinking.
The Art of Understanding
The founder of the field of computational mechanics James Crutchfield defines “fully-automated “artificial science” as the one “in which theories are automatically built from raw data.” The Purpose of this book is to prove that science is based on understanding that automatically builds new theories by superposing seemingly unrelated existing theories represented by low dimensional cognitive maps, schemata or conceptual metaphors as different scientists named them.
Only when such a low dimensional understanding of a new theory will be created (invented, discovered), only then can it be tested by applying it to raw observational data either available empirically in the environment or obtained by the organised unpredictability of a scientific experiment.
The low dimensional topology of the maps of understanding may be supported by hierarchical networks of asymmetrically synchronised (coupled) chaotic oscillators as mathematical entities which to date represent most accurately relevant processes in nature from hexagonal trimers of dimers of bacterial transducers to neuronal ensembles of hexagonal lattices in human brains.
Stable heteroclinic channels (SHCs as per Michail Rabinovich) or central pattern generators (as per Lewis Dartnell) of chaotic oscillators may represent not only different concepts but also different past episodes (causal states as per James Crutchfield) lowering down the extra memory cost usually incurred while inferring causes from effects. The inference will not be inverted by them. Rather they will enable the mathematical model of the system to travel back in time to the point where a cause preceded the effect and to infer the effect that already occurred from the cause that occurred even earlier.
So the model will not reverse the arrow of time but rather will fastwind time back and play-back the necessary episode forward from the causal state (distribution of different probable pasts) preceding the effect.
Mental Travel Across Different Pasts
Understanding is a mental search for the cause of something that has already happened, that already exists. Traveling into the past, even if mentally, is an extremely costly process in terms of the amount of necessary memory and processing power. It’s much more expensive than forecasts for the future.
The reason is causal asymmetry resulting from causal equivalence discovered by James Crutchfield. All variants of the past are equivalent if they lead to the same present.
You can create a million models of the past, but today is impossible to choose which one is true. One can only introduce some kind of an artificial criterion. For example, which model is more economical, that model is true. But this is not so.
Understanding is finding the true model in the past. To do this, we have Endel Tulving’s episodic memory that allows us to travel across a plane of our past, and not as along a line. One-dimensional and irreversible time becomes two-dimensional and reversible.
We no longer have to go back and forth on the arrow of time each time, checking each of the countless models of the past, each of which explains the present. We can see them all as routes on the plane and compare.
The shortest route will be true. It is also the most beautiful. It is the opened bit of the omega number according to Gregory Chaitin. It is the meaning unpacked into a “text” from a semantic vacuum according to Vassili Nalimov.
Computers and modern systems of artificial intelligence are not yet able to reproduce the process of understanding. Quantum systems seem to have no causal asymmetry, but this still has no applied value. Moreover, they can walk only back and forth along the arrow of time without a punishment for moving back. The ‘intelligence’ of even the simplest living creature can roam in the past along and across. And moving across is no less important than moving back. Among all life forms humans have probably the deepest and the widest plane of the past to explore — a cognitive map of the episodic memory.
We can also walk diagonally across time to check different possible causes of the same effect. Nevertheless, we often grab onto the first model we come across and begin to develop it logically. And it always turns out very well. But all the rationales are on a foundation which truth or falsity cannot be determined. Usually, conspiracy theorists are blamed for this approach. Although the official versions are no better if they are not based on understanding.
It is a mistake to think that it will be easier to understand the problem if we accumulate more data. Rather, the opposite. Understanding is an introspection of a prepared mind. There is no other way to understand. Attempts to replace understanding with computations only lead to the accumulation of errors, and not to their elimination. A critical mass of mistakes may lead to the dropout of the organism (either a microbe or a human being, or a human society) from its ecological niche. Only understanding differentiates us from the environment. Free energy isn’t free after all.
Ivan Petrovich Pavlov coined the term “reinforcement” to refer to the connection between two stimuli, which is formed in the psyche of an animal if both stimuli act on it simultaneously.
Thirty years later, Pavlov realised that the connections between various phenomena of the external world can form in the psyche of an animal even if these phenomena are not directly related to the animal. He called this method of forming an understanding of the world that wasn’t based on reflexes the embryo of science, because the behaviour of an animal that purposefully examines the environment was no different in his opinion from the behaviour of a scientist studying a scientific problem.
Here we call Pavlov’s ‘germ of science’ the art of understanding or the science-of-concrete depending on the context.
Pavlov, however, noted that science is increasingly gaining new knowledge from the already accumulated old knowledge but not from observations. So Pavlov’s ‘reinforcement’ over the course of a century went around the whole planet with a victorious march. Indeed, nothing more was needed for the exploitation of old knowledge. Reinforcement turned out to be a simple but effective method of learning things without understanding them. That’s why people forgot about understanding. What for do we need to understand again things that have already been understood?
But, it turns out, there is a reason. Each scientific paradigm begins with an understanding that is given by a metaphor that defines this paradigm. Over time, the old metaphor becomes transparent. An opportunity to use only literal definitions emerges. A paradigm shift — the development of science requires a new metaphor, and therefore a new understanding.
If the priests of the temple of science have lost the ability to understand, then the development of science has stopped. Instead of living and changing paradigms, only a set of dogmas remained.
Understanding is the basis of humanness. The shared cultural evolution of people is based on the ability to understand another person, not a message, not a set of symbols, but a human being as an individuality.
Therefore, the current ‘corona crisis’ was a crisis caused by the lack of understanding, the deficit of humanness as a fundamental cognitive property of a person as an individuality. Without understanding and humanness, people will not survive in the digital jungle. If we shut down the portal into the digital dimension that was opened for us by Alan Turing and John von Neumann, we’ll slide straight into the harsh reality of the Stone Age, not into the fictional idyll of the technological singularity.
