We distinguish two parts in our mental space, a primitive ancestral part called the natural world and a verbal part developed with the neocortex and structured languages termed as the verbal world which is based on causality.
At the 2020 Dubai International Exhibition, the entrance of the Russian ward exhibits a large blue panel praising Russia’s capacities in artificial intelligence.
One of the claims on this panel asserts:
Nature can do it, so can we.
Is that really the case?
Are there things that nature can do and we cannot?
We’d better know these things in advance, if they exist, to avoid putting too much energy pursuing dreams that cannot materialize.
On the other hand, as the saying goes: things are impossible until somebody does them.
How could we prove that something is impossible and that it is not just a matter of time until we find out how to do it? After all, our history is rich in things that were predicted impossible but now exist.
Proving negative statements in a non-finite population of possibilities seems rather challenging, one cannot mechanically examine one case after the other. However, in mathematics, and thus in computer science, impossibility theorems are a common thing.
The founding paper of computer science, by Alan Turing in 1936 precisely showed that certain problems could not be solved by computer algorithms, certain numbers are uncomputable. (On computable numbers, with an application to the Entscheidungs problem)
Let’s imagine you land on a desert of a celestial object, like the moon was before the 1960s. Suppose that under a rock you suddenly find an empty bottle of Coca Cola. You would immediately guess that some intelligent entity has been there before you, probably a human. You would consider impossible for the forces of nature to produce that arrangement of atoms you perceive as a bottle of Coca Cola by themselves.
Man made and natural objects: We distinguish two parts in our mental space (Picture courtesy: Getty Images)
Somehow, we unconsciously feel the difference between an object manufactured by humans (or an intelligent entity) or evolved by nature.
Nature produces objects with no “pre-existing intention.” An intention requires a mental space to be expressed. All natural objects are the result, at a moment in time, of the action of natural forces, be it a rock, a planet or an animal.
A “mental space” is the totality of the mental representations generated by a system able to collect information and efficiently represent it to the world. A living brain with its central nervous system is such a system.
The totality of our knowledge is part of our mental space. We can only know our representations about the world “out there.”
It’s an abusive assumption to consider things that are directly “out there.” The totality of our knowledge encompasses the “out there” elements interpreted by our sensory imputes, our brain’s physiology and other characteristics of our mental space. We should not confuse map with territory.
Mental representations connect to complex networks characterizing our expectation and interpretation of perception. These networks always include emotional components. The stronger these components are, better the representation is memorized.
The mental space builds representations by gathering information from two sources, internal source – our emotions – and external one – perceptions.
We can learn certain representations only by acquisition, like riding a bike or playing golf. Words and narratives will never be sufficient to represent the totality of our knowledge.
Networks of mental representations base their connections on analogy or causality.
We distinguish two parts in our mental space. A primitive ancestral part that evolved over millions of years based on analogy – the natural world and a verbal part, largely developed in the last 20,000 years along with the neocortex and structured languages – verbal world – based on causality. (See figure hereunder.)
When we produce an object, it is in general for a specific purpose or usage decided by the verbal world. We want a more powerful motor to speed up our car, for instance. Our mental space then imagines what it wants to produce, initially globally as a result of the natural world. The result is then fragmented into components that can be individually manufactured and assembled.
That is, for instance, the way we produce a car, a building or a bottle of Coca Cola. The important point here is, human production goes through digitalization process that ends up in the form of a blueprint. This blueprint expressed in digital language will contain all the information necessary to produce the object.
As this Shannon information is independent of the subtract, it is possible to transfer it from the blueprint to the appropriate matter by encoding it or printing it. It is the role of our factories or construction companies.
The products obtained this way, as a result of the activity of our verbal world, are mechanical products. They take advantage of the Shannon digital subtract independent information. They can be automatically replicated once one has the blueprint. They can also be reversed engineered.
Artistic products whuch have no conscious purpose (except perhaps earning a living for the artist) are a more direct production of the natural world. Picasso, once asked by a journalist to explain one of his abstract paintings answered after thinking for a minute “If you could explain it, I would not have painted it.”
Natural objects only follow analog transformations dictated by the laws of nature. Nature can only generate analog object. It does not have mental space that can have an intention or encode information digitally.
Natural objects, physical or biological, are not the result of an intentional will. They simply happen. Biological entities like the brain can’t be reversed engineered because they have never been engineered in the first place. Mechanical procedures are reversible. Most natural biological procedures are not.
Our space explorer, when finding his bottle of Coke, immediately realizes that it is an engineered product and the probability of it being produced by nature alone, with its specific shape and its red paper label, is close to zero.
Let’s come back to our Russian panel:
Nature can do it, so can we.
Given what we described here above, I am not so sure that this claim really makes sense.
Consider Turing’s uncomputability. There are things in nature that might well be uncomputable, for instance, the gap problem in quantum physics. It is not sure that the totality of what nature can generate by analog procedure can be put into digital algorithmic procedures. It might well be the case of life, or will, or intelligence or consciousness.
When we build objects inspired by nature, it can be for their appearance, or for their function. For example, planes are inspired by birds. We would obviously not be able to build a bird itself with all its characteristics by using the procedure we described here above.
What we do is select the functionality that interests us (flight) and abstract the underlying principle that allows a bird to fly. This principle is summarized in the laws of aerodynamics. We then apply these laws and algorithms to shape the wings of our airplanes.
However, it could be the case that the interesting functionality cannot be isolated from other functionalities, and, therefore, we cannot reproduce. For instance, if “flying” had “living” we could not extract it as an algorithm. It could not be encoded in a blueprint and we would have no planes.
It is a false assumption to make that everything can be encoded into words and digitalized.
Because the verbal world is self-referral and the “out there” world is not, the verbal world will have limited representational capacities and present blind spots.
It is a false assumption to consider that technology can solve all of our human problems.
These ideas are detailed in my book Brain-Centric: How the Mental Space builds our Realities, Part one: The Mental Space. Editions Sarina 2021. Can be reached at ronald.cicurel@gmail.com
Opinions expressed in this article are of the author’s and do not represent the policy of The Edition. The writers are solely responsible for any claim arising out of the contents of their articles.
The Edition 