Artificial intelligence in the relationship between science and the humanities

Indice

The debate on artificial intelligence has distant roots: the gestation of artificial intelligence is traced back to a time window between 1943 and 1955 and, from the very beginning, the importance of the relationship between the scientific and humanistic disciplines was understood.

Today, this relationship seems slightly unbalanced and nothing good can come from this imbalance.

The eight pillars

In his paper‘A Comprehensive Overview of ArtificialIntelligence‘ (December 2022), Kouassi Konan Jean-Claude clarifies some fundamental aspects of artificial intelligence and its emergence. In particular, he focuses on the eight pillars that, more than others, support the development of this technology.

  1. Philosophy
  2. Mathematics
  3. Economy
  4. Neuroscience
  5. Psychology
  6. Computer engineering
  7. Control theory
  8. Linguistics

These eight disciplines are the founding basis of artificial intelligence and enable its harmonious and timely development.

Philosophy

Philosophical input is essential to be aware of the relationship between thought and the external world, between the metaphysical and the physical. Even if in everyday life human beings tend not to fix their attention on the former, we know for certain that self-perception has a radical and essential impact on human thought and perception of reality.

Mathematics

The contribution of mathematics as a discipline capable of systematising logic is clearly indispensable. Without mathematics, the possibility of opening a procedural dialogue with any artificial system would be lost, but nevertheless, it has substantial limitations. This was said by Turing himself in Alan Mathison Turing’s book ‘The Indecibility of Life’:

Incompleteness is expressed by saying that no effective procedure can prove all arithmetical truths. Therefore, it is impossible to express all mathematics in a single system.

In Turing’s sentence, it is assumed that there is no single algorithm that can provide an answer in a given time on all the variables of a problem. Multiple procedures must be used for multiple problems and this requires the use of differentiated approaches.

Despite their remoteness from sense experience, we also have something akin to a perception of the objects of set theory, as can be seen from the fact that the axioms themselves force us to take them as true. I see no reason why we should have less faith in this type of perception, i.e. mathematical intuition, than in sensory perception, which leads us to construct physical theories and expect future sensory sensations to accord with them

Kurt Godel

Economy

Economics is not just the discipline used for financial analysis, indeed that aspect is perhaps one of the least interesting; economics actually is:

Rational use of money and any limited means, aiming to obtain the maximum benefit for the same expenditure or the same result with the minimum expenditure.

Source: Treccani(Link)

The attribution to the financial sphere depends on its Greek etymology οἰκονομία: οἶκος ‘dwelling’ and -νομία ‘-nomy’ (propr. ‘house administration’). This involves administering limited resources, which forces administrators to plan and use method in the distribution and allocation of resources. Consequently, economics revolves around methodological models that are studied on a case-by-case basis and applied according to ever-changing and highly complex logics.

Neuroscience

Neuroscience is the discipline that studies the human nervous system and, in order to do so, systems and integrates notions from anatomy, biochemistry, physiology and psychology. In essence, we speak of a ‘collector’ discipline, whose power is to provide new conclusions by combining information from other sciences. The contribution of neuroscience is essential: those who deal with artificial intelligence, for example, are used to reasoning by mentioning neurons even outside the biological domain. For the sake of clarity, here is a table comparing neurons within biology and those within computer science.

DescriptionBiological neuronArtificial neuron
Nature of the connectionOrganicArtificial
Number of neuron types10,000 specific typesless than 100
Average number of connections per neuron10.000-100.000less than 10,000
Average number of distinct active connections per neuron (with its other neighbours)10.000less than 10,000
Number of active connections between two neurons31
Number of receptors per connection5.0001
Source: “A Comprehensive Overview of Artificial Intelligence”, Tab.1 Page 6

It is impressive to note that the artificial domain is profoundly less complex than the biological one but offers equally surprising results.

Psychology

Psychology has become, especially in the last two hundred years, a ‘sociocentric’ and ‘anthropocentric’ discipline. The increasingly refined culture of Western societies (but also of Eastern ones) has refocused attention on the human mind and human thought. In this sense, it is necessary to point out that Western psychology differs in approach from Eastern psychology, which, however, in many cases, reaches the same conclusions. Certainly, Western psychology, developed in a context of industrialisation, was more integrated with the mentality of society and, nevertheless, had enormous difficulties in establishing itself due to prejudices. For the reader’s benefit, it is worth remembering that the Industrial Revolution was a historical period from 1760 to 1840, Freud lived between 1856-1939, Jung lived between 1875-1961, and Jung himself stated:

We need more psychology. We need a better understanding of human nature, because the only real danger that exists is man himself. He is the great danger, and unfortunately we do not realise it. We know nothing about man, or too little: we should study the human psyche, because we are the source of all evil to come.

Computer engineering

Engineering is the discipline that, above all, has changed the face of the planet over the last three centuries, but has always made a difference in the lives of mankind: we need only remember that the ancient Romans were excellent engineers and, thanks to their skills, made the difference between life and death. Computer development has been considerably propelled by the application of engineering (from the Latin ingenium) and today in computer engineering faculties, not only programming is taught but also disciplines such as: automata and languages, software engineering, automation and robotics, automatic controls, operations research, systems theory (from Uniroma2). Engineering has led to a gradual miniaturisation in the world of electronic components and this has favoured an increase in computing power.

Control theory

Explaining the theory of control is not difficult: the Treccani writes ‘methods for understanding, governing and modifying the behaviour of dynamic systems, natural or artificial, in order to guide them to achieve assigned goals’ and indeed it is a perfect definition. Artificial intelligence is to all intents and purposes a highly complex dynamic system: it is necessary to ‘move’ many aspects of the system in order to achieve the set purpose in the defined time and manner. Control theory is used to interact with the characteristics of dynamic systems, which are: states, parameters and relationships. These three characteristics can vary constantly, making the management of the system very complex.

Linguistics

Everything is communication, communication is everything; a living being is a social being and it was Aristotle himself who stated this in ‘Politics’:

Man is a social animal: he tends by nature to aggregate with other individuals and form societies.

Source: Politics, Aristotle, 4th century BC.

Communication enables human beings to experience the integrated reality in which they exist, and linguistics is nothing other than the scientific and systematic study of human verbal language and its structures, and it does so across historical eras and cultures. Linguistics is thus the basis of society and its ability to interact and survive over time: it is curious to reflect on the fact that linguistic extinction can occur even before the last native speaker passes away. The University of Tuscia, in its degree programme in ‘Linguistics of Societies’ wrote: “there are more than 7,000 languages used worldwide and linguists estimate that by 2050 half of these languages will be extinct”. Language is not just a means of communication: it marks the difference between the life and death of a culture.

Conclusion

There can be no artificial intelligence without integration between the scientific and humanistic domains. There is a wonderful phrase by Albert Einstein that goes something like this:‘imagination is more important than knowledge, because knowledge is limited‘. Fantasy belongs to the domain of thought, where knowledge is not yet formed but where knowledge exists. We must defend and encourage this mixing of knowledge, this multidisciplinarity without which it would be impossible to achieve such appreciable results as those achieved in recent years.