1. Introduction
What is the best way to use the dispersed knowledge among all people? This is the question posed by Hayek in his essay “The Use of Knowledge in Society”, which turns out to be one of the fundamental problems to be solved in Economic Policy. He goes on to say that “the degree of efficiency of these systems depends mainly on the most complete use of the existing knowledge that we can expect”1
From the Austrian School of Economics, the market is understood not as a place, physical or virtual, where goods or services are exchanged, but as a spontaneous, complex, and dynamic process. A network in constant motion, where people, according to their values, tastes, preferences, knowledge, and expectations of the future, exchange knowledge. Each person participates in the catallactic process in a relatively very small fragment compared to the immense network of dispersed knowledge in all humanity.
This knowledge is embodied, materialized, or made tangible in goods and services, which are ultimately what is exchanged. According to Mises, lato sensu, society and market are synonymous. Both are processes in which knowledge plays a fundamental role. The exchange of values, based on knowledge, is ultimately expressed —never measured—in what we call market prices.
As Benegas Lynch (h) points out, saying “free price” is redundant. Price, as such, when not intervened, reflects in the market price system the value structure of people. If it does not, then it is simply a number dictated by some governmental authority, which is of no use and indicates little or nothing2.
Now then, these subjective valuations are once again supported on a base of knowledge that human beings manage and which, following Huerta de Soto, we could classify into Type A and Type B knowledge. Type A knowledge is what we call practical, tacit, dispersed, and of unique events. It relates to know-how and is not articulable. On the other hand, Type B is that which can be centralized, articulated, and which we call scientific knowledge. The latter is the type of knowledge handled by Artificial Intelligences. Of course, both have reciprocal influence. Type A knowledge influences B, and B influences A.
A machine cannot feel, nor create new means or ends. Artificial intelligence does not naturally meet the prerequisite of human action, which is to feel a state of need. Therefore, the possible prerequisites must necessarily be previously loaded by programmers.
In Human Action, Ludwig von Mises illustrates that only the individual thinks and acts. Thinking is an individual task, and knowledge can only be acquired, made one’s own, by rethinking it anew. Mises states: “Only he knows how to arrange his multiple items of knowledge and experience into a consistent system.” He later continues developing his idea and explains that “Every action is always based on a definite idea about causal relations. In perceiving a causal relation, the subject formulates a theorem. Action without thought and practice without theory are inconceivable… It is always the individual who thinks… reasoning itself is invariably an individual task.”3
Finally, on the role of ideas, Mises says: “Tradition preserves and transmits ideas, urging later generations to continue the intellectual task. Nevertheless, man has no other way of appropriating the thought of his precursors than by thinking it over again.”4 Without a doubt, “The history of thought and ideas is a process that unfolds from generation to generation. The thought of later ages springs from that of earlier ages.”5
We will see below how the problem that arose in previous centuries, with catastrophic and anti-civilizational results, could today provoke the deification of Artificial Intelligence, thereby diluting progress toward a better world.
2. On the Problem of Extreme Rationalism
“The enthusiasm that Physics (and mathematics) generated in academia was excessive. Everything worked. Problems could be solved: mechanics, heat, waves, sound, light, magnetism, electricity—everything mastered by the same technique. Physics and mathematics were relentless and could solve any human problem. It seemed that the way to solve the world had been discovered. The frenzy was absolute.”6 Many times, the enthusiasm over certain discoveries, combined with the imagination and potential of brilliant minds, leads human beings to believe themselves omnipotent.
René Descartes sought a method that could absolutely dominate human knowledge, based on human reason. His method lay in mathematics, since for Descartes there was concordance between the real world and mathematical language. Therefore, any phenomenon would be predictable for one who knows the mechanical laws of mathematics and has the relevant information. This vision necessarily leads to a universal determinism, and predictions could be solved through mathematical equations. There arises an exaggerated faith in human reason, which now appears capable of solving all the world’s problems mathematically. It is often said, too, that mathematics is the language in which the universe is written.
On the other side, moderate rationalism tries to recognize the limits of the human mind. Immanuel Kant acknowledged the impossibility of knowing everything that can be known. Against this, idealism arose, as a Hegelian current prevailing in Germany, which rejected such limitations of human reason and believed that the human being is a sufficient source to know everything in its entirety. Curiously, Hegel himself said that the state is a supra-personal reality of which the individual is a mere accident.7
Later, continuity came through Auguste Comte and Positivism. This current gathers the intellectual errors of empiricism, extreme rationalism, and idealism. First, the excessive enthusiasm for the empirical sciences—Physics and Mathematics. From there, the urgent need to apply the method of Natural and Exact Sciences to the Social Sciences, despite the methodological error this entails. From Positivist thought derives the creation of a new religion, where God is society. A society that can be conceived, designed, and improved through the establishment of objective laws that govern its order and progress. Thus, human individuality and Liberty are destroyed. This constructivist vision believes, in short, that reason can build a better society, creating institutions out of nothing, ignoring the evolutionary character of Social Institutions. The scientist, now called a Social Engineer, is capable of creating a more efficient and just society. His source of unlimited knowledge may be historical facts and his own human reason.
Ultimately, following Meseguer8, if the foundation of the social engineer lies in historical laws, then we are facing Historicism, and its proponent could well be called a historicist social engineer. If his pillar is human reason, we are facing extreme or Cartesian rationalism, and we will call him a rationalist social engineer. According to Karl Popper in The Open Society and Its Enemies, all of these could ultimately be classified as historicists.
It can be observed how, in some way, everything ends up deriving into the scientific, articulable, and storable historical data—the type of knowledge handled by Artificial Intelligence.
The study of socialist epistemology, that is, the kind whose characteristic is based on a society whose economy is entirely centrally planned, makes us realize that at least, in this case, fertile ground was provided by the following issues: the exaggerated consideration of human reason as a creationist instrument; the lack of recognition of the importance of Evolutionary Social Institutions; the neglect of methodological individualism, whose omission removes the preeminence of the human being from the center of everything and leads to hypostases and anthropomorphisms; the mystery of human existence itself, which can cause anguish and the need for guidance, direction, or leadership in the face of the ineradicably uncertain character of the future and may drive humans to worship false gods.
These issues, which resulted in the most fervent socialism, still persist in different societies and serve as the same preconditions for the frenzy now aroused by the irruption of Artificial Intelligence.
3. The New False God: Artificial Intelligence
It is sensible to be astonished by the results and potential that Artificial Intelligence represents, in the same way that human beings marvel at seeing a Boeing 747 of almost 450 tons take flight. But one thing is the natural amazement that technological innovations can provide, and quite another is to elevate one of them to the level of a deity. In the case of AI, we must first understand that, unlike the human mind, it works with centralized, articulable, and mathematized knowledge. And this is not the only type of knowledge that human beings master.
Type A knowledge is that subjective, practical, dispersed, tacit, non-articulable knowledge of unique events. This knowledge is linked to human experience and life events. Type B knowledge is objective, scientific, centralized, articulated, and of class, as previously mentioned. Artificial Intelligence is nothing more than a computer program that works with Type B knowledge. When the engineer-programmer must design the inputs and outputs, they must determine, categorize, conceptualize, and define their limits and scope. This is knowledge that necessarily must be articulable in order to later transfer it into AI software. Technical or objective knowledge (Type B) helps the human being, serving as a catapult9 to multiply Type A knowledge, and thus create more and better means and ends. In absolute terms, AI cannot surpass the human being and does not have creative capacity in a dynamic sense. This does not mean it cannot generate music, text, or images, but from this it does not validly follow that it can create new means and ends in the dynamic sense10.
At present, the real problem is not so much this, but rather the belief of people themselves about what AI is and will be capable of, and the attitudinal consequences for human beings when facing this tool.
4. The Problem: “Because AI said so”
In addition to what was mentioned above, regarding the epistemology that provides fertile ground for elevating AI to the divine plane, part of the problem arises, first, from not understanding how Artificial Intelligence truly works and what type of knowledge (of those mentioned above) it handles.
Another part of the problem is the repetition loop it generates, which will reinforce certain beliefs in popular knowledge. For this second issue, two serious and equally important things happen. On one hand, there is the matter of human reason and formal logic, which serve to determine which reasoning is valid and which is not. Only the human being can determine whether a system of reasoning and beliefs is coherent and consistent as a whole, or else recognize invalid points. At the same time, human beings continue to incur reasoning errors such as fallacies. In this case, argumentum ad verecundiam, the fallacy of authority becomes relevant when not knowing—or rather believing—that Artificial Intelligence contains all human knowledge concentrated. Of course, this is not so.
AI contradicts itself permanently. For example, one might ask it to explain the benefits of Social Justice, and it will provide an elegant response with the appearance of coherence, but which, for the truly scholar, cannot withstand deeper analysis. At the same time, it can contradict itself if asked again and be made to explain why what it just said is not so. Thus, it will depend on who uses the tool, how far they go with it, and above all, whether they truly know its functioning and the historical character of the information that serves as inputs.
AI cannot create new knowledge in a dynamic sense. At this point, one might ask: Could it have discovered The Regression Theorem of Money on its own, as Ludwig von Mises did?
We must remain attentive, alarmed, and allow ourselves to doubt every time that, after a declaration or value judgment, someone states: “because AI said so.” This is nothing more than another fallacy of authority. It will be argued that such reasoning or declaration is true or correct because it is supported by someone (in this case AI) with prestige or intelligence. Worse still, it will be considered that AI itself possesses all current and centralized human knowledge, and therefore its response is the product of the collision of confronted human knowledge. This idea will be unfortunately reinforced when the so-called Singularity becomes more tangible or real. That is, an Artificial Intelligence capable of everything, overcoming the obstacles that the current one presents, and considered absolutely superior to the human mind in every possible aspect.
The second issue, which relates to the problem of the repetition loop, can be framed in what the Spanish philosopher and writer José Ortega y Gasset defined in his work The Revolt of the Masses: the mass man.
As human beings increase their material well-being—undoubtedly thanks to technological progress embodied in goods and services that facilitate the achievement of their ends—they tend to rely more on them and trust in their results. This, in many cases, ends up making them forget, or even never learn, certain knowledge. For example, with the incorporation of calculators, many people have forgotten how to divide by three digits. The solution: there is no other alternative but to trust the result presented by the device.
In this same sense, Ortega y Gasset expresses that the mass man lives a passive life, settling for what is given11. He argues that it has nothing to do with a matter of social class, but rather with an attitude toward life itself. The problem lies in the fact that material well-being makes him feel satisfied and secure with what he has, and little by little he renounces values related to demand, effort, and creativity—those that ultimately drive progress. The mass imposes a uniform, mediocre vision, where the creative tension that generally arises from the collision of ideas disappears. The danger lies in the fact that more conformist human beings exist, where the normal replaces the excellent.
Of course, this does not contradict the idea that any progress allowing a task to be performed more efficiently frees up resources and creativity to expand the frontier of possibilities. The real problem lies in adopting a conformist attitude toward what is given and, in essence, losing critical capacity and the values of self-improvement. The mass man is satisfied with his life, his current level of knowledge, and is inclined—or even compelled—to believe everything that technology and progress in general serve him on a platter. Little by little, he loses his skepticism and his capacity to question.
“It is here where asking questions proves essential for exercising a healthy mind. To question is to cast doubt on what someone has stated; and for that we use reasoning, evidence, foundations, new questions, and hypothetical scenarios, among other things. The word to question comes from quaerere, Latin for ‘to seek the truth.’ But what truth would we seek if everything is assumed to be valid and true?”12
5. Ideas and Beliefs
If human beings find themselves in a position of comfort, inclined to believe, then we must understand the leading role of ideas and beliefs and how they structure people’s lives. Jordan Peterson, in his work Maps of Meaning: The Architecture of Belief13, states that beliefs organize human experience and guide human action. They are interpretative frameworks without which the world would be a chaos of disconnected and impossible-to-process data. Beliefs not only describe reality, they structure it; they give direction to individual life. Without them, according to Peterson, human beings would be disoriented. Friedrich Hayek explains in The Sensory Order14 that if there is no classificatory capacity, what is observable—or reality—would be nothing more than meaningless information.
With these ideas from the mentioned authors, we could then consider beliefs as the structure of a building, which allows it to contain and support a certain amount of knowledge depending on its weight and shape. According to this, knowledge may either not fit into a given architecture of belief, or it may end up breaking it, giving rise to a new one.
The next question arises: does this explain the permanent tendency of human beings to seek certainties and the need to believe in certain things? Very likely.
In this sense, Allan Walstad, professor at the University of Pittsburgh, states that: “… these are unconscious neurological processes acquired through evolution and genetic inheritance, whose purpose is to facilitate the development of our actions, reducing the anguish of having to decide on the basis of incomplete knowledge, which facilitates our decision-making capacity and gives us a sense of security in an uncertain and changing environment…”15 Believing in something can be a mechanism to reduce or mitigate the anguish of human existence itself, uncertainty, and the impossibility of knowing everything.
But now, many believe that this “everything” is concentrated in Artificial Intelligence itself.
Another major problem closely tied to beliefs arises when no one questions anything, someone imposes censorship, or everyone thinks the same. The result may be the same: intellectual stagnation. And therefore, the advancement of human civilization comes to a halt. A point of connection can indeed be reached between Freedom of expression and the fact of not questioning, where all people come to believe in the same “truth.” For humanity to enjoy good health and well-being, John Stuart Mill lays out four reasons in his essay On Liberty why Freedom of opinion and expression must never be curtailed. Mill says:
“Any opinion that has been silenced (censored, as we would say today—#cancelculture) might be true, and to deny it is to claim infallibility. Furthermore, even knowing that the censored voice may fall into error, it could still contain a portion of the truth. And Mill argues, only through the collision of opposing viewpoints can the rest of the truth come to light. But Mill goes further, adding that even if the opinion of the majority (the one with the power to crush the minority group) were the whole truth, if it is upheld as the sole voice, it risks becoming prejudice and thus turns into a claim with little real understanding of its rational foundations—unless it is constantly challenged by opposing reasons. In other words, it risks becoming a mere doctrine, which then weakens and ends up being professed only formally.” He then concludes: “In this way, such truth, by being unquestionable, is ineffective for the growth of any real, sincere conviction based on reason.”16
Conviction weakens, reason ceases to be exercised, and the comfortable tendency to live and believe without questioning is reinforced. Without a doubt, language plays a central and transversal role in all this. That’s why AI developers aspire to dominate this field and make their machines capable of processing natural language in the same way humans do. Here too lies another danger, one related to language. In his 1895 lecture, Henri Bergson said: “One of the greatest obstacles to the Freedom of the spirit, we said, are the ideas that language offers us already made.”17
6. Theorem and Conclusion
When a company leverages itself operationally or financially—depending on whether through fixed costs or third-party debt—the Leverage effect shows how the operating result changes when variations occur in the level of sales. We say that profit increases or decreases more than proportionally, amplifying the result when sales rise and provoking greater losses when they fall.
Following this concept, one could formulate the following analogy: Leverage-AI. Artificial Intelligence will undoubtedly be able to solve many of humanity’s problems. But it will address problems that lie within what Huerta de Soto conceives as the static dimension of efficiency. This, in turn, frees resources and leaves human beings to resolve the other side of the coin—the dynamic aspect. That is, the possibility of employing the innate creative and entrepreneurial capacity of human beings to discover new ends and means, and thus expand the frontier of possibilities to inconceivable limits.
However, if societies rely—leverage themselves—too heavily on Artificial Intelligence without caution or safeguards, it may well be that the positive outcome of such leverage is greatly amplified, or conversely, that social disaster is profoundly magnified.
The solution is not to be found in regulating AI, as the European Union intends to do, but rather in elevating all those tools and characteristics that make us human. In the last century, human civilization came to know the ultimate consequences of extreme rationalism and constructivism—both of which revealed the true statist essence (whether Soviet socialism or fascism). Bearing this lesson in mind, human beings must become aware of the enormous power that AI represents and rely on moderate reasoning, recognizing its limitations, on the coherence of the systems formulated, and on a system of values that define the framework of human action.
According to Benegas Lynch (h)18, Liberalism is a necessary but not sufficient condition for a better world, and it is a conception that requires understanding certain basic moral principles. Similarly, Professor Huerta de Soto states that: “there are a number of ethical and moral principles of great importance for the dynamic efficiency of social processes, in relation to which the following paradoxical situation arises: their lack of observance at the individual level carries an extremely high cost in terms of dynamic efficiency, but, on the other hand, attempting to enforce them through the coercive power of public authorities also generates severe inefficiencies from the dynamic point of view.”19
With the rise of mass media, information can reach hundreds of thousands of people within seconds. The benefit provided by the underlying technology is very high—just as the associated risk is. Humanity had to learn to be cautious and skeptical of everything consumed through mass media. In the same way, the Mass Man cannot relax into a lifestyle marked by credulity and the absence of questioning—in sum, by the acceptance of what is given. Quite the opposite. AI will not suffice to create a better world.
After all that has been laid out, one may formulate a possible theorem, which could be the following: if new discoveries and innovations emerge based on the use of technology leveraged by Artificial Intelligence, it will be necessary for human beings to make an even greater effort to leverage and reconnect with their roots—relying on moderate human reasoning and on an ethical and moral system—if they wish to reap the advantages of new technologies and ensure such benefits are realized with caution and without setbacks.
Notes
[1] Conflicts of interest Conflict of interest
The author declares that he has no conflict of interest.
[2] Hayek, F.A. (1945) “The use of knowledge in society”. American Economic Review XXXV N° 4. In reference to possible economic systems.
[3] Benegas Lynch (h), A. (2011) Fundamentos de Análisis Económico. Grupo Unión. Buenos Aires.
[7] Quoting Huerta de Soto in: Di Grigoli, B. (2022) “The Dynamic Laffer curve” Procesos de Mercado: Revista Europea de Economía Política. Vol. XIX N° 2 2022.
[9] Martínez Meseguer, C. (2009) La teoría evolutiva de las Instituciones. La perspectiva austríaca. Unión Editorial. Madrid.
[11] For a further develop of this idea: Di Grigoli, B (2024) Artificial Intelligence vs. Human Intelligence: An Analysis of Creative Processes. Unión Editorial, Madrid.
[12] Ortega y Gasset, J. (2006) La rebelión de las masas. Ed. Austral. España.
[13] Di Grigoli, B (2021) “¿Why Questioning?” https://questionar.com/por-que-questionar/.
[14] Peterson, J (1999) Mapas de sentidos. Ed. Ariel. España.
[15] Hayek, F (2011) El Orden Sensorial. Unión Editorial, Madrid.