Plato in the land of thinking machines

The vitality of Plato's thought is a matter that is beyond doubt. The explanations for this fact and the ensuing postulates can be interpreted in various ways. Today, the most radical and at the same time the most grotesque figure in this context is Alexander Dugin, who claims that the so-called Platonic minimum should serve as a mandatory foundation for universal education applicable to all members of the civil service, state administration, from management personnel to traffic police (Dugin, 2019). Although there is an evident exaggeration in this, it must be acknowledged that Plato still has much to say to us, and the emergence of Artificial Intelligence tools in the public domain only confirms the need to return to the roots of thinking.

A fascinating and simultaneously obvious fact is the diversity and multiplicity of knowledge that a human can acquire. By using knowledge for the most practical issues, such as guessing causes, making decisions, and predicting consequences and outcomes of actions, we lose sight of the multitude of possible perspectives, positions, views, opinions, and conjectures regarding true knowledge. In a world of rapid decisions and constant information flow, we face a binary situation: you know or you don't know. We most often reach for knowledge as a reservoir of data and information intended to ensure effective decision-making or judgment.

Two main approaches to knowledge that have existed since the dawn of civilization in scientific and philosophical reflection can be defined by the Latin terms: dialectica and demonstratio. Demonstrative knowledge, characteristic primarily of the exact sciences, assumes that there is an unassailable objective truth, and the task of science, or more broadly, scientific thinking, is to present this truth as accurately as possible and to showcase it in the form of knowledge. A prime example here is mathematics, which deals with providing proofs of true statements. In demonstratio, we primarily talk about proving. A proof has the characteristic that its task is to establish the relationship between a given statement and the corresponding state of affairs as accurately and reliably as possible. Mathematical statements are closely linked to proofs that anchor them in the realm of truth. This is the level of precision to which all so-called natural sciences aspire. Doubt and alternative ways of looking at a given problem are not always an advantage within their frameworks. Certainty of knowledge is the greatest value we can obtain here. To illustrate, one might refer to the film Dekalog (Kieślowski, 1989), in which a scientist, trying to measure the thickness of ice on a skating rink to ensure the safety of his child playing on it, seeks to apply scientific calculations to the empirical world. The director, outlining a tragic scenario of events in the film, aimed to illustrate the struggle of science on the difficult path to achieving certainty about the states of affairs present in reality. Reasoning under the banner of demonstratio concerns such issues where the slightest contradiction or uncertainty is perceived as a defect that must be eliminated.

Although voices regarding the so-called mathematical nature of the universe are becoming increasingly frequent (Tegmark, 2014), this is merely a hypothesis that most likely exceeds human capabilities in terms of its potential verification. One could venture to propose a somewhat contrary thesis that some aspects of reality cannot be mathematized because they inherently belong to a different order of thinking. This alternative to demonstrative reasoning is dialectical reasoning.Dialectica concerns issues where conjecture and uncertainty reign. A significant part of the reality inhabited by the human species surpasses its complexity the possibilities of objective knowledge. Without delving into the details of the centuries-old dispute between relativism and objectivism, I will only invoke a film that is striking in its message and simple in form titled Rashomon(Kurosawa, 1950) depicting the murder of a samurai, which from the perspective of different witnesses appears as a weave of completely different stories connected by minor details. In such a reality, where perspectives, viewpoints, and aspects create a tapestry woven into an uncertain reconstruction of facts, dialectics comes to the fore with its instruments. These primarily include: argument, difference, similarity, contradiction, etc. One cannot help but feel that in dialectical reasoning we not only discover some part of reality but also influence its perception and method of interpretation. The knowledge gained through dialectical reasoning is uncertain, and its nature is dynamic. Dialectics is a process in which truths cannot be established forever. Where possible, there is room for demonstrative reasoning. The key assumption of dialectics regarding the category of truth is summarized in the words of Thomas Mann: "For what is true is not truth. Truth is infinitely distant, and all conversation is infinite." (Mann, 1997).

The distinction between demonstratio and dialectica has very rich traditions in philosophical reflection (Reeve, 2011). The reason for numerous disputes surrounding this distinction is that it has a significant impact on the practical aspects of science and its social consequences. I bring up this issue to use it in an attempt to understand what, from the perspective of the categories of knowledge, truth, and methods of reasoning, Artificial Intelligence is. Attempts to automate thinking and decision-making processes are older than one might assume. An emblematic example from the history of civilization is the biblical Urim and Thummim, which was a device helping to make decisions in accordance with the will of God (Ex 28:15-30). The efforts to create thinking machines were very intense and occurred at various times and in different circles of civilization. They manifested not only in the creation of objects or devices but also in changing the way of thinking that would resemble automated thinking. The philosophies of T. Hobbes and G. W. Leibniz are the most significant examples here. In the continuation of this text, I will want to consider the instruments of artificial intelligence as another incarnation of the approach known throughout history under the banner of demonstratio. My argument regarding the assessment of AI proceeds as follows. When I talk about artificial intelligence, I will focus on chatbots, which, regardless of the manufacturer's brand, present the same logic of operation. They differ in quality and accuracy but do not differ in the convention within which they operate. My argument is that by considering the emergence of AI chatbots as a revolutionary breakthrough of thinking or even conscious machines, we are conflating two orders. This error lies in attributing the scope of demonstrative reasoning to the tasks belonging to dialectical reasoning. The chat-based incarnation of artificial intelligence achieves a masterful, unattainable level of demonstrative reasoning for humans. Bots resemble outstanding scholars or even savants adept at performing calculations of monstrous scale. In demonstrative reasoning, knowledge must be precise, and the certainty of results indicates the quality of reasoning conducted. It is worth recalling here the fundamental difference between analytical philosophy and continental philosophy (closer to classical philosophy). Within the former, we can expect to reach conclusions bordering on one hundred percent certainty but concerning a very narrow scope of reality. Classical philosophy aspired to the boundless expanses of abstract thinking, uncovering previously unexplored concepts and imaginations, whose connection to reality and relation to objective truth are questionable and far from certain (Russell, 2010). The same goes for

Mój argument w sprawie oceny AI przebiega w następujący sposób. Gdy będę mówił o sztucznej inteligencji, to skupię się na chatbotach, które bez względu na markę producenta prezentują tę samą logikę działania. Różnią się jakością i dokładnością, ale nie różnią się konwencją, w ramach której funkcjonują. Mój argument polega na tym, że uznając pojawienie się chatbotów AI za rewolucyjny przełom myślących, lub nawet świadomych maszyn, dokonujemy pomieszania dwóch porządków. Błąd ten polega na przypisywaniu rozumowaniu demonstratywnego zakresowi zadań przysługujących rozumowaniu dialektycznemu. Chatowa inkarnacja sztucznej inteligencji osiąga mistrzowski, niedościgniony dla człowieka poziom rozumowania demonstratywnego. Boty przypominają wybitnych uczonych, lub nawet sawantów biegłych w dokonywaniu monstrualnych w swej skali obliczeń. W rozumowaniu demonstratywnym, wiedza musi być ścisła, a pewność wyników świadczy o jakości przeprowadzonych rozumowań. Warto przywołać tutaj podstawową różnicę między filozofią analityczną, a filozofią kontynentalną (bliższą filozofii klasycznej). W ramach tej pierwszej możemy liczyć na osiągnięcie wniosków graniczących ze stu procentową pewnością, ale dotyczących bardzo wąskiego zakresu rzeczywistości. Filozofia klasyczna swą ambicją sięgała do niezmierzonych połaci myślenia abstrakcyjnego, odkrywając nieeksplorowane wcześnie pojęcia i wyobrażeń, których związek z rzeczywistością i relacja do prawdy obiektywnej są podważalne i dalekiej od pewności (Russell, 2010). Podobnie jest z demonstratio and dialectica, although the range of demonstrative reasoning is broader than that of analytical philosophy.

One of the better examples that captures the difference between questioning a chatbot and live argumentation is the Oxford debates. In this format, two sides clash, and the quality of their arguments is judged by a jury or the audience. Importantly, not every thesis can be discussed within this format. If a thesis were to be debated that is uncontroversial in reality and its status is definitively established, then the advantage would be automatically assigned to the side tasked with arguing for the thesis. If the answer to a doubt could be determined directly through empirical investigation, such a debate would also be meaningless. Therefore, debaters in Oxford debates typically discuss normative issues or submit questions related to the public evaluation of a given issue. In a dialectical dispute, many unexpected aspects of the problem are revealed, and the most important effect of the clash is progress – the development of lines of argument. Let us now consider this issue in the context of AI and analyze it in two steps. First, could chatbots engage in discussions amongst themselves in the framework of an Oxford debate? Let’s assume this is possible. Secondly, could chatbots achieve the goal outlined above, that is, to develop lines of argument not based on what they knew previously, but based on what they criticized or accepted from the opponent's point of view? I will leave this question unanswered, but I hope this example illustrates the difference between questioning and arguing.

Another area where one can find fascinating manifestations of this problem is student papers written using AI. These texts utilize demonstrative reasoning excellently. They present data and information in a relatively reliable and coherent manner, enabling any student to "write" a paper or text on any topic. The fruits of demonstrative reasoning, regardless of the topic, will share the same flaw. Artificial Intelligence can help solve many problems, but – as far as I know while writing these words at the end of 2025 – it cannot independently pose them. Dialectical reasoning, which perceives the flow of thought primarily as a process, rather than the shortest path to specific knowledge, has always been and remains a source of creativity that allows us to recognize problems and formulate conditions and methods for their resolution. Works directly based on the results of questioning chatbots can be easily recognized without using complex tools. They lack the dialectical factor. Instead, we receive knowledge and information – the tip of the iceberg visible above the water's surface. The dialectical part, consisting of the paths leading to this knowledge, remains unknown – to readers, authors, and the AI tools themselves, for which data remains the primary raw material.

At this point, Plato enters the scene with his immortal philosophy, which is entirely based on dialectical thinking. This thinking is primarily focused on asking questions and, to a lesser extent, on obtaining answers. This is evidenced, for example, by the dialogue "Euthyphro," where the main element is aporia – an irreconcilable contradiction, a problem that cannot be definitively resolved. Its echoes have been heard for centuries in the intellectual tradition of the West and reached the 20th century, where they constituted the main problem of post-war legal philosophy.Eutyfron, gdzie głównym elementem jest aporia – sprzeczność nie do usunięcia, problem którego nie da się jednoznacznie rozstrzygnąć. Jego echa były słyszalne przez wieki w tradycji intelektualnej Zachodu i dotarły również do wieku XX, gdzie stanowiły główny problem filozofii prawa okresu powojennego.

Creating an AI chatbot that speaks as Plato would, as far as it is technically possible, makes no sense, because Platonism is based on a dynamic conception and on meditation over aporias. Without going into details, it should be noted that, according to Plato, the role of the philosopher is to pursue the world of ideas and the return journey to the world of earthly things. This journey is not a one-time trip but a way of life for a thinking person. Interestingly, this postulate of Plato has been realized in his own case. The oft-repeated slogan stating that all Western philosophy is merely footnotes to Plato seems to well reflect the state of affairs. It also confirms my thesis, which states that dialectics gives us creativity and illuminates problems while creating conditions for their resolution. The recipe for creating such a far-reaching philosophical approach is precisely to anchor it in dialectical reasoning. In conclusion, it can be noted that Plato is still doing well in the realm of thinking machines, and dialectical reasoning can still serve as an answer to the command: "Prove that you are not a robot."dialectica daje nam kreatywność oraz naświetla problemy i tworzy warunki ich rozwiązywania. Receptą na stworzenie tak daleko oddziałującego podejścia filozoficznego jest właśnie zakotwiczenie go w rozumowaniu dialektycznym. Konkludując można zauważyć, że Platon w krainie myślących maszyn wciąż ma się dobrze, a rozumowanie dialektyczne wciąż może służyć za odpowiedź w poleceniu: „Udowodnij, że nie jesteś robotem”.

Perlikowski, Ł. (2025). Dialectical Reading of Plato. Toruń: UMK Scientific Publishing., Ł. (2025). Dialektyczne czytanie Platona. Toruń: Wydawnictwo Naukowe UMK.

Sources:

Dugin, A. (2019). Political Platonism: The Philosophy of Politics. Arktos.; , A. (2019). Political Platonism: The Philosophy of Politics. Arktos.; Kieślowski, K. (1989). Decalogue I-X.; , K. (1989). Dekalog I-X.; KurosawaKurosawa, A. (1950). Rashomon.; MannMann, T. (1997). Joseph and His Brothers. Lower Silesian Publishing.; Reeve., C. D. C. (2011). 'Demonstration and Dialectic', Practices of Reason: Aristotle's Nicomachean Ethics, Oxford Academic.; Russell, B. (2010). The Problems of Philosophy. Cosimo Classics.; Tegmark, M. (2014). Our Mathematical Universe: My Quest for the Ultimate Nature of Reality, Knopf.

#7 Ł. Perlikowski (26.12.2025). Plato in the Land of Thinking Machines. https://lukaszperlikowski.pl/blog