Abstract:

In this presentation, we would like to outline the need of developing the concept of the pedagogy of artificial intelligence (AI).

In the era of the fourth industrial revolution basing its development on artificial intelligence, all spheres of life of modern man have been changing. We have all recently witnessed how the introduction of ChatGPT revolutionized dozens of
domains, from education to journalism or even healthcare, almost overnight. Therefore, science must also change and adapt, including machine learning and the emergence of "self-learning" machines. Since the birth of pedagogy as an academic science, it has been dedicated to the education (training and upbringing) of man. It encompasses a whole range of issues that support the all-round development of the individual, from childhood to seniority. But what if humans were no longer the only object of interest of pedagogy, and education has ceased to be exclusively a human affair? According to the theory of constructivism, AI, and in the future most likely also artificial general intelligence (AGI), will need access to sources and experiences from which it can infer, rather than a large amount of data injected through programming. It is the data and content that a self-learning algorithm has access to that have a key
impact on changing the parameterization of the algorithm itself and, consequently, the rules of its operation. Such an algorithm makes inferences like a human being, but it is up to us humans to determine what experiences, beliefs, or knowledge are provided to it in the course of training. 
Such a direction of thinking opens a wide field for analysis related to, among other things: the place of machines in the human world, the role of the teacher in the process of machine self-learning, the importance of the relationship between humans and AI, the consequences of this relationship and, most importantly, the direction of AI "education". The emergence of Artificial Intelligence Pedagogy (AIP) is a response to the need to enrich the methods of machine learning, and can be seen as a starting point for discussions on introducing a humanistic dimension to robotics, or the development of artificial intelligence technology. In this presentation, the authors have made a joint effort to bring together their very distinct fields of expertise and put forward the argument for the need to be concerned about AI pedagogy, supported by the subject literature.
The methodology of choice was the targeted, critical literature study. The results show that indeed, the efforts towards establishing the discipline of AI pedagogy have not been sufficient enough. As a result, the authors state how significant it is for the scientific community to engage with the pedagogy of AI, and give a set of recommendations concerning what could be done to do so.

Keywords:

AI, artificial intelligence, education, machine learning, pedagogy.