15 de enero de 2026
Origins of AI, Disciplines Behind it, Competences, and Classroom Vision
Marco Antonio Feria Uribe
Magíster en docencia de la matemática
Docente de la Maestría en Educación
Facultad de Ciencias de la Educación
Universidad Externado de Colombia
María Fernanda Téllez Téllez
Docente de la Maestría en Educación
Facultad de Ciencias de la Educación
Introduction
This text presents an overview of Artificial Intelligence (AI) origins to understand how it is nourished by several disciplines. In addition, it shows how teachers can establish a connection between AI and competences. Lastly, it illustrates the importance of having a positive attitude towards AI in the classroom.
An Overview of the Origins of Artificial Intelligence
To understand AI genesis, it is worth knowing about its evolution and advancements in its field of study. First, in 1943, Warren McCullogh and Walter Pitts invented a “mathematical model of an artificial neuron” (Norman, 2004-2025, para. 1). Second, in 1950, Alan Turing created the Turing test as a model to simulate human beings’ behavior, establishing the bases of computing theory (Oppy & Dowe, 2021). Third, John McCarthy invented the AI term in 1956 (Cordeschi, 2007; González Quiroz, 2019). Fourth, Herbert Simon and Allen Newell created the first program of AI with logic theory (Newell & Simon, 1956 as cited in Maldonado Granados, 1998). Then, they developed a “GPS (General Problem Solver)” (Newell et al., 1974, p. 335). Finally, “In 1957, Frank Rosenblatt” invented “a neural network capable of information storing and, more importantly, learning and improving” (van Assen et al., p. 5), being able to recognize neural patterns such as “Joseph Weizenbaum’s chatbot ELIZA” (Weizenbaum, 1966, as cited in Adami, 2021, p. 134), “a natural processing program” (London intercultural Academy, 2025, para. 1) that “simulate[s] a conversation with a Rogerian psychotherapist” (London intercultural Academy, 2025, para. 6). Those advancements established the basis for the creation of robots.
During the 1970s and 1990s, AI had stagnation periods due to computational trustworthiness and budget limitations (Couto, 2025). Nonetheless, after the 1980s, with Edward Feigenbaum’s studies about “expert systems” (programs that simulate experts in different areas) (Feigenbaum, 1980, para. 3), commercial AI emerged like MYCIN, which helped identify blood infections (Copeland, 2018) and XCON, that allowed practitioners to set up computer systems (Kraft, 1984). Additionally, Japan started the fifth generation of computers (Feigenbaum & Howard, 1993). In the 1990s, AI lost popularity and some other type of machine learning emerged (Mundlamuri et al., 2025). After AI decline, there was a new development with Judea Pearl, who designed the Bayesian networks that fostered AI reaction to face uncertainty (Gregersen, 2025). During that period, it was also fundamental the exponential computing power (Moore’s law) (McKee, 2025) and the availability of big data, fostering the next AI generation, which was called Deep Learning. This revolution of learning encouraged the use of deep capacity of neural networks, big data, computing power, and language processing programs.
Disciplines Behind AI
In 2025, AI was considered a panacea by many users. However, it started its development and rise in the 1990s (Ufinet, 2024). This technology was performed by experts dealing with programming algorithms in the creation of internet and computer software exclusively. Nonetheless, AI is currently spread all around the world and used due to its free access, which allows people to be autonomous concerning its application and multiple functions. In the academic arena, AI has an interdisciplinary ground with some interconnected disciplines: mathematics, logic, languages (Swarno, n.d) (programming or formal languages), neuroscience (Yang et al., 2025), and cognitive psychology (Assuncao et al., 2025).
First, in relation to mathematics, there is a relationship that deals with processes and procedures which allow people to find connections among practice, thinking, and AI. Mathematics stands as the universal language and the base of any item regarding AI (Universidad de Sevilla, 2024). Second, logic is related to AI (Familia, 2023) through the axiomatic propositional method (induction, deduction, abduction) as it is a way to build knowledge. Likewise, it is worth mentioning that Pierce (as cited in Soler-Álvarez & Manrique, 2014) attested that there are three linked ways to reasoning: induction, deduction, and abduction. Inductive reasoning involves obtaining a general conclusion from specific information. Also, it “refers to inferences from the observed to the unobserved, or to general laws” (Kang et al., 2025, p. 16322). Deductive reasoning entails “drawing specific conclusions from general principles under certain conditions” (Kang et al., 2025, p. 16320), and abductive reasoning embraces “deriving an explanation for a finding or set of facts” (Organization for Economic Co-operation and Development (OECD), 2021, p. 57). Thus, this axiomatic method permits dual logic that produces generalizations and inferences based on evidence and it validates propositions. In addition, it allows for the construction of grammatical structures that unveil new categories of knowledge. Third, neuroscience is the study of the brain and its connections or neural networks. Lieto (2018, as cited in Sarrazola-Alzate, 2025) stated that “research in the broader field of neuroscience has focused on a deep understanding of the cognitive architectures present in the biology of the brain” (p. 49). Hence, this science follows the architecture and functioning of several models of AI used currently. Also, it inspired the development of algorithms for AI.
Correspondingly, although mathematics and logic are not completely related to each other, they depend on one another (Wang, 2023) and are essential for the development of artificial neural networks in AI (Ge, 2022). Besides, Lenci and Padó’s (2022) assert that “the last generation of neural network architectures has allowed AI and NLP to make unprecedented progress” (para. 3). Fourth, it is worth mentioning that looking back, as Barr (1980) explained, “New AI approaches to natural language processing were influenced by many scientific developments of the 1960s, including high-level programming languages” (p. 6). Programming languages or formal languages are fundamental because they develop programming models that are the framework in which AI is shaped to perform its algorithms. Concerning algorithms, Luckin (2018) states that “The difference in the various types of AI is mainly to be found in the way that the instructions of the algorithms are written, and the way that the information that is to be transformed by those instructions is structured” (p. 69). Consequently, understanding algorithms is essential in AI languages programming.
Moreover, Uchiyama et al. (2024) emphasize that “models trained on programming languages exhibit a greater ability to follow instructions compared to those trained on natural language datasets” (p. 18143). Likewise, these languages mentioned above engage three actions. Transforming entails organizing and classifying statistical data to identify and distinguish the relationship between words and contexts. In this regard, Chen et al. (2025) mentioned that “LLMs operate by encoding input text into high-dimensional vector representations, where the contextual and semantic relationships between words and phrases are captured through multilayer transformer architecture” (para. 4). Pre-training is a huge teaching process of big data volumes before being used in specific or general tasks.
In relation to this action, Uchiyama et al. (2024) state that “The generalization ability of [Large Language Models] LLMs originates from pre-training on large text corpora” (p. 18139). Being generative involves a mechanism to elaborate original, well-structured, and coherent texts (natural languages) to answer the proposed and well-structured questions (prompts). In this sense, “The quality of these responses depends on factors such as the input prompt, which shapes the context and specificity of the responses” (Chen et al., 2025, para. 4). Fifth, cognitive psychology studies “how people perceive, learn, remember, and think about information” (Ignau People’s University, Unit 1 Cognitive psychology: structure, Block-1, n.d., p. 17), among others. Thus, AI seeks to reproduce brain functioning and human cognition.
As AI recognizes statistical patterns of words and phrases, it can transform formal languages into natural languages to develop cognitive processes like generalization, categorization, particularization, deduction, induction, and abduction. Furthermore, Lenci and Padó (2022) explain that “systems (e.g., the GPT family) are typically trained with huge computational infrastructures on large amounts of textual data from which they acquire knowledge, thanks to their extraordinary ability to record and generalize the statistical patterns found in data” (para. 3). Therefore, it is evidenced that AI can translate the different programming languages to produce new languages based on enormous amount of data.
Undeniably, mathematics, logic, neuroscience, programming languages or formal languages, and cognitive psychology are present in the construction of AI. Each one of these interdisciplinary grounds contributes to the body of new knowledge to nourish AI.
AI and its Competences
A competence embraces “knowledge, skills, and abilities an individual possesses” (Regan et al., 2023, p. 3). Accordingly, concerning the use of AI in any teaching space, teachers need both, to foster students’ competences regarding it and to know how to evaluate their use, being aware of the positive aspects it offers and its limitations.
Thus, the main concern is what competences students require to foster their responsible and critical use of this tool for lifelong and meaningful learning. UNESCO (2024) proposes “12 competencies across four dimensions: Human-centred mindset, Ethics of AI, AI techniques and applications, and AI system design” (p. 4) for teachers to foster in students. According to this organization, the dimensions are the following: “Human-centred mindset” fosters “students’ values, beliefs and critical thinking skills, …how humans should interact with it, and what responsibilities individuals and institutions should take on to contribute to the building of safe, inclusive and just AI societies” (p. 22). In relation to this dimension, students must reflect critically on their decision making and consequences for themselves and their context. In relation to ethics, this competence “represents the ethical value judgements, embodied reflections, and social and emotional skills students require to navigate, understand, practise and contribute to the adaptation of a growing set of principles and regulatory rules” (p. 23). Therefore, students must apply this tool by being critical, responsible, and aware of its ethical issues and by showing respect and care for others, the context and the environment in which they are immersed.
Likewise, “the ‘AI techniques and applications’ aspect[s] represent the intrinsically linked conceptual knowledge on AI and associated operational skills, in connection with concrete AI tools or authentic tasks” (p. 24). It embraces the students’ skills and abilities to understand its foundation, functions, purposes, and uses by integrating the former dimensions. Finally, “[t]he aspect of ‘AI system design’ focuses on the systemic design thinking and comprehensive engineering skills required for problem scoping, design, architecture building, training, testing and optimization of AI systems” (p. 25). This implies that students must be aware of AI possibilities and limitations and know how to evaluate its use in different contexts, supported by interdisciplinary and ethical accuracy.
Regarding how to evaluate AI’s competences, it is worth mentioning that in literature, the main tendency focused on evaluating these competences as if they were implicitly learned rather than being fostered in students. On the contrary, they require to be developed and learned, and this implies teachers’ own knowledge, competences and familiarity not only within its application, but also in the way they can evaluate students’ competences, for example, having knowledge of data bases analysis, information about networks, use of dual logic (true or false), and polyvalent logic (many factors in the results of a proposition; diverse variables are involved), among others. Flores and Chiappe (2024) conclude that it is mandatory “teacher training in the development of digital competencies, skills, and proficiencies” (p. 5). Similarly, Setiyawan et al. (2025) point out that there is still a bridge to build concerning teachers’ knowledge, their competences and pedagogy for applying digital and AI technologies in the classroom. Therefore, teachers need to overcome these limitations concerning their own knowledge and competences.
In relation to AI’s constraints, this technological advancement has allowed people not only to be autonomous cognitively, but also to be skilled in pragmatic knowledge to solve some daily problems. However, there is a huge risk because individuals might gain skills and others might lose them through their use. In this respect, Rock (2025) reflects on the positive and negative aspects of using AI for many tasks like making summaries from any conference: Is this action really helping people’s metacognitive process? In this view, there is an epistemological query that emerges: Are some competences and cognitive processes like generalization, categorization, particularization, deduction, induction, and abduction lost or gained when using AI? If AI becomes a black box, individuals may lose skills. On the contrary, if they know about the AI processes, they may use it critically. According to Luckin (2018), “[t]hese neural networks are ‘black box’ machines that cannot explain the decisions they make or the actions they take. And this lack of transparency is a considerable problem, one that limits the usefulness of these smart technologies” (p. 72). Thus, it can be inferred that AI is a tool that helps human beings in many tasks. Nevertheless, OEDE (2021) attests that “it still cannot perform the full range of tasks of humans and lacks some basic human skills” (p. 21). In sum, this tool has its benefits and limitations, but it cannot surpass human beings’ skills entirely.
A New Vision of AI in the Classroom
It cannot be denied that digital and AI technologies are advancing by leaps and bounds. Consequently, teachers require a positive attitude towards any of these kinds of technological advances and their application in the classroom. Vieriu and Petrea (2025) consider that “The integration of AI in academic environments raises critical questions related to … the evolving role of traditional teaching methods” (para. 2). To change this traditional view, it is key to tackling this reluctancy being supported by pedagogy and didactics to select the appropriate pedagogical model or approach for AI implementation and evaluation. Setiyawan et al. (2025) affirm that “AI tools offer new possibilities to automate and enhance visualization, feedback, and assessment” (para. 4). Consequently, teachers must be prepared to handle AI appropriately by being conscious that it cannot replace pedagogy and didactics in the teaching and learning processes, but that they require to learn about it and know how to teach it effectively, as AI will remain in the classroom for a long run.
In this regard, Flores and Chiappe (2024) highlight that “The integration of technology into classroom environments is continuously increasing, resulting in digitally enriched pedagogies becoming the global norm rather than the exception” (p. 3). Thus, there is no going back. Teachers as well as students must prepare to use new technologies enhancing critical thinking and cognitive competences in which both have the possibility to discern critically and to know about the ethical concerns and thorough use of these technologies. Accordingly, this new vision of teachers should be mediated by their critical, pedagogical and didactic loop when using AI as a prospective to teach students any subject by fostering their AI competences. Chiappe and Flores (2024) state that “Realizing the potential of these technologies in classrooms hinges on teachers skillfully using them in practice” (pp. 4-5). In this respect, teachers are called upon to assume a new role in the face of this technological advancement to guide students towards its critical implementation.
Conclusion
To conclude, this article shows a synopsis of AI genesis to understand its evolution. Additionally, it highlights how teachers can foster students AI competences. Besides, it invites teachers to challenge themselves and be updated regarding AI to teach their students critically in its use and have a positive atmosphere and stance towards AI in the classroom. An analogy that can be useful for this purpose is the following: AI is a powerful search engine, like a ship. It can travel fast and far. However, it is crucial to have a guide or a captain (a teacher) with mapping knowledge (pedagogy and didactics) and a crew (students) to navigate (competences) whether the ship engine fails or loses its way in the sea of information.
References
Adami, C. (2021). A brief history of Artificial Intelligence research. Artificial Life, 27(2), 131–137. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1162/artl_a_00349
Cordeschi, R. (2007). AI turns fifty: Revisiting its origins. Applied Artificial Intelligence, 21(4–5), 259–279. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1080/08839510701252304
Assuncao, G., Castelo-Branco, M., & Menezes, P. (2025). Self-emotion-mediated exploration in Artificial Intelligence mirrors: Findings from Cognitive Psychology. AI, 6(9), 220. https://doi-org.basesbiblioteca.uexternado.edu.co/10.3390/ai6090220
Barr, A. (1980). Natural language understanding. AI Magazine, 1(1), 5-10.
Chen, B., Zhang, Z., Langrené, N., & Zhu, S. (2025). Unleashing the potential of prompt engineering for large language models. Patterns, 6(6). https://doi-org.basesbiblioteca.uexternado.edu.co/10.1016/j.patter.2025.101260
Copeland, B. (2018, November 21). MYCIN. Encyclopedia Britannica. https://www.britannica.com/technology/MYCIN
Couto, E. (2025, abril 6). El invierno de la IA: así fue el período en el que nadie apostaba por la inteligencia artificial. Muy Interesante. https://muyinteresante.okdiario.com/historia/invierno-ia-escepticismo-inteligencia-artificial.html
Familia, R. (2023, diciembre 26). La lógica como paradigma de la programación en inteligencia artificial. UNIBE Portal Espacio Docente. https://docentes.unibe.edu.do/la-logica-como-paradigma-de-la-programacion-en-inteligencia-artificial/
Feigenbaum, E. A. (1980). Expert Systems in the 1980s. Stanford University CA. https://stacks.stanford.edu
Feigenbaum, E. A., & Howard, S. (1993). The Japanese national Fifth Generation project: Introduction, survey, and evaluation. Future Generation Computer Systems, 9(2), 105–117. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1016/0167-739X(93)90003-8
Flores, W., & Chiappe, A. (2024). Integrating AI into education: Preparation factors and Teachers’ digital competencies. Revista Colombiana de Educación, (97), e20825, https://doi.org//10.17227/rce.num97-20825
Ge, X. (2022). Brief introduction to artificial neural networks. Culture Sciences de l’Ingénieur, 1-11. https://eduscol.education.fr/sti/si-ens-paris-saclay
Gregersen, E. (2025, January 1). Judea Pearl. Encyclopedia Britannica. https://www.britannica.com/biography/Judea-Pearl
González Quirós, J. L. (2019). La Inteligencia Artificial y la realidad restringida: Las estrecheces metafísicas de la tecnología. Naturaleza y Libertad, 12, 127-158. Dialnet https://dialnet.unirioja.es
Ignau People’s University. (n.d.). Unit 1 Cognitive psychology: structure, Block-1. In MPC00, Cognitive psychology, learning and memory (pp. 5-18). https://egyankosh.ac.in/bitstream/123456789/67216/1/Block-1.pdf
Kang, J., Li, X., Xu, L., Liu, Q., Chen, X., Tu, Z., Chu, D., & Sui, D. (2025). Exploring deductive and inductive reasoning capabilities of Large Language Models in procedural planning. Findings of the Association for Computational Linguistics: EMNLP, 16320–16341.
Kraft, A. (1984). XCON: An Expert Configuration System at Digital Equipment Corporation. In P. H. Winston & K. A. Prendergast (Eds.) The AI business: Commercial uses of Artificial Intelligence (Chapter 3).The MIT Press. https://doi.org/10.7551/mitpress/1165.003.0005
Lenci, A., & Padó, S. (2022). Editorial: Perspectives for natural language processing between AI, linguistics and cognitive science. Frontiers in Artificial Intelligence, 5,1059998. doi: 10.3389/frai.2022.1059998. PMID: 36406471; PMCID: PMC9671705.
London Intercultural Academy. (2025). The Story of ELIZA: The AI that fooled the world. https://liacademy.co.uk/the-story-of-eliza-the-ai-that-fooled-the-world/
Luckin, R. (2018). Machine Learning and human intelligence: The future of education for the 21st century. UCL IOE Press.
Maldonado Granados, L. F. (1998). Línea de inteligencia artificial y procesos de razonamiento. Tecné, Episteme y Didaxis: TED, (3). https://doi.org/10.17227/ted.num3-5705 https://revistas.upn.edu.co/index.php/TED/article/view/5705
McKee, A. (2025, July 15). Moore’s Law explained: Past, present, and what comes next. DataCamp. https://www.datacamp.com/tutorial/moores-law
Mundlamuri, R., Gunnam, G. R., Mysari, N. K., & Pujuri, J. (2025). The Evolution of AI: From classical Machine Learning to modern Large Language Models. IEEE Access, Access, IEEE, 13, 178302–178341. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1109/ACCESS.2025.3621344
Newell, A., Simon, H. A., & Seoane, J. (1974). Simulación del pensamiento humano. Teorema: Revista Internacional de Filosofía, 4(3), 335–378. https://www-jstor-org.basesbiblioteca.uexternado.edu.co/stable/43045829
Norman, J. M. (2004-2025). McCulloch & Pitts Publish the First Mathematical Model of a Neural Network. Jeremy Norman’s History of Information.com Exploring the History of Information and Media through Timelines.
Oppy, G., & Dowe, D. (2021). The Turing test (Winter 2021 Edition). The Stanford Encyclopedia of Philosophy Edward N. Zalta (ed.) (First published Wed Apr 9, 2003; substantive revision Mon Oct 4, 2021). https://plato.stanford.edu/entries/turing-test/
Organization for Economic Co-operation and Development (OECD) (2021). AI and the Future of Skills, Volume 1: Capabilities and Assessments, Educational Research and Innovation, OECD Publishing, Paris, https://doi.org/10.1787/5ee71f34-en
Regan, J., Keepers, H., & Straehle, M. (2023). Executive Bulleted Research Brief (EBRB) Competencies & capabilities. AERE Assessment, Education, and Research Experts. https://www.aerexperts.com/pdf/Competency-vs-Capabilities.pdf
Rock, D. (2025, December 1). AI and Machine Learning What’s lost when we work with AI, according to neuroscience? Harvard Business Review. https://hbr.org/2025/12/whats-lost-when-we-work-with-ai-according-to-neuroscience
Sarrazola-Alzate, A. (2025). Cognitive dual-process theories applied to Artificial Intelligence. Revista EIA, 22(44), Reia4432, 1-11. https://doi.org/10.24050/reia.v22i43.1854
Setiyawan, A., Soeharto, S., Wijaya, T. T., Korenova, L., & Lavicza, Z. (2025). Measuring teachers’ competencies for AI integration: Development and validation of the AI-TPACK in vocational education. Computers and Education Open, 9. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1016/j.caeo.2025.100319
Soler-Álvarez, M. N., & Manrique, M. H. (2014). El proceso de descubrimiento en la clase de matemáticas: los razonamientos abductivo, inductivo y deductivo. Enseñanza de las ciencias, 32(2),191-219. http://dx.doi.org/10.5565/rev/ensciencias.1026
Swarno, S. Si. (n.d). Mathematics: The Essential Foundation of Artificial Intelligence. Binus University. Faculty of Humanities. https://pgsd.binus.ac.id/2024/03/05/mathematics-the-essential-foundation-of-artificial-intelligence/
Uchiyama, F., Kojima, T., Gambardella, A., Cao, Q., Iwasawa, Y., & Matsuo, Y. (2024). Which Programming Language and What Features at Pre-training Stage Affect Downstream Logical Inference Performance? Proceedings of the 2024 Conference on empirical methods in natural language processing, 18139–18149 November 12-16.
Ufinet. (2024, January 23). Inteligencia artificial: definición, historia y evolución. https://www.ufinet.com/inteligencia-artificial-definicion-historia-y-evolucion/
UNESCO. (2024). AI competency framework for students. https://unesdoc.unesco.org/ark:/48223/pf0000391105
Universidad de Sevilla. (2024, July 16). Inteligencia artificial y matemáticas, el binomio que está cambiando el mundo. 9º Congreso Europeo de Matemáticas en Sevilla. https://www.us.es/actualidad-de-la-us/inteligencia-artificial-y-matematicas-el-binomio-que-esta-cambiando-el-mundo
van Assen, M., Muscogiuri, E., Tessarin, G., & De Cecco, C.N. (2022). Artificial Intelligence: A century-old story. In: De Cecco, C.N., van Assen, M., & Leiner, T. (eds) Artificial Intelligence in Cardiothoracic Imaging. Contemporary Medical Imaging. Humana, Cham. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1007/978-3-030-92087-6_1
Vieriu, A. M., & Petrea, G. (2025). The impact of Artificial Intelligence (AI) on students’ academic development. Education Sciences, 15(3), 343. https://doi.org/10.3390/educsci15030343
Wang, H. (2023). Analysis the nature of Logic: The distinctions between Logic and Mathematics. Communications in Humanities Research, 5(1), 417-422. https://www.researchgate.net/publication/373922451_Analysis_the_Nature_of_Logic_The_Distinctions_Between_Logic_and_Mathematics
Weng, X. (2025). AI competency as a catalyst for creativity and entrepreneurship: Insights from the Big Five personality traits. Interactive Learning Environments, 1–16. https://doi-org.basesbiblioteca.uexternado.edu.co/10.1080/10494820.2025.2554997
Yang, S., Huang, X., Bernardo, D., Ding, J.-E., Michael, A., Yang, J., Kwan, P., Raj, A., & Liu, F. (2025). Foundation and Large-Scale AI Models in Neuroscience: A Comprehensive Review.