The advance of applied artificial intelligences, mainly Generative AIs based on Pre-Trained Large Language Models (PT-LLM) on Artificial Neural Networks (ANN) under Machine Learning (ML) techniques, represents great opportunities for improvement in various fields of work and study. But it also entails significant challenges for the redesign of roles and jobs and the development of capabilities to allow a harmonious incorporation of these tools with the least possible collateral damage in the management of change.
In direct relation to the education system, the central issue revolves around how these AIs can be effectively integrated, either as agents that assist teachers or that accompany students in the learning process. AIs can also be introduced as a curricular content in what has come to be called “computational thinking” (CT) for early childhood and primary education, the teaching of computational skills in secondary education, and in programs of varying length and orientation in higher education. In addition, different types of AIs have the potential to improve the management of formal education, both by accelerating existing processes for large volumes of data and by setting measurement and monitoring objectives that were never before possible with existing technologies.
We are clearly witnessing the transition towards an Education 4.0, where artificial intelligence plays a crucial role in the personalization of part of the learning process, the automation of repetitive tasks, and the optimization of educational resources; however, the means to articulate all the components –human and technological– of this actor-network system are not so clear. To this effect, the recommendations of international organizations that have adressed the subject seriously and in detail, such as UNICEF and its team of specialists (2021 onwards), the documents of the Beijing (2023), Cartagena (2024) and Montevideo (2024) summits, and the initiatives and tools made available by the World Bank, the IDB and CAF, are very helpful. However, for real innovation to take place –not just in education and training with and for AI– a radical change in the way “artificial intelligence” is presented and perceived is still necessary.
Thus, a first key objective is to reframe the culturally established narrative on AI stemming from popularization outlets and the non-specialized media, which attribute mental capacities, agency and a sort of indeterminacy and free will to AI. It is imperative to provide greater technical precision on the real scope of AIs from a perspective centered on the relationship between person and technological object rather than on interpersonal relationships. In this regard, and recognizing the effect of language on this reframing of the narrative, we propose a double shift from the singular to the plural. On the one hand, by not using “artificial intelligence” in the singular, for, although a useful generalization for a theoretical framework, it is insufficient for practical purposes and does not reflect the enormous differences between subdisciplines and products of AI, which are not limited to generative AIs. In this way, anthropomorphic attributions to AIs can also be offset: it’s not you! On the other hand, an emphasis on the plural introduces a collective understanding –it’s us!– that focuses on collaboration and shared responsibility in AI integration, involving industry, academia, government and third-sector organizations. Additionally, the notion of a “joint ownership” over AIs reinforces that these should be developed and implemented in an interdisciplinary and contextual manner.
Another important aspect of the “irruption” of AIs on the educational landscape is the opportunity they provide to fully expose the exhaustion of the encyclopedic educational model of the modern Enlightenment. Although it was temporarily effective in organizing a system in a now distant context of extreme illiteracy and profound cultural change, its current results have been negatively assessed for some time now, leading to initiatives in several societies to overhaul this legacy that for too long has resisted in its decadence. From a practical point of view, the capabilities of AI applications demonstrate that the focus of the educational experience cannot lie on something that these can easily do: the learning by rote of information and its textual formalization. In this regard, the possible outcomes of incorporating AIs appear as extremely divergent: either we stick with the model by automating its errors to a tragic extent, or we seize the change as an opportunity to reconfigure the system and gear it towards a holistic educational experience.
Lastly, a critical analysis of the unintended outcomes of previous “technological revolutions” shows that, due to its unprecedented speed and extent, the ongoing transformation driven by the “4.0” or “AI revolution” cannot be absorbed and managed through recourse to past models and experiences of technological adoption. A paradigm shift in technological adoption is needed that takes into account the social impact and mitigates undesired effects to a minimum. It is undeniable that these “revolutions” have brought concrete benefits and improvements in vast areas of human well-being, but a deeper analysis of the global ethos over the last hundred years –to which valuable thinkers of the most diverse backgrounds have contributed– shows that, in addition to the unfulfilled promises of these movements of change driven by technology, certain aspects of experience have taken a turn in the opposite direction, degrading the quality of life. Thus, in addition to the paradigm shift, a collective effort is required to ensure that AIs do not become an exclusive tool of technological elites, but a democratizing force capable of bridging educational –and therefore social and cultural– divides. To this end, it must be recognized that we are already witnessing a hyperdigital divide, due less to unequal access than to the enormous disparity in the understanding of emerging technologies and their effects.
Miembro de:
