Keywords
.inteligencia artificial, educación superior, universidad, innovación docente, debate, podcast, Artificial intelligence, higher education, university, teaching innovation, debate, podcast.
En busca de lo que nos hace humanos: Confrontar Inteligencia Artificial y natural a través de tres metodologías docentes
Abstract
El presente texto nace en el marco del proyecto de innovación docente con referencia 2223-02 titulado "HUMAN-CHAT, Inteligencia artificial versus natural: tres metodologías docentes (debate, experimento y podcast) en busca de lo que nos hace humanos", concedido por la Universidad Pontificia Comillas a la autora de este capítulo. La llegada de la inteligencia artificial (IA) a nuestras vidas es evidente: disponemos de asistentes virtuales, empleamos herramientas como ChatGPT o los algoritmos determinan el contenido que consumimos en Internet. De hecho, el impacto de la IA sobre la sociedad es tan elevado que se denomina “Cuarta Revolución Industrial” a la era actual, donde los sistemas inteligentes se integran en las organizaciones humanas. Así, surgen distintas visiones sobre las consecuencias que puede tener para nuestra especie, por lo que parece un momento idóneo para reflexionar sobre qué nos hace humanos mediante una comparación entre las capacidades de la IA y las capacidades, por encontrar un término opuesto, de la inteligencia “natural”; esto es, las capacidades cognitivas humanas. No se trata de establecer una competición entre ambos sistemas, sino más bien de volver a un debate pausado, socrático y filosófico sobre nuestra capacidad de contribución al mundo. Además, dado que esta revolución se produce a escala mundial pudiendo afectar a todas las disciplinas, y no solo a las específicamente tecnológicas, surge una oportunidad de trabajo interdisciplinar en los contextos educativos, invitando a alumnos de distintos grados a esta reflexión trascendental. Llegar al alumnado universitario es relevante porque son los actuales usuarios, futuros beneficiarios y potenciales creadores de los nuevos avances. Además, establecer un trabajo colaborativo entre ramas del conocimiento permite realizar un trabajo interdisciplinar cuando la compartimentalización del conocimiento no responde a la realidad global que vivimos. Así, en esta ponencia explico el desarrollo de una programación académica anual en dos asignaturas de dos grados (“Cognición Humana” en el grado Matemáticas e Inteligencia Artificial; y “Pensamiento y Lenguaje” en el grado de Psicología) con el objetivo común de conocer mejor el alcance de la IA actual y reflexionar sobre qué nos hace humanos. He programado una serie de actividades siguiendo tres metodologías docentes: debate, experimento y podcast. Por una parte, en IMAT y Psicología se hará un campeonato eliminatorio de debates por equipos donde cada uno defenderá una postura asignada al azar sobre si determinados procesos cognitivos son exclusivos de los seres humanos o en cambio pueden ser replicados en IAs (por ejemplo, dudar o pensar). La evaluación de estos debates se compartirá entre pares (30%) y la profesora (70%), y corresponderá a un 30% de la asignatura. Los equipos que pierdan podrán realizar ejercicios complementarios de aumento de puntuación continuando la reflexión sobre el tema. Por otra parte, inspirados por el test de Turing, en ambos grados cada equipo llevará a cabo 3 preguntas a la aplicación ChatGPT y a tres personas, y analizará las respuestas reflexionando sobre cómo los humanos categorizamos y expresamos la información y cómo puede este proceso trasladarse a las máquinas (10% de la asignatura). Después, se expondrán las respuestas obtenidas; se votará cuál consideran no humana y se reflexionará en conjunto. Por último, todos los debates serán grabados en formato podcast en clase y los dos trabajos de ChatGPT con mayor puntuación en cada grado serán resumidos y también grabados. Así conformaremos una serie de episodios sobre lo que nos hace humanos que podrán ser divulgados para que nuestra reflexión llegue a la sociedad. A modo de conclusión, es posible establecer una reflexión sosegada sobre el impacto de la tecnología empleando varias metodologías y fomentando la colaboración entre diversos grados universitarios sin requerir demasiadas infraestructuras o materiales.
The present text arises within the framework of the teaching innovation project with reference 2223-02 entitled "HUMAN-CHAT, Artificial Intelligence versus Natural: three teaching methodologies (debate, experiment, and podcast) in search of what makes us human", granted by the Universidad Pontificia Comillas to the author of this chapter. The arrival of artificial intelligence (AI) in our lives is evident: we have virtual assistants, we use tools such as ChatGPT, and algorithms determine the content we consume on the Internet. In fact, the impact of AI on society is so significant that the current era is referred to as the "Fourth Industrial Revolution", where intelligent systems are integrated into human organizations. Thus, different views emerge about the consequences it may have for our species, making it an ideal moment to reflect on what makes us human by comparing the capabilities of AI with the capabilities, to find an opposite term, of "natural" intelligence; that is, human cognitive abilities. It is not about establishing a competition between both systems, but rather about returning to a slow, Socratic, and philosophical debate about our ability to contribute to the world. Furthermore, since this revolution occurs on a global scale and may affect all disciplines, not just technological ones, there is an opportunity for interdisciplinary work in educational contexts, inviting students from different degrees to this transcendental reflection. Reaching university students is relevant because they are the current users, future beneficiaries, and potential creators of new advancements. Additionally, establishing collaborative work among branches of knowledge allows for interdisciplinary work when the compartmentalization of knowledge does not respond to the global reality we live in. Thus, in this paper, I explain the development of an annual academic program in two subjects of two degrees ("Human Cognition" in the Mathematics and Artificial Intelligence degree; and "Thought and Language" in the Psychology degree) with the common objective of better understanding the scope of current AI and reflecting on what makes us human. I have designed a series of activities following three teaching methodologies: debate, experiment, and podcast. On one hand, in IMAT and Psychology, there will be an eliminatory championship of team debates where each team will defend a randomly assigned position on whether certain cognitive processes are exclusive to human beings or, on the contrary, can be replicated in AIs (for example, doubting or thinking). The evaluation of these debates will be shared among peers (30%) and the professor (70%), and will account for 30% of the subject. Teams that lose can carry out complementary exercises to increase their score, continuing the reflection on the topic. On the other hand, inspired by the Turing test, in both degrees each team will ask 3 questions to the ChatGPT application and to three people, and will analyze the answers reflecting on how humans categorize and express information and how this process can be transferred to machines (10% of the subject). Then, the obtained answers will be presented; they will vote on which they consider non-human and will reflect collectively. Finally, all debates will be recorded in podcast format in class and the two ChatGPT works with the highest scores in each degree will be summarized and also recorded. Thus, we will create a series of episodes about what makes us human that can be disseminated so that our reflection reaches society. In conclusion, it is possible to establish a thoughtful reflection on the impact of technology using various methodologies and fostering collaboration between different university degrees without requiring too many infrastructures or materials.
The present text arises within the framework of the teaching innovation project with reference 2223-02 entitled "HUMAN-CHAT, Artificial Intelligence versus Natural: three teaching methodologies (debate, experiment, and podcast) in search of what makes us human", granted by the Universidad Pontificia Comillas to the author of this chapter. The arrival of artificial intelligence (AI) in our lives is evident: we have virtual assistants, we use tools such as ChatGPT, and algorithms determine the content we consume on the Internet. In fact, the impact of AI on society is so significant that the current era is referred to as the "Fourth Industrial Revolution", where intelligent systems are integrated into human organizations. Thus, different views emerge about the consequences it may have for our species, making it an ideal moment to reflect on what makes us human by comparing the capabilities of AI with the capabilities, to find an opposite term, of "natural" intelligence; that is, human cognitive abilities. It is not about establishing a competition between both systems, but rather about returning to a slow, Socratic, and philosophical debate about our ability to contribute to the world. Furthermore, since this revolution occurs on a global scale and may affect all disciplines, not just technological ones, there is an opportunity for interdisciplinary work in educational contexts, inviting students from different degrees to this transcendental reflection. Reaching university students is relevant because they are the current users, future beneficiaries, and potential creators of new advancements. Additionally, establishing collaborative work among branches of knowledge allows for interdisciplinary work when the compartmentalization of knowledge does not respond to the global reality we live in. Thus, in this paper, I explain the development of an annual academic program in two subjects of two degrees ("Human Cognition" in the Mathematics and Artificial Intelligence degree; and "Thought and Language" in the Psychology degree) with the common objective of better understanding the scope of current AI and reflecting on what makes us human. I have designed a series of activities following three teaching methodologies: debate, experiment, and podcast. On one hand, in IMAT and Psychology, there will be an eliminatory championship of team debates where each team will defend a randomly assigned position on whether certain cognitive processes are exclusive to human beings or, on the contrary, can be replicated in AIs (for example, doubting or thinking). The evaluation of these debates will be shared among peers (30%) and the professor (70%), and will account for 30% of the subject. Teams that lose can carry out complementary exercises to increase their score, continuing the reflection on the topic. On the other hand, inspired by the Turing test, in both degrees each team will ask 3 questions to the ChatGPT application and to three people, and will analyze the answers reflecting on how humans categorize and express information and how this process can be transferred to machines (10% of the subject). Then, the obtained answers will be presented; they will vote on which they consider non-human and will reflect collectively. Finally, all debates will be recorded in podcast format in class and the two ChatGPT works with the highest scores in each degree will be summarized and also recorded. Thus, we will create a series of episodes about what makes us human that can be disseminated so that our reflection reaches society. In conclusion, it is possible to establish a thoughtful reflection on the impact of technology using various methodologies and fostering collaboration between different university degrees without requiring too many infrastructures or materials.
Información detallada
Materias, derechos, colecciones e identificadores