Artificial intelligence (AI) has entered the educational and academic sphere at a pace few specialists had predicted. In response to this transformation, the University of La Laguna is working across various areas to address the questions, challenges, and hurdles posed by the use of AI-based tools. Simultaneously, research is beginning to integrate AI into its methodologies, leading to new lines of work aimed at leveraging its full potential.
The Information and Communication Technologies Service (STIC) at the ULL continuously monitors advancements in AI to assess their applications and risks. José Carlos González, head of STIC, warns about the importance of understanding where data goes and how it is stored when using these tools, as well as the risks of information leakage and reliance on unverified applications. Therefore, he considers it crucial to promote training on the responsible use of AI.
In line with this, the ULL is developing outreach and training initiatives open to the public. The University Extension courses on AI, led by Juan Albino Méndez, a professor of Computer Engineering, are attracting significant interest. These non-regulated training courses welcome individuals of diverse ages, including seniors adapting to a complex digital environment. The fundamentals, models, tools, and ethical implications of AI are explored.
The Director of Secretariat for Digital Teaching, María Belén San Nicolás Santos, highlights the positive reception of the AI for University Teaching Conferences. AI also holds a significant place in the Teacher Training Plan and in micro-credentials, both for the workplace and academic activities. The ULL and the University of Las Palmas de Gran Canaria are collaborating on new micro-credentials for university faculty, covering teaching, research, technical aspects, and ethics.
STIC is developing a user guide focused on security and is negotiating centralized licenses for AI tools. Pilot projects with local AI servers have been launched at the ULL to ensure data privacy, running medium-capacity AI models locally. These systems utilize high-performance hardware and open models for secure data processing.
Collaboration with the Cajasiete Chair of Big Data, Open Data, and Blockchain, directed by José Luis Roda, is key. The Impulsa IA agreement facilitates a research environment to analyze AI strategies and models. Isabel Sánchez, a management team member of the Chair, notes that Impulsa IA has carried out actions for companies, administrations, and citizens, including projects with the Cabildo of Tenerife to improve its electronic office and mobility chatbots.
The BOB Chair has promoted projects such as an automatic Spanish Sign Language translation system and the recreation of historical scenes from San Cristóbal de La Laguna using diffusion models. Work is also underway on anonymizing sensitive financial data and developing prototypes to optimize the management of electronic office requests. The chair advises on the technical feasibility of AI solutions.
AI is also transforming teaching and learning processes. Virtual classrooms, complemented by interactive resources and AI, enable more agile and personalized learning experiences. Learning analytics, analyzed by the EDULLAB research group, help identify at-risk students and inform teaching and institutional decisions. The SAPA Project combines virtual campus data and academic information for early interventions.
Within the university, there are diverse stances on tools like ChatGPT. María Belén San Nicolás Santos believes AI's opportunities outweigh its limitations, and the challenge is to teach critical and responsible use. AI is also applied in educational research, with advances in natural language processing and generative AI tools accessible to society.
Juan Albino Méndez researches AI applications in energy communities to predict consumption and generation, optimizing collective self-consumption. These models are being tested in ULL buildings as pilot projects. Other research lines include astrophysics, for telescope control optimization, and medicine, for predicting parameters like morbidity or hospital stay.
Silvia Alayón Miranda, from the Medical Image Analysis Group (GAIM), leads projects for glaucoma diagnosis using explainable deep learning and the automation of justice records. She emphasizes the challenges of explainability, privacy, and ethics, particularly in medicine and justice. The glaucoma tool aims to support clinical judgment, not replace it.
The researcher reflects on the fear that AI might replace human reasoning, while celebrating student interest in applying AI to medicine and other socially impactful fields. AI is seen as a tool to enhance professionals' work, not replace it, by validating the information it provides.
