Is Artificial Intelligence an Independent Science? Implications for Higher Education
Ioannis Pitas
Artificial Intelligence is rapidly evolving into an autonomous scientific field, with strong interdisciplinary links to Computer Science, Neuroscience, Psychology, Linguistics, and Engineering. The global rise of undergraduate AI programmes demonstrates that traditional Computer Science and Electrical Engineering curricula can no longer fully accommodate its scope.
The lecture examines how the digitisation and mathematization of knowledge are transforming all academic disciplines and creating the need for a fundamental redesign of education. It argues for broader access to quality AI education, the cultivation of critical and algorithmic thinking, and the establishment of new academic structures, such as Schools of Information Science and Engineering. Particular emphasis is placed on the impact of AI on Higher Education, the Humanities, Law, Health Sciences, and the need for human-centred technologies informed by Ethics, Sociology, and Psychology.
Applications of Artificial Intelligence in Culture and Health
Sotiris Goudos
This lecture presents two research-based applications of Artificial Intelligence developed at the Aristotle University of Thessaloniki, highlighting AI’s growing role in culture and health sciences.
The first application concerns the use of Convolutional Neural Networks for the recognition of Byzantine hymns, a complex task due to the distinctive musical and acoustic characteristics of the Greek Orthodox tradition. The second focuses on the use of AI and Computer Vision in Assisted Reproduction, through the development of software designed to identify the most viable embryo and support successful pregnancy outcomes.
Artificial Intelligence in Materials Science
Prof. Joseph Kioseoglou
Artificial Intelligence and Machine Learning are transforming materials science by accelerating property prediction, optimization, and the design of new materials. Modern approaches, including deep learning, graph neural networks, generative models, and autonomous experimentation platforms, enable researchers to analyse large datasets, propose new crystal structures, and guide materials discovery more efficiently.
The lecture outlines the main AI methods used across the materials pipeline, from data extraction and property prediction to inverse design and automated experimentation. It also addresses key challenges, including biased datasets, limited transferability, interpretability, physical consistency, and the need for reliable benchmarking and data provenance.
Moderator Panos Patsalas
