Strategic planning in education needs to consider the contemporary challenges and the type of decision-making context. Today’s education has been significantly impacted by two major unprecedented events which have had a strong transformative effect: the COVID-19 pandemic and the rapid spreading of AI.

To better understand the decision-making in such contexts, it is useful to look at them through the lens of the Cynefin framework. As explained by Snowden (2002, p. 104), “Cynefin creates four open spaces or domains of knowledge all of which have validity within different contexts“. In the Cynefin framework, contexts are “defined by the nature of the relationship between cause and effect” (Snowden and Boone, 2007, p. 2). In the four contexts, simple, complicated, complex, and chaotic, leaders need to recognise the situation and act appropriately.

The COVID-19 context was a chaotic one, where it was not possible to identify the relationships between cause and effect and there were no patterns, so leaders had to act immediately and work to shift from chaos to complexity. Once the world has shifted to a complex context, another challenge emerged: the mainstreaming of AI. Both challenges have had a significant impact on education today, on the global, national, and institutional level, as well as on the level of an individual learner and educator.

Strategic decision-making today is therefore double-burdened. On the one hand, it has to consider the possibility of new major global threats like a pandemic or a war, and their possible tremendous effect on education. On the other hand, careful thought should be given to AI, which can be a powerful assistant to deal with threats like this, but also a hidden enemy to ethical and meaningful teaching and learning.

Previous research has shown that the agility and fitness-for-change of educational systems and their main actors is essential in responding to and coping with major challenges (Svetec and Divjak, 2021). To respond to the requirements of the contemporary complex context, educators should be continuously strengthened and motivated to harness the potential of learning design (Divjak et al., 2022), as a possible universal language, and innovative pedagogies. Meaningful learning design and implementation of future- and learner-oriented teaching and learning practices can be enhanced by the ethical and creative use of AI. Sensible use of AI and minimising the black-box effect require changes in curricula, as well as initial education and continuous professional development of educators (Rienties et al., 2023). Finally, it is crucial that decision-makers are ready to adequately respond to the given complex context, being agile, collaborative and relying on relevant expertise and evidence provided by learning analytics (Divjak et al., 2023). Both educators and decision-makers need future-literacy to be able to imagine possible futures, as well as practical skills to streamline education towards the desired future.

Prof. Blaženka Divjak, PhD, is a Full Professor of Mathematics and Information Science at the University of Zagreb, Faculty of Organization and Informatics. She served as Vice-Rector for students and study programs at the University of Zagreb (2010-2014) and was the Croatian Minister of Science and Education (2017-2020). She chaired the EU Council of ministers for education and for research and space during the Croatian presidency (2020). She has been a coordinator/researcher in over 30 (inter)national projects. Currently, she coordinates the "Innovating Learning Design in Higher Education" Erasmus+ project, as well as the “Trustworthy Learning Analytics and Artificial Intelligence for Meaningful Learning Design” project funded by the Croatian Science Foundation. She has led the research group designing the concept and the accompanying Balanced Learning Design Planning tool, used in more than 30 countries. She is a Vice-President of SOLAR and a member of the Advisory Board of GRAILE. Her area of expertise, besides mathematics, includes curriculum development, e-learning, learning design, assessment, learning analytics, strategic decision-making and AI in education.

In this talk, I'll share my personal journey navigating the intersection of AI and education, reflecting on how large language models (LLMs) has significantly shifted the landscape. Beginning with Project Halo, reflecting Paul Allen's vision of an all-knowing AI system and tutor, and prototyped as a "knowledgeable textbook" for students, we found educational gains but at an unsustainable manual development cost. The arrival of LLMs dramatically changed the landscape, allowing us to develop Aristo, the first AI system capable of passing elementary science exams, and later able to also explain its reasoning to students. This evolution led to a prototype called TeachMe, where students learned by "teaching" a LLM when it made mistakes (following Feynman's maxim: if you want to master something, teach it). Finally, I'll describe our current work on an AI Research Assistant, where a researcher and AI system work collaboratively together to solve research questions, learning from each others' strengths. Through this talk, I will share the lessons, surprises, and the evolving opportunities I've encountered, and offer a personal perspective on how LLMs have changed our approach to AI, including in education, and where I see the evolution heading.

Peter Clark is a Senior Research Director and founding member of the Allen Institute for AI (AI2). He leads AI2's Aristo Project, aiming to build the next generation of systems that can systematically reason, explain, and continually improve over time. He received his Ph.D. in 1991, has published over 250 papers, and has received several awards, including four Best Paper awards (AAAI, EMNLP, AKBC), a Boeing Associate Technical Fellowship (2004), and Senior Member of AAAI.

Artificial Intelligence in Education (AIED) possesses the capacity to augment individuals' ability to address educational challenges across various levels, from system-wide policies at the country level to supporting individual students with tools like intelligent tutoring systems. In my talk, I will begin by showcasing some of the outcomes my group has achieved at the student level, specifically in enhancing students' motivation and engagement in STEM subjects through gamification, which involves the application of game elements in non-game contexts. I will then explore the broader implications of AIED on public policy, emphasizing the critical need to bridge the gap between AIED research and its practical applications to inform the design and implementation of educational policies. Using the Brazilian context as a reference point for the challenges faced by countries in the Global South, I will introduce the concept called "AIED Unplugged," my group's innovative methodology designed to create and integrate AI technologies into existing educational ecosystems without requiring infrastructural modifications, consistent internet access, or advanced digital literacy. This approach has been applied to reform an education policy in Brazil, aimed at improving students' writing skills and mitigating some of the adverse effects of the COVID-19 pandemic on students' learning. Our findings demonstrate a reduction in the time, cost, and complexity of policy implementation, alongside a substantial positive impact on more than 500,000 students across 7,000 schools nationwide. I will conclude this talk by offering the community some ideas for the strategic application of AIED research and development as a means to augment equitable learning opportunities for all.

Seiji Isotani is a Visiting Professor of Education at the Harvard Graduate School of Education and a Professor of Computer Science and Learning Technology at the University of Sao Paulo, Brazil. He earned his Ph.D. from Osaka University, Japan, and was a postdoctoral researcher at Carnegie Mellon University. For more than 15 years, Isotani has dedicated his research career to advancing the science of how people learn with interactive/intelligent educational technologies and exploring potential mechanisms to ensure every student receives the personalized support they need for fulfilling and meaningful educational experiences. He is renowned for his work in the fields of Gamification in Education, Intelligent Tutoring Systems (ITS), and Artificial Intelligence in Education (AIED). Since 2017, he has served as a technical/scientific advisor to the Brazilian Ministry of Education, designing and implementing public policies related to educational technologies. He was a key contributor to the development of norms for the K-12 Computer Science Curriculum in Brazil and to the design and implementation of educational policies that have significantly influenced over 50 million students nationwide. Examples include the policy to evaluate, purchase, and deliver books to every student in the country, the establishment of the Brazilian National Hybrid Learning Network, and the Policy for Learning Recovery. The latter policy was acclaimed at the 2022 World Economic Forum as a groundbreaking post-COVID-19 innovation.

The advent of advanced AI systems, particularly generative pre-trained transformers (GPTs) and large language models (LLMs) like ChatGPT, Gemini, Claude, has significantly changed the landscape of artificial intelligence and its applications across various domains. GPTs have demonstrated impressive capabilities in generating human-like text and solving complex problems. Educators are finding creative ways to deploy GPTs systems in their practice [1]. Many schools are struggling with the question of whether to allow or ban the use of AI, because of the potential risks (overreliance, academic integrity). The anticipation of change in the education system is palpable. What does the future hold? This raises important questions about the role of artificial intelligence in education (AIED) research. This panel will discuss the reasons why and ways in which AIED research should or should not remain essential in these transformative times for education, as well as ways in which it should change.

Panelists: Giora Alexandron, Ig Bittencourt, Rafael Ferreira Mello, Danielle Thomas, Olga C. Santos

Panel organisers: Julita Vasileva and Alexandra I. Cristea

Danielle R. Thomas is a Systems Scientist at Carnegie Mellon University and the Research Lead at Personalized Learning Squared (PLUS). PLUS is a research project led by Prof. Ken Koedinger that aims to double math learning among marginalized students. Dr. Thomas’ research focuses on enhancing student learning and promoting educational equity through the advancement of AIED, the development of hybrid human-AI tutoring systems, and understanding how inequity affects student learning. Dr. Thomas brings valuable experience from her past role as a principal in a school grappling with significant educational challenges such as poverty and limited access to technology. Her firsthand experiences drive her commitment to synergizing human and AI to efficiently enhance learning among underserved students, including this year’s AIED paper, “The Neglected 15%: Positive Effects of Hybrid Human-AI Tutoring Among Students with Disabilities.” Recently, she has published in conferences, such as AAAI, EDM, LAK, L@S, and the International Journal of AIED.

Rafael Ferreira Mello has a Ph.D. in computer science with research interests that span across learning analytics, generative AI and natural language processing. He is a Professor at the UFRPE and CESAR in Brazil, where he holds the position of Chair of Education Technology Adoption and Data Analytics. In the last few years, Dr. Mello has supported the adoption of Learning Analytics techniques and tools in several institutions in Brazil. One recognition of his contributions is that Dr. Mello is the leading researcher at two national-level projects, funded by the Brazilian Ministry of Education, aiming the development of (i) natural language processing and learning analytics techniques to enhance students written productions and (ii) Learning Analytics Dashboards to support decision-making regarding the national education plan. Moreover, he has worked on several multinational research projects, which involve institutional and organizational partners in Europe, Australia, and Latin America. He has published in the leading international journals and conference proceedings in the fields of his research. He also served as an assistant editor and reviewer for several journals and conferences, including the International Learning Analytics & Knowledge Conference, Journal of Learning Analytics, Computers and Education, British Journal of Educational Technology, and Internet and Higher Education. 

Giora Alexandron is an Assistant Professor at the Department of Science Teaching at the Weizmann Institute of Science and the head of the Computational Approaches in Science Education (CASEd) research lab. His research primarily focuses on three areas: First, on Teacher-AI partnership, spanning from modeling and algorithmic foundations through learning engineering, to studying how human factors shape acceptance of and trust in educational AI systems. Second, on educational NLP, particularly in non-English languages. And third, on AI education for school students and AI literacy for science teachers. In all these areas, he is particularly interested in working in authentic learning environments and closing the loop with the field to bring actual value into schools.

Ig Ibert Bittencourt is a psychologist and computer scientist, an Associate Professor at the Federal University of Alagoas (UFAL), a Visiting Researcher at the Harvard Graduate School of Education, and a recipient of the Brazilian National Order of Educational Merit medal. His career is dedicated to the United Nations Sustainable Development Goals (SDGs), especially SDG4 and SDG5. He is devoted to education, working on the design, development, and experimentation of intelligent educational technologies to promote equitable learning and well-being. 

Olga C. Santos is an Associate Professor in the Department of Artificial Intelligence at the School of Computer Science of UNED, the Spanish Distance University. She is also the founder and director of the Research Center on Physical User Modeling (PhyUM) at the same university. Active in the AIED community since 2003, she is currently the president of the International AIED Society. In her visionary paper published in the Journal of AIED in 2016, she introduced the SMDD framework for developing intelligent psychomotor systems. Following this framework, she has overseen the development of several prototypes (and a patent) in various domains, including martial arts and sports, rehabilitation, and active aging. Some of these prototypes, such as KSAS, KUMITRON, and iBAID (at AIED 2024), have already been disseminated within the AIED community. Dr. Santos has received numerous awards and distinctions, including the Award for Best Doctoral Thesis from the IEEE Spanish Chapter of the Education Society, the Young Researcher Award from the IEEE Learning Technology Technical Committee, and the Archimedes Engineering Award from the Ministry of Education, Culture, and Sports of Spain.