by Zac Woolfitt, Inholland University, The Netherlands. This article on the Media and Learning conference in Leuven 18-19th June 2025 provides a detailed description of many of the highlights from…
by Zac Woolfitt, Inholland University, The Netherlands.
This article on the Media and Learning conference in Leuven 18-19th June 2025 provides a detailed description of many of the highlights from this year’s well-attended event.
Editor’s note: Long Read – Reading time 10 – 15 minutes.
The annual Media and Learning conference took place in Leuven on 18th and 19th June with the tagline Educational media that works. It was organised jointly with KU Leuven’s Learning Lab. There were more than 250 attendees from 20+ countries, of innovative educators, educational media production and audiovisual services as well as researchers and policymakers. Disclosure, I am on the advisory panel for the conference and have been attending since 2015.
There were so many interesting sessions and workshops, and I could only attend a few. This article is by no means comprehensive, it can’t be. I’m sorry if I missed your great session and it’s not included here.
By far the highlight of the conference for me, was being on the panel discussion with Richard E. Mayer, Distinguished Professor of Psychological and Brain Sciences, University of Santa Barbara, California. In case you are not familiar with Dr. Mayer, he is the number 1 ranked educational psychologist in the world, being ranked the ‘most productive’ and ‘most cited’. He is ranked in the upper 0.01% of scientists in the world based on total citations by PLoS Biology and is ranked in the top 100 research psychologists in the world by research.com (#27 in the United States and #50 in the world).
Waving to Dr. Richard E. Mayer live streamed from California
Over the last fifty years, Dr. Mayer has conducted research into the role and design of video for learning. His Multimedia Principles continue to form the basis for much of the work carried out in our sector. He joined us live from his home in California. If you are not familiar with the principles, then you can take a look before your next presentation.
Dr. Mayer took us through his key principles and how they relate to cognitive load theory. Then the panel had a chance to ask questions and discuss some ideas. His multi-media principles are based on multiple empirical experiments and tested in multiple contexts over many years. So, we know that each principle works. Everyone who presents a combination of spoken text, written text, and images should use them. That means, all educators!
One of the multimedia principles states that we learn better in multimedia presentations when we are ‘pre-taught’ important vocabulary. The conference programme included a list of 20 key terms from his presentation that we could study in advance, to help us learn better from his session.
I asked Dr. Mayer, why, even though his principles are tried and tested, and we know they work, that this theory is not mandatory in all teacher training courses around the world? He said there was some progress but still a long way to go. I also asked him whether we should be focusing more on the impact of AI, rather than multimedia presentations? He accepted the importance of that, even though his multimedia principles remain valid. Effective teaching can be done in collaboration with AI. One idea that emerged during the session to develop an AI tool that could scan any media presentation and assess how well it follows the principles. After the conference, I was very glad to see that Serge de Beer, video specialist, set up an initial prompt to start this process.
With thanks to Lucy Kendra of Heriot-Watt University moderated the panel, and Lana Scott, MIT, Elke Van Ael, KU Leuven. It’s the first time that Media and Learning have live-streamed an external speaker to our conference, but we could not have picked a better expert or subject for our first effort.
Thank you Dr. Mayer, for getting up early to join us!
Why do some images, words, or experiences remain vivid in our memory, while others vanish quickly? In a fascinating demonstration, Thomas Espeseth from the University of Oslo, shared his research into how memory works and how this can be used in making educational materials easier to remember. Referring to the Gorilla and Basketball video, and showing the Who-Dunnit Video, inattentional blindness and divided attention show how easily we can overlook important details. We often don’t see something if we are not looking for it.
‘Attention is a crucial component of consciousness but uses cognitive resources. What we attend to is often shaped more by stimulus properties and task demands, than by deliberate choice’
The relevance for educators is to understand what makes learning content ‘stick’ and how to leverage this as instructional designers. With some quick fire images (flashed to the audience in a second or so), Thomas demonstrated how we often only see what we are looking for, and miss a lot of other things during attentional blindness or the ‘attentional blink’ . He referenced Schacter, 2001, seven sins of memory: The-seven-sins-of-memory-Insights-from-psychology-and-cognitive-neuroscience.pdf: transience, absent-mindedness, blocking, misattribution, suggestibility, bias, and persistence.
And looking back to Ebbinghaus (1885) and the ‘forgetting curve’ he showed how we can make text and images ‘more memorable’ which can result in learners remembering more easily, and for longer.
Ebbinghaus would have loved how AI can optimize for memorability
The model he developed predicts which text will be more easily remembered and works on using AI to adjust it. Principles to make images and text more memorable are sometimes obvious, sometimes not. In many of his tests, the most memorable image (for both men and women) was of a ‘Bikini’. Training AI on large-scale memory data has great potential for predication and generation of high impact content to complement learning strategies. I found it a fascinating presentation and encourage you to follow Thomas’ work.
Since 2015, I’ve been writing my blog, called ‘Video Teaching’. I’m interested in how ‘teaching through the screen’ impacts the learning process. So, I was glad to moderate this session with five expert panellists on the ins and outs of Lecture Capture. We examined Lecture Capture (LC) from three perspectives: the organization, the educator, and the learner. We started looking at the practicalities and reasons for LC but ended up with a much wider reflection on teaching models, paradigm shifts, and the essence of teaching and learning.
Some universities legally require recording of lectures for accessibility while some focus on hybrid virtual classrooms. We discussed what control each educator has in what is recorded, and when they can stop recording, for sensitive in-class interactions. And there are still fears of ‘losing our job if we are recorded’, which I first wrote about in 2015. Has lecture capture improved the live performance of lecturing, or acted as a reflective tool? How do the post-production processes of tagging, subtitles and archiving support the learning experience? And does anyone every actually watch all this recorded material stored on servers for so many years? These were just some of the points in our discussion.
Although this session was not ‘lecture captured’ (there is no recording for you to view), we had a lively discussion with critical input from the audience and panel. With many thanks to Carlos Turró Ribalta, Universitat Politecnica de Valencia, Timo Nogueira, virtUOS, University of Osnabrück, Jean-Marie Cognet, EdTech France, Lars Binau, Technical University of Denmark (DTU), and Leif Kristiansen, Norwegian Agency for Shared Services in Education and Research, Norway (SIKT)).
According to Tine Baelmans, Vice Rector for Educational Policy at KU Leuven, ‘GenAI takes many to tango!’. KU Leuven, (currently celebrating its 600th year) hosted the conference and the Media and Learning office will move into premises in the University later this year. Tine has been a strong advocate of Media and Learning in her role as vice rector over the last few years. I had a chance at the pre-conference dinner to thank Tine in person for the strong support she has given M&L over the years and will step down from her role as Vice Rector later this year
We should engage, embrace, and not be afraid of AI. Tackling academic misuse cases is still challenging. Chat GPT only took 5 days to reach a million users and KU Leuven were one of the first movers to establish and publish guidelines: with ongoing support, training for staff. AI literacy was a main theme during the conference. Including working on the AI act at a European level to examine security, environmental, ethical aspects and how it impacts the educational paradigm. We should bridge the ‘AI gap’ and not punish those using AI but be aware of the risks. Recent research warns we may become ‘less capable’ as humanity.
Maarten Francq from the Royal Institute for Theatre, Cinema and Sound (RITCS) in Brussels is pursuing a PhD in “AI in Live Entertainment”. His innovative research includes fully AI-generated music videos (check out the very cool video for Belgian artist DAAN). If we do things right, we can spend time relaxing on the beach while AI does all its work for us. Just like he does! Maarten’s lively, provocative and fun presentation woke up the audience as we reflected that we may be a lot further ‘behind the AI hype curve’ than we thought!
Will AI take my job? No. Our jobs won’t be replaced by AI, but we’ll be replaced be people who can work with AI better than we can. Maarten is considering building a clone of himself, so his clone can create the work he’s paid for today. By adding all that he has ever created in his life, to the model, he can set AI to work for him. We should not lose track of where the human is in the loop.
Maarten had us all running to create our own ‘core AI Agents workflow’ via an ‘MCP’ (Model Context Protocol). I had to look up what an MCP is for a good explanation. Basically, an MCP helps AI systems work together more smoothly, replacing the messy one-off fixes we often use today. It lets AI keep track of what it’s doing—even as it switches between different tools and datasets. As this technology develops, AI won’t lose its place when moving between apps—it’ll be like having a consistent memory across tools. The MCP connects to real-world tasks through a central workflow, making the AI more useful in practical ways.
For example, it can link Google’s VEO3 (Video Generator, used here by Imagine.Art and Heygen (AI text to Video Generator). The idea is that AI agents will be able to use these tools directly to create high-quality content. It’s still early days—it’s only June 2025—so let’s check back in six weeks to see how things are progressing.
It is now possible to easily create stories and make movies with CGI for very little money. Imagine livestreaming your favourite film and choosing which actor you’d like to play the main role, as it streams. Or taking a picture of a missing part of your bike, using AI to convert it to 3d code, and then 3d printing it in your home.
It is time to move beyond prompting and integrate AI usage in other ways. Today, most people are using AI tools to substitute what we are already doing, or to help with daily tasks. Now is the time to combine different AI tools to create completely new things. Overall, a very informative presentation (thank you Maarten) and it woke me up. Time to get more plugged in to these developments in order not to be left behind. We shall see!
Andy Thys, KU Leuven moderated the panel on ethical issues in AI. Sonia Hetzner, FAU took us through how AI caught higher education institutions off guard, and now AI-generated assignments pose significant academic integrity challenges. From the initial shock and pain, we have moved to toward experimentation and practical adoption. Using Kotter’s 8 step model of change. We are now entering a more strategic phase that focuses on AI as a structural force reshaping teaching and learning, requiring institutions to rethink curricula, develop AI competencies, and implement long-term, adaptive strategies informed by research and leadership. This includes using frameworks like Kotter’s change model and tools such as AI checklists, while also addressing ethical concerns like bias and discrimination. The “Furth Manifesto“(March 2025) is a guiding document for future pedagogical practices and policy design.
Cornelia Amon, IMC Krems University of Applied Sciences, Austria was involved in designing the iMooX course “AI! But how?”. It offers valuable insights into designing AI-integrated teaching while fostering critical media literacy. AI’s arrival was disruptive but should be treated as another tool in the evolving media landscape—comparable to past shifts with books, radio, and TV. The course encourages students to become critical, ethical users of AI, helping them evaluate tools not just by function but by their data practices, company backgrounds, and ethical implications. Interactive and scaffolded activities—like gamified quizzes, short exercises, and discussion forums—enable engagement, but reflections revealed that content format (e.g., videos) needs careful consideration. The key takeaway: we don’t have to use AI for AI’s sake—pedagogical value and ethical clarity should guide tool selection and course design.
Jessy De Cooker, Fontys Hogeschool, The Netherlands Four dimensions of AI literacy: the dos and don’ts in AI education in J-schools. In journalism education, the integration of AI challenges traditional roles, prompting educators to shift from authoritative knowledge-givers to facilitators of collaborative exploration. AI is shaping the media landscape—including generating fake book lists and mimicking editorial processes—there is an urgent need to define AI literacy within the field. Many educators feel unprepared to teach these tools, yet initiatives such as peer-led learning, flipped classrooms, and interdisciplinary approaches can help develop student and teacher competence. Does AI improve journalistic quality, and how do we manage the ethical and practical risks involved?
The audience asked several times ‘How do we know which AI is good or not?’ I suggested to the panel that we could develop an ‘AI Nutrition Score’ along the lines of NewsGuard (who do the same for media sites). NewsGuard have a thorough, logical and transparent rating system which we could use for AI tools.
Ivana Juraga is Team Leader for the Digital Education Ecosystem and Evidence, Digital Education Unit, at the EU. She zoomed out to the European level and looked at research and reports that are currently being compiled on this subject. She mentioned the (2020) Digital Education plan (2021-2027). Also, the outline of the 2030 Roadmap for the future of digital education and skills which is currently under preparation. Their 2022 publication, ‘Ethical guidelines for educators on using artificial intelligence’ is one of the most downloaded reports from the EU site (update pending in 2025). Ivana reminded us that there is an increasing digital divide. Only 20% of EU children attend schools with internet speeds above 100mbs, and, according to the ‘International computer and information literacy study’, IEA, (2023), 43% of 8th graders lack basic digital skills.
In another workshop, I role-played some additional dilemmas about AI in a business context. We were assigned a role in a software development company, with budget to invest in certain AI tools. Our team looked at which AI tools to buy, with which functionality, for which goals? This was a good way to explore the complex issues of making ethical choices in a fast-shifting world where there is little transparency. Lara Pedraz and Pein Leeuwenburg from Leiden University guided us expertly through the session. I would have liked more time, but their original 90-minute concept had to be squeezed into 60. Such is conference life.
Thierry Koscielniak is President of France Immersive Learning / Executive Advisor, Arts et Metiers Institute of Technology. He showed examples of how they are using AI and technology to create immersive learning experiences. Combining the power of virtual, augmented, and mixed realities transcends traditional learning media, offering deeply engaging, experiential learning which can enhance retention and engagement.
Polling the conference audience, Theirry saw that about 20-30% of us had a VR headset at home. He stated that when this reaches 90%, then VR will finally start to connect in education.
He explained the JENII project which creates integrates ‘digital twins’ and ‘immersive virtual environments’ to train future engineers. They can practice using tools they will encounter in dangerous and complex industry settings. In a digital twin of a steel foundry, learners wear VR glasses and use tools and hand controls to practice from the safety of their own space, in what in the real world, is a highly dangerous place. In this VR world, you can ‘get burned’, ‘cut your finger’, or ‘drop acid on the floor’. He introduced us to the term Paleo Futuring (or Retrofuturism), in which we examine what people in the past thought the future would be. We are rarely right so maybe we should not be too confident about what is coming next.
André Rosendaal, University of Groningen moderated the final session I attended in which research and case studies of various XR projects were explained. The group of students from Leiden University, Sofie Delansay, Ula Andrulyte & Zsofia Egri, gave an impressive presentation on their research. They clearly and confidently shared their provisional findings while nuancing their scientific approach. Creating an immersive VR world, their project featured a “1,000 Cuts” VR journey. In this world, you step into the physical body of someone else, you look in the mirror and see you are an African American. Then through various scenes, you experience first-hand what it is to experience racial abuse in different contexts. With pre- and post-surveys and thematic analysis, 50 codes emerged, to identify relevant factors. While IVR was shown to enhance immersion and complement traditional learning, there were challenges regarding classroom logistics, scepticism from faculty, and a lack of intersectional perspectives. They emphasized the importance of post-experience discussion and proposed incorporating gender, class, and race dynamics to deepen impact. This was an enjoyable, informative and clear presentation and a good case of how VR can enhance the learning experience.
Eline Wydooghe, VIVES University of Applied Sciences, showed how they are using VR to practice laboratory experimentation on animals. In veterinary training at VIVES University in Belgium, they tackle ethical and educational challenges in laboratory skills development. Eline highlighted the “Three Rs” of laboratory experimentation on live animals (Replace, Reduce, Refine). Their VR simulations use Meta Quest 3 to reduce reliance on live animals, with students practicing realistic tasks, like injecting model mice.
Practicing handling lab mice in VR
Survey results showed increased engagement and perceived learning capability, particularly among those with prior VR experience. While the approach was seen as effective and ethical, the team found more introductory sessions could help. Participation was optional and there were some issues of nausea and physical strain. This was an excellent example of how VR can support learning and while reducing the number of animals used in laboratory experimentation.
Florence Mei Kuen Tang, The Chinese University of Hong Kong uses VR to help medical students practice various emergency medical treatments on virtual manikins. Active, simulation-based education through VR can engage learners. The team developed a VR ECG (electrocardiogram) training simulator using Unity, incorporating gamified elements to help students move from theory to practice. Using interactive VR scenarios, students could remember concepts better and apply them meaningfully supporting the move towards more dynamic, memory-retentive pedagogy. There were also issues of nausea and motion sickness, but the general response was positive.
Emerald Grimshaw, Durham University shared the VR world, with touch and sound elements, developed for younger children to use VR to visit the Silk Roads in an Immersive VR. She focused on VR’s potential in early education, particularly for engaging children with immersive historical experiences. Using Meta Quest headsets, they recreated the Silk Road and ancient Chinese trading cities, allowing children to explore 3D-scanned artefacts virtually (this included creating a model of a camel’s hump covered with carpet, which could be physically touched during the experience). Their research highlighted the growing prevalence of VR in households and underscored the importance of educational institutions understanding and leveraging this trend. The use of VR not only fostered curiosity and cultural learning but also made traditionally inaccessible artefacts more tangible, supported by clear and engaging visual materials. I had some serious questions about the ethics of using Meta headsets with very young children, and although they said the users were anonymous, there is still a vast amount of data being sucked up by Meta. See my recent reflections on this here. My compliments to Emerald for the clearest slide deck I’ve seen in a long time; Dr. Mayer would have been impressed.
John Walker from SURF and Npuls ran a session to examine values in XR design. His engaging and high-energy approach made this fun and informative. What should XR technology protect or promote? Is it safe for students? How do we balance innovation with ethical concerns?
In groups, we worked through scenarios using the ‘Polder Perspectives XR ’ card game designed to make ethical dilemmas in XR technology more tangible. This is a complicated subject, and we only had time to scratch the surface. For example, see the recent case of Meta’s Smart Ray-Ban Glasses that were used to secretly film unsuspecting women on the street. How should we as educators manage the ethical issues that XR raise in our educational contexts?
John Walker covers some ethical challenges of XR
Nynke Kruiderink is Project Leader for the Centres for Teaching, Npuls, in the Netherlands. Nynke took us through the Dutch approach to Centers for Teaching & Learning at a national level. Npuls is an 8-year plan funded by the Dutch government to build an education sector that supports a resilient society, broad prosperity and a strong economy CTLs support five core functions: professional development, knowledge sharing, educational innovation, support, and research. If set up and deployed correctly they can create a positive meeting point to empower educators and encourage learning organizations. CTLs can use the SECI knowledge model to move between explicit and tacit knowledge. There was a pre-conference workshop on CTLs that was well received, though I could not attend.
The CTL landscaped as seen by Npuls
Nynke referenced research by Ruth Graham that showed that for most of their careers, teachers in universities are not rewarded for their teaching achievement. There is a significant ‘stagnation’ of reward. Only at the start and end of our careers is performance rewarded. The CTL can support improvements to change the system, so we facilitate an attitude and behaviour of excellence in teaching.
One of the most engaging aspects of the conference was the group of finalists who join for the annual Medea awards. This year they joined from Australia, the US, Spain, Germany and several other countries. It is inspiring to see the top quality and original educational content created through media. The winner this year was from Supercampus – 3Cat – Contents, Programation and Cultural Area in Spain.
It was a packed, energetic and fun conference. Many thanks to Sally, Mathy, Dovi, Chloé, members of the advisory committee, and colleagues at KU Leuven for a well-organized and enjoyable conference.
Conference delegates and Ed Tech suppliers networking
And of course, thanks to the sponsors of the conference, see below, without whom it would not have been possible!
Conference sponsors
In general, the text is written by me. In a very few places, I used ChatGPT to simplify some text to make it more readable.
Thank you for reading this far. If you’ve got questions, corrections, or suggestions please email me at zac.woolfitt@inholland.nl
Zac Woolfitt, Inholland University of Applied Sciences, The Netehrlands
Editor’s note: This article appeared first on Zac’s blog which you can check-out here.