Author Dr Jim Turner Learning Technologist and Current Chair of ELESIG
“Educational video policy is up for review, better see what’s new. Hold up!”. That was my internal monologue a few days ago when looking at that week’s jobs, one of which is to review our current policy. For years, lecture capture has been implemented at universities to increase access and flexibility for students. This routinely involves recording the live presentation of materials, slides, audio and lecturer. Complex infrastructures have been built, expensive technology rollouts undertaken, and even more complex negotiations held over video policies. But with the rise of AI, we have to ask – is lecture capture still relevant?
Lecture capture has helped make lectures available anytime, enabling flipped classrooms and broader access. Whether you were opt-in or opt-out or something else, institutions have utilised this flexible technology at scale to seemingly meet many external and internal pressures (Ibrahim, Howarth and Stone, 2021). But in a sense, lecture capture took a surface approach to educational technology. It digitised lectures, but didn’t fundamentally reimagine teaching and learning. While lecture capture provided some student benefits, it was arguably a sticking plaster solution. The deeper, more difficult work of leveraging this technology to truly transform pedagogy rarely happened. Change was incremental rather than revolutionary (Morris, Swinnerton and Coop, 2019).
Now with the rise of AI services, we are beginning to see new ways of creating learning focused videos. With the speed of development, any mention of a particular tool will become rapidly out of date. However, an interesting example is Synthesia, UCL start up company which achieved ‘unicorn’ status in June 2023. This system allows you to develop talking head video content driven through text. Yes, things are a little wooden, but as the current saying goes ‘this is the worst it is going to be’. It doesn’t take much imagination to think through what might be coming in the near future. So let’s indulge in a little speculation.
A little speculation
Ownership: during the past 10 years careful negotiations around performance rights have helped distinguish lecture recordings from say lecture slides. One, being owned by the ‘performer’ the other, just part of normal work. The script for AI video creation, I’m guessing, would probably fall into that second category. But the video image and audio might end up under licence. Synthesia have tried to make their rights and data protection clear in their terms and conditions. This may become more complex as companies try to monetise their technology. Resolving these issues will require new policies and legal frameworks. However, as I review my university’s educational video policy, I’m conscious that AI-generated video disrupts the status quo.
Authenticity or efficiency: The mirror of AI content might help us identify the uniqueness of our own creative and quirky selves versus the current more ‘wooden’ AI self. Students might value the distinctive human qualities. However, taking a cue from Meyers research into video enabled learning perhaps there will be a more scientific data-drive and efficient learning process on offer. This also links with the ‘uncanny valley’ effect and the importance of developing trust between tutors and students, and a reduced sense of value in this type of media.
Rethink lecture capture’s role: This is a pivotal time as we rethink lecture capture’s role. Thoughtfully applied, AI-generated lectures provide scalability and access. But we must balance innovation with protecting what makes learning profound – the human connection between educators and students. If we keep sight of this core principle, then the AI revolution offers a chance to evolve into a new, human-centred era for teaching and learning.
Student experience: Students might have reached peak video consumption. This recent tiktoc shows how a student is trying to create their own efficient revision notes from transcripts of lecture videos run through ChatGPT. All of these stages are highly questionable, but it illustrates a frustration with learning from “boring lecture recordings” and balance other aspects of their life. For students, AI-generated lectures can increase flexibility even further. Videos could be customised to individual learning needs, with adjustable length and examples. Passively watching lectures, while sometimes necessary, is inherently less engaging than live participation.
Active learning: However, we have to consider the implications for student engagement and learning. Watching a video, no matter how slickly produced, is inherently more passive than participating in an interactive live lecture. There are concerns students may retain less and feel less connected to instructors and fellow students. Does this now open up more possibilities for video assessment, student-generated content, and open course resources to support learning beyond the institution.
Is video just too static: Except for very specific cases, such as demonstrations, will the AI assistant, possibly enhanced through visual avatars, just mean the talking head video is replaced with the AI talking head. This is probably a little further off, I think, he says tentatively, but at the very least it offers some level of interaction beyond play/rewind.
So just as AI has changed our relationship with text, we now have to address its impact on video creation. But was lecture capture technology always a bit of a lazy answer to a deeper, more purposeful use of technology that complex institutions couldn’t achieve at scale? Perhaps it’s time to go back to first principles. My view is a balanced model combining the best of both approaches is ideal. Use AI for foundational concepts, freeing up resources for human-centred sessions focused on discussion, problem-solving and collaboration. While AI-generated videos offer many benefits in terms of flexibility, cost and scalability, challenges remain. As with any technology, it is not a magic solution. Thoughtful integration and policies around rights, ownership and student experience will be needed to successfully leverage AI video in education. “Right blog over, must get back to that policy”
Ibrahim, Y., Howarth, A. and Stone, I. (2021) ‘Lecture Capture Policies: A Survey of British Universities’, Postdigital Science and Education, 3(1), pp. 144–161. Available at: https://doi.org/10.1007/s42438-020-00102-x.
Morris, N.P., Swinnerton, B. and Coop, T. (2019) ‘Lecture recordings to support learning: A contested space between students and teachers’, Computers & Education, 140, p. 103604. Available at: https://doi.org/10.1016/j.compedu.2019.103604.
Dr Jim Turner Learning Technologist (LJMU) and Current Chair of ELESIG
Alex Spiers, Build and Visual Design Manager, King’s College London