Higher education needs a united approach to AI

If universities’ response to AI and education is as fractured as the sector’s adoption of blended learning, we may well find ourselves in a similar position in 20 years’ time with duplicated costs and missed research opportunities, writes Sara de Freitas

Sara de Freitas's avatar
Birkbeck, University of London
28 Jun 2024
bookmark plus
  • Top of page
  • Main text
  • More on this topic
Laptop with warning sign
image credit: iStock/Parradee Kietsirikul.

You may also like

Rather than restrict the use of AI, let’s embrace the challenge it offers
5 minute read
A robot and human hand put two jigsaw puzzle pieces together

Since the information revolution ignited in the 1980s, with the development of the personal computer and internet, education has been undergoing what could be called a “quiet revolution”. At its heart, this has been an information-driven revolution, which has seen extensive changes to the education and research landscape. It’s been a revolution, in that huge changes have happened over the 40-year period, and it’s been quiet because it’s been driven mainly by a bottom-up evolution of education systems.

In education, the quiet revolution has allowed for: teaching models that co-evolve with students, varied delivery models and modalities, new roles for teachers/tutors, student focus and greater personalisation, new curricula and global reach through online education networks and infrastructure. This has widened access to first-in-family students, improved opportunities for students studying overseas and supported global development. In the research context, we have seen capabilities transformed to also provide wider access to research through digital libraries, coupled with the open access and open educational resource movements, which have fuelled a research explosion that has seen, for example, growth in access to more PhD theses, faster-to-market innovations, and advanced-research findings/patents. 

Ironically, one research area that has been left behind is educational research. Unfortunately, historical bias rests here with both practitioners and researchers in the field, to the extent that in some countries its study is called “scholarship” rather than science. All biases are problematic, but this one has held back funding opportunities and could even be seen as a bulwark against cross-disciplinary and institutional innovation and research. The bias has been self-reinforcing, and underfunding has meant fewer opportunities for longitudinal or substantial scientific studies and gaps in the knowledge base. 

On the plus side, universities have put in place robust pedagogic and technical infrastructures capable of delivering online education and research services and introducing education technology applications. This considerable investment in hardware, software, technical knowledge and training was worth it, in retrospect, particularly in light of the pandemic, making our educational institutions more resilient and capable of delivering 100 per cent online education when it was needed, and significantly improving lost learning globally. 

However, it has meant uneven sectoral development, different models of delivery and fewer chances for sharing practices. The costs of this digital revolution fell first on the “early adopters” and then on the rest of the sector, which grew localised institutional/departmental learning and teaching capabilities and collaborated with edtech and open educational resources providers one at a time. But this has meant universities took on more costs. If the considerable investment that each institution put in had been better supported by centralised government agencies and research support, a huge amount could have been saved at universities over the 20-year period – about £1 million per university per year in additional staffing costs for learning and open access alone, I estimate. It’s another example of moving more cost to universities, a trend that seems to be accelerating. As an additional challenge, individual institutions have to curate the available scientific research and then find the best ways to implement it.

In the 20 years since my first report on blended learning, and in light of recent Higher Education Commission (HEC) recommendations, costs would be reduced and quality improved through better research coordination. To achieve this dual benefit, the addition of a new UKRI research institute in education sciences, as the 10th institute, would allow us to strengthen our science base for education, explore pressing challenges, such as generative and responsible AI, dynamic analytics and the ethics of education, while leveraging our cross-disciplinary and institutional excellence and saving the sector a significant chunk over the next two decades. 

Undeniably, the research base, as fragmented as it was, turned out to be critical for progress in this area, especially during the global pivot to online. For example, in the UK, we were fortunate to have had two decades to test and evaluate online education, and to deliver sound methodologies and curriculum approaches, helping our valued transnational and international students, as well as benefiting on-campus students. However, many thousands of studies replicated the finding that “no significant difference” existed between physical (face-to-face) and online education – but when a combination of physical and online was studied, a “significant difference” was found, which showed better outcomes for blended than either online- or physical-only delivery. Despite being well referenced, this is still not widely accepted or fully understood by some senior executives.

While it has been 20 years since my 2004 report on blended learning for the now-closed Learning and Skills Research Centre, the UK HEC report published in April does acknowledge the need to share practices, structure industry engagement and put in standards, but if our adoption as a community of AI and education is as uneven as our adoption of blended learning, and if we don’t start to bring our findings together, we may well find ourselves in a similar position in 20 years’ time with AI and education. Having spent probably a lot more on AI technology, is it time we took stock and put educational research at the centre of our research landscape rather than eking out an existence? We need to work together quickly in this space, as already the options for education research in AI are drying up. 

For me the next stage of the “quiet revolution” will be noisy, but if we can put in the shared investment, the sector will benefit and outcomes for students will improve rapidly. 

Sara de Freitas is visiting professor at the Open University and University of South Wales. She is governor of the board of the University of Sunderland.

If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.


You may also like

sticky sign up

Register for free

and unlock a host of features on the THE site