- This blog was kindly authored for HEPI by Ian Pickup, Pro Vice Chancellor, Students, at The Open University, and builds on opening and closing remarks made at a recent HEPI round table event, held in partnership with Chegg.
As I was leaving home to attend a recent HEPI roundtable event, my A-level-year-daughter asked where I was heading. On learning about the focus of the discussion, we chatted about the use of Generative AI tools amongst her peer group, and she wondered why my contemporaries were so flustered by this latest advance in technology. Her views chimed with the findings shared in a recent HEPI policy note which suggested that the majority of students consider it acceptable to use Generative AI as a support for learning. “Catch up, Dad”, was the encouragement I needed!
Thinking through the implications of the advance of Generative AI-powered technologies for universities has been a cause for personal reflection. I was amongst a generation of undergraduate students who submitted essays, and sat final exams, with a pen on paper; for whom a limited supply of recommended textbooks was held in the campus library with a waiting list deployed at peak study times; who subsequently experienced and benefited from new technology in the early phases of professional lives (in my case as a schoolteacher). It is strange to remember a time before I communicated so habitually through email, the first time I used an academic search engine, or ventured into a virtual learning environment. For many, the daily use of video conferencing was only accelerated during the Covid-19 pandemic, and experimentation with Generative AI tools may have recently shifted beyond the most frivolous of hobbyist prompt writing.
In this way, there has always been a next generation technology; I have always been riding the wave of technological change.
In his recent book, Mustafa Suleyman (co-founder of Inflection AI, and previously DeepMind) draws parallels with the experiences of Ned Ludd and contemporaries to remind us that humanity has a wonderful tendency to adapt. Whilst significant disruption is a threat (as it was for the textile workers in the early 19th Century) Suleyman argues that we must claim the benefits of the wave, whilst protecting ourselves and society from the potentially negative impacts that the same wave may bring.
I am fortunate to benefit from the wisdom and expertise of colleagues at The Open University who have been working on AI and large language models for some time. Our open-access, distance learning model has applied the use of AI to support students for over 10 years. We deploy an in-house tool that supports tutors in identifying students at risk of withdrawing, using weekly predictions to enable timely educative interventions. This can support a significant uplift in pass rates, retention rates and reduction of awarding gaps amongst our diverse student body. We are now seeking to develop this approach further, whilst testing the possibilities of AI-enabled educational co-pilots, personalised instructional design, and an ever-more-individualised approach to learning, teaching and assessment at scale. Through the diligent work of academic leaders and a vibrant community of practice of over 200 OU staff, we published our first statement about the use of Generative AI in learning, teaching and assessment at The OU in 2023, along with guidance for students and staff. We have been careful to foreground parallel work that facilitates inclusive and authentic assessment strategies, whilst ensuring that our teaching, learning and assessment best prepare our students to achieve their study and career goals.
The use of Generative AI in professional service delivery is also in scope, where we see potential to drive efficiencies in operating models, personalise communications and curate our wonderful learning materials. We host CORE, the world’s largest open-access articles collection, which welcomes 20 million unique monthly users, curates 34 million full-text research papers, and 260 million metadata records. The opportunity to foreground cutting-edge research within our learning materials, deploying the power of Generative AI to do so, is an irresistible proposition.
With such tools, reach and opportunities at our fingertips, we are conscious of the need to take the risks of AI deployment very seriously, and to counterbalance techno-exuberance with concern for digital inclusion, guard against potential for bias, protect our ambitions for financial and ecological sustainability, drive positive societal impact, and safeguard academic integrity. We are therefore balancing risk prevention with an enabling approach, optimism with care, and endeavouring to make decisions centred on what we should do – not what we could do. We see the need to enable experimentation and do not underestimate the importance of ensuring that colleagues and students, with diverse experiences and skills, are at the centre of our work. More than simply having ‘humans in the loop’, we proactively welcome opportunities to challenge, reflect and ask each other hard questions at all stages of development and implementation.
My own perspectives are centred on the optimistic notion that Generative AI can be a powerful tool for enabling learning, and that this learning can make a transformative difference to the lives of individuals, their communities and society at large. When I was learning to be a teacher (with pen, paper and books) I was encouraged to ‘start with the child’, and today our considered use of powerful technology in higher education must be equally focused on the needs, motivations and support requirements of each learner. Rather than persevering with an educational model that seeks to trammel and ‘police’ learning, let us commit to creating the means through which all of our students are enabled to demonstrate their knowledge, understanding and skills across lifelong learning journeys, as they ride their own successive technological waves.