GenAI is here to stay. Now let’s get on with it!

Authors: Mike O’Dea, University of York and Xianghan O’Dea, Kings College London.

What has happened

Most educators are now very aware of the impact of Generative AI (GenAI) on education. Since the appearance of ChatGPT in November 2022, GenAI has become one of the fastest adopted technology in higher education globally. Unlike other technology, such as VLEs, lecture recording, and the flipped classroom, the introduction of GenAI  has been driven by the students. Consequently the technology has had a somewhat uncontrolled incorporation into learning and teaching practice. Leading educational organisations, such as QAA, JISC and Advance HE have organised advisory workshops, produced guidance literature and have led and/or supported pedagogical research projects on various topics relating to the use of GenAI in higher education. Alongside this, academic blog publishers, such as Wonkhe, have published articles regularly aiming to increase our knowledge and understanding of the rapidly changing technology landscape and to help us keep pace with some of the latest techniques and skills to enable us to adapt. 

We are now at a stage where most academic institutions have guidelines and principles in place or are at an advanced stage of doing so. So the time for talking and debating is ending. GenAI is like the shiny new car in academia’s driveway. It’s now time to stop reading the reviews and looking at the spec, we need to get in the car and drive it.

What we should do now

Is it just about getting on and using GenAI? In higher education, we what needs to be done, we need a concerted AI literacy programme for all, and it needs to include topics such as ethics, good AI academic practice, responsible use, verification and prompt engineering. This isn’t the full picture though.  

Firstly, to date, there has been one important voice missing from the debate, industry. We know that industry has been adopting and incorporating the technology on a much more rapid scale, compared with education. Even though businesses have similar issues and debates as with the education sector, industry adoption of GenAI continues apace alongside these debates and is now extremely widespread and further increasing. However, the higher education sector has been so focussed on addressing our own issues, it has not yet fully engaged with industry. As a result, we don’t fully know what industry needs from our graduates in terms of AI knowledge and skills, and hence cannot be confident that our graduates are well prepared for working with GenAI in their post graduation world. Universities need to work more closely with industries to figure out the most industry oriented AI skills. 

Secondly, universities need to understand how AI or GenAI literate students actually are, and their attitude towards the tools, at both the local and national levels. Many universities have conducted surveys locally, however, until now, there has only been one UK-wide survey (HEPI) exploring the use of GenAI among the UK university students. It is equally important to gather the data at the national level, because it helps provide a broader picture of trends and patterns across the entire HE sector. The HEPI survey results provide some valuable and insightful results, although the questions were related to assessments mainly, rather than AI and/or GenAI literacy and the size of the survey was fairly small  with 1,200 participants (undergraduate students), compared with the whole undergraduate population in the UK (1,734,805, 2024). 

Application of AI in learning and teaching

Until now, the barriers to the adoption and integration of GenAI in learning and teaching can be grouped into two main areas: knowledge and access. Knowledge refers to staff and student knowledge of the technology, the comprehensiveness and clarity of the guidelines for its use and the existence of a body of work demonstrating the use of it. Access refers to the availability of the software and equality of access for all.

With the national and institutional level guidance and policies in position, our attention perhaps should now be shifted more to focus on solutions, for example, developing practical and nuanced solutions and guidance at the programme and module level, following the more generic level policy. It may also be helpful to produce case studies, and examples for individual disciplines and to address different types of assessments. Interaction with industry may also prove fruitful in areas such as developing innovative assessments.

We understand access of GenAI is not equal. Many universities in the UK are exploring the opportunity of integrating Copilot (MS) and Gemini (Google) into their existing business platforms. OpenAI also offers an enterprise licence. Universities will soon need to consider whether to buy into these. It will certainly address the access and equality issue. Additionally, the private versions of the new software claim that an organisation’s data can be prevented from leaving its boundary, see Microsoft’s example here. This, if correct, should address some of the real concerns regarding IP and information and data security.

However, such an integration should not be solely the result of peer pressure. Universities need to be clear about the purpose, and staff and students need to be formally informed of the availability of the tool. More importantly, clear guidance, and support should be offered to help enhance teaching practices and learning experiences. There needs to be a clear and concerted effort to improve AI literacy. We should also consider how we apply and integrate GenAI into learning and teaching more effectively so that it not only addresses academic issues but also prepares our students for using it in industry.

To conclude, the higher education sector is now at the point where we have to make decisions as to how we move forward with AI and GenAI in particular. We know what we need to do and we have a pretty good idea of how we should do. Let’s get into that shiny new car and start really using it now.