The problem with technology in education is the difficulty of working out how to integrate it successfully. This is complicated because the challenge varies according to the task, the setting, and many other factors. Using AI to help with lesson planning or creating assessments presents different challenges to using it to mark assessments or interpret assessment results. This, in turn, is very different from using AI to support teaching and learning. The only common factor is that, in each case, it is the teachers and leaders of schools and colleges who are best placed to work out how the technology can most usefully be deployed.
This is what makes AI in Education exciting. This is not a project being driven by technology companies with products to sell; it is not a research project driven by academics with theories to test; it is a project initiated by and led by teachers who want to work out how they can best use the astonishing capabilities of AI to support their pupils.
Not all teachers have responded positively to the disorienting speed with which AI is developing. The arrival of large language models such as ChatGPT created a good deal of alarm. ‘How do we stop children cheating?’ was the first question many asked. In New York, schools decided to ban it - an odd reaction to the invention of a machine that provides access to most of human knowledge.
But such guardedness is understandable. There are many problems with the way AI works. Reliability is a major issue. You do not have to spend much time with ChatGPT before you discover that it is not only extremely knowledgeable and thoroughly plausible, it is also very prone to make things up. In most situations, this is a major drawback. But not in all situations…
For example, if you were looking for a way to develop a student's critical thinking skills - if, say, you needed or a way to assess how they might cope with a clever but dishonest interlocutor - in that situation, ChatGPT might be your ideal tool.
As with all technology, the question we need to ask is not whether it is good or bad - the question we need to ask is how we can use it in a way that is beneficial. That is the question that AI in Education seeks to answer.
Since AI in Education was launched, there has been a huge amount of interest and enthusiasm from those working in AI and education. The depth of knowledge and experience on the Technical Advisory Panel is testament to that. We approach the task with a sense that education will be transformed by the capabilities of AI and with an awareness that, although the potential for benefit is huge, a positive outcome is far from certain.
Achieving a good outcome will require collaboration between people working in technology and those working in education. It will require iteration, and experimentation, with much trial and the occasional unavoidable error. And it will take time to learn what works and what does not - something that feels in very short supply as the speed of technological change accelerates exponentially.
Education could be described as the process by which one generation passes on to the next its best understanding as to what the hell is going on. You could think of it as a global hand-over note with: ‘that’s as far as we’ve got’ scribbled on the bottom. It is a process that has been remarkably consistent over centuries even through huge technological upheavals. But there are good reasons to think that much will change in the face of AI. Education will need to adapt because AI offers new ways to communicate and learn; because it opens up new ways to understand ourselves and our universe; and because it presents new challenges to what we mean by human understanding. To be able to work with so many teachers and educational institutions that are engaged in these vital issues at such a remarkable moment in history is enormously exciting for me and the members of the Technical Advisory Panel.
