Artificial Intelligence in Education: Promises and Pitfalls

Focused Discussion
Gerard Houlihan  

In recent years, no educator could have remained immune from the wide-ranging pronouncements in the media about the looming impact of Artificial Intelligence (AI). Whilst these have typically been delivered with gusto, there has been scant detail as to what this will look like in schools, leaving most educators anxious and devoid of the necessary information to make informed decisions around AI. And whilst most schools already access commonplace applications, such as with mathematical and linguistic tutoring, the question remains as to what are the broader implications for more nuanced applications of AI as with student wellbeing? Drawing on Gerard’s work in the Oxford Artificial Intelligence Programme and at St Michael’s, this session will carefully unpacks these questions, first starting with the history of AI in context of parallel developments in education. Gerard will explore some of the ways in which machines learn, identifying those areas with most potential for schools. A range of case studies will show how St Michael’s uses data visualisation software to create student ‘dashboards’ and the use of algorithms as ‘intelligent agents’ to identify issues and trigger interventions. We look at how recent work in neural networks mimics the human brain, and provides inspiration for our understanding of students think and learn. We explore the impact of AI in the future of work. Lastly, we will consider the ethical and legal conundrums of AI as it applies in an international context to the education sector, focusing on replication, privacy, and inbuilt decision-making biases.