Richard is a Professor and head of the Artificial Intelligence research group in Computer Science at the University of York, he is also the AI Lead for SAINTs (UKRI AI Centre for Doctoral Training in Safe AI Systems).

What is your role in SAINTS, and what does it involve?
I am AI lead for SAINTS. This is a very general role, which consists of input and advice for any part of SAINTS which touches on artificial intelligence. For example, it involves input into project selection, recommendations for supervisors in the area of AI and advice to teams on AI technologies. I also oversee the two AI courses which are part of SAINTS.
Tell us about your research interests. What do you find most interesting or enjoyable about your work?
My main area of research is in machine learning. In particular I work on machine learning for graphs and networks (that is to say, data described by the relationships between things). Some common examples are social networks, protein interactions and chemical structure graphs. I also have interests in computer vision. Recently, I have been involved in projects on brain networks for Alzheimer’s disease, a vision system with mobile phones for vehicle insurance and driver safety, and the analysis of the lifecycle of cells in microscope images.
Like many academics, my job involves both teaching and research. On the research side, the most enjoyable part of the job is the discovery of something new, some new algorithm or a new understanding of a problem which nobody else knows. When teaching, it is a great pleasure to get involved in deep technical discussions of a challenging topic with engaged students.
What working achievement or initiative are you most proud of?
One of the highlights of my academic career so far was my involvement in an EU Framework 7 research grant. This project involved some brilliant academics from some of the top institutions in Europe, including ETH Zurich, University of Venice and IST Lisbon. The project was a tremendous success and a great experience in terms of visiting these institutions and learning from the scientists working there. The project studied the fundamental topic of similarity-based pattern recognition.
What’s next on the research horizon for you?
I have a number of projects in the early stages. Two are on fundamental algorithms for machine learning with graphs (one on generative models and the other on graph spectral methods for deep learning). I am also involved in collaborations looking at generating chemical compounds as potential drugs for disease treatments, and protecting sensitive data in machine learning, which is at the intersection of machine learning and cryptography.
Can you share some interesting work that you read about recently?
I have been recently reading some work on student-teacher networks, where the learning from a teacher network is distilled into another network, which is usually smaller and more efficient. I have a PhD student who has just started in this area. The key is to identify the most relevant concepts learnt by the teacher, and is in some ways related to the interpretation of deep learning networks.
What are your thoughts on the future of AI?
There was a recent interview with Geoffrey Hinton for Newsnight on the BBC; my thought are quite similar to those he gives in that interview. In the near to medium future, AI will be capable of human-level performance in most technical and creative jobs. This is likely to lead to massive disruption to society which is based economically and socially on the value of having a job. We may need to fundamentally rethink the way society values what people do.
What one piece of advice do you have for SAINTs postgraduate researchers?
Postgraduate research is difficult and challenging; it is important to be committed to your work and so you should make sure you research something you really care about.











