General Profile:

After completing an MSc in Physics, I spent three years working for the NHS as a trainee clinical scientist. In this role, I had the opportunity to work alongside healthcare professionals from a variety of specialisms, and to work with different medical technologies and software tools to support hospital service delivery and local research. Through this work, I became particularly interested in machine learning and artificial intelligence technologies; how they could be applied and distributed to help solve real-world problems, and the legal and ethical challenges surrounding their adoption and use. My experiences working in healthcare have motivated me to pursue research investigating how the development of AI systems can be informed by human interaction and expert guidance.

Research Project Summary:

Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) have led to the development of systems which can match and even surpass human performance on certain tasks. As digital technologies become pervasive, there is an incentive to develop systems that can work alongside people to support individual and social needs in day-to-day life. However, state-of-the-art AI systems still struggle to adapt to unfamiliar tasks, and to learn by observing or interacting with humans and other AI agents. Recent work at the intersection of Affective Computing and Multi-Agent Systems has highlighted how these issues may be addressed by incorporating models of “social intelligence” into AI systems.

In humans, social intelligence allows us to reason intuitively about ourselves and others, with an awareness of how our decision-making may be influenced by various psycho-social factors, such as personality and affective state (transient emotions and longer-term mood states). This awareness plays an important role in our unique ability to cooperate, collaborate and learn from one another to solve complex problems.  

In my research, I am interested in investigating how behavioural cues from human interaction (with other humans or autonomous agents) can be used to better understand the underlying principles of human decision-making, and how AI systems can be trained to exploit this information to reason about humans and other agents in their environment. My aim is to develop and evaluate techniques for modelling social behaviour through interactive agent-based simulation in “mixed-group” game environments, designed to support player populations of artificial agents, humans, or a combination of both. These environments provide a useful experimental testbed for empirical study of human behaviour, as well as for testing of novel agent models of social intelligence, within a consistent task setting.