Emotional AI: Human Emotion and the Impact of AI Technologies
The Emotional AI Lab is an international research group that examines the ethical, societal and cultural impact of artificial intelligence technologies in relation to data about human emotion, moods and affective states. The lab is led by Prof Andrew McStay at Bangor University and Dr Lachlan Urquhart, Lecturer in Technology Law at the University of Edinburgh, is an investigator in the research group.
‘Emotional AI’ refers to technologies that use affective computing and artificial intelligence techniques to sense, learn about and interact with human emotional life. Techniques to try and sense and discerns people’s emotions and expressions include analysis of online sentiment, voice analytics, eye-tracking, and the use of virtual reality and augmented reality.
While there is much value in studying Emotional Al, in that it promises better experiences of services, devices and technologies, there are also wider considerations, such as: Is it desirable that emotions are machine-readable? What racial bias is there in computer vision and training data? Are protections adequate? Is tracking ethical, such as at-home emotion tracking?
With work funded by the UK’s Economic and Social Research Council (ESRC) and Japanese Science and Technology Agency, the group seeks to generate conversation among UK and Japanese academics, industry, artists, NGOs and regulators about these technologies. Most recently, this has been in two major projects:
ESRC AI and Society Call funded Cross-Cultural Conversation on Emotional AI: Japan and UK. This project involved the team running comparative multi-stakeholder workshops in Tokyo and London. This led to a detailed report (in English and Japanese) mapping out areas of crossover, divergence and emerging themes for further research in Japan and the UK.
UKRI-JST funded Living well with Emotional AI in Smart Cities. This project began in Jan 2020 with investigators at Edinburgh, Bangor, Northumbria, Asia Ritsumeikan Pacific, Chuo and Meiji Universities. It is a 3 year programme of work to understand a wider range of issues around emergence of emotion sensing infrastructure in urban life. This will range from impacts on civic life and commerce to use in policing and security, and the various governance and design challenges this can raise. The team are conducting a programme of empirical research with industry, the public, police and wider stakeholders. We are keen to speak to anyone working on emotional AI or smart cities. If you are interested, please get in touch with Lachlan.firstname.lastname@example.org