UoA ML Seminar: Alistair Knott – Oversight and Regulation of AI Systems in Social Media

Short Bio
Alistair Knott is Professor of Artificial Intelligence at Victoria University of Wellington. He has been an AI and computational linguistics researcher for 30 years. He studied Philosophy and Psychology at Oxford University, then obtained MSc and PhD degrees in AI at the University of Edinburgh. He then moved to New Zealand, working first at Otago University and now at Victoria University of Wellington. Ali’s AI research is in computational modelling of cognition, most recently with the New-Zealand-founded AI company Soul Machines, where a longstanding project is the development of a model of embodied baby cognition, BabyX. Currently, Ali’s work mostly focusses on the social impacts of AI, and on AI regulation. He co-founded Otago University’s Centre for AI and Public Policy, where he worked on Government uses of AI, and on the impact of AI on jobs and work. He now works on social media governance at the Global Partnership on AI. Ali has also contributed to the Christchurch Call’s Algorithms Workstream, the Global Internet Forum to Counter Terrorism, and the Forum for Information and Democracy.

Abstract
Social media platforms are one of the main gateways through which AI technologies reach the global public. 59% of people in the world were social media users in 2023; of those people, the average user spent over 2.5 hours on social media per day. Social media platforms run on AI: recommender systems are responsible for pushing content to users, and harmful content classifiers are responsible for moderating content that’s deemed unacceptable. These AI tools are having profound impacts on the way information flows in the world. But we know far too little about these impacts. We don’t know enough about how they are trained, or tested. And we don’t know enough about these systems affect platform users, as individuals or as collectives. A new uncertainty is how AI-generated content diffuses through social media platforms.
Studying how AI systems impact citizens through social media is best done in international projects: the technologies in question are deployed by large multinational companies, and many aspects of regulation are international in scope. I co-lead a project on social media governance for the Global Partnership on AI – an international grouping of AI researchers. In this talk, I’ll outline three topics we are working on. Regarding recommender systems, our main focus is on how we can enable access for external researchers to the methods companies use themselves to study the effects of recommender systems on platform users. Regarding harmful content classifiers, our main focus is piloting an idea about how the training of these systems can be moved outside companies, into a more public and accountable domain. Regarding AI-generated content, our main focus is on how to build detection tools that can reliably identify such content. This work has involved advocacy with EU and US policymakers, and pilot studies in India.