GESDA Global 2023: ‘Shaping the future of AI’ panel
October 2023, Geneva: panel with Amandeep Gill (UN Tech Envoy), Pascale Fung (Hong Kong University of Science and Technology), Alessandro Curioni (IBM) and myself, chaired by Alexander Illic (ETH AI Centre).
National Academy of Sciences and Royal Society: session on the role of data institutions in data access
September 2023: session chaired by Sir Nigel Shadbolt, with Hyon Kim (Data.gov), Margaret Levenshtein (ICPSR), Meredith Goins (World data system) and Sylvie Delacroix
Keynote on data trusts at the Foresight Centre, Estonian Parliament
November 2022: Keynote on data trusts followed by Q&A.
Habitual Ethics: discussion panel organised by the Alan Turing Institute
October 2022: Panel chaired by Timandra Harkness, with Neil Lawrence , Emran Mian and David Leslie.
Dubai Future Foundation: ‘How much will your data be worth in the future?’
October 2022: Panel with Ashta Kapoor, organised by the Dubai Future Foundation at the Museum of the Future
Habitual Ethics?
October 2022: University of Birmingham inaugural lecture chaired by Prof Lisa Webley with reactions / comments from Prof. Heather Widdows and Prof. Chris Baber
Information Flows, Data Trusts, and Privacy Enhancing Technologies, Liberté! Talks
October 2022: Panel chaired by Alek Tarkowski in the context of the 'Freedom Games' 2022.
Data Trusts as bottom-up data empowerment infrastructure
July 2022 UoM Institute for Data Science & AI keynote
The inherent instability of top-down valuation methods: bottom-up data trusts and their political, economic and social potential
June 2022 presentation on the potential inherent in bottom-up data trusts in the context of the Institute for Mathematical and Statistical Innovation (University of Chicago) workshop on 'Data value: assessment and evolution'
Algorithmic Habits and Precluded Transformations
March 2022 invited talk at the Montreal Speaker Series in the Ethics of AI, based on the last chapter of my forthcoming Habitual Ethics?: https://www.bloomsbury.com/us/habitual-ethics-9781509920419/
Preserving the interpretability of automated systems over time
December 2021 COHUBICOL Interview, conducted by Laurence Diver.
Years of societal transformations can negatively impact the interpretability of some ML systems for two types of reasons. These two types of reasons are rooted in a truism: interpretability requires both an interpretable object and a subject capable of interpretation. Not enough attention has been paid to the 'subject side' of the interpretability challenge.
Full paper: https://ssrn.com/abstract=3728601
How can we unlock bottom-up empowerment structures?
July 2021 conversation about data trusts as bottom-up empowerment structures to create a better future which benefits all, no the few.