University College London (UCL) is hosting a series of weekly webinars to hear from cutting edge research developing theories and methods for understanding values in AI research.

The series runs from Thursday 13^th November to 27^th November, and is organised as part of the ESRC Public Values in AI Research (PAIR) project. PAIR is exploring and explaining the potential of mapping emerging AI trajectories with a view to informing government funding and regulation. Using scientometrics and cutting-edge text analysis and complemented by qualitative methods, the project assesses the feasibility of strategies for understanding the values currently implicit in AI research and development and gaps with public values.

The project is funded by the UKRI Metascience programme.

WEBINAR 1, Thursday 13th Nov. 4-5.30pm GMT: William Agnew (Carnegie Mellon University): The Values Encoded in Machine Learning Research Chair: Cian O’Donovan (UCL) . This semiar will also include an introduction to PAIR from Jack Stilgoe, who regular readers will know from his fundamental work on responsible innovation.

WEBINAR 2, Thursday 20th Nov. 4-5.30pm GMT: Dror Markus (Universität Zürich): Societal AI Research Has Become Less Interdisciplinary Sergio Peláez Sierra (Inter-American Development Bank): Large-scale text analysis using generative language models: A case study in discovering public value expressions in AI patents Chair: Tommaso Ciarli (UNU-MERIT)

WEBINAR 3, Thursday 27th Nov. 4-5.30pm GMT. Shazeda Ahmed (UCLA): Field-building and the epistemic culture of AI safety Atoosa Kasirzadeh (Carnegie Mellon University): AI safety for everyone Chair: Noortje Marres (University of Warwick).

Register here to attend any or all of the three sessions.