ART-AI Seminar
We are pleased to have Séb Krier, who is a Policy Development & Strategy Manager at Google DeepMind, join us for this ART-AI seminar entitled ‘AI in the next 5 years: what to expect’.
This seminar will take place online on Zoom on Tuesday 14th May 2024, 12.15pm-1.15pm (BST). Jack McKinlay will be chairing. For joining instructions, please e-mail [email protected].
Title
AI in the next 5 years: what to expect
Abstract
The rapid evolution of artificial intelligence capabilities in recent years presents the dual challenge of harnessing truly transformative opportunities for individuals and society alongside mitigating potential novel and extreme risks. AI’s growth has elicited diverse global regulatory responses, ranging from countries adopting risk-averse approaches to those focused on pro-innovation interventions. This rate of change is ever increasing and, over the next five years, we may see models with stronger multimodality, agency, and increased reasoning, planning, and memory capabilities. This underscores the importance of advanced AI labs’ safety, testing and evaluation work, as well as research into future risks and impacts. Several significant policy questions largely remain under-appreciated. These include how to bolster AI’s defensive capabilities against offensive, harmful threats; how to effectively delegate activities to AI agents, which may be capable of pursuing complex goals with limited supervision, while limiting the risks they present; and whether these technologies could accelerate growth through automating parts of the labour market. In this talk I will explore the capabilities we can expect from AI over the next few years, and address these important policy questions including outlining the work of Google DeepMind to further develop our understanding.
Bio
Séb Krier is currently a Policy Development & Strategy Manager at Google DeepMind, specializing in the intersection of research and strategy to shape Google DeepMind’s policies on frontier AI governance issues. Before joining DeepMind, he served as a Senior Technology Policy Researcher at Stanford University’s Cyber Policy Centre. Earlier, Séb advised various governments and organizations on AI policy and governance. His involvement in AI policy started at the UK Government’s Office for Artificial Intelligence in 2018. In the earlier stages of his career, Séb practiced law at international law firms, including Bryan Cave Leighton Paisner and Freshfields Bruckhaus Deringer. He holds a Masters in Public Administration from UCL and a Law degree from King’s College London.
