Getting Aligned On AI Alignment
October 31 - November 1, 2024
Embedded Files
Overview
As AI becomes increasingly capable and autonomous, it is of paramount importance that their goals, behaviors, and decisions align with human values and societal norms to ensure their safe and ethical deployment. This workshop will bring together experts from academia, industry, government, and nonprofit sectors to explore the latest developments, and open challenges in AI Alignment.
Through a series of presentations, panel discussions, and collaborative sessions, participants will engage in deep-dives into the technical, ethical, and philosophical aspects of aligning AI with human values. Key topics will include robustness and safety in AI systems, value alignment, interpretability, fairness, security and privacy. The workshop aims to foster interdisciplinary collaboration for advancing AI alignment, so that AI can truly bring about social good.
Invited External Speakers
Invited External Speakers
Invited Internal Speakers
Invited Internal Speakers
University of Maryland
University of Maryland
University of Maryland
University of Maryland
University of Maryland
University of Maryland
Workshop Schedule
Workshop Schedule
Thursday, October 31, 2024
9:30am - Breakfast
10:00am - Plenary Session: When do generative models memorize copyrighted data, and what can we do about it? -- Tom Goldstein
10:25am - Plenary Session: Centering Community Perspectives in AI Evaluation -- Remi Denton
10:50am - Plenary Session: AI's Certain Uncertainty Problem --Jordan Boyd-Graber
11:15am - Break
11:35am - Lightning Talk: World Alignment via Contrastive Imitation and Rule Learning -- Tianyi Zhou
11:55am - Lightning Talk: Auditing and Mitigating Data Bias in COVID-19 Case Forecast Models -- Vanessa Frías-Martínez
12:15pm - Lunch Break
1:30pm - Invited Talk: Concrete Problems in AI Deception: From Evaluation Gaming to Cyber Attack -- Ruiqi Zhong
2:00pm - Invited Talk: AI-assisted Safe Deployment of AIs -- Shi Feng
2:30pm - Lightning Talk: Why we need AI alignment: A case study of reconstructing the historical memory of Colombia -- Margret Vilborg Bjarnadottir
2:50pm - Lightning Talk: A.I. as Educator? Foregrounding Ethics and Authenticity in the Classroom -- Campbell Frank Scribner
3:10pm - Lightning Talk: Aligning AI with Teams -- Susannah B.F. Paletz
3:30pm - Break
4:00pm - Breakout Session
5:00pm - Adjourn
Friday, November 1, 2024
9:30am - Breakfast
10:00am - Plenary Session: What exactly is alignment? A big-picture view of the challenges and opportunities of AI alignment -- David Jurgens
10:25am - Plenary Session: On AI Alignment in the Provision of Social Services: Opportunities and Challenges -- Sanmay Das
10:50am - Plenary Session: Aligning AI on the Fly -- Furong Huang
11:15am - Break
11:35am - Panel: Remi Denton, Furong Huang, Ruiqi Zhong, Sanmay Das, David Jurgens (Moderator: Sanghamitra Dutta)
12:30pm - Lunch
Page updated
Report abuse