( events)   Timezone: »  
Tue Dec 14 04:00 AM -- 01:45 PM (PST)
Political Economy of Reinforcement Learning Systems (PERLS)
Thomas Gilbert · Stuart J Russell · Tom O Zick · Aaron Snoswell · Michael Dennis

Sponsored by the Center for Human-Compatible AI at UC Berkeley, and with support from the Simons Institute and the Center for Long-Term Cybersecurity, we are convening a cross-disciplinary group of researchers to examine the near-term policy concerns of Reinforcement Learning (RL). RL is a rapidly growing branch of AI research, with the capacity to learn to exploit our dynamic behavior in real time. From YouTube’s recommendation algorithm to post-surgery opioid prescriptions, RL algorithms are poised to permeate our daily lives. The ability of the RL system to tease out behavioral responses, and the human experimentation inherent to its learning, motivate a range of crucial policy questions about RL’s societal implications that are distinct from those addressed in the literature on other branches of Machine Learning (ML).

Pre-show meet and greet (Gather town session)
Welcome (Brief introduction)
Culturing PERLS (Plenary presentation)
Audience Q+A for plenary presentation (Live Q+A)
5 minute break (Break)
V&S | Theme and speaker introductions (Brief introduction)
V&S | RL Fictions (Presentation)
V&S | Assumptions of Making Things Computable (Presentation)
V&S | Panel discussion (Live panel discussion)
10 minute break (Break)
LAF | Theme and speaker introductions (Brief introduction)
LAF | "Legitimacy" in the Computational Elicitation of Preferences in Mechanism Design (Short presentation)
LAF | The Role of Explanation in RL Legitimacy, Accountability, and Feedback (Short presentation)
LAF | Evaluating Reinforcement Learners (Short presentation)
LAF | Panel discussion (Live panel discussion)
45 minute lunch break (Break)
Poster session for accepted papers (Gather town session)
5 minute break (Break)
TD | Theme and speaker introductions (Brief introduction)
TD | Antimonopoly as a Tool for Democratization? (Short presentation)
TD | Reinforcement of What? Shaping the Digitization of Judgement by Reinforcement Learning (Short presentation)
TD | Metrics are Tricky (Short presentation)
TD | Panel Discussion (Live panel discussion)
Closing remarks (Brief conclusion)
Deciding What's Fair: Challenges of Applying Reinforcement Learning in Online Marketplaces (Poster)
Robust Algorithmic Collusion (Poster)
Power and Accountability in RL-driven Environmental Policy (Poster)
Calculus of Consent via MARL: Legitimating the Collaborative Governance Supplying Public Goods (Poster)
Demanding and Designing Aligned Cognitive Architectures (Poster)