As the action space range and the optimization power of autonomous agents grow over the years, researchers can foresee specific new challenges in making the future of AI what designers, customers, and society would like it to be. Topics covered include the types of emergent issues, technical research threads aimed at preventing or mitigating these issues (e.g. increasing contextual awareness, avoiding reward-gaming, safe exploration, establishing scalable control, computational deference, statistical-behavioral trust establishment, and many others), and the ethical, economic, and societal challenges that these systems will pose and what is being done about those.
Speaker: Richard Mallah, Future of Life Institute
Contact:Website: Click to Visit
Save this Event:iCalendar
Windows Live Calendar