Themes for the Call
Policy Considerations: As the pace of development of AI technologies continues to accelerate, there is considerable debate as to the effects and potential responses across sectors and policy domains. CIFAR seeks workshops that examine a diverse range of policy considerations related to AI systems, such as urban planning, transportation, education, work, and law.
Global Dynamics of AI: AI technologies are increasingly being developed and deployed across borders and jurisdictions, despite significant regional differences. CIFAR seeks workshops that will explore questions about AI on a global scale, and as it relates to globalization, trade, development, international relations, culture, competition, inequality, and governance.
Discovery and Creation: While prediction and analysis are the primary functions of most AI systems, some AI technologies are creating new works of art, music, science, and journalism. CIFAR seeks workshops that will examine questions related to the nature of intelligence, creativity, authorship, art, bias, accountability, transparency, and scholarship.
Summary of Approved Proposals
Indigenous protocol and AI - Two workshops in March and May 2019
How do we broaden discussions about the role of technology in society, and how can these conversations be informed by Indigenous knowledge and culture? These workshops will bring Indigenous ways of making and maintaining relationships with human and non-human kin to bear on the question of how to build ethical relationships with AI. Workshop participants will come from a variety of disciplinary backgrounds, including machine learning, design, symbolic systems, cognition and computation, visual/performing arts, philosophy, linguistics, anthropology, and sociology.
Proposal Team: Jason Edward Lewis, Concordia University, Canada; Angie Abdilla, Old Ways. New. Indigenous Knowledge Consulting, Australia; ʻŌiwi Parker Jones, Oxford University, United Kingdom; Fox Harrell, MIT, United States
Generation AI: Reducing inequality and enhancing digital inclusion via smart design and developmental science - Two Workshops beginning in May 2019
Though children comprise one-third of the world’s internet users, online algorithms have not been constructed to consider this especially vulnerable population, particularly those exposed to high levels of poverty and inequality. These workshops will bring together an interdisciplinary team from machine learning, pediatrics, psychology, and cultural anthropology as well as proposed collaborators in educational technology, communication sciences, and computational biology. They will focus on AI as a potential amplifier of inequality and inform a research, policy, and communication agenda that will bring these issues to the forefront of scientific, policy, legal, and public discussions.
Proposal Team: Candice Odgers, University of California Irvine, United States; Anna Goldenberg, University of Toronto, Canada; Ronald Dahl, University of California Berkeley, United States; Mizuko Ito, University of California Irvine, United States
AI-powered information ecosystems and democracy - One workshop in June 2019
The potential risks AI poses to democratic institutions are not well-understood, in part due to the tension between potential benefits and harms to society. This workshop will engage diverse groups of researchers from academia and industry with practitioners and civil society representatives, to encourage collaborations between those developing tools and those who with expertise in policy, civil rights, and democratic values.
Proposal Team: Derek A. S. Ruths, McGill University, Canada; Seda Guerses, University of Leuven, Belgium; Alexandra Olteanu, Microsoft Research, United States; Joris Hoboken, Vrije Universiteit Brussels and University of Amsterdam, The Netherlands
Regulation of defense and security AI technologies: Options beyond traditional arms control - Two workshops beginning in July 2019
Debates about AI arms control have focused predominantly on autonomous weapons systems, been aimed principally to determine whether a total ban is appropriate, and taken place in the UN Convention on Certain Conventional Weapons, a traditional arms control setting. The workshops will bring together AI technologists, arms control practitioners and researchers, and other experts in regulatory and control regimes to develop a better understanding of arms control in the age of AI.
Proposal Team: Kerstin Vignard, UN Institute for Disarmament Research (UNIDIR), Switzerland; David Danks, Carnegie Mellon University, United States; Amb (ret.) Paul Meyer, Simon Fraser University, Canada