In our morning tutorial, we’ll introduce participants to the basic methods used by modern
machine learning researchers (supervised, self-supervised and unsupervised learning and
reinforcement learning); the framework used to simulate and analyze multi-agent systems
composed of artificial agents; the formal tests used by machine learning researchers to
evaluate algorithmic bias and develop explanations; and the techniques used to construct
the recommender systems that shape social media, online commerce, and more.
2:00 pm - 4:00
AI Tutorial 2
In our afternoon tutorial, we’ll discuss research in machine learning that invokes
institutions and organizations that are well known to the SIOE audience. We’ll discuss the
insights available for AI from incomplete contracting theory, how contracts might be
deployed in multiagent settings, what we learn about the tragedy of the commons, norms and
culture from multiagent simulations, how algorithmic bias and explainability challenges
might impact organizational design and regulation, and how democratic processes might be
devised to improve the alignment of algorithms and recommender systems with social
welfare.
4:00-5:00 pm
Registration
Location: Atrium
5:00 pm - 6:15
Keynote #1 - Room J250
(in-person) Siwan Anderson (UBC): "Unbundling Female Autonomy"
(in-person) Gillian K. Hadfield, “Judging Facts, Judging Norms: Training Machine Learning Models to Judge Humans Requires a New Approach to Labeling Data”
(in-person) Joel Leibo, “Deep reinforcement learning models the emergent dynamics of human cooperation”
(in-person) Marzyeh Ghassemi, “Just following AI orders: When unbiased people are influenced by biased AI”
10:15 am - 10:30
Break
10:30 am - 12:30
Parallel sessions #4
12:30 pm - 1:30
Lunch
1:30 pm - 3:00
Parallel sessions #5
3:00 pm - 3:15
Break
3:15 pm - 5:15
Parallel sessions #6
5:30 pm - 6:30
Keynote #3: Presidential address - Room J250
(in-person) Robert Gibbons (MIT): "Visible Hands Building Equilibrium: Naming (and Framing?)"