Welcome to the SFU AI and Society Reading Group! We meet regularly to discuss research papers related to the societal implications of AI, AI ethics, and everything in these spheres. Our group is open to folks at SFU from any background.
2026 meeting details: We meet Wednesdays at 11:30 AM Pacific, in person in TASC 9204 unless otherwise noted.
Please note that next week’s session (Wednesday, August 5 at 11:30 AM) will move to TASC1 9408. We don’t have a set topic, so it will be a session for networking and informal conversation.
Human interactions form one of the most important complex dynamic systems we know. This meeting introduces how social scientists model those interactions and explores how AI tools—including large language models, reinforcement learning, and causal graphs—could make social science more powerful and accessible.
Two paper previews about the goals and values shaping AI research. The presenters invited critical feedback on the first paper, which was under review at the time of the meeting.
Calls on the AI technical community to diversify its collective imaginations of AI beyond replacing humans. The paper identifies limits in the mainstream imagination of AI and demonstrates an initial method for constructing alternatives in two AI tasks.
Audits highly cited NeurIPS and ICML papers to identify the mainstream values guiding AI development and examines how those values can support the concentration of power.
Wagstaff’s 2012 ICML keynote paper argues that machine-learning research has lost touch with consequential problems in science and society. It highlights limitations in commonly used datasets and evaluation metrics, weak communication with originating domains, and the need to conduct research for greater real-world impact.
Group planning and an open discussion of current events in AI and society.
Focused paper discussion
Focused paper discussion
Discuss jailbreaking benchmark paper and the OpenAI Model Spec
Discuss recent events: Paris Summit, new models, reasoning
Focused paper discussion
Focused paper discussion
Focused paper discussion
Discuss papers related to training data (copyright, consent, etc.) and other legal aspects of generative AI
Focused paper discussion
An initial get together to share research interests and goals for the group
To join the mailing list or ask a question, contact:
Nick Vincent
Email: nvincent@sfu.ca