Fireside Chat: Societal Impact of AI Agents
Fireside Chat: Societal Impact of AI Agents
AI agents are rapidly moving from experimental tools to active participants in how we work, build, and make decisions. As agents begin to operate with greater autonomy and coordination, what does this shift actually mean for society?
In this fireside chat, William Zhu、Weiyan Shi and Raphael Shu will share what they are currently working on and reflect on the broader societal impact of AI agents — from real-world applications to emerging questions around responsibility, coordination, and long-term change.
Event Details
📅 Date: February 11, 2026
🕒 Time: 12:30 PM PT
📍 Location: Online (Zoom)
Join Zoom Meeting:
https://us05web.zoom.us/j/81553565068?pwd=bSCheXTfMRu93rEn9DMJ8v6RBwa7JK.1
Speakers
William Zhu
Senior Data Scientist; Founder,AI Discussion Club
Weiyan Shi
Assistant Professor, Northeastern University
Raphael Shu
Founder, OpenAgents
Agenda
12:30 – 12:35 | Opening & Session Framing (5 minutes)
12:35 – 12:50 | Speaker Introductions & Current Work
William Zhu: current work and focus
Weiyan Shi: current work and focus
Raphael Shu: current work and focus
12:50 – 13:50 | Fireside Chat (60 minutes)
A moderated conversation among the speakers on the societal impact of AI agents
13:50 – 14:20 | Open Community Discussion & Q&A (30 minutes)
An open, interactive discussion with the audience
Who Should Attend
Builders, researchers, and practitioners interested in AI agents, multi-agent systems, and their implications for society, work, and governance.
Speaker Bios
William Zhu
https://william-wei-zhu.github.io/
William Zhu is a Senior Data Scientist focused on helping organizations strategically apply AI to better align with their purpose. He holds a Master’s degree in Computational Social Science from the University of Chicago, where he studied competitive strategy, marketing strategy, and consumer behavior at Chicago Booth, alongside machine learning and generative AI applications.
Beyond his professional work, William is deeply interested in how technology shapes human behavior, organizations, and collective decision-making. He actively hosts community conversations through bi-monthly TED Talk discussion salons and a monthly AI discussion club in Washington, D.C., bringing people together to reflect on learning, strategy, and the broader societal impact of AI.
Weiyan Shi
https://wyshi.github.io/index.html
Weiyan Shi is an Assistant Professor at Northeastern University, jointly appointed by the ECE and Khoury Computer Science departments. Her research spans Natural Language Processing, Human-AI Interaction, and AI Safety, with a focus on developing trustworthy and socially beneficial intelligent systems.
Her key projects include mitigating LLM mode collapse via Verbalized Sampling, strategic dialogue systems for diplomacy and safety, and investigating how LLMs encode harmfulness. She also explores AI's societal impact and privacy-preserving methods.
She earned her Ph.D. in Computer Science from Columbia University in 2023, following a postdoc at Stanford NLP. Her work has been recognized with honors such as the AI2050 Early Career Fellowship, MIT Technology Review's Innovators Under 35 (2024), and multiple Outstanding Paper awards at ACL. She actively serves as an Area Chair for top-tier conferences (ICLR, ACL, EMNLP, NAACL) and organizes workshops on multi-agent interaction and AI safety.
Recent highlights include her paper on Verbalized Sampling being featured by Forbes, grants for AI safety research, and organizing the inaugural Workshop on Multi-Turn Interactions in LLMs at NeurIPS 2025.
Raphael Shu, PhD
https://www.linkedin.com/in/raphael-shu/
Raphael Shu is the founder of OpenAgents and a leading researcher in multi-agent systems and AI agent collaboration. He earned his PhD in Computer Science from the University of Tokyo in 2020, receiving the Dean’s Award, and was a visiting researcher at NYU under the supervision of Yann LeCun.
After joining the AWS AI Lab, Raphael focused on generative models and agent systems. In 2022, he led the Dialog2API project, enabling LLM agents to understand conversations and invoke external APIs. From 2023 onward, he drove the development of multi-agent capabilities for Amazon Titan, and later served as Senior Science & Engineering Lead for Amazon Bedrock Agents, where his team launched the first commercial-grade multi-agent collaboration platform in 2025.
In 2025, he founded Acenta AI and released OpenAgents as an open-source initiative to advance agent interoperability and collaboration. He is also an active contributor to the research community, founding the WMAC (Workshop on Multi-Agent Collaboration) series at AAAI and co-organizing the COLM Workshop on AI Agents.