professional, social, intellectual society of diverse leaders

The Boston Hub Society was delighted to host Professor Munther Dahleh of MIT for a lively discussion on his new book, Data, Systems, and Society: Harnessing AI for Societal Good.

The evening began Munther signing copied of his book for the members, and then with a warm introduction by Jinane Abounadi, BHS member and Executive Director of the MIT Sandbox Innovation Fund. Jinane recalled meeting Munther when she was a PhD student and he was a young faculty member—then added with a laugh that it’s not every day she gets to introduce her husband of 35 years. “At least this time,” she joked, “I get to brag about his professional accomplishments.”

Munther opened with the story of how this book came to be—born out of his experience launching MIT’s Institute for Data, Systems, and Society (IDSS). Writing for a general audience, he admitted, was its own experiment: translating deeply technical work into language and examples accessible to anyone curious about the forces shaping the data-driven world.

To ground the discussion, Munther traced the history of artificial intelligence, placing today’s rapid developments in context—AI was born in the 1950s not Nov 2022 . He then explored the full AI cycle, emphasizing that data is the world’s most valuable commodity—and, perhaps, its least properly valued. He unpacked the complex interplay between data, algorithms, decisions, and feedback loops—revealing how these ingredients can amplify biases of all kinds: data, algorithmic, confirmation, and selection.

Through vivid examples, he illustrated how agent-based decision systems can introduce their own unique errors—sometimes surpassing human ones. The audience was quick to engage, diving into questions about AI-driven investment, fraud, misinformation, and the growing influence of generative AI. Munther demystified these technologies, explaining how they can—and must—be guided to serve societal goals rather than undermine them.

He contrasted reliable agentic systems like self-driving cars, which steadily improve through feedback, with generative language models, which remain less predictable and prone to inaccuracy.

The conversation extended well beyond the formal Q&A—proof of an evening that sparked both curiosity and conversation. The Boston Hub Society was thrilled with the turnout, the energy, and the thoughtful exchange that followed—a true meeting of data, systems, and society.