Rebooting Democracy in the Age of AI

What can the World Cup's controversial VAR replay system teach us about deploying AI in government? In a new Reboot Democracy piece, Anirudh Dinesh argues that VAR's failures, such as addressing the wrong problem, ignoring stakeholder input, and eroding the human experience, are the same failures civic AI keeps making. Sarah Sachs and Tal Roded describe how the Block Party team turned 20 years of Manhattan Community Board 3 resolutions into a searchable public archive: democratic AI in practice. And in two new conversations, Beth Simone Noveck argues that universities should become laboratories for democratic problem-solving, and that AI should be judged not by efficiency alone but by whether it improves lives, expands participation, and helps institutions solve public problems.

Local government generates enormous amounts of public knowledge, but much of it remains buried in disconnected PDFs and difficult-to-navigate archives. This post explores how the Block Party team used AI, semantic search, and human expertise to build a searchable archive of 17 years of Manhattan Community Board 3 resolutions, making local government decisions, institutional memory, and civic participation more accessible to residents, journalists, and public officials.

As AI transforms how we work, learn, and govern, what role should universities play? In this wide-ranging conversation, Beth Noveck argues that universities should become laboratories for democratic problem-solving, where students work with communities to address public challenges. She outlines a vision for "democratic AI" that puts public purpose, participation, and impact at the center of technological innovation.