📣New today! In "Democratizing AI," @profmgilman offers evidence-based recommendations for better structuring public participation processes for AI, and underscores the urgency of enacting them. 1/5 datasociety.net/library/democr…

@profmgilman Public participation is difficult to do well, and its mechanisms can backfire if they’re not carefully designed. Luckily, as @profmgilman shows, policymakers don't need to design public participation for AI from scratch. 2/5

@profmgilman .@profmgilman's brief builds on a comprehensive review of evidence from public participation efforts in policy domains like anti-poverty programs and environmental policy to offer practical advice for incorporating meaningful public participation into AI. 3/5

@profmgilman There is no one-size-fits-all model, & embedding public participation in AI governance is no panacea for keeping AI safe. But particularly in areas where the tech's use holds the greatest risks to civil & human rights, well-designed public participation can make a difference. 4/5

@profmgilman We hope these guidelines can promote opportunities for the public to truly affect and improve the technologies shaping our lives. 5/5 datasociety.net/library/democr…