In a congressional hearing about disinformation on Thursday, Representative Debbie Dingell, a Michigan Democrat, asked Facebook CEO Mark Zuckerberg to respond to a claim he once made about his own company: that the more likely content posted to Facebook is to violate the company’s community standards, the more engagement it will receive.
Is this, she asked, still accurate?
Dingell cited a recent investigative report by MIT Technology Review’s Karen Hao based on interviews with former and current members of the company’s AI team. The story dove into how AI models that drive Facebook’s recommendation algorithms allow misinformation and abuse to continue to thrive on the site. As Hao wrote, and Dingell paraphrased, “A former Facebook AI researcher who joined in 2018 says he and his team conducted ‘study after study’ confirming the same basic idea: models that maximize engagement increase polarization.”