WASHINGTON—The Supreme Court began recording its oral arguments in 1955. Unfortunately, the court’s most celebrated case—Brown v. Board of Education—was decided in 1954, meaning there is no audio record of the hearing where the constitutionality of school segregation was argued, or the day 70 years ago when Chief Justice Earl Warren read from the bench his entire opinion declaring that, “in the field of public education, the doctrine of ‘separate but equal’ has no place." That historic near-miss long has frustrated Jerry Goldman, a Chicago law professor who spent decades cataloging, transcribing and publishing Supreme Court audio files on the educational website Oyez.org.
But at the theater last year, Goldman found an unlikely muse: a computer-synthesized clone of President Biden’s voice commanding patrons to silence their cellphones. Inspired, Goldman set out to use similar artificial intelligence technology to replicate the voices of Warren, lawyer Thurgood Marshall, who was himself appointed to the court in 1967, and other key figures from the case. Now, coincident with Brown’s 70th anniversary, synthetic audio of the arguments and opinion announcement joins more than 12,000 hours of flesh and blood recordings.
“You’ve heard of deepfakes. This is a truefake," Goldman says. “When you listen to Frankfurter, man—it’s Frankfurter!" adds the self-described Supreme Court nerd.
Read more on livemint.com