Jude Joffe-Block for NPR:
One of the things being litigated in this presidential campaign is whether the crowds at rallies are even real.
At a Detroit aircraft hangar last week, Democratic presidential nominee Vice President Kamala Harris and her running mate, Minnesota Gov. Tim Walz, stepped off Air Force Two greeted by thousands of supporters. NPR’s Tamara Keith was there to see it.
There were 15,000 people at the rally, according to the Harris campaign. Photos and videos by attendees and media organizations captured the crowd from many angles.
But former President Donald Trump and his supporters have falsely claimed the crowd seen in a photo of the rally in front of Harris’ plane was a product of generative artificial intelligence. On Sunday, Trump made the nonsensical claim that the very real crowd at the event was a fabrication….
The refusal to accept basic, verifiable facts has some observers concerned about a repeat of 2020 false claims of a stolen election if former President Donald Trump loses.
Scholars who study deepfakes have pointed out that the existence of the technology means people can try to claim authentic videos and photos are fake. Back in 2018, law professors Robert Chesney and Danielle Citron even coined a term for this phenomenon, calling it, “The Liar’s Dividend.”
Hany Farid, a professor at University of California, Berkeley who specializes in forensic images, ran the Harris campaign rally photo through two computer models to see if there were any signs of patterns consistent with generative artificial intelligence or manipulation and found none.
“This is an example where just the mere existence of deepfakes and generative AI allows people to deny reality,” Farid said. “You don’t like the fact that Harris Walz had such a big crowd? No problem. Photos are fake. Videos are fake. Everything’s fake.”…
It’s a problem for citizens of a democracy to have a blurred understanding of what is real and staged, said University of California, Los Angeles law professor, Rick Hasen during a panel discussion hosted by the Wikimedia Foundation on Monday.
“If voters tend to disbelieve anything they see, and think that whatever they see might be faked, then they’re going to distrust their own instincts as to what the truth is to be able to make competent decisions,” Hasen said….