|
In 2017, a Redditor uploaded a video that had been manipulated using artificial intelligence: a deepfake. Ever since, Simon V.Z. Wood writes, “the potential for next-gen ratfuckery had been obvious: release a clip of a politician saying something she didn’t, wreak havoc.” In our conspiracy-addled political atmosphere, as Americans lose their sense of consensus reality, politicians, digital scholars, and journalists have aimed to draw attention to the risks deepfakes pose to democracy—and to basic epistemology. But when he scoured the internet, Wood turned up few videos designed to deceive, and many more intended to raise awareness about the ability to deceive. “Searching for evidence that bad actors were weaponizing artificial intelligence for political gain,” he writes, “what I found instead was an emerging field of detection firms, government grantees, startups, academics, artists, and nonprofits that seemed to depend on one another to sustain interest in deepfakes. Call it the deepfake-industrial complex—or, perhaps, a solution in search of a problem.”
The result is a media ecosystem characterized by constant possible delusion. Sam Gregory—the program director at Witness, a nonprofit focused on the threats and opportunities of emerging technologies—tells Wood that deepfake-awareness stories “imply the scale of visual deception around us is far greater than it is.” The victors in this scenario, Wood discovers, are “the people who exploit fears of deepfakes to create plausible deniability about real-life events.” In academic circles, that concept is known as “the liar’s dividend.” All the while, the largest market for deepfakes—much less discussed in the press—reflects a general rule of the internet: “Roughly 95 percent of the known universe of deepfakes, according to a report published in 2019, consists of the heads of celebrities being grafted onto the bodies of porn actresses.” —Betsy Morais, managing editor
READ MORE »
|
