Director Adam Bhala Lough didn’t got down to make a documentary a couple of digital simulacrum of Sam Altman.
However after about 100 days of texting and emailing the OpenAI CEO for an interview—with no response, he claims, and with financiers hounding him to make good on his authentic pitch—Lough was at his wit’s finish.
He’d exhausted nearly each angle. “As soon as I reached that time, I gave up and I pivoted to gate-crashing OpenAI,” he says. Although he’d employed an analogous tactic in his Emmy-nominated 2023 documentary Telemarketers—a chronicle of industry-wide corruption within the telemarketing enterprise—it wasn’t a filmmaking model he felt all that comfy with. “It was a fortress. I used to be capable of slip by means of the gate, and instantly safety grabbed me and bodily eliminated me from the premises.”
So begins Deepfaking Sam Altman, Lough’s portrait of how AI is reshaping society and his quest to speak to the person behind it. When his authentic plan fell by means of he drew inspiration from Altman himself. “The Scarlett Johansson controversy erupted,” he says. In 2024, the actress publicly known as out OpenAI for seeming to repeat her voice for its new AI voice assistant Sky. “It was at that time the place I bought the concept to do the deepfake.” (In a Could 2024 assertion, Altman apologized to Johansson and stated Sky’s voice was “by no means meant to resemble” hers.)
What initially begins out as a easy voice clone balloons right into a full deepfake of Altman known as Sam Bot, which Lough travels to India to have created. This being a Lough movie, although, nothing goes in line with plan. With out spoiling an excessive amount of, Sam Bot finally turns into its personal entity, and the movie takes an excellent stranger—and revelatory—dive from there. “There’s parallels between this film and Terminator 2: Judgement Day, however there’s not one of the violence,” he says. Lough grew up throughout what he calls the “AI 1.0 period.” His obsession with James Cameron’s Terminator 2 was a significant affect on his craft.
Deepfaking Sam Altman, which is predicated partially on the New York Journal story casting Sam Altman because the Oppenheimer of our age, options commentary from former OpenAI security engineer Heidy Khlaaf, who tells Lough, “We’re beginning to see OpenAI dip its toes in army makes use of, and I can’t think about one thing like Dall-E and ChatGPT getting used for army assists. That basically scares me, given how inaccurate these techniques are.”
