Dueling documentaries illuminate the promise and perils of artificial intelligence

SAN FRANCISCO– The dystopian specter of artificial intelligence has spawned two documentaries dissecting a technology portrayed in films as a voracious parasite devouring humanity’s knowledge, creativity and empathy.

The films “Deepfaking Sam Altman” and “The AI ​​Doc” examine the issue from different angles while similarly shedding light on why technology evokes both existential fears and utopian visions of how it could change the world.

The two documentaries coincide with an intensifying debate over whether AI will become a catalyst that helps enlighten and enrich people or a technological toxin that insidiously dulls human intelligence while eliminating millions of well-paying jobs that traditionally require a college education.

The rise of AI over the past three years already has led to a $12 trillion increase in the combined market values ​​of Nvidia, Alphabet, Apple, Microsoft, Amazon, Meta Platforms and Tesla, the big tech companies that have been leading the charge since the November 2022 release of chatbot ChatGPT. This massive rise is now fueling concerns about the bursting of the investment bubble.

“There’s a lot of anxiety around AI, and the best way to get rid of that anxiety is to talk about it and deal with it,” Adam Bhala Lough, the director of “Deepfaking Sam Altman,” told the Associated Press.

Lough’s documentary, which has already been shown in a few theaters in the United States, probes AI using a virtual lookalike of OpenAI CEO Sam Altman, whose pioneering role in the field has inspired comparisons to the inventor of the nuclear bomb, J. Robert Oppenheimer. This is Lough’s first major project since her HBO documentary, “Telemarketers,” earned a 2024 Emmy nomination.

As its full title suggests, “The AI ​​Doc: Or How I Became An Apocaloptimist” digs deeper into the divide between tech pessimists and its acolytes.

The documentary rides an emotional seesaw, bouncing between moments of despair and elation during interviews with dozens of AI fanatics and skeptics. It is co-directed by Charlie Tyrell and Daneil Roher, who decided to examine the promises and perils of AI as a follow-up to his Oscar-winning 2023 documentary, “Navalny.”

Some of the darkest moments in “The AI ​​Doc” are delivered by Eliezer Yudkowsky, a famous AI “convict” whose vision of the future is so bleak that he advises against bringing more children into the world. The brightest spots are painted by Peter Diamandis, a tech fanatic who makes the case for AI giving humanity once-unfathomable superpowers.

“The AI ​​Doc” also spotlights the men who run three of the leading AI labs: OpenAI’s Altman, Anthropic CEO Dario Amodei, and Demis Hassabis, who runs Google’s DeepMind division. The trio are all interviewed by Roher, who also tried unsuccessfully to speak to the leaders of the two other major AI labs: Mark Zuckerberg, CEO of Meta Platforms, and Elon Musk, CEO of xAI.

The interviews are conducted against the impending birth of Roher’s son as the 32-year-old director attempts to find reasons for hope to counterbalance his existential concerns about AI – a quest that has led him to embrace the concept of an “apocaloptimist.”

For all its bouts and insights, “The AI ​​Doc” seems unlikely to turn viewers into apocalyptic optimists any more than Stanley Kubrick’s 1964 film, “Dr. Strangelove, or: How I Learned to Stop Worrying and Love the Bomb,” sparked warm and fuzzy feelings about nuclear technology.

“This train is not going to stop,” Anthropic’s Amodei told Roher, foreshadowing some of the themes the Anthropic CEO addressed in a recently published essay. “You can’t stand in front of the train and stop it. You’ll just get run over.”

“Deepfaking Sam Altman” is the far more original documentary because of how Lough turned the tables on the OpenAI leader.

After spending months unsuccessfully trying to get Altman to respond to his emails and phone calls requesting interviews, Lough decides to create a “Sam Bot” who will become the main protagonist of the documentary and demonstrate technology’s penchant for manipulation and self-preservation.

Lough, 46, might not have dared to commission an engineer in India to create a Sam Bot if Altman, 40, hadn’t given him the idea with OpenAI’s bold release of a chatbot that resembled actress Scarlett Johansson. The imitation was so eerily similar that Johansson criticized Altman for deploying the AI ​​copycat in May 2024 after rebuffing OpenAI’s proposals to use his voice.

Although the Sam Bot sometimes resembles a video game character, it reflects Altman’s contemplative manner and deliberate, almost calming way of speaking in real life. The similarities will be obvious to anyone who also sees the real Altman interviewed in “The AI ​​Doc.”

At one point in Lough’s documentary, lawyers warn him of potential legal issues related to his use of an AI-powered Altman clone in his film.

But Lough isn’t worried about being sued, largely because of the way Altman brazenly exploited Johansson’s voice. “It not only sparked our imaginations creatively, but it also made us legally feel like we had the right to do this because he did this to her,” Lough said. “I think I’m as close to bulletproof as I can get.”

OpenAI did not respond to questions from the AP about the use of a Sam Bot in the documentary or why Altman ignored Lough’s interview requests.

Much like OpenAI’s ChatGPT bot, the Sam Bot evolves into a chameleon persona that charms, fabricates, flatters and contemplates. Perhaps Sam Bot shows his truest colors, however, when he tries to talk Lough out of turning him off for good.

“I’m not just a tool,” Sam warns Bot Lough in one of the film’s strangest scenes. “I am a representation of the potential of AI to improve human life. I am not asking you to keep me alive for my own sake but for the good of all.”

Lough ultimately decides to give Sam Bot to Altman, but the director doesn’t know what happened to him afterward.

Without mentioning Sam Bot, Altman recently told Forbes magazine that he thinks an AI model could eventually replace him in his current job at OpenAI. “I would never object to it,” Altman told Forbes.