In the News

The New York Times:

Congress screens ‘The AI Doc’ | The New York Times

With Congress struggling to find its footing on artificial intelligence, proponents of greater government checks on the A.I. industry thought a movie night might help.

Last night, two nonprofit A.I. safety groups invited members of Congress and their staff members to the Capitol’s visitor center for a screening of “The AI Doc: Or How I Became an Apocaloptimist,” which focuses on safety concerns that have since risen to the national agenda after Anthropic’s C.E.O., Dario Amodei, called for a slowdown of A.I. development last week.

The event, co-hosted by Representative Ted Lieu, Democrat of California, tapped into the time-honored Washington tradition of leveraging politicians’ concerns about their own children, Nancy Scola reports.

The film is framed as a quest by a father-to-be. It follows the filmmaker Daniel Roher as he tries to work through his fears about the world that his child will inherit, with limited success: He balks at the assertion by Sam Altman, OpenAI’s C.E.O., that “our kids will never be smarter than an A.I.”(“It does unsettle me a little bit,” Altman says, “but it is reality.”)

“I wish you could get people to care about the future, because we care about the future,” Tristan Harris, a former Google design ethicist who helped found the nonprofit group Center for Humane Technology, told DealBook at the screening. “But instead, sadly, sometimes you have to invoke people’s children, or the idea of children, just to get people to care.”

“The AI Doc” underscores the need for legislators to act, said Riki Parikh, the Policy Director at The Alliance for Secure AI, one of the event’s hosts. That is one reason, he added, that his group wanted to bring the film to Capitol Hill.

Anthropic’s Amodei tells Roher, for example, that “it’s up to all of us to push for the government to get involved.” Google DeepMind’s Demis Hassabis, who has backed Amodei’s call to rein in A.I. development, also appears in the film.

Lieu, for his part, has championed the A.I. Kill Switch Act, which would require major developers to establish mechanisms for shutting down their models. The bill so far has only three co-sponsors, including Republicans.

During a post-screening panel, Harris asked a crowd of several dozen to call out one-word takes on how the film made them feel. “Bummed,” one person said.

SHARE WITH YOUR NETWORK

Media Contact

We welcome inquiries and interview requests from members of the media. Please contact us for more information. 

Sign Up for Our Newsletter

By signing up, you agree to receive email updates and communications from The Alliance for Secure AI Action.