Daniel Roher, who won the Academy Award for Best Documentary for Navalny in 2022, has a new film in U.S. theaters as of March 27. It's called The AI Doc: Or How I Became an Apocaloptimist, it runs 103 minutes, and it features what might be the most stacked interview roster in documentary history: Sam Altman, Dario and Daniela Amodei, Demis Hassabis, Yoshua Bengio, Ilya Sutskever, Timnit Gebru, Eliezer Yudkowsky, Yuval Noah Harari. The full cast list on IMDb runs to 35 names.
The hook is personal. Roher and co-director Charlie Tyrell both learned they were becoming fathers in early 2024 and built the film around a single question: what kind of world are their kids inheriting? Producer Ted Tremper spent the production developing over 100 contacts inside the major AI labs, with pre-interviews running up to 20 hours each. The directors ultimately filmed more than 40 experts, generating 3,300 pages of transcripts, according to a Variety feature published this week.
The rolodex is the point
Getting Altman, both Amodeis, and Hassabis in the same documentary is no small feat. Toss in Sutskever (who has barely spoken publicly since leaving OpenAI), former OpenAI researcher Daniel Kokotajlo, and critics like Gebru and Emily M. Bender, and you have a film whose cast functions as its own argument for relevance. Distributed by Focus Features and produced by the teams behind Everything Everywhere All at Once (Daniel Kwan, Jonathan Wang) and Navalny (Shane Boris, Diane Becker), the pedigree is calculated to cut through a crowded documentary market.
Whether that access translates into insight is another question. The title's neologism, "apocaloptimist," comes from one of the interview subjects, and Roher used it at the Sundance premiere in January to describe a third path between doomsday panic and tech boosterism. It's a good coinage. But some critics think the film's commitment to that middle ground is exactly its weakness.
Critics are split, and the split is telling
On Rotten Tomatoes, the film holds an 89% approval from 35 reviews, which sounds strong until you check Metacritic, where it sits at 60 out of 100 based on nine weighted reviews. That gap tells you something. Most critics think it's worth watching. Fewer think it's actually good.
The Variety review by Owen Gleiberman called it essential viewing, praising its editing and sheer informational density. The RogerEbert.com take was warmer on the emotional framing but flagged a real blind spot: the film imagines AI tutors and AI doctors in every home without interrogating who actually gets access to those tools. Not everyone is Elon Musk's kid.
The negative reviews are more pointed. One Metacritic critic argued that Roher's willingness to accept his interview subjects' claims at face value leaves the film feeling toothless. Another called it a well-intentioned but aggravating soup of information. And a Letterboxd reviewer went further, calling the entire film a 104-minute ad for Anthropic.
So should you see it?
If you've spent the past two years reading AI coverage, the information here won't surprise you. Several critics noted this. The film acknowledges its own potential obsolescence, with one interviewee warning that any examples shown will seem dated by release. That's honest, at least, even if it doubles as a preemptive excuse.
What the film does offer is a concentrated dose of the people making decisions about AI sitting in a room and talking about what they think they're building. That alone has value, even if Roher doesn't push back as hard as some viewers might want. The fatherhood angle gives the whole thing an emotional pulse that pure tech documentaries usually lack.
The film is playing in U.S. theaters now through Focus Features and is available digitally on Fandango at Home and Apple TV. It's rated PG-13 for language.
