The AI question arrives in The AI Doc: Or How I Became an Apocaloptimist not through a server room, a product demonstration or a graph forecasting what comes next, but through a filmmaker preparing for a child. Daniel Roher, a director and the documentary’s on-screen first-person guide, makes his impending fatherhood the frame for an inquiry into powerful artificial intelligence: what kind of world is taking shape, and what does it mean to bring someone new into it?

That starting point gives the film a domestic register. Its attention is trained on a father-to-be attempting to understand a subject that can otherwise seem remote, technical or too vast to grasp.

Roher’s search organizes the documentary without turning it into a simple personal answer. The premise is a father trying to work out what the arrival of powerful AI could mean for the child he is about to have. The scale of the issue is large, but the stake is deliberately ordinary and immediate: the prospect of raising a child in a world where the rules may be changing before anyone can fully describe them. That personal thread is the route into a wider investigation, rather than a detour from it.

The AI Doc: Or How I Became an Apocaloptimist is a 104-minute English-language feature documentary from the United States arriving on Netflix. In it, director Daniel Roher investigates artificial intelligence while preparing to become a father, speaking with figures associated with the development of advanced AI systems as well as figures who have raised concerns about their consequences.

The result is built around disagreement. Rather than treating AI as a single question with a settled response, the documentary gathers voices from industry, research, ethics and technology journalism whose assessments do not neatly align. Roher’s first-person investigation connects those competing accounts to the uncertainty of parenthood, making the film less a technical explainer than an attempt to understand what adults owe to children when even experts differ on the shape of the world ahead.

Its title provides a name for that unsettled position. “Apocaloptimist” fuses “apocalypse” and “optimist,” describing a posture in which dread and wonder are held at the same time. The film does not present that word as a clean resolution to the conflict it explores. Instead, it identifies the tension facing Roher as he listens to people who view the technology’s possibilities and risks in sharply different ways.

Among the people building or leading major AI efforts are Sam Altman, a featured subject associated with OpenAI; Ilya Sutskever, a featured subject and AI researcher formerly associated with OpenAI; Demis Hassabis, a featured subject associated with Google DeepMind; and Dario Amodei and Daniela Amodei, featured subjects associated with Anthropic. Their presence places leading figures from the current AI industry within a documentary shaped by a more intimate question than corporate ambition: what should a prospective parent make of the systems being built now?

The film also includes voices identified with warnings, critiques and debates around AI’s social and existential implications. Yoshua Bengio, a featured subject described as a deep-learning pioneer and AI-safety voice; Eliezer Yudkowsky, a featured subject and AI-risk theorist; Timnit Gebru, a featured subject associated with AI ethics; and Emily M. Bender, a featured subject and computational linguist, bring differing perspectives to that side of the conversation. So do Yuval Noah Harari, a featured subject, historian and author of Sapiens; Tristan Harris and Aza Raskin, featured subjects associated with the Center for Humane Technology; and Karen Hao, a featured subject and technology journalist.

These are not presented as a single bloc with one shared forecast, any more than the film’s builders are presented as one unified camp. The documentary’s subject is the absence of consensus among people with deep involvement in, or close scrutiny of, artificial intelligence. It assembles separate interviews into a film in which competing views can sit beside one another, allowing the arguments to register without claiming that one side has definitively answered the other.

That structure matters to its fatherhood frame. AI risk is not treated solely as a distant scenario or a specialized dispute; it becomes part of the ordinary, unresolved work of imagining a child’s future. Roher’s investigation asks the people shaping, studying and challenging AI to speak to a question that remains open: whether the world being built will be one a parent can understand well enough to prepare a child for.

Director Daniel Roher brings an access-driven approach associated with his earlier documentary Navalny, which won the Academy Award for Best Documentary Feature, to material that depends on proximity to people shaping the field. Here, however, access is not used to establish a single authoritative account. Roher’s encounters are part of an inquiry whose central difficulty is that proximity to AI does not produce agreement about it.

Co-director Charlie Tyrell helps give that reporting an essayistic form, assembling the film around the friction between its interviews rather than converting them into a settled argument. The structure leaves room for competing assessments to accumulate: confidence, alarm, skepticism and uncertainty can all be present without being reduced to a common conclusion. That is an especially consequential choice for a film anchored in an expectant father’s attempt to make sense of the future. The documentary does not suggest that gathering eminent voices removes the uncertainty; it makes the uncertainty visible.

Its production team includes producer Daniel Kwan, who co-directed Everything Everywhere All at Once, alongside producers Jonathan Wang, Shane Boris, Diane Becker and Ted Tremper. Their credits sit behind a feature that moves between a personal first-person account and a broad collection of interviews. The film’s method is less about appointing a winner in a public argument than about showing why the argument persists when its participants have unusually direct knowledge of the systems at its centre.

That approach arrives during a 2026 AI moment marked by frontier labs racing ahead, an existential-risk debate that has entered the mainstream, and public concern about work, truth and children. The split mapped by the documentary is not simply between specialists and an uninformed public. It is a schism among people with deep involvement in AI’s development, research and critique, including figures who have helped build the technologies now under debate.

“Apocaloptimist” gives that condition a name without making it a solution. The title’s fusion of apocalypse and optimist acknowledges that wonder and dread can coexist, particularly when the possible effects of powerful AI remain difficult to define in advance. For a parent-to-be, that uncertainty is not an abstract matter of competing forecasts. It becomes part of thinking about what preparation, responsibility and reassurance can mean for a child.

The documentary’s route to Netflix also changes its reach. After its world premiere at the Sundance Film Festival on January 27, 2026, the film received a U.S. theatrical release through Focus Features on March 27, with Universal Pictures handling its international theatrical release. It arrives on Netflix in the United States on September 15, taking a documentary first seen in festival and theatrical settings to a substantially wider streaming audience.

Variety’s 2026 review described the film as “A Scary, Dizzying and Essential Deep Dive Into the AI Revolution” while the IMDb title page lists an audience user rating of around 8.2. Those measures reflect attributed critical and user response, rather than a resolution of the questions raised by the film itself.

A gathering of leading voices cannot make the future knowable. It cannot turn sharply different accounts of powerful AI into one reliable prediction, nor can it settle whether the technology’s possibilities will ultimately be understood as benefit, danger or both. The people closest to the machine do not offer a single answer.

That is the uncertainty left with the parent at the film’s centre: not certainty that AI will save or end the world, but the knowledge that even those building, studying and challenging it remain divided over what lies ahead. The question stays open, as does the task of imagining a child’s place within that unsettled future.