Apocaloptimism at Sundance: The Director of 'Navalny' Tries to Survive in an AI World

Imagine you're standing on railroad tracks, and a high-speed train is hurtling toward you. You don't know if it's carrying a cure for cancer or a nuclear warhead, but one thing is certain: this machine has no brakes. That's exactly the feeling that opens the new documentary "The AI Doc: Or How I Became an Apocaloptimist," presented at the Sundance festival. Director Daniel Roehr, who recently accepted an Oscar for the film "Navalny," suddenly discovered that the main threat to our time is not politicians in suits, but lines of code that write texts and draw pictures better than he does himself.

It all started with personal anxiety. Roehr, like many of us, began tinkering with tools from OpenAI and quickly moved from the stage of "wow, how cool" to "wait, what am I going to do now?" His film is an attempt to make sense of an industry that changes faster than we can update the tabs in our browser. Daniel Kwan, one of the creators of the wild "Everything Everywhere All at Once," was brought into the production, and you can feel it: the film doesn't look like a boring lecture—it's more of a visual investigation of our collective confusion.

At the center of the plot are big players and big fears. Sam Altman appears on screen, calmly talking about a future where AI will solve all of humanity's problems. But Roehr doesn't let him simply sell us a bright future. He confronts Silicon Valley optimism with harsh skepticism from critics and personal experience of artists. The film asks a question that concerns everyone right now: is AI an existential threat or a historic opportunity? And, to be honest, the answer "both" sounds increasingly frequent.

The term "apocaloptimism," featured in the headline, perfectly captures the current state of the tech industry. We all understand that the old world is crumbling, but at the same time we can't take our eyes off the possibilities that are opening up. Roehr shows how neural networks are already changing film production now, turning complex tasks into a mouse click. For him, as a documentarian, this means a crisis of trust in the image. If we can generate anything, how will we believe the truth?

Why does this matter right now? Because we've passed the hype stage and entered the integration stage. While politicians argue about regulation and corporations compare model parameters, ordinary people—from screenwriters to designers—are trying to find solid ground. Roehr's film doesn't offer ready-made recipes for survival, but it captures a historical moment when humanity first seriously questioned its own uniqueness.

The main point: The train really isn't going to stop. The only question is whether we'll manage to get seats in the car or remain standing on the tracks, wondering what that was. Claude 4 on the roadmap?