Pete Ohs and Jeremy O. Harris Resist AI, Release Erupcja
Director Pete Ohs and writer Jeremy O. Harris debut the indie film Erupcja, a 71-minute character drama starring Charli XCX that opens in theaters on April 17. The collaborators frame the movie as both a modest travelogue shot in Warsaw and a creative statement about presence in an age dominated by generative AI. They describe production choices that privilege human improvisation, co-writing with the cast, and staging a narrative about disconnection that literalizes a desire for a digital detox, most starkly by discussing time spent inside a Japanese prison as a forced reboot. On the broader cultural front they voice skepticism about Silicon Valley AI culture, invoking figures associated with rapid commercialization of AI while arguing for artistic spaces that remain human-first.
What happened
Director Pete Ohs and writer Jeremy O. Harris released the indie film Erupcja, a 71-minute drama starring Charli XCX that opens wide on April 17. The film was shot in Warsaw and developed collaboratively, with Charli participating in the script; it centers on a relationship strained by a natural disaster and a protagonist drifting back to an old life. The filmmakers frame the project as an intentional counterpoint to tech-driven creative workflows, discussing resistance to AI, the role of Silicon Valley figures in shaping public conversation about automation, and a personal account of a Japanese prison stay portrayed as an extreme digital detox.
Technical details
The production emphasizes hands-on, low-footprint filmmaking and improvisational writing choices. Key production features highlighted by the team include:
- •Collaborative authorship, with the script co-written by Pete Ohs, Charli XCX, and the cast to allow spontaneity on set
- •Location-driven shooting in Warsaw over the summer of 2024, traded for slick VFX or AI-assisted post workflows
- •A short-form structure at 71 minutes, favoring mood and character over plot-heavy exposition
Context and significance
The creative stance here matters because it is a visible example of artists articulating practical and ethical alternatives to AI-first production. The filmmakers do not reject technology wholesale, but they reject a default model where generative systems and AI-first platforms substitute for human labor, improvisation, and embodied performance. That stance maps onto a broader cultural debate: artists and technicians are negotiating when to adopt AI tools that accelerate production, and when those tools erode the skill sets and serendipity that produce distinctive work. The anecdote about a Japanese prison as a place of enforced disconnection is useful as a metaphor for what some creators now seek deliberately, a space free from ambient tracking, recommendation engines, and rapid iteration cycles driven by algorithmic feedback.
Why practitioners should care
For machine learning engineers and product teams, the film and the creators conversation are a reminder that creative users will choose workflows based on values, not only on efficiency. If AI tools ignore craft-sensitive touchpoints such as improvisation, co-authorship, and on-set decision making, they risk limited adoption among creators who prioritize those qualities. This also suggests a potential product opportunity: tooling that augments rather than replaces live creative collaboration, which respects provenance and centers human performance.
What to watch
The film's release could sharpen debates around ethics in creative tooling and content provenance, especially if prominent artists adopt explicit anti-AI positioning. Expect follow-up interviews and festival conversations to push on how to design AI systems that are augmentative, auditable, and opt-in, rather than default replacements for creative labor.
Scoring Rationale
This is primarily cultural news with limited technical novelty for ML practitioners. It matters for AI ethics and creative workflow design, but it does not introduce new models, tools, or regulations. The visibility of the artists gives the story modest influence on public debate and tool adoption.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problemsStep-by-step roadmaps from zero to job-ready — curated courses, salary data, and the exact learning order that gets you hired.
