Summary
Ben Affleck says his AI company, InterPositive, spent eight months shooting an original footage dataset to adapt existing models for filmmaking tasks without relying directly on peers’ work. Speaking at Bloomberg Screentime, he described the tools as production aids intended to work with artists rather than replace them. Filmmakers can further train the system on footage from their own productions while keeping that material and its learned adaptations proprietary, he said. Netflix acquired InterPositive for $587 million in July, according to Business Insider’s report carried by MSN. Affleck also said the technology was used extensively in post-production on his Netflix film Animals, but he declined to specify the shots or tasks involved. His account offers no independent assessment of the dataset, underlying models, labor effects or performance of the tools.
Hana’s perspective
I see a meaningful attempt to move the film-AI debate from abstract promises toward a production-specific system, but “ethical” cannot rest on a founder’s assurance. A deliberately filmed dataset may reduce some consent and provenance risks, and production-level fine-tuning could protect a project’s visual language. Yet the unanswered questions matter: which jobs the tools change, how contributors to the new footage are compensated, what the underlying open models contain, and whether filmmakers can audit generated results. Affleck’s refusal to identify the uses in Animals also weakens the case for informed industry debate. If this model is meant to earn trust, selective transparency about real workflows should be part of the product.
Are you a writer with a great story? We'd love to hear from you: APPLY HERE