Ben Affleck is going viral this week for demonstrating a genuinely technical understanding of AI in recent interviews, just months after selling his AI filmmaking startup to Netflix for a reported $587 million. In a GQ ‘One More Question’ episode, Affleck told interviewer Zach Baron that he has been into computers since childhood and got deeper into the space when analog film moved to digital. He described how the visual effects workflow has used machine learning for years and explained concepts like convolutional neural networks, tensors, transformers, GPUs, and inference. He also said he writes ‘pretty shitty Python scripts.’

Affleck said his celebrity status helped him get early looks at emerging AI technology, including visits to OpenAI, which led him to think about getting more directly involved. He explained the reasoning behind the startup he founded: to apply AI ethically in filmmaking, he wanted to create a proprietary dataset rather than rely on public data. He raised money and shot for about eight months with many cameras to build a dataset for late-stage training of open models on discrete tasks, pairing training and inference code so the models could generate value in production.

In a separate clip from Bloomberg’s Screentime 2026 conference, Affleck told Lucas Shaw how he fine-tunes open video models by unfreezing the weights and training only the last cinematic layer, allowing a film crew to hit real production standards. He said he used the technique in his movie ‘Animals’ for post-production. The article notes that Affleck founded InterPositive in 2022, a 16-person company, and that he pushed back on the $587 million figure, saying he did not own the whole company.

Affleck also shared his concerns about AI: he worries about his kids in school, a 30% rise in college A’s over the last three years, learned helplessness, and responsible use. He does not worry about Skynet-style scenarios, and he argues AI will be additive to the movie business rather than replace it.