Sora: OpenAI's video model that shocked Hollywood

2025

When OpenAI released Sora in 2024, the response from Hollywood was not excitement; it was existential anxiety. The model generated 60-second videos with consistent characters, realistic physics, and production quality that rivaled professional cinematography.

"A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots." The resulting video looked like it was shot by a professional crew on location.

What made Sora different from earlier video models was temporal consistency. Characters maintained their appearance throughout. Objects obeyed physics. Camera movements were smooth and intentional. Scenes had narrative coherence, not just visual quality.

For the film industry, the implications were profound. Pre-visualization became trivially easy. B-roll footage could be generated on demand. Entire sequences could be prototyped before committing to expensive production.

Independent filmmakers gained superpowers: the creative constraints of budget evaporated for visualization purposes. A student with a script could see their vision rendered at near-Hollywood quality.

The response from creative unions was predictable and justified: this technology, if unchecked, could displace thousands of production workers. The industry began negotiating AI clauses in contracts, establishing new norms for a world where creation and generation were converging.