The film director winning awards with AI | Ben Hansford

Summary

Ben Hansford makes the case that AI is evolving from a novelty into an autonomous production partner that can execute creative missions across directing, VFX, and publishing. He says he began using generative AI about four and a half years ago, quickly found it useful for winning commercial jobs, and then brought it into USC teaching, where students used it to produce award-winning work. The conversation tracks Hollywood’s shift from open hostility during the writers strike era to a more pragmatic search for answers as budgets shrink and distribution becomes harder. Hansford also details a comic-book platform, Block Block, that combines AI-assisted creation with the infrastructure needed to actually ship and share a product. A standout example is a “talking raccoon” workflow that went from idea to polished result in about 30 minutes using tools like Live Portrait, ComfyUI, Move AI, and 11 Labs.

Notes

• AI is framed as a mission-oriented creative crew, not a passive feature set.
• Hansford says the key value is autonomous execution, even when the output is imperfect.
• He says he started using generative AI about four and a half years ago, before mainstream tool adoption.
• Early AI use gave him a direct commercial edge in treatments and style frames.
• He says he used AI to produce style frames and image manipulation that clients could not easily trace.
• He claims USC’s AI teaching program started earlier than the university itself realized.
• He says he began teaching AI at USC almost four and a half years ago.
• He describes a cultural shift from hostility toward AI to studios actively asking for help.
• He links the studio reset to falling budgets and harder distribution.
• He argues younger audiences are drifting away from legacy studios.
• He describes Block Block as a safe place to develop, create, share, and explore comic books online.
• The platform targets a comic creation bottleneck: many storytellers, very few illustrators.
• Block Block includes AI-assisted layout, word balloons, and sound design.
• He says the product is meant to create comics instantly and then let users share them.
• He emphasizes that the real product challenge sits below the visible creative layer.
• He calls databases, publishing, and security the hidden 90% of software creation.
• The example underscores an investor-relevant moat: workflow orchestration, not just generation.
• A “talking raccoon” test case went from concept to finished output in about 30 minutes.
• He used Live Portrait and ComfyUI to generate the raccoon’s facial performance.
• He used Move AI to extend the performance into wide shots.
• He treats AI performance as a craft with its own rules and tuning knobs.
• For the raccoon character, he says the eyebrows and head motion had to be exaggerated.
• He says voice transformation in 11 Labs is part of the performance stack.
• He rejects the idea that AI is the real threat; he says shrinking budgets are.
• He says filmmakers need tools that supercharge them when they must do the same work with less money.
• He argues AI can build momentum and function like a director, crew, or Dungeon Master.
• He says some critics miss the highly personal projects students are building with AI.
• He cites a student project using a dead grandmother’s voice recording as an example of personal use.
• He says the main constraint is human attention, not model capability.
• He uses agents to review their own work before escalating to him.
• He says his agents even check his Oura sleep score before bothering him in the morning.