Aniket Bera, director of Purdue University's Idea Lab, has worked on two very different projects in the AI field. One is the restoration of an 1899 film fragment, believed to be the oldest surviving footage of India, and the other is an early AI-based experiment with Satyajit Ray's Pather Panchali.
“AI softens the shadows and contrasts that were central to the movie's atmosphere. AI doesn't understand symbolism, it just infers patterns,” he says. Vera says every step required human review to ensure the results were true to the original. “AI often 'improves' things by hallucinating details and changing the visual language, putting us at risk of rewriting history.”
For Mukherji, AI has helped him realize his filmmaking vision. How else could they have cast two deceased actors? The AI recreated Uttam Kumar's voice throughout Oti Uttam. However, he emphasized that the project relied heavily on human input for everything from writing the script, collating archival footage, obtaining legal permissions, and vetting the AI's output.
AI tools are rapidly evolving, raising many regulatory and ethical issues. Mukherji advocates optimism. “Humans should get used to AI instead of panicking,” he says. “Tame it, master it, harness it. It's not an android-like monster trying to devour your creativity. It complements it, it doesn't replace it.”
But for some, the limits of AI are still evident. Chandu now shares what he learned on set in the classroom, teaching a course on AI in film at a university. In one module, he invites students to create two films. One uses ChatGPT and AI video tools, and the other uses completely traditional technology.
