003: AI Filmmaking, Part 2

How generative AI tools can change filmmaking in general?

Let’s brainstorm a little bit about how generative AI tools can change filmmaking in general.

The most terrifying extreme example is that it will totally automate the creation process of content for a specific audience member or group, in accordance with their own taste and mood, and the AI-generated content will be created instantly on demand.

Imagine this future scenario with some more details:

Our main character is Richard, 41, who works for a tech company. He is married to Jane, 40, who works from home as an artist. Richard arrives home after a day of work. Jane is out having dinner with friends. Although there are humanoid robots available for daily jobs, they still prefer to handle most of the household tasks themselves, believing that it motivates them to move more within the house. Besides, they think it’s kind of intimidating to have a walking humanoid robot around the house, never being sure whether it sends any of their private data elsewhere for any purpose. Yet they are not totally anti-tech and still embrace new technologies in moderation.

Richard likes to listen to a podcast while prepping the meal. So, by asking his AI agent with a voice command, a podcast on a specific topic can be generated in seconds. This agent is already familiar with Richard’s interests and can create the podcast in alignment with them, or he might ask for a specific topic and even set the tone—informative, fun, or fictional. It really doesn’t matter; the content is instantly created. If he would prefer to have a video for the same podcast content, that is also not an issue with AI generation tools. The podcast can easily be generated as a dialogue between Richard’s favorite real people and/or fictional characters. It might have any kind of setting as Richard pleases: a radio station with big microphones in front of them like “The Joe Rogan Experience” or “Conan Needs a Friend” style, perhaps a panel on a TED stage, or an educational documentary with high-quality video sequences like “Cosmos: A Spacetime Odyssey.” Maybe he even peeks at the video podcast through his XR glasses without getting too much distracted because he is slicing the onion at the same time. Both Richard and his AI agent know approximately how much time he needs to prepare the meal, so the episode of the podcast takes exactly that long.

Image generated with Copilot

He then sits in front of his big TV with his plate and has time for a one-and-a-half-hour movie. Yes, the XR glasses are cool, but they still cannot provide the comfortable viewing experience yet. 🙂

He might have a short conversation with his AI agent to set the genre and pick the main actors—whose appearances and voices Richard’s AI generation service has rights to use. Maybe the same streaming platforms of today like Netflix, Prime Video, Disney, etc., can still be the ones that own those rights. Perhaps having paid memberships to more than one service can even allow users to use fictional characters from totally unrelated parallel universes, like "Lord of the Rings" from Amazon meets "Star Wars" of Disney. The movie can again be instantly generated to his taste.

Let’s turn the knob for “Black Mirror”-ness to eleven and imagine that Richard’s reactions are also being monitored while watching the film. His heart rate, eye movements, facial expressions, and maybe even his brain waves might be sent to the service, which can understand whether Richard loses interest in the movie at some point. So, why not give the service the ability to change the content in real time? Once the system realizes that Richard is bored, the content is adapted to increase his interest. Maybe some cool visual effects to mesmerize him or sophisticated jokes among the characters to make him laugh.

I know it's starting to sound a little bit creepy for many who read this far, but there is good neurocinematic research that I have been sharing with my students in recent years, which might imply that this kind of content generation is exactly what great film directors and studios have been aiming for. According to this research, the movie “Bang! You’re Dead” by Alfred Hitchcock—who proudly considered himself a master of manipulation and orchestrated the responses of his viewers while building tension and suspense with his writing, cinematography, and editing—causes greater simultaneous brain responses in audience members at almost the exact same spots in the brain and with similar rates. In comparison, segments from the classic movie “The Good, the Bad and the Ugly” by Sergio Leone, “Curb Your Enthusiasm” by Larry David, and random video shot in a park cause, in that order, less and less simultaneous responses in the viewers' brains. So, I believe that the above scenario will happen to some degree, and we will witness a lot of backlash and never-ending discussions. Yet the entertainment industry will eventually adapt to these changes sooner or later.

Credit: New York University

What’s the worst-case scenario? I believe it would be giving individuals the power to entertain themselves infinitely in an addictive way while escaping reality completely. Especially for those who might be on the spectrum of some mental health disorders, this might even make some people give up on moving and feeding themselves, like in some examples of online gamers who lost their lives after uninterrupted gaming sessions for a few days.

As someone who is still working as a gaming startup advisor and game design department lecturer, I will talk about the effects of AI tools in the game industry in another post. Yet the basic principles that I have mentioned here—like instant tailor-made content creation—are still going to be relevant for the games and game-like experiences with the addition of keywords like interactivity, immersion, and presence.

Back to Richard 🙂 Once the movie is finished, he might share some feedback and suggestions with the service provider, and maybe the revised version of the movie can even be submitted to the library of the platform, which might attract its own audience. Maybe this could create another model of user-generated content creation economy, where the main creator can be paid in accordance with the demand for their content.

After a lovely dinner with friends, Jane arrives home. She’s a little bit hungover and would like to cuddle on the couch with Richard while watching another short episode of a romcom TV series they created together before going to bed. 😴

Image generated with ChatGPT

Thanks for reading this far. Feel free to share your thoughts, and take care.