When people hear about AI in the film industry, they almost always think of AI stealing the jobs of screenwriters. As a writer, I tried using AI for writing out of curiosity, but the results it produced did not feel threatening to me at all.
AI is not imaginative or creative. But it can be a powerful tool for making certain creative processes more efficient.
I saw that clearly when I was casting actors for my films. There was often a delay with self-tape auditions because actors needed someone else to read the other lines. Some hardworking actors even recorded the other lines themselves and played them back during the tape. It always struck me as an unnecessary hurdle. And, to be honest, some self-tapes threw me off because the off-screen scene partner gave a very poor read.
As a screenwriter, I was dealing with another issue. It can take a real-life table read to find out that a scene simply does not work. I often found myself wishing for a shorter feedback loop, some way to hear an early draft of a scene out loud. A real-life table read is expensive, difficult to organise, and usually something you only do when the script already feels in good enough shape.
I’m also a software developer. Last year, me and my partner Robin Edman entered the World’s Largest Hackathon. We had both been impressed by how realistic AI-generated voices had become, and that led us to an idea.
What if we created an app that actors could use to record self-tapes without waiting for a flatmate or friend to read the other lines? They could use an AI actor as a scene partner for their self-tape instead. Also, I had heard enough tapes with flat, robotic reads from the actor’s scene buddy to feel that the AI could genuinely do better.
At the same time, I saw how the same technology could help with my other problem as a writer. What if the app could also work as a table read tool for screenwriters who want to hear their scenes acted out while they are still drafting? It would not replace a real-life table read. But if the script is being tested and refined as it is written, the eventual table read can be far more valuable when the time comes.
That is how Table Read Studio began.
Table Read Studio went on to win the Uniquely Useful Tool award in the World’s Largest Hackathon. With more than 100,000 participants, that felt like huge recognition.
Later, we split the product into two apps: Table Read Studio for screenwriters and Self-Tape Studio for actors. It became clear that actors and screenwriters needed slightly different kinds of support.
To me, this is where AI is most useful in film: not in replacing artists or generating finished work, but in helping writers and actors keep moving when budgets and deadlines are tight.
Vaiga Perkauskaite is a Brighton-based filmmaker, writer, comedian and software developer.
