Screenwriting with AI – Exploring Storytelling with GPT
The project “Screenwriting with AI” is a scientific-artistic investigation into the collaborative process of screenplay writing with Large Language Models (LLMs). This research aims to explore the potential of LLMs in enhancing human creativity, examining the dynamics of human-machine interaction, and understanding the implications of AI-generated content on the creative industry.
The project involves a mixed-methods approach, combining artistic research with scientific inquiry. The artistic research component includes the development of a screenplay using a collaborative human-AI approach, while the scientific inquiry focuses on analyzing the process, dialogue, and tools of co-writing with LLMs.
The research questions guiding this project are: How can human and machine creativity engage in a fruitful exchange to produce innovative outcomes in the form of a narrative for a film? What are the benefits and limitations of using LLMs in screenwriting, and how can they be effectively integrated into the creative process? What are the implications of AI-generated content on the creative industry, and how can we ensure that LLMs serve as a beneficial collaborator rather than replacing human creativity and jobs?
This project explores the intersection of human and machine creativity in screenwriting, examining the potential for fruitful exchange between humans and machines to produce satisfying narratives. It also investigates the benefits and limitations of integrating machine learning models into the creative process, with a focus on ensuring a harmonious collaboration between humans and machines.
The project will looking into current uses of AI in moving images, and dive into the question whether machines can “think” and be truly creative. Its methodology involves a writers’ room situation, where human writers collaborate with LLMs to develop, write, and refine narratives. The LLMs used in this project is GPT-4, which are evaluated for their ability to generate “satisfying” narratives. The project’s outcomes includes a screenplays for a cinema film, which demonstrate the potential benefits and limitations of using LLMs in screenwriting. The findings highlight the importance of human oversight in ensuring the quality, coherence, and ethical integrity of AI-generated content. Another output is a short film made with Runway ML based on the screenplay’s narrative.
The project is part on the ongoing dialogue about the role of AI in arts and storytelling, suggesting a collaborative model that leverages the strengths of both human and machine intelligence. The research emphasizes the need for a balanced integration of AI in creative processes, ensuring that AI enhances rather than replaces human creativity.