Rewriting the Past for the Present: A Computational Study of Sitcom Presentism using Fine-Tuned Language Models

This paper explores sitcom presentism by using fine-tuned language models to rewrite 1990s Seinfeld plots as plausible 2025 versions while preserving their core comedic beats and character dynamics. Through a three-stage computational pipeline, the study shows that the model can produce culturally updated plots with strong semantic continuity, low lexical overlap, and high diversity, framing presentism as adaptation rather than simple repetition.