AI-generated podcasts: Synthetic Intimacy and Cultural Translation in NotebookLM's Audio Overviews

TL;DR

This paper examines AI-generated podcasts by Google's NotebookLM, revealing how they standardize language and cultural context, thus shaping new forms of media interaction and public sphere formation.

Contribution

It provides the first analysis of AI-generated podcasts as media, highlighting their structural template, language translation, and cultural default settings.

Findings

Podcasts follow a fixed structural template.

Language is translated into a standard American accent.

Cultural translation defaults to a white, middle-class American perspective.

Abstract

This paper analyses AI-generated podcasts produced by Google's NotebookLM, which generates audio podcasts with two chatty AI hosts discussing whichever documents a user uploads. While AI-generated podcasts have been discussed as tools, for instance in medical education, they have not yet been analysed as media. By uploading different types of text and analysing the generated outputs I show how the podcasts' structure is built around a fixed template. I also find that NotebookLM not only translates texts from other languages into a perky standardised Mid-Western American accent, it also translates cultural contexts to a white, educated, middle-class American default. This is a distinct development in how publics are shaped by media, marking a departure from the multiple public spheres that scholars have described in human podcasting from the early 2000s until today, where hosts spoke to specific communities and responded to listener comments, to an abstraction of the podcast genre.