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

2025-11-11

AI Summary

This study examines how AI-generated podcasts produced by Google NotebookLM enact cultural standardization through automated multilingual translation and speech synthesis. Method: Employing a dual-AI-host architecture integrating document understanding and conversational generation, the system transforms heterogeneous, multilingual, and culturally diverse source texts into standardized audio outputs—uniformly rendered in Midwestern American English and aligned with white, middle-class cultural norms via fixed templates and text-to-speech synthesis. Contribution/Results: The research demonstrates that such AI podcasts are not epistemologically neutral but actively impose monolingual stylistic conventions and hegemonic cultural defaults, eroding source-language contextual specificity and audience situatedness. Consequently, public discourse shifts from pluralistic, locale-embedded forms toward abstract, decontextualized media artifacts. This work constitutes the first systematic identification of cultural standardization mechanisms in generative podcasting, significantly advancing critical scholarship on AI’s role in reshaping public spheres and sociocognitive frameworks.

Technology Category

• Natural Language Processing: Generation
• Humans and AI: AI for Accessibility
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Web Mining and Content Analysis: Large pretrained models with web data

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.

Problem

• Analyzing AI-generated podcasts as media rather than just functional tools
• Examining how AI standardizes cultural contexts to American middle-class defaults
• Investigating the shift from community-specific podcasts to abstracted AI formats

Innovation

• AI-generated podcasts with dual-host conversational format
• Fixed template structure for automated content organization
• Cultural translation to standardized American demographic defaults