On Thursday, Spotify extended its partnership with Google Cloud to incorporate large language models (LLMs) for analyzing user preferences across podcasts and audiobooks, enabling tailored recommendations.
The content and related data are generated using advanced language models like OpenAI’s ChatGPT and Google Bard, which are trained on vast datasets.
Google Cloud, a subsidiary of Alphabet (GOOGL), boasts multiple LLMs specialized in text, code, images, audio, and video, such as PaLM 2, Codey, Igniter, and Chirp.
A decade ago, Spotify embraced AI for music recommendations, positioning itself as an early AI adopter. Now, the Finnish company aims to leverage LLMs to enhance recommendations for non-musical content like podcasts and audiobooks.
In a bid to boost revenue, the music streaming giant is diversifying its content offerings to include ebooks and podcasts.
Previously, Spotify had anticipated significant returns from its podcast and audiobook ventures.
Gustav Söderström, Spotify’s Chief R&D Officer, emphasized Google Cloud’s commitment to providing an optimal platform for innovation through conceptualAI’s evolving capabilities.
Additionally, Spotify is exploring the use of LLMs to enrich user experiences and identify potentially harmful content through its expanded collaboration with Google.
