PARIS: The music streaming platform Spotify is seeking to develop the use of artificial intelligence across its various services.
Thanks to Google Cloud’s large-scale language models, Spotify intends to enhance the user experience by proposing content discovery through personalised recommendations spanning its entire audio offer, including podcasts and audiobooks.
The aim is for these algorithms to analyse all the metadata of this content, from the title to the artist or author’s name, not forgetting the ratings and other details appearing in the search results.
Similarly, certain sensitive content can be excluded from certain profiles. In the case of podcasts and audiobooks, the AI even analyses all spoken content. The aim is to make the recommendation process even faster and more efficient than it is today.
Ultimately, Spotify’s use of AI to organise its recommendations has several advantages. It can help users discover new and interesting content, and thus possibly consume more of it. It also aims to reduce the time spent on searches, improving the user experience on the platform.
Earlier this year, Spotify launched an AI tool specifically dedicated to music, a sort of smart DJ capable of creating personalised playlists.
Recently, Spotify also unveiled an AI-powered voice translation tool, capable of offering certain podcasts in several languages with a cloned version of the podcaster’s own voice. The idea is to overcome the language barrier so that this content can one day be listened to and enjoyed all over the world.
The aim is to soon be able to offer creators all over the world, and not just English speakers, the means to bring their content to a wider audience.
Currently, over 100 million people regularly listen to podcasts on Spotify, a figure that is sure to grow.