Dream - The music streaming platform Spotify has carried out a massive layoff. The decision is said to be due to Spotify's desire to advance artificial intelligence (AI) on its platform. Spotify has carried out layoffs three times in a year. A total of 590 employees in January, 200 employees in June, and another 1,500 employees in December.

According to CNN, Spotify's investment in AI to increase margins for its podcasting and audio book division seems like a total overhaul in a strategy believed by Wall Street to be successful. "Spotify utilizes AI across its platforms, launching AI DJ, simulating traditional radio experiences in 50 additional markets, and launching AI Voice Translation for podcasts," said Justin Patterson, equity research analyst at KeyBanc Capital Markets, in a research note.

Spotify has joined other technology companies in making savings due to decreased demand in the pandemic era. Spotify also had to pay over US$1 billion for podcasting, most of which was used for deals with celebrities to create podcasts that never materialized and acquire podcast studios that were later closed.

In November, Spotify announced a partnership with Google Cloud to revamp the way this platform recommends audiobooks and podcasts by using one of Google Cloud's language models, Vertex AI Search.

Spotify introduced "AI DJ" in February and began using the voice translation tool "Whisper" from OpenAI to translate certain episodes of English-language podcasts into Spanish, French, and German. A representative from Spotify stated in an email to CNN that the company plans to further develop this technology in the future while awaiting feedback from creators and listeners.

Spotify has personalized its experience for users for about a decade. Spotify is able to add that personal touch after acquiring the music analytics company, The Echo Nest Corp, in 2014, to combine machine learning and natural language processing. Spotify's technology builds a database of songs and artists by recognizing musical tones and tempo and connecting artists' works in the same cultural context.

Metadata such as release date and metrics such as volume, duration, and how likely a song can make someone dance also determine which songs are suitable for users' preferences. From here, playlists such as "Daily Mix" and "Discover Weekly", "Time Capsules", to "On Repeat" are born.