All rights reserved: La Stampa/Worldcrunch – By Marco Tonelli
Spotify is constantly and continuously evolving. Take the agreement with Universal that grants exclusive access to its albums to Spotify Premium clients, or Spotify’s exploration of allowing songs to be listened to in hi-fi mode, with the same sound quality as CDs. The near future of the Swedish giant might also be written in the technologies developed by the startups it has acquired so far in 2017 — from the use of blockchain to attribute licensing rights of songs to the implementation of artificial intelligence in the system for searching recommended songs. There is also a new feature that recognizes music through smartphone microphones.
Artificial intelligence to recommend music
Spotify has long been betting (successfully) on customizing music recommendations. Launched in 2015 and 2016, the Discover Weekly and Release Radar playlists are followed by tens of millions of users, and are based on personal listening habits. The first comes out every Monday and provides a weekly selection of songs. The second is available on Friday and is composed only of new releases.
In light of this, the recent acquisition of the French startup Niland was hardly a surprise. The Paris-based company has developed a technology that uses artificial intelligence to recommend songs. Niland “shares Spotify’s passion for surfacing the right content to the right user at the right time,” the company said in a May 17 statement announcing the acquisition.
This means that in the future, the streaming service will be able to predict even more accurately what songs we are prone to like. All this thanks to machine-learning and the ability to use collected information to anticipate possible user choices.
At the same time, the system that allows the search and selection of songs may also change thanks to the technology developed by Mighty TV. Acquired in late March, the New York-based company has created an application that allows, with an interface that closely resembles Tinder, to choose a favorite video clip between two suggested options. In addition to integrating the startup, Spotify has hired Mighty TV’s founder Brian Adams as its vice president of technology.
Spotify as Shazam?
In addition to Tinder, Spotify is also taking inspiration from Shazam. In the coming months, we may see the release of a new feature that can use a smartphone microphone to recognize songs and simple tracks of audio. Indeed, the English startup Sonalytic (acquired by Spotify in early March) has developed a technology that can track a song by analyzing a single portion of a music clip. This allows you to identify songs by piecing together multiple audio tracks. Spotify has provided few details, but it may only be a matter of time before you can take advantage of this feature in new application updates.
An open database to recognize rights
On March 17, 2016, Spotify signed an agreement in the United States with the National Music Publishers Association to pay $20 million in compensation for not having properly attributed proceeds to the record labels and copyright holders on some streamed songs. The company had defended itself, stating that it did not have the information necessary to properly identify all of the license owners. But above all, Spotify was not yet able to set up an archive that could catalogue all the existing rights to songs and albums.
More than a year later, the blockchain network could solve the problem. Acquired last April, the New York startup Mediachain has developed a technology that protects, traces and identifies artwork on the Internet. This open and decentralized database allows for content to be catalogued and attributed to the legitimate owner. Mediachain has also created a search engine to find and identify rights holders. It also has released a virtual currency (using the same mechanism as bitcoin) to pay owners each time their work is displayed or played. Immediately after the acquisition, Mediachain announced that it will adapt its technology to the needs and specificities of Spotify.