The use of artificial intelligence is becoming more and more widespread every year. This achievement of the human mind is actively used in the music industry. In particular, music streaming services are actively using artificial intelligence to create effective music recommendations for users, as well as many more options.
And the Swedish music streaming service Spotify is among the leaders in the use of artificial intelligence. It is not for nothing that Spotify has been the undisputed leader of the global music streaming market for many years in a row.
The new How Bad Is Your Spotify app uses artificial intelligence to gauge how good your taste in music is. With the help of AI, it will be possible to analyze your musical tastes and the percentage of one or another musical content and performers in the music listened to and listened to by users. The system will also give information about those performers whose number of auditions is simply beyond common sense.
Based on the data obtained, it will be possible to give clearer and more detailed musical recommendations for each user. Note that this feature, as you might expect initially, is only available to subscribers of the premium version.
Spotify has also filed a patent application for an AI model that can look for plagiarism in songs. According to the patent documentation, the tool is able to find borrowings at the level of notes, melody, chords and texts. This model will allow you to immediately check the content on the plagiarism detector. She was trained on songs already stored on the Spotify server.
The model looks for borrowings based on melodic similarity, sampling, or rhythm similarity. The graphical user interface (GUI) is able to highlight specific pieces of the work and pick up a list of corresponding tracks that could be the original source. Potentially, such software can even calculate the similarity value.
The company claims that the new tool will be useful for composers themselves, as it will allow them to check the composition for originality in almost real time. To prevent a song or melody from being weeded out as plagiarism, musicians will be able to immediately give references to certain compositions.
Such a system will work much more efficiently than manual selection. Now moderators have to listen to several thousand songs, and in many borrowings cannot be determined the first time. When implementing the model, they will only need to make the final decision on the composition.
The new system could heat up the music industry. Since note combinations are inherently limited, they are often repeated in different compositions, and this can provoke litigation between labels and artists.
A little earlier, Amazon Corporation received a patent for some developments, the main purpose of which is to protect music on the streaming service from piracy. The developers of the corporation have applied AI algorithms, in which personal information can be automatically added to user search queries, in a visible or hidden way.