It is fascinating how technology has fuelled the growth of the music industry at each and every juncture in the last 40 years. From the nostalgic days of vinyl records and cassettes to the times of CDs and MP3s and the current digital disruptions through music streaming platforms, the music industry has benefitted from rapid technological advancements.
Deeper mobile penetrations, cheaper data and the emergence of Artificial Intelligence (AI) and Machine Learning (ML) have created a great ecosystem for streaming players to thrive and grow. Today several players dominate the music streaming industry and are leveraging advanced technology to provide listeners with an almost immeasurable collection of songs from artists all around the world. This bodes well for the music industry, that for long has suffered due to piracy and has witnessed plummeting revenues. According to the Recording Industry Association of America® (RIAA), the music industry at its peak was estimated around $21.5 billion in 2000 but soon hit the decline that lasted for 15 years. But the mass adoption of smartphones and the rise of music streaming apps since 2005 have since lifted hopes for the industry.
RIAA estimates that for the first time in this millennium, the music industry has posted an increase in revenues in two consecutive years and possibly in 2018 as well. From revenues to user behavior, the music streaming industry, in more ways that one, are changing the way we used to listen and enjoy music. Gone are the days when the only way to listen to music on one’s phone was to painstakingly download and save the individual songs on it. At the swipe of a finger, music lovers now have access to a wide range and genre of music. In a highly competitive marketplace, the key brand differentiator would be how these music streaming platforms make the best use of g algorithm-based AI engines and machine learning, to create a seamless experience for the users. Learning the specific tastes and preferences of every individual, streaming players are now working to build a better, more intuitive playlists. These algorithms take into consideration the music listening history of the user, analyzing genres, tonality, chord progression, length, pitch, tempo, vocal styles, instrumentals used, and more, to suggest similar songs, albums, and artists that he or she might like. As a result, users are regularly presented with personalised suggestions and playlists full of new music, keeping in line with their specific tastes, thereby creating an endless market for music consumption.
However, these music streaming services are still at a nascent stage when it comes to leveraging technology to improve their offerings. While most of the major brands use machine learning, which analyses the songs, artists, and albums a user plays over time, to learn what appeals to him, others also use recommendations from actual musicians and music editors to create daily playlists and suggest similar tracks. The most successful practice, however, is a mix of both, with human editors analysing the data gathered by the AI engines to further supplement the algorithms. This method is used by a few of the current market dominators, but the application of AI in music still has a long way to go.
While music streaming brands have come up with innovative value added services like voice assistant, automated one-touch personalised playlists, song mixing, and more, the real power of AI lies beyond such features. Extensive research is currently being carried out on the potential of using state-of-the-art AI to analyse music, not just using metadata, but by analysing the actual song itself, so as to gain a deeper understanding of the melodic medium. This presents an incredibly boundless prospect for its application in the music industry, and the domain, as a whole.
Creating music using AI is now the next step in this transformational journey, although the use of AI in music creation is definitely not something new. Several musicians have, over the years, designed and created AI programmes to mix and master songs recorded in the studio, or even create compositions and lyrics based on inputs. For instance, multi-Grammy award winning legendary musician David Bowie collaborated with Ty Roberts, technological innovator and founder of Gracenote, to create a tool known as the Verbalizer. Bowie was able to input up to 25 sentences and word groups into the tool to create potentially significant lyrical combinations.
Music streaming services are currently leading the market in this domain of AI adoption, constantly experimenting with various applications for this rapidly transforming futuristic technology. Personalised music recommendations have not only been a boon for consumers, but for artists as well. Serving as a platform for countless budding musicians to showcase their work and present it to the world, it allows many musicians to have a direct audience, even before becoming a breakout star with mass acclaim. As a result, users have often discovered talented musicians on streaming platforms, who have later gone on to win critical acclaim, signifying the creation of a symbiotic avenue for growth in the industry, both for the services and the artists.
This presents an extremely optimistic prospect for the future of AI in the music streaming industry, as AI driven data and marketing can open up entirely new strategies for brands. In the years to come, the vast amount of user data that is being generated and analysed will help businesses enhance their marketing strategies, through deeper insights into user consumption. This, in turn, will help them create a perfect harmony between musicians, consumers, and the music industry, while simultaneously driving profits and innovation.
Prashan Agarwal is CEO at Gaana.