Rage Against The Machine: AI and The Music Industry

With major advancements in Artificial Intelligence (AI) technology, its potential and effect on various spaces is becoming evident. Music not an exception to this trend as technology has entered the realm of song consumption in subtle ways over the past decade. We live in an era of hyper-personalisation, in which algorithms gathering user data are constantly improved upon to generate playlists that are tailored to each listener’s tastes. Spotify’s Discover Weekly, Apple Music’s My New Music Mix and Deezer’s Flow Mix are prime examples of the power of algorithmically generated playlists, as the popularity of streaming such selections is ever increasing. Google Play goes one step further by aiming to use location, activity and weather data input to suggest appropriate tracks for the user’s mood. As computers are increasingly able to complete tasks that were previously believed to require human judgment and input, the question arises of whether we are moving towards a future of computer generated original music. Can song creation ever become a fully automated process facilitated by AI? The technology to achieve this already exists, and the business implications of even a simple score produced by AI are far reaching. Platforms such as Jukedeck and Amper allow users to generate a custom, royalty free, easy to edit track within seconds. [1] That is not to say that such tracks are of higher quality to those produced by humans. They may be described as lacking ‘soul’: an elusive, unmeasurable quality which lays at the heart of an art piece. But the reality is that AI generated tracks do not need to compete. People working under significant financial constraints will not seek the highest quality soundtrack to be the backdrop of a short film project or a YouTube video. The market for AI produced tracks will thrive at places where terms such as ‘quality’ and ‘soul’ are irrelevant and practicality is of the essence. It may not be the finest produced art, but it will be ‘good enough’ for advertising, TV, social media and other projects where the originality and quality music is not the primary concern. [2] AI produced music means quick turnover, no royalty payment, no need for recording equipment, contracts with individual artists. The terrifying reality is that if AI music companies control this market space, the earning potential or creators and songwriters will receive a heavy blow. Human musicians will be pushed to the margins due to limited availability of projects, having to lower the price of their work to be able to compete with the AI track licensing fees. Humans do not seem alarmed by this prospect. In a study conducted by MusicStats.org, 45 seconds of a fully AI produced song was played to 50 millennials. [3] One group was given a fictitious emotional backstory for the artist and track, whereas the second group was informed that the song was created by an AI algorithm. Interestingly, when measuring the perception of the song by both groups no statistical differences were observed. Knowing that the song was generated by AI rather than produced by a human did not alter the values of ‘Purchase intention’ and ‘Satisfaction Towards the Performance’. Ultimately, AI has the potential to reinvent the way we create and consume music. While fear is a natural instinctive reaction to when the homeostasis of an ecosystem is altered, it is not always the appropriate response. We can choose to look at these innovations with excitement, as an opportunity to evolve and expand our creative talent within this emerging dynamic.

Image placeholder

Tim Hegarty

Tim was signed to Warner Chappell and wrote a Top Ten for D:ream and cuts for Gloria Gaynor and World's Apart. He also was signed as an artist and his band Baby June had a number of dance hits back in the 90's. He formed one of the wordls first online music portals called Songsforsale and now is building Evershare.io a Decentralised Digital Assets Exchange where fans can buy a share of a song.

Get Started

Find out more about Evershare