In the second of Synchtank’s two-part analysis, Ben Gilbert takes a deep dive into the subject of AI to explore the increasingly central role technology has to play in the influx of information surrounding digital music. Click here for part one, where we consider how these tools are reshaping the craft of songwriting.
As decreed by unwritten hip-hop lore, Jay-Z has a myriad of fake monikers, guises to mask the humdrum reality of his birthname, Shawn Carter. At various moments across a landmark career, these have included Young Hov, Jiggaman, J-Hova, Jazzy and, more obscurely, President Carter, Lucky Lefty and Iceberg Slim. For music fans, such an elusive cast of characters creates the canvas on which a music star paints the narrative of their art.
But for music professionals, dealing with the explosion of information prompted by the rise of digital technologies, even Carter’s simple negligence towards the name bestowed upon him by mother Gloria presents a complication, a quandary that the industry is still to fully tame. “How do we know with certainty that Jay-Z, Jaÿ-Z, Jay Z and JAY:Z are all in fact Shawn Carter? And for an artist who is not as famous as he is, will there be an effort to reconcile those differing names?” questioned a Billboard op-ed in July.
“It’s not only the act of music creation that is set to be revolutionised by the exponential power of automation and machine learning but the associated management of data that has and will accompany it.”
This is an issue that artificial intelligence intends to solve. Because it’s not only the act of music creation that is set to be revolutionised by the exponential power of automation and machine learning but the associated management of data that has and will accompany it. At the centre of this process is the Music Modernization Act, which passed in 2018 and introduced measures intended to improve the distribution of royalty payments. These included the Mechanical Licensing Collective (MLC), a non-profit designed to build a public database of ownership information for musical works.
Unclaimed music royalties amount to $2b worldwide
The imperative here is underlined by the gargantuan numbers that govern this area of global culture. According to 2014 census data, copyright-intensive industries account for 5.5% of total US GDP, with recent estimates suggesting unclaimed music royalties amount to $2b worldwide. AI music startup Boomy, where thousands of original songs have already been created, predicts the music business should get ready for a “head-spinning” influx of new material comparable with photo uploads on Instagram and Facebook.
“The industry needs to start preparing for a world of 10m original songs per day,”
– Alex Mitchell, Founder and CEO of Boomy
“The industry needs to start preparing for a world of 10m original songs per day,” founder and CEO Alex Mitchell told Synchtank. One of the platforms designed to cope with this eventuality is Exactuals, a payments company that works with labels, publishers and creatives to streamline metadata processes to ensure the right people are compensated. One of its core products is RAI, an open API that uses machine learning to resolve conflicts within this welter of data. Chris McMurtry, Head of Music Product, believes that to truly understand the role of AI, pragmatism rather than suspicion is fundamental.
“The concern, regardless of the field, is almost always around the question ‘is AI going to replace such and such?’” he explained. “The right questions are those that keep everything moving forward – ‘how can I use this technology to assist with…..?’; ‘is there an element of my craft that I hate that I can delegate or pass off to the technology?’; “is there something about the technology that helps makes me better at what I love to do or what I’m good at? The technology is a tool. These are the types of questions we’re asking in our pursuit to get rights-holders paid accurately and more efficiently.
“Our goal is attribution and payment. That is why we hire not just the best developers on the planet, but also the best musicologists and publishing administrators or what we refer to as subject matter experts who know the intricacies of metadata better than anyone.”
– Chris McMurtry, Head of Music Product at Exactuals
“Our goal is attribution and payment. That is why we hire not just the best developers on the planet, but also the best musicologists and publishing administrators or what we refer to as subject matter experts who know the intricacies of metadata better than anyone. They train the algorithm, making it better, and the algorithm in turn reciprocates, making them better – making us better,” he commented.
“The most important problems in the history of the music industry”
Within this process of automation, McMurtry is quick to point out the enduring role of humans and the impact of his “amazing team” on everything they do. “When you think about great innovations, they are often the result of two worlds colliding. We have been fortunate to target the mastery of two unique and seemingly different domains (in our case, musicology and payments). When a team has such an opportunity to marry two worlds and then use that unique combination to focus on serving a customer with a real need (in our case the rights-holders) that is a recipe for innovation,” he said.
McMurty believes Exactuals and other emerging technology firms have the potential to fix “the most important problems in the history of the music industry”. For example, the aforementioned royalties surplus. How can RAI resolve this? “Right now the technology has to be applied in the middle of the supply chain, because they are the ones feeling the brunt of the problem and are being held responsible. But as the technology continues to move toward the poles, where the flow of data begins (creators) and its destination (consumers), the ability to account for the data will accelerate right alongside the increase in production,” he explained.
“Right now the technology has to be applied in the middle of the supply chain, because they are the ones feeling the brunt of the problem and are being held responsible.”
– Chris McMurtry, Head of Music Product at Exactuals
This is an unpredictable, dynamic and futuristic ecosystem. Indeed, we are living in an era when music technologists take inspiration from the likes of Uber, Google and Tesla in roadmapping infrastructures and devising the next phase of user consumption and future iterations of tools like playlists. In 2018, Spotify revealed their “self-driving music” strategy, which harnesses data analysis “to mould the discovery experience autonomously to an individual user’s tastes and behaviour, across multiple devices and environments.”
The essential collaborative exchange between human and machine
Music Story, like Exactuals, operates in the field of metadata, offering “fine human tuning” across an expansive portfolio of songs totalling in the region of 16m and more than 200,000 artists globally. Working with the likes of MTV and Deezer to improve digital music services and increase audience engagement, CEO Jean-Luc Biaulet believes there remains an essential collaborative exchange between human and machine.
“Of course, humans have to be helped by machines to produce quality metadata and spread them on a large coverage at the same time, through assisted curation for example. But we have experienced that, to reach a high level of quality of metadata, which is mandatory to really enhance the user experience, a minimum of human touch is required.”
– Jean-Luc Biaulet, CEO of Music Story
Echoing the sentiments of McMurty, he told Synchtank: “Of course, humans have to be helped by machines to produce quality metadata and spread them on a large coverage at the same time, through assisted curation for example. But we have experienced that, to reach a high level of quality of metadata, which is mandatory to really enhance the user experience, a minimum of human touch is required.
“AI will be more and more capable of recommending accurate decisions but up to a point, important choices will have to be taken by people in charge,” Biaulet added, before concluding: “On the other hand, the ability to integrate AI in operational processes will certainly make a difference in the future and this may accentuate the current reshaping of our industry, already driven by technology.”
This is the second of Synchtank’s two-part analysis exploring the impact of AI on the music industry. Click here for part one.
1 comment
[…] to make and release creative works at scale (one AI music startup claims the industry should start preparing “for a world of 10m original songs per day”), look for this issue to heat up in a hurry – […]