What Was AI Made For? AI’s Place in the Music Industry
Joseph Perla, the Founder & CEO of Hangout FM by Turntable Labs, cuts through the clutter to offer a music tech insider’s look at the good and the bad of music AI.
by Joseph Perla, Founder & CEO, Hangout FM by Turntable Labs
All of a sudden, artificial intelligence (AI) is all anybody wants to talk about in the music industry. It makes sense – AI has the potential to be a game-changing tool. But what does it do best?
Reflecting on my time at Princeton University, where AI was my major research focus, I’ve come to understand its capabilities. AI excels in pattern recognition, making it ideal for solving various music industry challenges. However, not all uses of AI are equally effective.
1. Fraud Detection
Starting on a positive note, AI’s ability to find patterns is perfect for identifying and stopping streaming fraud, currently one of the major issues impacting the music industry. Fraudulent streams could account for 5% to 10% of all plays on streaming services, according to a recent Music Ally article, which also quotes Beatdapp Co-CEO Andrew Batey saying that the number is often between 20% and 60% of all streams for some distributors month to month. However, streaming fraud can be identified via several tell-tale signs – for example, a user who streams music 24 hours a day or a large number of accounts that all stream only a small number of the same songs. These are patterns that AI can detect and eliminate, weeding out bots and hacked accounts and preventing scammers from profiting off the work of others.
2. Production Tools
AI’s impact is most visible in production tools. Ghostwriter’s “Heart on My Sleeve” track, featuring AI-generated vocals in the style of Drake and The Weeknd, got a huge reaction, with many fretting that AI could put musicians out of work. But here’s what you need to remember: AI is not creative, it can only recognize patterns. Any music generated completely by AI will never be innovative, just a rehash of what has come before. For that reason, humans will always be at the forefront of music creativity, as they are the only ones who can push it forward into uncharted territory.
However, AI can be an incredible tool for musicians. WAVE AI makes several tools that can help musicians with writer’s block generate lyrics and melodies that serve as the starting point of inspiration, allowing them to meet the pace of the moment without giving up quality. In addition, forward-thinking musicians like Grimes and Holly Herndon have created AI-generated versions of their voices that artists can use in their tracks in exchange for royalty fees. The Beatles even managed to release one final song after AI was able to remove the background noise from John Lennon’s demo of “Now and Then.” The possibilities are endless and, more importantly, are a reason for excitement, not fear.
3. Music Discovery
Here’s where AI gets into trouble. Given its affinity for pattern recognition, there is a temptation to outsource music discovery to AI. This might work in the short term, giving users a stream of songs that sound similar, match a certain mood, or share certain qualities, like BPM (beats per minute) for example. However, this does not account for the uniquely human yearning for something new and different. After a while, someone who is being consistently fed rock songs will want to listen to something else, and human curation is the best way to find your way forward.
Say you’re into hard rock and are looking for something country. AI can help you find something in the genre that has louder guitars than the usual country song, but humans can recommend an artist that captures the spirit of what you’re looking for if not the sound – like Johnny Cash. What we need is more social-powered music discovery, not AI-generated playlists. It’s something I’m working on with Hangout FM, which will be launching in Summer 2024.
In the end, if the music industry can embrace AI as the tool that it is, we stand to make some significant advances in fraud detection and music production. Just remember, AI can never replace human spirit, and it is important to lean on other humans when it comes to innovation and discovery.
Joseph Perla is an entrepreneur and music tech expert currently serving as the Founder and CEO of Hangout FM by Turntable Labs, Inc., a social platform connecting users through the discovery and curation of music. He got his start during his years attending Princeton University, when he launched several social networks tailored to his interests, including Zandigo in 2005 for college students and Labmeeting in 2007 for scientists. He was also, in 2011, the founder behind Turntable, which introduced the idea of social listening. Perla also spent time working at Facebook (now Meta), where he led product and engineering for Facebook’s iPhone News Feed; co-founded Terminal.com, where he helped to secure over $4 million in funding and which was acquired by Udacity in 2017; and spearheaded Passenger and Driver Growth at Lyft during its most significant expansion phase. After raising over $20K from fans in a crowdfunding campaign, he brought ideas from Turntable back to create Hangout FM. Negotiations are ongoing with all three major labels and Merlin to build a dedicated Hangout FM music library, as well as music publishers, PROs, and other rights-holder organizations. A true polymath, outside of his professional endeavors, Perla is a certified private pilot, sailing captain, and skydiver. As an amateur violinist, he received classical training in Vienna.