The Art and Science of Music Created By AI
February 18, 2019
Is your favorite music created by a machine? It just might be. Meet Artificial Intelligence – the newest artist to hit the music scene.
Siri and Alexa are the obvious faces of AI, but where else does Artificial Intelligence (AI) live?
Think about your personalized entertainment experiences, emails that auto-filter in your spam folder, and the ETA on rideshare apps. AI is everywhere in our day-to-day lives, including music.
While you may not find AI on the latest Adele or Coldplay track, it could be present in any beat-driven, techno-heavy song. In the past five years, AI has erupted as a collaboration tool to help artists and composers alike.
With how quickly AI has progressed in other industries, it's only a matter of time before AI takes off in the music realm.
Here’s everything you need to know about the art and science of music created by AI.
The Evolution of AI in the Music Industry
AI and technology are a well-established pair. Your recommended movies on demand? That’s the work of AI. Digital assistants like Alexa and Siri? Also AI. But, AI and music have even been in business together as early as the 1990’s.
It all started when David Bowie helped create the Verbasizer - an application that regenerated sentences to form lyrical songs. The basis of this technology is simple: scientists fed the machine a series of literature, and the machine transformed it to form unique lyrics.
The same concept can be seen in today’s technologies. In 2016, Sony Music unveiled the Flow Machine. Rather than feeding the machine literature, they trained it to learn Beatles songs. The result? The first machine-composed, Beatles-style song called “Daddy’s Car”.
The Flow Machine was just the beginning of AI software. Since then, programs like Logic, Jukedeck, and Amper have emerged to help musicians create musical patterns. More recently, Amper has helped artists like Taryn Southern to create the first AI-composed album titled, I AM AI.
AI music production looks something like this: users choose preferences (like BPM, instrumentation, mood, key) and the platform generates a tune accordingly. From there, users can adjust any aspect of the song, making it as much or as little as their own as they please. Artists may then choose to record sounds live or use machine-created sound to replicate instrumentation. When listening back to a track, AI’s presence is virtually unrecognizable.
The AI Impact
So, can AI completely replace human musical creativity? Not quite. Even the best technology has its restrictions. At the core of any strong musical piece lies emotion. Unfortunately for AI, emotion can’t be generated (yet). So far, AI has produced tech-heavy, moody music like in the songs Break Free by Taryn Southern or Mr. Shadows by Sony. This may be the most significant drawback to AI in the music industry in addition to ambiguous copyright laws. AI in the music industry is still relatively new. As so, laws have yet to be established regarding authorship rights. We know that AI must be fed certain material to regenerate it into something new. So, do the authors and artists of the prior material receive credit for the new? It’s still unclear.
Luckily for artists, the obscure legal parameters and lack of musical emotion means they’re not out of a job - so long as they continue to build raw, relatable songs for which users will be willing to pay. But, that doesn’t mean AI can’t act as a tool for artists throughout the creative process. Most notably, AI can help artists reduce time and money spent on music creation. Be it a producer, composer, musician, or songwriter, AI can act as the whole team. What does this mean for the music industry? More affordable resources for more artists to contribute more music to the industry. Best of all, AI cuts production time by lending a hand when artists are suffering from writer’s block (or in this case, composer’s block).
The Future of AI Music
What lies ahead for AI-composed music? Perhaps more emotion, collaboration, and legal parameters. The truth is, only time will tell. What we do know for sure is AI is about to take the music industry by storm.