Is AI The New Mozart Of This Era? Future Of Music With AI
By Kamal Jacob
Every time we talk about artificial intelligence, we only tend to think of some particular fields because over time, we have built a notion that AI is all about transforming the technological sector. However, that is not completely true. Instead, the days are gone when AI’s capabilities were limited –today this sought after tech has moved past that phase and is gradually marking its presence in several spaces. The music industry is one of those sectors that AI has off lately invaded. The music industry has also witnessed tremendous transformations done by AI over the past couple of years. Whether you talk about mobile music streaming applications or AI-based music composition, the industry is booming with AI being a part of it.
How Music Apps Are Using AI To Recommend Music
Gone are those days, when we used to have an external storage card just to save music on our phone. Today, in this technology-driven era there is an application for every single thing, and music is no exception. Now you all must be wondering, what role AI plays in a mobile phone music app. Let me tell you all, it does play a vital role in the form of recommendation engine or recommender system.
The recommendation in music streaming applications has gained tremendous traction in recent years. For applications without a recommender system, it gets really tough to make a mark in the market. So, a recommendation engine is nothing but a type of information filtering system. Simply put, a recommendation system in a music app is a system that tries to understand the user based on the type of songs s/he listens and recommends songs that the user would like to hear. And it all this is possible because of these technologies like machine learning and artificial intelligence.
This is one aspect of AI’s role in the music domain. But today, AI has gone way far more than recommending songs. In this article, we will also look at some of the most amusing use cases of this sought after in the music industry.
A New Musician In The Industry
Even after seeing AI being used in multiple verticals, many still find it hard to accept the idea of using AI to produce music. But AI seems to be less bothered about what others think; today, there is music-making AI software and they are so advanced that they can and is being used by producers to help in the creative process of music composition.
In 2017, American singer-songwriter Taryn Southern released a song — a big, moody ballad song Break Free. And as soon as the song was out, it became the talk of the town. The song was produced by an AI. To the surprise, it’s not just one song, Taryn Southern created an entire music album, “I AM AI” and it was the first LP to be entirely composed and produced using AI.
But how did Taryn Southern did this? She used an open source AI platform called Amper Music to create the stems of “Break Free.” Developed by Amper, a New York City-based music software company that uses artificial intelligence to create music, Amper music has the ability to create music across a range of genres and moods. It is designed in such a way that it gives users creative control regardless of expertise.
So, for each track of “I AM AI”, Southern plugged in the genre, the instruments she wants to use, and beats per minute. And in return, Amper churned out disjointed verses that are rearranged into a song and layered beneath her vocals.
Talking about more features of Amper music, the tool is constructed from scratch and it has a proprietary library that has many individual samples. And what keeps Amper ahead of the curve is the fact that even after being an AI-powered music making tool, it doesn’t sound synthetic, and Taryn Southern’s songs are proof to it.
Furthermore, it is not just Amper Music that is disrupting the music industry, tools are Jukedeck, Nsynth Super, IBM Watson Beat, etc. are also playing a vital role.
When AI Goes Death Metal
Have you heard about Dadabots and Relentless Doppelganger? No? Let me tell you —Dadabots is a band on YouTube that has been live streaming a 24/7 death metal called Relentless Doppelganger, and this is what their ‘About’ reads on YouTube, “Neural network generating technical death metal, via live stream 24/7 to infinity.”
You got it completely right; it is actually a neural network that is generating the music, even though it might sound human-made to the untrained ear. The brainchild of two music technologist CJ Carr and Zack Zukowski, this Death Metal version has become a talk of the town. Now, let’s go for the elephant in the room — how Dadabots is able to do this? This is all possible because of SampleRNN, an unconditional end-to-end neural audio generation model.
To generate this live stream to infinity death metal song, the deep learning models are trained on samples of a real death metal band called Archspire. And the collected snippets of audio or raw audio are fed into the model using SampleRNN. And till date, the music technology duo has created six albums this way.
“This early example of neural synthesis is a proof-of-concept for how machine learning can drive new types of music software. Creating music can be as simple as specifying a set of musical influences on which model trains,” stated the Carr and Zukowski. “We demonstrate a method for generating albums that imitate bands in experimental music genres previously unrealized by traditional synthesis techniques (e.g. additive, subtractive, FM, granular, concatenative).”
While many are still skeptical about the capabilities of technologies like artificial intelligence, it is on a spree of doing remarkable things in this world. AI over the years have proved that its use case can never be limited — whatever job a human is doing, soon AI will also join the humans in every domain.
It doesn’t matter anymore, how much people criticize AI. There was a time when Auto-tune and things like digital workstations in the music industry were not much popular, but today if you look at it again, they have become mainstream in music creation.
Talking about AI disrupting the music industry, we have seen “Break Free” (created using Amper music) and Relentless Doppelganger (created using the neural network) and as far as capability is concerned, these two examples are enough. However, there are many other uses cases of AI in music is available. And the rate at which this sought after the technology is transforming industry after industry, it seems the days are not so far when AI-powered music making machines will become a must-have or must-use for artists.