ML in Music

We as a whole think about the marvels of AI and ML, and the disruption it has brought in the entertainment industry. Machine learning algorithms are changing the way we tune in to music.

On the off chance that there’s one thing, I can’t survive without, isn’t my telephone, workstation or my ride, It’s music and losing all sense of direction in it!!

On the other hand, another thing that gives me creeps is Machine Learning. ML gives me the chills about the capabilities of the machine which was built for helping humans.

Machine Learning, on the other hand, may give intense time to songwriters, lyricist and musicians because the algorithms are definitely getting better and better at music generation. Composing music using AI or ML isn’t something new around the bend and will be there as startups are flooding in this area of the industry.


The famous Swedish startup which almost started a decade ago has the most advanced mood-sensing algorithm.

The Weekly discover playlist made by the algorithm is worth every penny paid for…

Last year Spotify celebrated it’s first Spotify ML day on July 9th in Stockholm, Sweden. The general purpose of the meetup sorted out was to portray the procedure behind such AI assisted compositions and point to current directions of work.

Another spotlight stealer in this industry was

Hello World — Skygge (First music album composed by AI + artists)

In 1958, AI was used to compose Bach-like chorales (the Illiac suite). Since, huge progress has been made in AI technologies, including in machine-learning, combinatorial optimization, statistical inference, and related areas.

This album was created using artificial intelligence technology Flow-Machines. This album is the story of musicians who met in the lab, took control of the AI to compose the music they had in their souls.

There are many ventures which aim at developing AI that can compose professional-quality music!

How good is the music created by using ML??

Recently, there are many open source projects including huge players like Google who are working on this stratum of technology to create music by using ML algorithms. All these projects aim at creating music with minimal user input and further closing the bridge of music composition between human composers and machines.

In fact, the algorithms have improved so much over the past year that it is really difficult to differentiate between the music composed by humans and machine learning algorithms.

Magenta by Google

Magenta was started by some researchers and engineers from the Google Brain team, but many others have contributed significantly to the project. They use TensorFlow and release their models and tools in open source on GitHub.


A result of hackathon by Ji-Sung Kim is used to generate jazz music and is open source. It uses deep learning and AI to produce compelling work, which can be considered as deeply human.


A generative model for raw audio, it was developed by researchers at Google’s DeepMind. It is not open source but a few samples are available here.


The goal of FlowMachines is to research and develop AI systems which are able to generate music autonomously or in collaboration with human artists. The website notes that the music can come from individual composers, for example, Bach or The Beatles, or a set of different artists who are using the system.


This AI for classical music is a project by Feynman Liang of Cambridge University. It is presented in the form of a challenge on the website which provides small samples of music either extracted from Bach’s own work or generated by BachBot.

Most of them are open source!!!


Based on the facts, It is easier to say that the music generated by ML algorithms is quite close and indifferentiable from human composers. As listed above, the samples generated by different projects are quite similar to songs composed by humans.

However, there are certain challenges with these tools.

Most of the music generated by these models consist only a single stream of notes. Nonetheless, times aren’t far ahead when researchers would minimize the human effort and input as much as possible. A perfect ML composer isn’t a distant reality and is in the nearby future.

read original article at——artificial_intelligence-5