Music production is constantly evolving. And these past couple of years things have really gotten crazy.
New technology is shaking up the way we create and consume music. From awesome plugins to super-specific algorithms, these tecnologies are expanding the frontiers for musicians across the board.
Personally, I believe that if we fail to incorporate these new techologies, we risk being replaced by folks who master AI assistance to fast track all their creative processes.
One of the most exciting areas of AI and Machine learning in music production is personalization. Imagine music created or altered specifically to match your mood or preferences.
Finding out where do us humans stand in this new wave is one of the challenges we’re about to face. Jobs can (and will be) displaced, lawsuits will fly around like crazy.
Whatever your moral ground regarding this tools is, there is people bound to use them for evil and for good in equal amounts. I think it’s important to be familiar with what is going on so you can choose what will work in your advantage instead of just being swept away.
Understanding the tools helps you be prepared.
Would you believe me if I told you the first attempts at AI in music production date back to the late 1950’s?
The earliest example of computer-composed music is the Illiac Suite for String Quartet (1957) by two Americans, the composer Lejaren Hiller and the mathematician Leonard Isaacson. It was a set of four experiments in which the computer was programmed to generate random sequences of notes based on certain rules and constraints. and was later interpreted by real human musicians.
In the years following these few first experiments, the advancements in thechonoly increased what was possible. In the late 1980’s, researchers at MIT developed the "Experiments in Musical Intelligence" system, which used early versions of what we now call AI to compose music in various styles. In the 2000s, companies such as Amper Music and Jukedeck began to offer AI-powered music composition services for commercial use.
Current state of AI in music
2022 was the year where everyone and their grandma was talking about artificial intelligence.
The rise in popularity explains why today we have many tools avaliable including some new ai-powered music generators and analyzers for mixing and mastering.
These tools can help producers automate some processes, although they still have a way to go before they can make the same things a human can.
Here are a few areas where AI is being developed and used in music production.
- Composing: Companies like AIVA and Amper Music provide composition of “original” melodies based on style or trained on a specific artist and output a song with the required caracteristics.
- Mixing and mastering: AI powered mixing tools such as Landr can analyze and optimize audio tracks using machine learning to analyze the audio data of a track and make adjustments to the levels, EQ, and other aspects of the mix.
- Discovery: AI and ML can be used to understand the audience’s preferences and create personalized recommendations. Spotify has been doing this for a few years.
The future of AI in music production is vast and exciting.
Let’s talk about the good things first:
Ai powered music could be used to create personalized music just for you and whatever mood you are in. This could involve using machine learning to analyze a listener's music preferences and listening habits, and then generating music that is tailored to their specific tastes.
Discovering music you like could be easier than ever with alorithms trained with your preferences and specific tastes. As well as making modifications or adaptations of music to match your taste and mood.
And finally, the most exciting application of AI tools to me is music education. AI and ML could provide personalized feedback to students instantly to learn and train faster. Additionally, AI-powered music generators could be used to generate personalized practice exercises and exercises that would help students to improve where their skills are lacking the most.
Now here’s the dark side.
As any tool ever invented, people are going to use it both for good and evil.
In the future knowing how much human involvement was a part of the creation of music will be a big problem. As well as job displacemets for tasks that can be automated, algorithms controlling what we listen to, and marketers having even more information about how to sell to us fueled by data-driven insights.
As AI generators become more advanced, this also raises ethical and legal questions about copyright and participation in the making of tracks.
Regarding creativity, this could go both ways. On one side, getting ideas will be 1000% easier. On the other one, AI could lead to a lack of originality in the music produced, as music generated by AI is based on pre-existing models and data, it could result in everything sounding the same.
AI has been on my mind non stop lately.
What do you think? Are we done?
Some people embrace the change, like the University of Rochester who invested 1.8 millions into building a new echosystem that involves AI in the music careers. Others are fiercly against it.