Oren Sharon & Arielle Spivac
What's Going On In The World Of AI And Music Today
For the past few decades or so, the extent to which we can use artificial intelligence has been rapidly increasing, with new high-end technology being developed daily.
The music industry in particular has seen the emergence of innovative tools that can benefit listeners and artists alike.
In this article, we will delve into the ways in which artificial intelligence technology is revolutionizing the music industry.
AI music composers
AI music composition works by training machine learning algorithms on large datasets of existing music, in order to create music that sound similar to the ones in the dataset. The user chooses the mood, genre, and tempo, and the AI generates a brand new track. The main players in this field today are:
Lack of musical depth, creativity, and complexity that human composers bring to their work.
Lacking originality, in some cases, the AI technology will generate new music that sounds similar to a previous generated track.
Audio quality isn’t high enough and sounds a bit synthetic.
Copyright issues: Since it’s a new field and the generation of the music is based on already existing tracks, the question of who owns the rights to AI-generated music is largely unresolved.
Background music and beats: This could be a solution for someone who looks to reduce licensing fees and use it for background music for a video or as music beats to rap on.
AI Music Production
This use of AI is all about helping music creators, rather than generating music from scratch.
Some AI-powered tools can benefit artists throughout their creative processes. AI platforms can identify key changes and chord progressions and suggest melodies based on the musician’s preferences.
Many musicians have been able to beat their creative blocks and take their music to places they never would have reached had it not been for AI.
The main players in this field today are
New Ideas - artists can overcome creative blocks with the help of AI.
Lack of creativity and originality in music created by human composers
Legal Issues: There may be legal issues surrounding the ownership and copyright of AI-generated music.
AI Sound Changer
This is the most impressive feature we’ve encountered; sound changers in real time could be a game changer in the production process. In collaboration with Google, Magenta can analyze a melody and replace it with a different instrument. For example, you can upload a vocal and change it to be played by a violin or a trumpet. The results are overwhelming, and this could be a game changer for music producers. Think of it this way: instead of finding a trumpet instrumentalist or violinist, the producer can generate a sound similar to an instrument based on a melody or another instrument. Yes, there are good VSTs today, but none of them can analyze a melody and create a similar melody from a different instrument.
It also comes in a plug-in version for your DAW.
We haven’t found any disadvantages to using this technology, it’s pretty awesome.
AI-powered mastering tools can analyze audio files and automatically adjust levels, EQ, and other parameters to create a balanced and polished sound.
This can save time and money for producers and artists and improve the quality of the final mix by automating many of the tedious and time-consuming aspects of the mixing and mastering processes.
There are already great mastering plugins for DAW, but they all require previous knowledge of sound and adjustments.
AI-based mastering companies on the Web are learning the uploaded audio and generating the right level, EQ, and compression adjusted for the track.
Companies working in the field today are:
Sound can be over processed in the audio. This can result in a final mix that sounds over-compressed or overly processed, which can be detrimental to the overall sound quality.
Many AI mastering tools offer a set of pre-defined settings or parameters that cannot be customized beyond a certain point. This means that the user may not have as much control over the final sound as they would with a human engineer.
Fast and cheap mastering treatment for your audio—this could come handy to upcoming artists at the beginning of their musical careers.
AI music separation
Another brilliant feature we’ve encountered is AI music separation. Users can upload their audio or a famous song that they like and separate the instrumentals from the vocals. This could come in handy if you are a singer who wants an instrumental version of a song you like to sing on, or if you are a rapper looking to rap on a famous song.
Players on the field:
The results are not perfect and are a bit synthetic.
Track separation is pretty impressive and can be valuable for singers who want an instrumental version of a famous song or track to rap on.
The internet is flooded with AI lyrics generators, from Jarvis to Rytr, These Lyrics Do Not Exist, and more. You can enter keywords, feelings, or moods and ask the AI to generate lyrics for you. These are tools you can use for first aid if you are stuck without lyric inspiration, but we wouldn’t recommend basing your whole song on them.
Players in the field are
Tools can help overcome writer's block by suggesting a word or a sentence.
In most cases, results are simple, a bit predictable, or even childish in some cases.
Additional usage of AI in music
AI is heavily embedded in streaming services such as Apple music or Spotify, it is embedded in their algorithm and it is used to generate personalized playlists such as Discover Weekly, Release Radar, Daily Mix and more.
Artificial intelligence is additionally making waves in the live music scene. Using techniques such as virtual and augmented reality alongside live performances creates a more immersive experience for concertgoers, helping them connect with the performer like never before. This can include AI-powered visualizations, responsive lighting, and interactive live shows.
AI can also have a significant positive impact on the marketing and distribution of music. Thanks to AI marketing platforms, companies can analyze patterns in data on audience preferences in order to understand how best to market and distribute their shows. This allows for more efficient marketing, which can lead to higher attendance and revenue.
Another field we are working on at One Submit is AI music submission. which is all about finding a playlist, blog, or YouTube channel based on the characteristics of a song. With a collaboration with MyPart, we’re about to introduce it in the next few months.
You may already think, with everything you’ve just read, that AI is hugely revolutionary in the music world, but the truth is, we’ve only barely scratched the surface.
From analyzing the tools available today, we can say that AI in the music industry could be very effective when there’s no creative process involved, for example in mastering, sound replacement, or sound separation, but when it comes to creativity, we are still far away from replacing a human composer or songwriter.
Should we fear AI?
While AI has been hugely beneficial, it wouldn’t feel right not to mention the fear that comes with it. It has been of concern that AI’s ability to perfectly create new pieces of music could decrease demand for human musicians in the future.
Furthermore, there have been conflicting discussions surrounding the potential loss of intellectual property rights for musicians whose music is used to create the AI’s algorithms.
To fight this, organizations such as the RIAA and BPI have demanded protection against AI music infringement.
The UK government was met with strong protest when it proposed that AI training be allowed to use creative works without permission. This clearly demonstrates the caution we should use AI with in the future in order to avoid the infringement of rights for artists whose music is ‘stolen’ by an AI platform.
A similar issue occurred in the late 90s when Napster was sued for reasons of copyright protection. This was because the music sharing platform had become so popular that it led to a decrease in the sales of artist merchandise such as vinyl records and CDs.
To avoid another crisis of this level, resulting in a demand decrease of musicians, AI companies must find a balance between dominating the music industry and simply aiding human musicians in their creation and marketing processes.
We must ensure in the future that this intelligence is used responsibly and ethically so that it does not completely swallow the music industry as we know it.
This article is brought to you by One Submit, a music promotion platform.