MUST READ: 3 ways to use AI in music and none is really acceptable

AI and music

In the past few years, an enormous discussion has started about the use of AI. This article claims that almost 40% of music released in July 2026 was created with AI worldwide. That is a gigantic amount of music that’s been created with the new AI tools. I think it’s necessary to say that this relates to generative AI, as AI itself is a very broad subject. I will dissect the use of AI in music into three distinct ways, and none of which is really acceptable. These ways are composing with AI, processing with AI, and getting feedback from AI.

Composing with AI

With this, I mean AI has a role in creating the song. Maybe AI has made the song entirely, or only certain sounds, or that AI created the structure or foundation. In any case, there are many problems with this. First, the ‘creator’ is not creating anything. They are a sort of overseeing the process but not exactly involved in making anything. Second, AI is taking the works from others without their consent and spits out a song for you that differs vastly from composing something yourself. Third, the resources it takes to have AI work are disproportionate to the output. With this, I mean that the costs of AI do not outweigh the benefits in terms of the quality of the product, meaning to society, and effectiveness. Fourth and last, it doesn’t make you a better artist. I would even say that it makes you worse. It has been proven that critical thinking skills deteriorate with the use of AI, which is an absolutely necessary part of creating art. When I relate this to the other blog posts I posted before, like ‘Why I create, and you should too’ or ‘What is art? Five ways to look at the meaning of art’, this practice goes right against the elements that make ‘art’ art. The meaning behind the piece and the craft is lost, so the receiver, by definition, has nothing to relate to.

Processing with AI

Basically, this is the same as the point above. What I mean with processing with AI is, for example, using a mastering service that masters your track with AI. What people who use these kinds of services don’t realize, is that mastering (or processing anything) is an art and craft in itself. A master engineer does have a distinct way of working, knowledge, preferences, basically all the things that you have as a musician, too. By outsourcing this to AI, you essentially have the same four problems as composing with AI: you do nothing, you take work from others, a waste of resources, no art value.

Feedback from AI

Asking feedback from AI is the only thing I see as having some value in the creation process. As in this case, you, as an artist, have the availability to show critical thinking skills. The AI doesn’t take other people’s work and plagiarizes it. It compares your work to theirs and gives you suggestions to get closer to the quality of the work you want to approach. You, as an artist, then have to evaluate if that feedback is valid or not, and you still have to do the work. This saves your artistic integrity. The only thing that’s a problem is the resources used to provide this feedback. In terms of efficiency here, I would suggest asking a friend or two to provide feedback on your song and/or lyrics. Using it one time or two times is not a problem, of course, but it gets debatable after a while. Overall, I wouldn’t rely on AI feedback for your songs.

These are essentially the way to use AI in music, and none of it is actually efficient or effective. Stay away from AI in music or art.