Music is the language of the soul and an expression of our emotions. It has always inspired people throughout the world and creates a universal connection between us. But, with technology advancing and artificial intelligence (AI) coming into use in the music industry, the question arises: can a machine ever replace the ingenious and unpredictable songwriting of an artist with a will of their own? In this article we will look at how AI can support musicians and why the creativity and spontaneity of free will remain irreplaceable.
The more predictable the structure of a music track, the less human input it requires
Artificial intelligence has brought significant progress to the music industry, especially when it comes to simple and recurring tasks such as generating drum beats. These beats follow clear musical rules and patterns. Frameworks such as beat and tempo are an area of application with a clear structure; something is either a 4/4 beat at 120 bpm or it isn't. Many music production tools already use machine learning algorithms to automatically create percussion foundations. This supposedly saves time and effort (if the producer can operate the AI confidently) and speeds up the production process. Examples of this are the programs "AIVA" (Artificial Intelligence Virtual Artist) and "Melodrive", both of which are able to generate realistic and varied drum beats.
However, creativity presupposes a (free) will and thus a personality
While AI is effective at simple tasks like generating drum beats, it can never replace a musician's creative genius. According to Immanuel Kant's philosophy, the ability to make free, spontaneous decisions is the core of every real artistic achievement. An artist, whom Kant calls a "genius," possesses both the technical knowledge and skills (chords, fingering techniques, strumming) and the ability to think creatively, detached and unhindered (creative songwriting) and to implement their ideas.
Ultimately, music is more than the stringing together of notes. Music is also the power of the performance of its author. Musicians as personalities are always part of their own songs.
AI can fall back on existing patterns and rules and reproduce them. But it is unable to think creatively and create anew or innovative ideas on its own. True artistic masterpieces therefore do not come from algorithms, but from the heads and hearts of musicians who have the free will to do so.
AI always creates music from very narrow, logical cause-and-effect chains. By definition, it generates what it has learned from existing rules and structures. But creativity is the unexpected, the surprising and the beauty of the effect without explanation.
Nevertheless, there can be areas of application in music apart from beats and rhythms.
Wherever two conditions are met, artificial intelligence can probably still support musicians and producers. Or at least speed up the arranging:
- Where there is an existing basis that artificial intelligence can build on, such as a reference or an unfinished song that "only" needs to be completed.
- Where there is a clear set of rules that the artificial intelligence can follow and a producer who can feed the AI the right instructions.
1. Application of AI in music: Mixing
One of the most notable achievements of AI in music production is the ability to mix songs. The mixing of music is a complex task that requires a lot of experience and know-how. It's about matching the different tracks of instruments and vocals, ensuring they complement each other in harmony and achieving the desired sound quality. With the development of AI technology, it has become possible for computer programs to perform this task in an automated and precise manner. By analysing pieces of music and through algorithm and parameter learning, AI systems can now quickly and effectively mix freshly recorded tracks from the recording studio. This not only saves time and money, but also contributes to higher quality and consistency in the production of music.
The combination of human creativity and AI technology could take music production to a whole new level – a collaboration that can produce very high sound quality even with few resources.
2. Application of AI in music: Mastering
In addition to mixing tracks, AI also has the ability to take over the mastering of songs. Mastering is an important step in music production that involves optimising the sound and volume of a piece of music to prepare it for optimal playback on different devices. Here, AI can play a helpful role, especially if there are already existing references. By analysing existing recordings and using algorithms, AI systems can help ensure that albums sound consistent and have uniform sound qualities.
In addition, AI enables the optimisation of the sound volume, the removal of unwanted disruptive noise and the adjustment of the sound balance for different playback systems. Not only can this AI technology help music producers work more efficiently, but it can also result in higher sound quality for solo artists and newcomers recording at home, as well as a better listening experience for audiences.
Although AI technology offers impressive precision in music production, it must not be forgotten that music is more than just mathematical calculations. Human intuition is as important as technical perfection to preserve artistic vision.
3. Application of AI in music: Extrapolation
Imagine you have an idea for a riff. You grab your guitar and play it quick and dirty. Maybe straight onto the metronome. We do have aspirations, don't we? In any case, AI can extrapolate other instruments to your recorded riff; put a bass line underneath it, add a drum beat, add a few fills here and there. In other words, AI can serve as a recording assistant that can help human musicians quickly and easily develop new song ideas and pour them into demo recordings. A complete beta version of a piece of music can be created. Once the riff, bass track and drum beat are set, you might play a solo over it. Here too, AI can then fill in further tracks accordingly.
Just don't confuse that with real songwriting. AI builds a suitable, average bass track to go with your guitar solo. A no-brainer. No more, no less.
It can also help musicians push their creative boundaries and explore new genres and styles.
4. Application of AI in music: Removing disruptive noise
Another useful application of AI technology in music production is removing disruptive noise from existing music tracks. While recording pieces of music (despite Noise Gate), noises such as background noise, hum or clicking that affect the listening experience often occur. AI technology can help identify and remove these disruptive noises to improve the sound of the music. The AI technology analyses the piece of music and identifies the disruptive noise using algorithms and patterns. AI technology can then automatically remove this noise without affecting the sound of the music itself. This helps to improve the quality and sound purity of music, something that is particularly important for live recordings and concerts.
Today, AI is already much better at recognising patterns than humans are. Removing unwanted sounds may be the safest and simplest application for AI in music.
5. Application of AI in music: AI equaliser and AI compressor
Good mixers work real wonders on equalisers and compressors. But they also invest many hours in practice, testing, good software and hardware and great studio monitor speakers. AI mixes and compresses in seconds, but can only do so on the basis of an example. If it has this reference, however ("Apply this track's equaliser mix to the entire album"), it can save a tremendous amount of time.
Four reasons that offer hope for the long-term future of human musicians
So, artificial intelligence can do a lot. And that's particularly gratifying when it does the work for us. But wait a minute – is what we musicians do actually work in the traditional sense?
1. Making music is a bit different from breaking rocks – we don't do it out of compulsion
Music is fun. It's a nice feeling to record the tracks until you get the perfect flow. It's liberating to gradually complete a song, to test it, to tinker, to craft it until a complete work of art has emerged. This applies both to the instrumentalists as well as to the mixing of the music and the mastering.
Automating work is good – but music is passion, craftsmanship and personal development. Giving that to AI seems absurd.
2. Giving music over to AI is a loss of control
It's a bit like the difference between ordering a good meal and cooking it yourself: AI is very good at delivering what you think you want. But what ultimately comes out of artificial intelligence is always a bit open. Do you want the bass line to be a little juicier? "Re-instruct" the AI again. Is a fill in the bridge too funky for you? Or not progressive enough? Get rid of it and press the button again. Or do you just resign yourself and let the AI have its own way?
AI does roughly what we tell it to do. But often only very roughly. The danger is that we sink to the low level of AI out of convenience.
3. Creativity through "happy accidents" becomes less likely with AI
Unexpected things, creative accidents, improvisations and spontaneous ideas happen when musicians work together and inspire each other. This strengthens your band and gives you confidence. Whereas another musician is a sparring partner, AI is just a better punching bag, at most.
AI doesn't challenge us as much as bandmates with strong characters or demanding mixers.
4. At its core, music is the interaction of different people
Music is communication. In part it's a back and forth between the musicians. Anyone who's ever made music – or even improvised – in the rehearsal room or on stage through eye contact with others will confirm this. With AI in music, making music may become a solo pursuit, isolated in quiet home studios. Some artists may like that, but it often robs us of the magical, emotional togetherness of bands.
Another perspective is music as a form of communication between musicians and their audience. If AI performs a more or less significant part of a song, who then speaks to the audience? An algorithm? The computer? A code? How authentic and believably do they resonate with the listeners? And are you still perceived as serious musicians?
When we say that music is always communication, we mean that it works between people on completely different levels.
Conclusion: We still need creative geniuses who make unique music. However, AI does offer interesting possibilities, especially for the solo artist.
Headergraphik: Adobe Stock | Sarah