top of page


Increasingly, over the decades up to the present day, every aspect of our lives, with music being no exception, is incrementally incorporating the use of AI in one form or another. In music, whether in composition, automated mastering or personalised playlists, AI is here, and it is here to stay, with already a few companies leading the way, all of which impact the industry.

“AI is changing the way music is created and heard” (Schroer, 2020). But how deep will AI infiltrate us in music? Will we ever see a machine improvising music at the standard of a talented human? Will machines be able to emulate human touch and emotion and the sensitivity that makes us engage with music?

Will they ever have those feelings for real? Well, they are working on all of the above as we speak. Will there ever be a blues legend robotic Stevie Ray Vaughn performing a live concert? But more importantly, will it be sold out?

All this is theoretically possible because a human musical performance can be translated into mathematical parameters, music follows patterns and can be learned from thousands of samples.

Perhaps one day machines, if they aren’t already, will get so good at understanding what makes a successful record or at reading our musical intentions, that they will become absolutely relied upon, for instance, to create a missing perfect orchestral arrangement or to manipulate song structures against our suggestions, resulting in masterpieces. Their involvement will be to such a degree that maybe one day, not too far ahead, some musicians will have to readapt and expand their skills to find work when we go back to a post COVID 19 scenario.

Even the concept of AI Music Industry will become an independent term. A point for debate is, can we call computers creative? We are used to computers following our instructions to our desired outcomes; they cannot give birth to a product out of “nothingness”, we need to feed them data in order for them to re-arrange the data and “create” something that was priorly inexistent, can those outcomes be considered art?

It is still early days, but from what we have seen so far, we can have a grasp of what the future will be like when the musician’s fear of fears (being replaced by machine) strikes, let us be confident about the fact that even though AI development will continue, it cannot and will not replace humans entirely. It will most likely be a symbiotic world, where humans will incrementally be enhanced by AI but never replaced.


It appears that humanity has not much else to do but to evolve; we are clearly seeing the early stages of the involvement of AI in aspects of music that we never witnessed before and were exclusively human jobs; such as composition and production of music, the question is not whether AI will incrementally replace more of these human functions in music, the question is when, and at what scale.

Since the invention of computers, humans have been digitalising their lives ever since; Drott (2020), describes this as the“digitisation of human knowledge”, over the past few decades, humans are relying on the transcription of our information into computer databases, for the sake of augmenting human comfort and reducing effort.

As we digitalise our lives, we simultaneously are humanising computers by the minute. Isn’t that an accurate definition of AI? The ability of computers to perform human tasks? Computers have many human behaviours now with what we call “machine learning” capabilities, from which they analyse data and then act on it.

Computers do not create anything new yet, but if we dig in this thought a little deeper, we find that most humans create in the same way computers do, out of concepts that already exist, or concepts that could have already been created, they learn, and then they create. Most humans learn by listening, analysing and understanding the data first; only after sufficient understanding, they can then start their musical creations. The main difference between humans and computers is that computers can learn instantly and do not need years of training, they can also be fed quite a lot of data, but they will only function with input data.

Most humans need years of training, but some humans are the exception. Some humans do not need much data, and they use their intuition to create music or learn incredibly quick, like all those prodigies so talented at such a young age. The incredible ability that computers have reached so far is a form of “scientific narcissism” describes Cole (2020); if computers could reflect on what they are able to do and how perfect they can remember data, they will most definitely be narcissistic about it. They have been becoming our multi-assistants so undeniably now that we cannot practically function without them in the society of today. They are always present, so it is normal and embedded to maximise human creativity using computers (Carnovalini & Rodà, 2020).


The subject of music AI might not be invading our thoughts and flooding our conversations entirely just yet, but it is definitely not new; it has been around for some time. Intelligent and curious scientists with time and resources in their hands have been trying for decades to play around with the intertwining of music and AI, like Alan Turing, “the godfather of computer science” (Chow, 2020), when in 1951 created a machine that could generate three simple melodies.

According to Google AI designer Li Chong (2019), this new paradigm started in the fifties with the invention of electronic musical equipment. At that time, it became the new way of making music, and today it is vital in any musical work. Chong describes music AI in three major eras;

The early stages; AI was focused on algorithmic composition, and two major key events occurred; the first happened in 1957 when the first musical work completely written by AI was programmed at the University of Illinois by composer Lejaren Hiller and mathematician Leonard Isaacson. In 1960 the first paper on algorithmic music was published by Russian researcher R. Kh. Zaripov. The breakthrough; AI was starting to understand music, and in 1975 the first intelligent music perception system was seen metering tempo and duration of notes while musicians played on a keyboard, thanks to the MIT Experimental Music Studio. In 1980, David Cope invented the EMI (Experiments in Music Intelligence) System to enable the AI to analyse existing music and then create a new piece based on it. Of course, we famously know that David Bowie in the nineties was one to try an application that reorganised words to come up with new combinations or lyrics.

The current research; This stage focuses a lot on composition; AI can already do an amazing job analysing music; it can already transcribe music and recognise different instruments, structures and even perform emotional recognition. There are several events in between the ones mentioned above that state many names nobody has ever heard of.

However, in 2019 Bjork and Microsoft, partnered up to create AI-generated music called “Kórsafn” based on the sun's changing weather patterns and position (Freeman, 2020). The brief historical description above is not emphasising names and dates in detail. The intention is instead to point out that historically, every two or three decades, a significant new development happens in technology. When the next thing arrives, it evolves and gets enhanced with accessories and perfected improvements in those twenty-plus years, until the following next important event arrives, and that becomes the cycle. One important and interesting point is that these developments were made mostly by Americans or Russians. Let us face it, musicians from the thirties had to have a much better brain for music since there was no help from computers. Even before that, when there were not any recordings, the level of intelligence required to write entire symphonies with music notation must have been quite impressive to see. It is still impressive today, let alone back then. It appears that evolving is as basic as human survival needs, no matter what phase the world is going through.

Despite major wars and the world still cooling off from them, the devastation of economies, hunger, poverty, social inequality, injustice, or pandemics. It seems that science will not seize to pursue the constant urge for the next thing. There is always somehow the necessary economic resources, so it appears. It is terrifying for some, and incredibly exciting for others, to imagine what will be normal fifty years from now.

Of course, this is assuming that a pandemic or nuclear Armageddon do not spoil the fun. To complement Chong's article, a fourth Music AI era should start to be accounted for starting from the rise of the omnipotent powerful and mighty silicon valley Tech giants. They are influencing our lives as never experienced before in such an extreme way in terms of AI. This is happening parallelly in music; the way music is consumed is now very much by social media standards. It is also essential to mention that these Tech Giant companies are the ones that are mainly funding the advancement in this field. There is evidently a new world after the rise of the Silicon Valley era; it is overwhelming to think where they will lead the world. They are the ones driving today. It is terrifying for anybody to have that amount of irrefutable and monopolised power.


In artificial intelligence, there are many companies out there already, funded by subsidiaries of more prominent companies, but to say it concisely, and if the money is followed, there are three major names when talking about AI pioneers.

Also, they are the majoritarian shareholders, and therefore the primary investors. We are talking, of course, about Google, Amazon and Microsoft (Shell, 2021).

Their massive investment is evident in the field, and there seems to be an urge for it too. It is being pushed forcibly and so rapidly that even people like Elon Musk, who is heavily invested in so many aspects of AI himself, are concerned about it. As described by him in one of his Joe Rogan interviews, he made suggestions to the people heavily involved in AI to “slow it down” and “nobody listened”, according to Musk. The average person has access only to the surface of filtered information about the field, so what does Elon Musk know for him to be concerned? (JRE Clips, 2018).

In music AI, the list of companies, apps and researchers in this field is quite extensive; in terms of music composition, the one that appears first on the list and everywhere is AIVA technologies. There are many others like Juckedeck, Lanrd, Shazam, and Spotify (Mohapatra, n.a.). In practicality, the most common example that we could all see today in our daily composition is in Logic Pro, when you can generate drum patterns in an improvised form. So this leaves us with no choice but to include Apple in our Music AI pioneers (Deahl, 2018). Music tech startup supporter Scott Cohen describes and confirms how our next evolutive stage of music generated by AI will be somehow symbiotic with humans to create something never seen before (Dredge, 2019). Movie editors are already, he describes, using AI to sync the music with the exact cut points of the images, so it is understandable if a large number of people consider AI of good help making our job easier, while perhaps other people rather enjoy doing the work manually. In May 2020, in the Netherlands, the first AI contest took place, organised mainly by Dutch broadcaster VPRO.

In a nutshell, the contest was trying to emulate the Eurovision contest, and the challenge was to create a song using artificial intelligence. The judges were experts in the matter, including the founder of the TikTok acquired AI music company Juckedeck (Ingham, 2020). The cultural connotation that is getting promoted with these contests with the statements made by the AI music experts in this particular contest is suggesting that humans should already be preparing for the idea that, in fact, humans and AI will be co-creators in the future. Also, that it is already here; and that it is cool, trendy, and fun. So again, no debate about whether it will happen or not since it is present already. It is about how long will it take to get as deep as it is intended to get. As well as how society will reshape itself not just in music, but in every aspect, to share our jobs wit AI. The research happening now is primarily focused on making the machine able to feel and respond appropriately to human emotions. The mechanical AI, as we see in industries, is made deliberate that way because it helps with high standards of consistency. At this stage, the thinking intelligence has also been reached in AI, like autonomous cars where the machines are basically emulating human intuition (Huang et al, 2019).


It is increasingly noticeable the number of movies about AI. It is not mainstream narrative, so it is not of public domain, and you will never see any member of the government or federal agency admitting to the fact that even though it is not an ascertainable official fact, it is widely researched by a number of people that a good way of getting society used, or on full predisposition to a new norm, is through movies (OrangeM, 2021).

The Hollywood infiltration by government agencies to prepare society for a new normative is known with the term “Preemptive programming”, according to English researcher and writer David Icke, who has written extensively about the subject. One role of these films, according to Icke, is, of course, entertainment. But another objective aimed at the subconscious mind, confirms Icke, is to acclimate society to whatever new normality is intended to be accepted and not repelled by society. You get the subconscious mind of people familiar with the world pretended to be imposed, so when the world does get to that, there is less resistance. If analysed deeply, this is actually an incredibly clever way of mind manipulation.

There are plenty of examples, but a very accurate representation of this is the movie Transcendence, where human conscience is relocated to another body. Again, this is a subject that we can relate to people like Elon Musk, who describes not only how this will be possible but also that it is in the making (JRE Clips, 2018). Another great example would be the film Irobot, where the robots are assisting humans in everyday tasks. The machines then develop their own conscience and eventually become evil and try to hurt humanity. Not all the facts in movies are completely accurate, and because these movies need to be cinematically compelling, there is a big element of drama, and that, will make them get some of the science facts wrong (Schults, 2015).


It will not be the case, but even if the entire music creation process gets replaced by machines, some jobs will stay irreplaceable and will remain untouched; like a singer, how can someone’s favourite singer be replaced? How can people connect emotionally with a machine as they do with their favourite singer? Machines are not able to engage in any emotions yet; therefore, they do not interlock with our human essence. Opinions vary on what jobs will prevail, some of which are arguable, apart from singers and athletes, but other obvious ones are; journalists, writers, priests, couriers until we have self-driving cars, maybe also marketing and sales (7 Occupations Irreplaceable by Artificial Intelligence (AI), 2018).

Even if machines replace a big part of human labour, the jobs of supervising and maintaining the machines will automatically be created or supervising and maintaining the machines that supervise and maintain the primary ones. Rather than replacing, machines will be assisting. For instance, wouldn’t it be fantastic that when an artist is experiencing a writer’s block, AI could unblock it? AI could finish the sketch, or the missing arrangement, or a more compelling ending, or the missing element that will unblock the entire creative process. If this was already hypothetically happening in a major way, we should note that the human element is still the primary source and very present. It will be pretty much like a business partnership; humans and machines will become co- creators. AI companies are already helping out with the music. However, these types of music AI practices are more useful with music novices without formal training or experience. The real question is, why would anybody serious about music not want to learn how to do it rather than allocating the work to a machine? Music is an amazing gift to have or develop; it is a great curative form of therapy also (Williams et al, 2020).

A study at the Northwestern University was carried out on eleven music novices. The study reveals how it is still early days for AI to be able to complement a started piece of music with the same intention that was originally meant. One of the participants stressed that while the machine contributed with musically coherent parts, and even though the notes where aligned, the emotional motif was gone, the essence had disappeared (Louie et al, 2020).

So we know that machines have no emotions. The important question and concern is, will they ever? If they ever do, then we can definitely forget about society as we know it, machines are better than humans, and if they had human emotions, we could see something that at this point is unimaginable with the amount of power they would have.


It is a fact that people who make a living making music for the game industry, advertisements and web content will not be able to rely on that income for a living anymore. We can already confirm this kind of negative impact with royalty free music websites. It is extremely easy to put a track together by having the computer do all the work quickly and be ready to upload in minutes; therefore, millions of tracks already exist to choose from. It is nowadays just dragging and dropping a few files into the DAW, then making a few tweaks here and there, and we are ready to bounce and upload the song.

There are so many uploaded tracks that cost so little to purchase, that it is hard to sell anything substantial to rely on the income, and this is not even involving AI yet. It is also true that more music than ever is on demand due to the massive growth of online creators, as well as the video game industry, so we will have to wait and see how this reshapes and balances itself out, like everything else in society (Chow A. 2020).

In most instances, the music required for these fields is quite simple and does not require deep artistic literacy or complicated musical variations or dynamics. It does not require either a high level of intimacy or storytelling; therefore, AI could be of some useful help in these cases. Of course, this is if none of the royalty free music already for sale, which is a vast number, is not up to the client's liking already. Clients can probably find what suits them already within those millions of tracks on sale for cheap. Many musicians who can create choose this profession because it is supposed to be fun and fulfilling. Even if AI can arguably do a better job a lot easier, why would any real musician go for it in the first place? It would stop being a channel for emotions and expel frustration and describe beauty. It would be fair to use AI to our benefit, but not to use it in our stead and stop expressing ourselves through art.

On the other hand, AI could be a handy tool for those who do not have musical abilities and do not wish to pay for music. There is also the argument of the copyright side of AI music, which at this stage is murky and unclear still. Drott (2020), describes how, in the UK, works done by a machine need to be allocated to their owners and not the computer itself.

The reason why we have copyright is to regulate human behaviour and can sometimes offer incentives. In the end, machines are not incentivised by money, so there is little interest in reforming laws to grant machines copyright entitlements. Legal issues could arise if an AI work resembles one of a human for sure. In the realms of streaming, in many cases, talented artists remain undiscovered because of the lack of means for marketing. It is expensive to get an artist known, so with a little luck, AI algorithms can help with the discovery in many streaming platforms of today. This is convenient for both the artist and talent scouting agents. A great number of fantastic musicians find it hard to stand out since, unfortunately, artists don’t start their careers with the same financial advantages. AI is probably, in some ways, not a bad thing, and it is just simply inevitable; it is helpful in many areas of human society.

In terms of music, it should probably be used to enhance the human mind. Machines will be around forever, but we will never lose our human spirit, our essence. There is a craving hunger for live music events that put in practice human connection. If Glastonbury was sold out in thirty seven minutes with no line up (Powell, 2019) in 2019, how long would it take to sale in 2021, after a year of being restricted of our very human needs? It would be sold out probably in less than five minutes.


Plenty of musicians, like Grimes, for instance, feel like this is the end of human creativity, it is spooky for them that something so personal, involving such human intimacy and redundantly humanism, could potentially be replaced by machines, and even more scarily, it could be done much better than a human can (Chow, 2020). Many other musicians, on the other hand, believe that human art is beautiful and irreplaceable and that the presence of AI in our daily creations will give birth to a new era of new creativity, and feel very positive about incorporating AI in the creative processes by default.

Maybe some kind of new musical genre will emerge with the merger of humans and machines, and it could be the next thing since most traditional musical genres have been deeply explored and experimented with. It is unclear which side of the argument will be the accurate prediction. Nevertheless, as the years pass, the only certain thing is that AI will be more and more accessible to everyone. That is most definitely the reason why the quality of musicianship in modern times is dropping.


To start concluding this argument. If we know that up to this point, machines do not have consciousness, and they theoretically improve our lives, why are some people so concerned about it? Why is there a massive argument happening if whether we should continue developing and investing in AI?

Perhaps it is because some people are struggling to trust in anybody with the amount of power as the tech giants that are so eager to invest in advancing this and knowing that there is a correlation between the size of a company and staying moral. As bigger as a company gets, they appear to become greedier, and desiring to obtain more control, and doing less for the benefit of the majority of the people. Or perhaps it is because of the fear that if computers will be able to think for themselves, could they turn evil? Can they turn biased? Or furthermore comes the ultimate question. Could they ever intend to harm us?

We are all unfortunately at the mercy of the big investors, it is up to them where they want to drive this, they could make a wonderful experience as well as a living hell, whatever the case, we are still light years away from AI feeling and responding like a human being, and even more in music. So will musicians, for gods sakes, just relax about filing for unemployment any time soon because of AI. It is a virus and not AI that is causing unemployment, but the pandemic will pass. So even if we see AI composing, recording and improvising music in our daily lives, as exciting as it might be at first, it will not be a long lasting relationship since we will not be able to connect emotionally.

Humans admire other humans because their passion and sensitivity inspires them. Part of our souls are reflected in our favourite artists, and they seem to be the closest connection to God, or so we like to think, we need to continue being human. The only inevitable is change, and we will reshape to it, as we always do in order to survive new parameters.

Let machines replace the soul-sucking jobs, so we may all transition into a more holistic approach to our economic systems. One that is not based on accumulating as much profit as possible, ignoring if people are passionate or if they know anything about how they are making their profits. There needs to be a more artistic way of finding meaning, so in the end, who knows, maybe AI is exactly what we need to realise it. BIBLIOGRAPHY Alfo Media, (2020). What Does Artificial Intelligence Music Sound Like?. Available at: (Accessed: 28 February 2021).

Bonnici, A., Dannenberg, R., Kemper, S. and Camilleri, K., (2021). ‘Editorial: Music and AI. Frontiers in Artificial Intelligence’, 4. Available at: 10.3389/frai.2021.651446/full. (Accessed: 28 February 2021).

Bughin, J., Seong, J., Manyika, J., Chui, M. and Joshi, R., (2018). Notes from the AI frontier: Modeling the impact of AI on the world economy. McKinsey & Company. Available at: frontier-modeling-the-impact-of-ai-on-the-world-economy# (Accessed: 28 February 2021).

Carnovalini, F., Rodà, A., (2020). ‘Computational Creativity and Music Generation Systems: An Introduction to the State of the Art’. Frontiers in Artificial Intelligence, V. 3. Available at: (Accessed:23 January 2021).

Chong, L., (2019). A Retrospective of AI + Music. Medium. Available at: https:// (Accessed: 25 February 2021).

Chow, A., (2020). Musicians Are Using AI to Create Otherwise Impossible New Songs. Time. Available at: (Accessed: 25 December 2020).

Cole, R., (2020). ‘The problem with AI music: song and cyborg creativity in the digital age’. Popular Music, 39(2), pp.332-338. Available at: popular-music/article/problem-with-ai-music-song-and-cyborg-creativity-in-the-digital-age/ F42497C10564D33C15ABF5495E46BEAC. (Accessed: 21 January 2021). Deahl, D., (2018). How AI-generated music is changing the way hits are made. The Verge. Available at: southern-amper-music. (Accessed: 21 February 2021).

Deahl, D., (2019). We’ve been warned about AI and music for over 50 years, but no one’s prepared. The Verge. Available at: algorithm-music-law-copyright-human. (Accessed: 28 February 2021). Dredge, S., (2017). AI and music: will we be slaves to the algorithm?. The Guardian. 6 August. Available at: intelligence-and-will-we-be-slaves-to-the-algorithm. (Accessed: 28 February 2021).

Dredge, S.,(2019). WMG’s Scott Cohen talks innovation: ‘What’s coming next, and what do we need to do?’. Musically. Available at: cohen-talks-innovation-whats-coming-next-and-what-do-we-need-to-do/. (Accessed: 27 February 2021).

Drott, E., (2020). ‘Copyright, compensation, and commons in the music AI industry’. Creative Industries Journal, pp.1-18. Available at: 10.1080/17510694.2020.1839702. (Accessed: 27 February 2021). Freeman, J., (2020). Artificial Intelligence and Music — What the Future Holds?. Medium. Available at: music-what-the-future-holds-79005bba7e7d. (Accessed:28 February 2021).

Huang, M., Rust, R. and Maksimovic, V., (2021). ‘The Feeling Economy: Managing in the Next Generation of Artificial Intelligence (AI)’. California Management Review, 61(4), p.46. Available at: f0e4-4ff5-ad59-a2075f9681c3%40pdc-v-sessmgr01. (Accessed: 28 February 2021). Ingham, T., (2020). Music made by Artificial Intelligence has started winning awards. Grammys next?. Music Business Worldwide. Available at: https:// winning-awards-grammys-next/. (Accessed: 27 February 2021).

Inside the score, (2020). Can Artificial Intelligence Make Good Music?. [video] Available at: (Accessed: 28 February 2021). JRE Clips, (2018). Joe Rogan - Elon Musk on Artificial Intelligence. Available at: https:// (Accessed: 27 February 2021).

Liebman, E., Saar-Tsechansky, M. and Stone, P., (2019). ‘The right music at the right time: Adaptive personalized playlists based on sequence modeling. MIS Quarterly, 43(3), pp.765-786. Available at: vid=7&sid=345a5098-aac2-4eb4-bf94-be5211d5cdc8%40sessionmgr4008. (Accessed:16 February 2021).

Louie, R., Coenen, A., Huang, C., Terry, M. and Cai, C., (2020). ‘Novice-AI Music Co- Creation via AI-Steering Tools for Deep Generative Models’. Conference on Human Factors in Computing Systems, Available at: 10.1145/3313831.3376739. (Accessed: 27 February 2021).

Lunny, O., (2019). 'You're Probably Going To Be Replaced' Digital Music Pioneer Issues A Stark Warning For The Industry. Forbes. Available at: oisinlunny/2019/01/28/youre-probably-going-to-be-replaced-digital-music-pioneer-issues- a-stark-warning-for-the-industry/?sh=31f341ef7b58. (Accessed: 28 February 2021).

Marr, B., (2019). The Amazing Ways Artificial Intelligence Is Transforming The Music Industry. Forbes. Available at: amazing-ways-artificial-intelligence-is-transforming-the-music-industry/?sh=4574549f5072: (Accessed: 28 February 2021).

Mohapatra, A., (n.d.). 8 companies that are coupling AI with music. Techseen. Available at: (Accessed: 16 February 2021).

OrangeM, (2021). David Icke Pre emptive Programming. Available at: https:// (Accessed: 27 February 2021).

Powell, K., (2018). This Massive Festival Just Sold Out In 30 Minutes with No Lineup. EDM. Available at: minutes/. (Accessed: 27 February 2021). Raval, S., (2018). AI for Music Composition. Available at: v=NS2eqVsnJKo. (Accessed: 28 February 2021).

RS Web Solutions, (2018). 7 Occupations Irreplaceable by Artificial Intelligence (AI). Available at: artificial-intelligence-ai. (Accessed : 14 February 2021).

Schroer, A., (2020). The AI music revolution is here and these 11 companies are leading the charge. Available at: (Accessed: 27 February 20212).

Shell, A., (2021). How to invest in artificial intelligence. USA Today. Available at: https:// (Accessed: 26 February 2021).

Shultz, D., (2015). Which movies get artificial intelligence right?. Science | AAAS. Available at: (Accessed: 18 February 2021).

Williams, D., Hodge, V. and Wu, C., (2020). ‘On the use of AI for Generation of Functional Music to Improve Mental Health’. Frontiers in Artificial Intelligence, 3. Available at: https:// (Accessed: 28 February 2021).

11 views0 comments
bottom of page