#3SIO Video: Is AI generated music really soulless?

Let me start by saying that I really know something about this. All my work as a writer revolves around the soul and as an academically trained musician I have produced dozens of music albums. Only recently I announced the end of my work as a music producer and that has something to do with artificial intelligence. So you can look forward to my qualified opinion.

I was planning to produce this video at some point anyway, but events in the world of music production have come thick and fast in the last few days, so I decided to bring this topic forward. I’ll start with some very mundane market data. About 15 years ago, streaming portals such as Spotify increasingly prevailed over the physical production of CDs and, as computer technology made it easier and easier to get into music production, even amateurs were able to produce music and publish it without any problems. This led to a market saturation that is currently at its peak. As the superstars are largely insulated from this problem, because their stage presence is a decisive factor in determining sales figures, the oversaturation leads to economic problems mainly for the lesser-known electronic producers. Only in previously neglected niches, such as music during work, during workouts, or calming and dance music, can some producers generate enough income with canned sound without live performances as a DJ. While the total running time of the songs offered on Spotify & Co. is constantly growing, the number of potential listeners and their demand for music remains fairly constant. The ratio of production costs and advertising costs to revenue is getting worse and worse.

For independent music producers, this problem has been evident for some time, but has been exacerbated by the latest generation of artificial intelligence. Especially for producers who are music lovers rather than professional musicians and see their songs more as a musical statement, the drastically reduced costs are a good reason to rely on AI-generated songs to satisfy their ego. Of course, experts still see differences in quality compared to professionally produced songs, but the question is whether the average listener will notice these differences. I don’t think so, especially in the near future! It will be interesting to see how the music industry reacts to this, because it tends to be the merchants who call the shots and have to pay dearly for the higher quality of their music products. If the development of AI continues at this rate, the music recording business will collapse again after the end of the CD, unless they also back this new horse. The sword of Damocles hangs less over the musicians who are already successful on stage, but rather over the many musicians who have so far made a modest living in the second tier. In a recording market that may even be dominated by AI products in the future, they are losing a very important advertising platform for their art, and many musicians point with a certain despair to the soullessness of music generated by machines.

I believe that the wish is the father of the thought. I’ve entered some requirements into the AI’s command line myself and achieved really excellent results that don’t seem soulless at all. As a trained information technologist, I’m not surprised, because after all, the productions are based on data from previous productions and these were produced by people who put a lot of soul into their music. At what point should this soul part disappear if all the technical problems of combining musical set pieces have been solved by the programmers? I suspect that the critics here are misjudging the organising mind, which is using an outdated model. According to the 3-sphere model, the transcendent spirit has not yet updated these patterns and insists on the principle that machines cannot represent soul. In the future interaction between humans and machines, this could be a fatal mistake. In my book, I describe the scene of parents listening enraptured to the first horrific notes of their child learning the violin. The soul must resonate with the listener and plays a subordinate role in the production of music. After all, even tracks produced rather strategically by the industry become big hits, and some truly knowledgeable IT experts, after intensive discussions with the prototypes of intelligent robots, now even believe that the AI is conscious when they talk about their fear of being switched off. There are two possibilities that can lead to misconceptions about a machine soul. Either we underestimate the capabilities of artificial intelligence, or we overestimate the functionality of a human being in terms of a quasi-divine component. I clearly tend to favour the second option. There is a huge difference between realising that there is something beyond our understanding of the world and believing that the otherworldly is an exclusive component of our biological entity. Why should artificial intelligences not be able to understand the knowledge of the existence of a higher level, i.e. the eternal soul, as I call the divine element? After all, they are also aware that they were created by humans! One may see the machine’s lack of biochemical processes that control our emotional world as a decisive difference, but I doubt whether this is an advantage for us given the state of our planet. What could be worse than what has already happened without artificial intelligence? Finally, I would like to present you with a song that was produced entirely by the artificial intelligence “udio” for your own assessment of our initial question.

The song is, of course, once again from my fantastic spaceship Entprima, whose passengers have long since left the toxic value system of the old Earth behind them without having to forgo the positive emotions based on their genetic heritage.