Who needs Tango DJs?
In the previous post, I was writing some thoughts about the potential impact of AI combined with other technologies in terms of new artists. Given that this is a blog about Tango, I bet you may have asked yourself… What about Tango DJs? Tango DJs are obviously more often in terms of playing music on a milonga than an orchestra. Even when an orchestra plays live, the DJs are there to fill in the gaps. Moreover, people seem to pay less attention to a DJ than an orchestra. I mean they are not the famous names that people usually expect to see in performances and also you can’t say they are… performing. So why not just have a computer… plug it in… run the Tango DJ software and start the milonga. No need for a DJ anymore…. right?
To enforce this argument even further let’s look at what a Tango DJ actually does. There are two things mainly. First, create Tandas with similar songs. Second, read the floor, and the mood of the dancers and respond with appropriate music. Engage in a dialog with them. The first one is already quite easy. The website El Recodo offers a utility in its app to create a Tanda given a specific song. So, the software to create Tandas is already there. How about reading the floor? Well, there are applications in the domain of security for example which can track how many people are on streets or specific places like an airport and automatically asses different parameters. So I guess an adaptation of software that could assess the mood of the dancers given some input from cameras… is also achievable. So… Let’s all say goodbye to Tango DJs! Right?
Copying and getting influenced
Again, I am not going to provide answers but mostly some thoughts and questions for us to discuss. Before going into them let’s have a look at some of the current features of AI like chatGPT or image creation AI.
Marques in a review video has tested those systems and expresses his own thoughts. One of the things he says is that such systems have been fed with a ton of information in order to be able and synthesize different pictures to give the final outcome. So the question that rises is if the authors of the original pieces of art should be somehow attributed credits. But then again think about it. Isn’t this more or less how human artists actually work? They gather input, information, and influences from already existing pieces of art and create new ones based on these influences. It’s what we call inspiration. Even the so-called avant-guard artists have gathered influences in order to try and break them, bend them, blend them, and come up with their innovative art. So what is the difference when this happens from a machine and not from a human brain? Why should we draw a line and distinguish between the same process coming from a human or a machine? In my opinion, we shouldn’t.
The power of human limitations
However, there is another crucial differentiation between the two processes. The current AI systems work based on huge data input. They have read tons of books, seen tons of photos, and heard tons of music. So much that a human mind would need many lifetimes to consume them all. Believe it or not… this is what I see as an advantage in this case. I mean, human artists consume a fraction of that input, but mostly they do so intentionally. They know what they want to read, look at, and listen to because they have a formed opinion and a taste. Each of the artists has a personality and a life trajectory behind them that influences their input and output as well. It’s not a machine with a limitless capacity that consumes tons of data without any judgment in order to be able to imitate it based on statistical analysis.
This limitation of the human mind (together with the practical limitations of time and space) is what leads to the building of character for each human artist. That is why the bandoneon of Troilo sounds different from Laurenz’s or why the piano of DiSarli is different from Pugliese’s. Small differences in their characters led to different outputs. Of course, a machine can imitate anyone and can play music like any kind of artist you tell it to. But this doesn’t mean it has a character. It just means that it can create or imitate one based on our choices and input.
So coming back to our beloved Tango DJs. Let’s say you plan to go to a milonga. Many people would choose to go or not based on the DJ. You see the character of the DJ, their education, their nationality, their experiences, and backgrounds in life are all more or less reflected on their DJ sets. You know what to expect beforehand. You know who you are dealing with and if it is according to your taste and mood of the day. So you make a choice. What if instead of a DJ… an AI tool like the one described earlier was playing. Would you know what to expect? The character behind the set? Would the music reflect any life story? Most probably not. But then again… Is this a big problem? Haven’t you gone to milongas where you didn’t know what to expect? So I think a lot of people wouldn’t say no to a milonga led by an AI Tango DJ.
Would that lead to the extinction of human Tango DJs? I don’t believe so. When we go dancing, knowing the character behind the music, we relate much easier and we connect with the DJ on a totally different level. When we select to go on a milonga where a more dynamic or a more melodic DJ is playing we do so because we relate with these parts of their character. We will always be seeking this kind of connection and although an AI-DJed milonga could be fun there will be nothing but the software on the other side to connect with. A software that is possible to form all kinds of connections with all kinds of people. Software without a character or maybe with too many characters to choose from.
Tonight’s Goodnight Tango
Speaking of character tonight’s Goodnight Tango is coming from my favorite character in Tango so far. A person with great character, strong beliefs, and passion for them. You can feel and see all these in his music.