The next step in the automation of rave? (Transcript from a workshop at Transmediale)

At this year’s Transmediale, I arranged a workshop together with Axel Gagge, Frida Sandström and Pedram Nasouri. Connecting to the festival’s larger theme, we invited the participants to discuss what a dancefloor can be. Our own interventions came from very different perspectives, such as physics, choreography and political activism. Here follows a transcript of my own presentation in the workshop.

In the history of popular entertainment, disco and rave represents a great leap towards more automation. Replacing the live orchestra with a discjockey makes the night club more productive, in strictly economic terms. In such terms, the orchestra and the DJ may could appear as just two ways to produce the same entertainment commodity – dance music. But we all know that there is more to the story than just capitalist efficiency. The dancefloor itself has become something different.

Disco and rave also marked the historical abolishment of the dancing convention typical of Western modernity, namely partner dance – dancing in a couple.
Instead of facing one chosen partner, we now have several alternative directions on the dancefloor. We may dance alone, in our own world, with the face turning in all directions as a part of the dancing body.
We may also dance in smaller groups of friends, facing each other.
Or we find ourselves standing in a line at the dancefloor, everybody facing forward towards the DJ, seeing mostly the backs of other dancers – just like at a conventional rock consert.
If everybody is facing forward, this brings a certain comfort to the dancefloor, because there is no risk of meeting the eyes of other people. We do not say that this is wrong, but we have been thinking about how it changes the way we are dancing, and the way we experience collectivity.

At some dancefloors, there is a sense that the dancing itself has become automated. Not just in the sense of being a slave to the rhythm, but of each body becoming just a part of one big machine. Techno as the highest stage of automation.

But there is also an apparently opposite tendency that we can observe at other dancefloors (or indeed at the same dancefloor, a few meters or a few hours away). Here the dancing is experienced as the immediate expression of individual feelings; each person bringing her own feelings into the common space, in order to share them with others. This is how I understand the concept of “radical self-expression” that is a cornerstone in the official ideology of Burning Man. But our experience from so-called burner raves is that this cult of individuality, in the end, turns out to be quite conformist. There is an unspoken norm for how you should look and how you should move when you are “expressing your true self”.

This difference in how dancing is experiences may correspond to different party environments or architectures.
On the one hand, think of the typical techno bunker; rough concrete walls that do not provide any visual distraction from the sonic experience. Because the bunker is not a complex space, it may allow for more musical complexity. But as a by-product, such spaces also tend to elevate the individual DJ to the center of attention. It simply does not make much sense to look in any other direction, just like it does not make much sense to do anything else than to dance.

In contrast, think about open-air environments that include natural complexity and secret spots, or think about a club filled with visual arts, labyrinthic passages and other surprises. Chill-out opportunities may draw people away from the activity of dancing. But complex environments also makes the DJ appear less central, and invites people to contemplate the collective effort that is behind every rave party.

Do ravers feel that they dance to music delivered by a DJ?
Or that they are invited to a temporary space, created through collective effort?
Or do they identify as consumers who, having paid twenty euro at the door, are entitled to entertainment?
This is a deeply political question, with repercussions far beyond dancefloors.
It surely has a long history. Even before disco, dancefloors could differ much in character.
But historically, the introduction of new machinery on the dancefloor has forced a renegotiation of roles and responsibilities.

The concept of “live music” was invented around 1930, in response to automation. First, cinema orchestras were replaced by sound movies. Soon after, many cafes and restaurants saw themselves economically forced to play recorded music instead of employing performing musicians. Facing a massive unemployment crisis, musicians themselves reacted in contradictory ways. (The situation in fact bear strong parallels to the contemporary debates over robotization and human obsolescence.)

Initially, many musicians reacted with resignation. After observing how machines took the place of human musicians in cinemas, many concluded that “the mechanization of music” would proceed swiftly towards its inevitable ending: a future with no musicians at all, only machines playing fully automated music.
As some tried to formulate a resistance to that scenario, around 1930, the concept of “live music” was established. Musicians and cultural critics initiated a campaign to educate the public about the magic that could only happen in the presence of real, human musicians. Liveness, to put it briefly, is the name of that which can not be reproduced by a machine.
The campaign was not anti-technology. It recognized that recorded music had legitimate uses, but aimed to limit it to particular social settings, letting live performances persist in a protected reserve.
To things about this campaign are worth pointing out here:
1. It was partly successful. In the early 1930s, many had though that automation would soon make dance orchestras obsolete. But this did not happen. Live orchestras would persist as the default source of dance music for another 40 years, that is, until disco arrived.
2. It was never simple to draw a clear line between “live music” and “mechanical music”. How much audio technology could be used by live musicians before their music lost its liveness? As programmable synthesizers began appearing on stage, liveness collapsed as a binary and became more of a spectrum. And on this spectrum, even discjockeys could claim a space as a new kind of semi-live musical performers. Be it that some of today’s superstar DJ’s do not do more on stage than press play and wave their hands. Their personal presence does still matter.

The historical experience of disco and rave also demonstrates how liveness can be reinvented, in response to the introduction of new technology. So far, it seems like a dancefloor needs someone present that can take responsibility for the music and act as a point of projection for the musical experience.

On the other hand, it is not hard to imagine how the automation of entertainment could take another leap, replacing the discjockey with an algorithm.

Unlike a human DJ, the robot DJ could analyze the qualities of all music that has ever been recorded. The same algorithm could also analyze the dancefloor in real time through various kinds of sensors spread out in the room.

From a commercial perspective it certainly would make sense to cut costs by replacing a human DJ with an algorithm. But how would dancefloors react to the disappearance of the DJ? Or to the appearance of surveillance technology? Or to the lack of a given center of attention? Would it force a visual or architectural rearrangement? Would liveness be reinvented in yet another form? What would be the status of the meta-algo-DJ, that is, the human responsible for the choice of a particular algorithm for a particular dancefloor?
Such questions do have a wider significance. I think they can help to reveal certain blind spots in the current debates over automation and artificial intelligence, both in the mainstream and in those parts of the left attracted to “fully automated luxury communism”. There is a widespread tendency to forget that automation is not evenly distributed and does not have one given direction, but also triggers various counter-movements.

Therefore I would now like to open the first of four group discussions by asking: How do you imagine the next step in the automation of rave? What would be a dancefloor at the maximum of automation – and what would its inhabitants care about?

16 kommentarer ↓

#1 monki on 4 February 2019 at 3:35 pm

Could the personal presence of the DJ move to that of a “host” for the musical experience in the middle of the dance floor? Are there examples of discos or raves with “dance leaders/ambassadors” on the floor? How common is it in the contemporary club scene to place semi-professional dancers in the crowd to set the mood? Disco was big on pro dancers in clubs, for what I know.

Maybe we will see a data-driven franchise of clubs similar to how Wework tries to create data-driven office spaces where data from their locations all over the world are supposed to influence decisions about constructing flows and networking in office environments.

Another example is Crossfit where the central organizations designs a “workout of the day” for all their thousands of Crossfit gyms to do at the same time and then use feedback from performance, injuries, engagement etc to design better optimized daily workouts. Imagine a franchise of clubs playing the same set in hundreds of locations and using data-driven methods to find out where certain tracks work or not. Data that would be valuable to labels and producers of course.

Sounds like horrible places to dance at though…

#2 rasmus on 4 February 2019 at 5:56 pm

monki: Thanks for those comments! I think that värdskap (being a host) could be a key concept here, and that the most interesting thing is that it allows for the exploration of collectivity, in a way that departs from the idolization of the individual DJ. One example that comes to mind is last year’s CTM night at Berghain with the Polish hardcore rave crew Wixapol S.A. that really did their work as a crew in special team colors etc.

I also completely agree about how horrible a “data-driven” night club would be, but whether such places will materialize or not, I think the hypothetic scenario has much to say about the evolution of other data-driven commercial spaces. I hadn’t thought about the parallel to exercise culture and didn’t know the details about Crossfit but that is truly something to look into.

#3 rasmus on 4 February 2019 at 6:59 pm

kommentar från kugg (som inte lyckades posta själv eftersom något verkar trasigt i blogghelvetet):

Raves became semi automatic with the mixtape. The dj signed copies of jungle and techno music in the early 90s turned the boombox owner in to a auto dj curator. The mixtape culture is the branch on which common music curation grows.
The .pls and later the built in playlists of spotify and soundcloud has become the curators outlet.
Collections of digitized playlists is a strong input source for a machine learning algorithm. Understanding what one sound has to do with another is a hard task that solved elegantly must go beyond shared sources of likes. But the training material can be derived from playlists or likes and then transformed in to pattern matches on instruments, velocities, bpm, vocal sources/languages, samples, tempo, tone saturation etc. with more factors in large datasets more unexpected but yet satisfying results. There is a distinction of the music you listen to and the music you dance to. For this reason sites like mixesdb does shazaam on entire catalogs of famous dance sets. Mostly interpreted by copycats but eventually by ai’s to reproduce the mood of a known dj.
The problem of where to put your face is easily solved with visuals or art. Cultural explanations and producers of context is a missing factor. Hiphop has natural ties within its elements and aestetics. Such contextual ques make out a cultural experience. A dj-less rave is very possible with its other ques and aestetics.
Surveillance a side, a dj today does not have the opportunity to understand and adapt to a crowd as machine learning algorithm can. Partly because a dj can only choose to switch tune/make a choice at most once per minute and cannot make a choice from a databank larger then their recollection/collection. Besides djs rarely read their audience as much as they claim. Th djs mind is focused somewhere else entirely: on the ego. The real question is, where should djs put their egos when they are replaced?

#4 Alaoglu on 4 February 2019 at 7:41 pm

As a further parallel to exercise culture the data-driven nightclub would also, by necessity, need to collect data on its visitors to feed into its algorithms. This data will then, of course, also find other convenient uses: some dancers might touch (= sexual harassment) and the police will hence need high resolution video of everything that takes place at the data-driven nightclub. This video material could also be employed to combat other ills of the old dumb dancing culture: some participants might be under the influence of banned chemicals and surely an algorithm could be furnished to comb the video feed in real time to identify these. And lastly of course: criminal elements might infiltrate the dance floor and facial recognition could be employed to identify these as well. Everything, of course, to bring you the safest possible entertainment experience.

#5 rasmus on 5 February 2019 at 3:17 am

the safest possible entertainment experience! how wonderful

#6 Magnus A on 6 February 2019 at 10:07 am

In the current Swedish folk dance scene, live music is very common, and recorded music is not (it can in certain contexts be preferable for the dancers to sing something themselves, especially for beginners). I have identified three things that make this possible and desirable:

* There are quite many talented musicians that are willing to play, mostly for free, but invited musicians and more concert-like gatherings are also common.

* The dual performance aspect of the dancers performing for the musicians as much as the musicians performing for the dancers, at the same time as nobody really performs for anyone.

* The musicians adapt in real time to the dancers, not just by changing what piece of music they play next, but actually adapt the rhythm based on what the dancers are doing, making the feedback loop dancer-musician perform as one unit, with no one as spectator.

One thing that is common to both Swedish folk dance (especially polska) and rave is the trance element, making repetitive movements seemingly automatically and arriving at a state where it actually requires an effort to stop dancing. This element has of course been seen as the work of the devil by authorities (see Hårgasägnen and current Swedish authorities reactions to rave parties).

#7 COPYRIOT | Rapport från CTM och Transmediale 2019 on 8 February 2019 at 5:30 pm

[…] ← The next step in the automation of rave? (Transcript from a workshop at Transmediale) […]