Pitting human performer against machine improvisor, bad decisions (2019) for Chaosflöte and AIYA shows a more playful and mischevious side of the machine. Three different scenes within the piece reflect different angles, different scenes of the 'identity' of the machine. It is referred to in a partially personified way, but ultimately its formlessness eludes our attempts to describe it. As the performer navigates the virtuality of the piece, she tries to make the machine “behave”…but bad decisions teases the notion that, rather, machine foul play and the unexpected disruptions brought on by technology can be a more interesting inspiration for a broader palette of chaotic expression.

composition/programming/performance by Melody Chua
in collaboration with XUAN, visuals

photo: Betty Fleck

3.4 bad decisions (aug. 2019)

bad decisions (August 2019) was the first piece I created and performed with AIYA. I kept much of the same mechanics as the previous version, but here, AIYA was put into a context of an audiovisual composition of a set length (the piece lasted 7 minutes 29 seconds). Here, I improvised with my live-generated electronic sounds and visuals alongside fixed media that framed the whole work. The distribution of fixed electronic elements vs. live was approximately 50/50.

Still not able to let go of my control issues, the on/off switch to the machine remained present in this work. The interesting and perhaps problematic part of the on/off switch to the machine was that it had the effect of reducing the machine from being an improvisation machine with perceived agency to resembling something more like a very complex sound effect module (or tool). It was because of the fact that I had such absolute control over it (turning it on and off), that the machine no longer retained any of its perceived agency.

I wondered to myself, is it such that in trying to fight against the feeling of vulnerability (due to lack of control), one is willing to compromise the agency of the Other? It seems somehow obvious that this was the case with my relationship with the machine improviser at this point in its development. Perhaps it is due to its nature as a machine, as a software patch I programmed myself, that I somehow felt that it was appropriate for me to sacrifice the machine's agency in exchange for a reprieve from this vulnerability. But how willing would I be to have this same exchange and reprieve, if it were to be with another human improviser? Do I, and other musicians, already do this, just disguised in the form of extensive rehearsals together, or in the form of eye contact and the attempt to read a cue or give a cue when we want to make an entrance or end the improvisation? Isn't our desire as chamber musicians to be synchronized in action and intent, to eliminate the vulnerability of musical "glitch" as much as possible?

It is of course not possible to literally turn a person on or off as one might do with a machine, but if we had the power to do so, would we do it if it would musically enhance the performance? Perhaps this goes back to the issue of trust and developing a musical history together...as the next section continues.

Framing: Between composition and improvisation
On the one hand, the precomposed fixed media helped to drive a consistent narrative of the performance, which might have otherwise been less concretely conveyed with an improvisation alone. The mix of fixed and live elements seemed to achieve a decent balance between structural reliability of the piece and the excitement of a spontaneous performative moment. On the other hand, I became somewhat conflicted about the nature of the work and its ambiguousness to the audience between what was done with fixed media and what was done live. The overarching narrative of the work was meant to “reflect different angles, different scenes of the ‘identity’ of the machine.” If the improvisation machine is by nature the opposite of fixed media, why would I use fixed media elements to attempt to convey its identity?

(above) signal flow of this version of AIYA, as used in bad decisions
(above) the face of Bec Plexus in bad decisions
(above) full recording of bad decisions

Putting a face to the machine
bad decisions was also the first piece where I used a visual medium to attempt to represent the improvisation machine. I worked with a visual artist (XUAN), who gravitated towards anthropomorphic visual elements, with the middle section of the piece even borrowing the face of her friend, singer and composer Bec Plexus.

The narrative of the piece was meant to show the improvisation through three different angles that reflected a sense of play, mischievousness, and "chaotic expression." When discussing these elements with XUAN, it was interesting that one of her more intuitive inclinations was to use pictures of Bec Plexus enacting various different facial expressions, as if it would be more difficult to represent these human-constructed "character traits" without the language of anthropomorphic elements.

In the end, despite our efforts to represent these human-constructed character traits in the machine, I feel that my own character traits—my playing aesthetic as a musician—still seemed to overpower the aesthetic of the machine. No matter how colorful and seemingly playful the visuals were, my classically-trained "virtuoso" mannerisms seemed to overshadow the attempted machine-focused narrative elements of the work. As the program notes state, the work was indeed "pitting human performer against machine improviser." The compositional efforts of XUAN and I to represent the machine were in direct opposition to the live improvisation efforts of the human performer to assert her own representation.

next ►◄ back