During the ALMAT residency I was exposed to new ways of thinking and doing, and appreciated the conversations that occurred during my time in Graz and Bergen. The conversations in the research catalogue as well as in person were stimulating and challenging for me. Conversations change who we are in the world, and I'm happy to have spoken with everyone involved and to be changed a little for it. For instance, the conversations themselves often involved questions of how our practice can become more radical, what is the correct way to attempt this in our own practices: for myself my radicalization lie in how a feminist materialism can affect the algorithmic through the implication of the body and the human in my technological assemblages/algorithms. As usual in my practice, the most influential and generative moments resulted from very simple, matter of fact statements and experiences, that were simply open enough for me to reflect upon. I will document these here.
Pre-Residency: I had been thinking a lot about began putting up some videos of ASMR on the research catalogue. I wasn't really sure where I was going with the ASMR stuff but it was an avenue I was interested in embodiment strategies for LSTM algorithms that could be "felt" through the voice ↗. It was a pretty new avenue for my work, but in general I was (am) attracted to ASMR as a means of aurally experiencing algorithms through a set of sound practices that developed in part through social media's big data, and also through the abstracted networked body's autonomous reactions. I thought that the amateur, feminine nature of ASMR was a nice counterpoint to so many digital aesthetics, and also felt ASMR to be a really "slow" medium that would allow one to experience optimization (or de-optimization?) through the affect of the voice. I also find the amateur nature of the practice attractive: ASMR can be practiced by anyone with a smart phone and network connection.
David had some interesting points about how the practice itself was woven through the big data of social media - it made me think about how the creators themselves exist at a curiously voice/voiceless threshold between the optimization of the practice through standardized techniques/practices established by community voting, and the voice that ASMRtists produce which is valued as residue of the personal and intimate. This voice/voicelessness is also at the threshold of the embodied optimization of big data ↗ in the reacts of viewers, likes and upvotes, and the embodied intimacy of whispers, clicks, and non-semantic content. I had this idea that I could combine the ASMR with the BioSynth research in a new work, but at a certain point I realized that I didn't have time to develop both the biodata concerns as well as the ASMR work. I made some tests while at ALMAT but didn't talk about it much for the rest of the residency. So if you see some things related to ASMR on the RC then this is why, I mostly dropped the conversation entirely, in fact deleted a lot of it because I didn't see it as relevant anymore. But it was there, this is my note about that.
First ALMAT meeting - The conversation about codes, rules and thresholds - I was thinking a lot about communication at this time, how the word "emotion" really conjures up aesthetic expectations that are both rapid (think about how quickly "I am sad" communicates) but constricting. Also, Western Classical music has cemented a lot of emotional tropes in music that are pervasive - minor keys for negative valence, major keys for positive valence. I know that I have been interested in working with emotional data, but see that working with these tropes was really frustrating as it isn't how the data or the body itself works at all. So I came into these first meetings thinkin about how rules both create language and patterns that require active making and breaking. What surprised me was how later in the residency these ideas broke away from music and started extending to the technology and code itself. So it's not the tool in itself that is to be villainized, but more to realize that we need to (in the words of Haraway) stay with the trouble of our tools, as their usefulness becomes less useful with time and they become more habitual than effective, like a dulling blade. I think about how I was once much more optimistic about open source code sharing as a means of making diversity and openness a part of technological culture. But with the acquisition of GitHub by Microsoft and the hardware dominance of Arduino microcontrollers, I think about how Maker Culture has become more of a thoughtless default, and how the very effectiveness of a technological practice (library, platform, microcontroller) might become in some ways so effective that it makes building other practices seem pointlessly difficult ("why rebuild the wheel?"). The answer isn't of course to stop making tools effective, but rather to note that the rise of the Maker movement and various open source communities should make some wary - even how one comes across certain tools and tutorials that are based on clicks, likes, and upvotes that we all depend on implies a tacit faith in search engines as a means of ranking our access to our tools ↗. These things are all embedded in a community of users, and I at this point I reiterate that conversations and encounters are mportant, that every conversation changes us. This interest in communication, practice, and how algorithms shape very meaningful communities online echoes my interest in ASMR practice as driven by these same factors: the way that algorithmic software online can steer human behavior in subtle and subconcious ways. This is already being discussed in some academic circles (this book "If...Then - Algorithmic Power and Politics by Taina Bucher was just released recently by Oxford University Press). It's a critical issue in contemporary life right now, this idea of the "filter bubble" or not having enough diversity in the way that information comes to you online, or enough questioning of tools: there is a flattening effect when things are always relevant, best. I think this is why art is important, as often only artists are crazy enough to try reinventing, or flirting with the ida of breaking, the wheel.
When music and speech, writing is renewed, it is a political act - I think about Jacques Attali who wrote in the 1970s that noise was the political economy of music, the scraping of power relations against one another. The idea is that the aesthetically comfortable, codified sound is resisted and transformed into something other through its interactions with other forces that dissent with the power given to these sounds, and that all transformation of music is political. In regards to algorithmic music, artists are working not only to map a dataset (or randomness itself) through code to create aural cohesion, but also to create this scraping, this difference. I observed during the residency we were all hunting for randomness and deviancy in our structures: maybe hoping our creations could speak back to us, say something we didn't already know. The irony of Narcissus gazing into a pond and looking for meaning in the small ripples in the water: wondering if the trick of the light might imply life in the limp reflection, when Echo's pleading voice exists just out of reach, ignored.
So even though I already "knew" these things, I was entering into the residency with an arpeggiator and a set of rules, thinking that more rules and more technique might create this noise in my embodied data. Every time I sat down to improve this system I was increasingly dissatisfied for reasons I couldn't articulate at the time, but are obvious to me now. So this set of thoughts really defined my first interactions with the residency.
David: Why not avoid If statements?
This was part of a longer conversation about categorization of data, but I found this half-joking suggestion an interesting exercise for experimentation in regards to mapping the emotional biodata into "musical" patterns. In particular the use of feedback and a kind of sympathy between data sets just made me more curious about how noise could be built into the structures rather than elaborate patterns and rules. So I began playing with timbres and filters that wouldn't really categorize the data so much as...allow it to define itself in a way I found interesting. I really also liked the suggestion that a good sonification draws attention to how you should listen to it.
Luc: "The hardware is a part of the musical algorithm"
I spent a good amount of time on the sensor hardware for respiration in my BioSynth system, about a week of tinkering and testing documented by a few notes and a completed circuit diagram. It's rare to see a In discussions about the potential meta structures for future BioSynth work, In particular it's difficult to explain concretely how I know that a signal is "working" or not - I was interested in Ron Kuivila's documentation during his residency, Hanns mentioned that a lot of his knowledge working with ultrasound sensors was tacit and based off of years of experimentation. Maybe this tacit understanding of certain forms of data articulates how some artists think of data through performance paradigms, or are more interested in behavior than end result. I suppose that Ron and I are both using our sensors in a way that is rooted in our experience watching the numbers float up and down - I thought about how this kind of "tacit relation" is a very artistic way of knowing data as a kind of performative movement rather than how some other disciplines might experience it more formally. So I'm curious if "knowing" the data in this way might also be generative. I hesitate to say that artists don't or shouldn't know anything about formulas or rules (I understand certain pratices and techniques very well!), I also understand that artists can bring many different ways of working with numbers and technologies, "wrong" ways of working (or accidentally correct ones?) or bring different goals to the table. There is potential in clumsy, untrained practices to manipulate data in new ways: I don't think it's something to glorify or aspire towards, more just that there can be value to developing one's own language or methods. It goes back to this idea of dull and sharp tools, under and overused methods: how do you know when it's the right time to reinvent the wheel?
During the residency I had the goal of "hardware mornings" and "software afternoons." I knew that these experiments with the hardware were a particular passion of mine and wanted to dedicate time to improving my software arpeggiator and rule system (which after a week I totally abandoned).
I had the idea of the hardware as divorced from the "music" but that the technology itself handled the biosignals correctly was of paramount importance in order to push embodiment. Luc Döbereiner (the next ALMAT resident) and I had a conversation about this concern of mine, and he simply responded that working on the sensor hardware was developing the musical algorithm. This resonated in a surprising way with me and a few things clicked. My engagement with electronic music, algorithms and new media in general has always been focused on radical physicality: something I see as a feminist-materialist resistance against the Enlightenment/Modernist fascination with the brain, "intelligence" and in particular "master works" (I don't really believe any of these things are as they seem, rather are more like patriarchal conventions that solidify power through aesthetics - Western musical expectations of emotion are shaped by these conventions and tropes as well). So I started working with my body(ies) and not against it/them, so to speak: trusting that the circuits and hardware would all speak on their own. I now understand my insistence on technological fidelity to the body as part of the musical/sonic work I make: if there are issues with the signal itself, it takes the user from the body and into a disconnected state I am not as interested in. Furthermore I feel resistant to imposing my musical ideas too strongly on the data before I give the body a chance to "speak", per se. In the past this has resulted in me somewhat feverishly working on the hardware to get it right, but if you can articulate this as part of your musical gesture, it actually becomes a part of the generative noise.
(Myself): "This sound is far less interesting when it isn't connected to my body."
Later in the residency, after the circuits and hardware had been refined, I began making tests using Max/MSP, as it was a far more flexible software system for experimentation than the hardware of the Teensy Audio system, which needed constant recompiling between each test.
I realized that I had this repeated experience of being satisfied sometimes in the compositional moment when I was attached to the biosensors and listening back to new code, but then listening back to recordings later, I realized the music wasn't interesting in the same way ↗. At first I thought this was just the typical compositional experience of not being able to listen to one's own work critically in the compositional moment, however after a while I finally realized the satisfaction I received while composing in this "wired" state was not a "musical" intrigue so much as an "embodied" one: the experience of the music was filtered through the connection to my body (as a musical instrument?). The rhythms that came from my amplified body were inspired my curiosity inasmuch as they were personal, and that the sensors worked properly, as it allowed an intrapersonal listening relationship to form. This relationship to sound more closely approaches the radical embodiment in sound and music I am interested in, and as I realized that the listening had to be intimately connected to the body, I also realized that the complexity of "music" I was previously attempting to create obscured this embodied relation through languages I was hoping to both use and defy. This kind of double-thinking/talking was far more complicated than throwing out my ego and making it simple again.
So I moved the sounds more towards "sonification," clearly making this connection to the body, although this embodied, extra-musical connection is difficult to document through sound, image or qualitative description. In this way I am interested now in how the human can enter the sonic algorithm by my shaping the terms of participation. This might relate a bit to Ron Kuivila's RC conversation on Hayles' cognitive assemblage. In particular I have been thinking lately about how relatively stupid our automated processes and AI are, and yet how effective contemporary algorithms on social media that sway elections are at using human intelligence to "close a loop" that AI cannot yet perform alone. When algorithms are combined with human participation, it seems that these cyborg machines can accomplish far more by voting, buying, and speaking to one another, than the dangerous AI that Elon Musk warns us about in the news.
{function: report}
This page is dedicated to Erin;s summary of her ALMAT residency period, her retrospective analysis of the experience in Graz and in Bergen and more general thoughts and considerations about the themes we touched during her residency.
{function: contextual}
Donna Haraway, Staying with the Trouble: Making Kin in the Chthulucene. Duke University Press 2016.
{kind: reference, persons: Donna Haraway}
{hhr, 10-Aug-2018}
this reminds me of a discussion a while back between me and David about the "tool status" of our software projects; what the merit of releasing something as open source is, and whether one should try to communicate a usable system or just the blueprint of it. Also related, some comments on the 'HAL' workshop last year in Brighton, in the Q & A with Palle Dahlstedt (⬀link).
{EG, 13-Aug-2018}
It was mostly a reaction against so much neural net work by scientists as well as artists that I've seen. We've all seen the machine learning artwork or viral video that demonstrates analysis through outputs trained on a data set...sets of photographs with labels, plays written by algorithms, percentages of certainty...the work typically expresses a humorous stance about how the analysis is inaccurate, or wrong in funny or uncomfortable ways. It's a bit nervous, we encounter the algorithm intellectually and evaluate its outputs, the unexpected outputs are often encoded by our reason as humorous or unexpected or novel but there's a judgement reaction here that I find boring and a bit human narcissistic. The learning itself is actually what's novel about machine learning, I'm most interested in artworks that demonstrate this process. Since ASMR asa
genre doesn't depend on narrative and is often noise/sound based, it can bypass this need for the listener to evaluate whether the outputs are good or bad or efficient or sensical, instead invites the listener to immerse their listening into the process and "feel" the style of learning in their body. This embodied focus gives the listener the patience to listen to the shifting connections of machine learning in a detached manner: the goal isn't really outputs/communication so much as embodied reaction.
{EG, 13-Aug-2018}
Yeah that's probably a misnomer - ASMR as a genre is optimized and refined through upvotes and clicks and comments that rank some practices as more effective than others - assumedly it's because certain videos make more "tingles" in users, so these "reacts" not only give feedback to individual ASMRtists but affect the rankings of videos in the YouTube search engine, ensuring that the first thing anyone sees is going to be well, populist. Which is hugely influential on what is produced...I'm not really sure how to articulate it, just I find it interesting that the resonance that media has on a large body of networked viewers creates this odd cybernetic system that reinforces itself. I don't think it's innocent (it actually reinforces many status quo systems of power) - I'm still thinking this through.
{EG, 12-Aug-2018}
By the same token I'm not used to/comfortable laying bare my source code for my artistic projects - I'm more in the blueprint camp when it comes to the art itself. Likely it's because much of what I'm doing isn't that complex on a technical level. So often people encounter new media work like a trick, asking "how does it work", learning about the trick, then they observe the system and evaluate its outputs. So new media art risks becoming something that is evaluated on this technical level that misses out on what an artistic encounter with an object in space is supposed to be (to me). If an artist approaches new media art formally and work with code on a plastic level that sharing code is useful and interesting but otherwise being exposed to the code buys into this output-evaluation stance I'm allergic to. I know there are arguments both ways and some artists are really interested in the code of others but I'd rather save these moments for like, one on one studio visits because exposure to code changes an audience's relationship one has to the work.
On the other hand I like to share the actual library with as much detail as possible because it's intended for other artists to reappropriate and use and develop.