Servus.at invited the artist jiawen uffline to participate in the art-science project Touching Thoughts (see also Davide Bevilacquas text), while their video essay »let yourself leak a little« will be presented in the upcoming STWST48x10 NOPE. For better readability, the title of the work is printed in plain text during the interview.
After Web 2.0, the internet is portrayed as a vessel that connects us with the world or helps us to find our »friends« in an online protected bubble, aka social media. However, it seems like it was »a matter of leak« to realize that our machines are already linked and constantly leaking. Would you like to tell us more about these two terms (link/leak)?
The word »link« has evolved from an early internet utopian dream to a sleek imagination of technology, facilitated by tech companies. This imagination is externalized and reinforced through visual representations of digital concepts and tech products. If I appropriate a term from McKenzie Wark, these »links« are vectorized, underpinned by business models and user personas. »Leak« is not as elegant as »link«, but for me, leak represents an unofficial form of connection -- unofficial because of a lack of access or freedom. Leak is a side channel, especially for marginalized groups, to have each other heard. Interestingly, the technology of imagination mentioned earlier also operates in a leaky way, and this leakiness is crucial to its functionality, transmission technology is so queer.
In your video essay »let yourself leak a little«, you say at one point that »we are leaking all the time as users and as citizens« and this is true. It seems that we have to leak in order to exist (both in physical and digital space) and we do it in many different ways, sometimes intentionally and sometimes unintentionally. What was the process of making this video essay and was there anything in particular that you were interested in or wanted to highlight?
(Bild: jiawen uffline)
Yes! The process of leak is so essential yet so hidden, In »let yourself leak a little«, I collected various instances of leaking. From the leaking Russian space station and Nord Stream gas pipe to surveillance in express houses, from Wikileaks and whistleblowers to radio signals, compression algorithms, communication protocols, and datasets, to illustrate how ubiquitous leak is in our daily life with or without us noticing. I actually had an Excel-file collecting all the leaking incidents happened during this period of time :)
It was very much inspired by Chun’s essay »Habits of Leaking: Of Sluts and Network Cards«. I compared digital leaks to bodily leaks, especially involuntary leaks when technology is used as a means of control, forcing people to leak. At the end, the investigation focus was shifted towards the periphery of the leak. Leaks become visible because someone captures and highlights them. Who are they? What about leaks that are »killed« by political »kill switches«?
For most people, the term leak has a negative connotation. Especially in the Internet space, it is associated with the unauthorised release of information. When it comes to users, women are particularly affected by the exposure of intimate data. However, in your video essay »let yourself leak a little« you show another side of leak. Would you like to tell us more about it?
»Leak« is often associated with a sense of brokenness, illegality, or shame, and is particularly linked to and weaponized towards women, both physically and in the sense of our »leaking habits.« But what is rarely talked about is that leaking is incredibly powerful. Leak is about making kin and caring, building sidechannels that circumvent authoritarian control, navigating to space and time that is hidden while it should not be, countering efficiency-oriented capitalism through messiness.
What I found very interesting (and what we find in your previous work »Filter Heuristics«) is the technical terminology used. Apart from leak, we have routers in promiscuous or monogamous mode, slutty network cards and dirty data. In Greece, we have a saying going like this: there are no dirty words, only dirty minds. What was your first thought when you first came across these terms and do you think we, as users need to abolish or reclaim these terms?
I have this feeling that naming significantly shapes our perceived reality, sometimes this perceived reality becomes our collective hallucination. Naming and terming are exercises of power, so observing those terms in technology/engineering is also looking into the power relationship, and the very prominent binary thinking in them, such as clean/dirty, pure/impure.
Reclaiming these terms is to take back the space rather than shying away from them, and letting the space become smaller and smaller. I observed people using a terrible term out loud, either until everyone realizes how messed up it is, or until it loses its nastiness.
Understandably, not all the data collected can be useful. The process of removing unwanted/dirty data from the datasets is called cleaning, because apparently to just delete them is not enough. In addition, the heuristic methods that are used to find these data, are based on the detection of certain »dirty« words in the datasets. However, a single word out of context can mean anything and nothing. Why is this method problematic and how do you approach/explore it in Filter Heuristics?
In the large scale datasets where entry-by-entry inspection is not feasible, they are usually cleaned by heuristic-based methods. A heuristic is a practical method that, not guaranteed to be optimal, perfect, or rational, but is »good enough« solving the problem at hand. Most of the heuristics are rule-based, or combined with previously trained detection models. But the concepts of »dirty«, or »harmful«, or »hate speech«, for example, are very nuanced and fluid. It depends on the context and on who is talking to whom. So rule-based bullet points or model from early years are hardly sufficient, if not more problematic with accumulated biases. In Filter Heuristics, I looked into 17 open or reproduced, thus open extraction-based, human-not-in-the-loop datasets and their papers, and asked whether a narrative of »cleaning« can emerge from technical papers. In recent times, companies have become increasingly reticent about disclosing the data used to train their models, mainly because of copyright issues which could lead to legal complications. So filter heuristics, in a way, has become a form of media archaeology research.
As individuals/users, we are given the option (or the illusion of it) to consent or not to the use of our data. Our decision is based on the information given at the time. We cannot foresee the future and of course there are scraping methods that can overcome this barrier. But when we browse the web, we are still citizens. Based on your research, what are your thoughts on the data collection (with or without consent)?
I used to be very critical on data collection and thought any kind of data collection was surveillance, especially when data is used to govern and mobilize people. Now I still think data itself is a static representation of becoming, no contingency or entanglement allowed to happen there. But I also have come to realize that if we want to make a change for our daily life, questions need to be framed more concretely and tangibly. For instance, who is collecting the data? Whose data is included and whose not? How is data collected? Is the collected data really relevant and necessary? Is the method to collect data pertinent and inclusive? Can individuals be identified from the collected data? Where is the data stored? Is the data encrypted? Who has access to it? With whom is the data shared? What are the purposes of data sharing? For how long will they be in use? When will the data expire? Data collection is so pervasive that every citizen or user needs to develop their own, what James Scott termed, »metis«, that is contextual, practical and flexible knowledge to deal with it.
You are clearly an artist with a research-based practice. On the occasion of your participation in the Touching Thoughts project, you had access to some physical/medical data. What motivated you to take part in Touching Thoughts and did you use the same research approach as in your previous projects?
In my research on leaks, I‘ve encountered connections between consciousness and leaking. I‘m very interested in the biological aspects of leaking - this is the initial connection with the Touching Thoughts project. The process for this project is very different from my previous work. We were provided with stunning medical imaging data, but my initial knowledge of the data was very limited. At first, I could only approach them visually, which made it difficult to use as a starting point. So instead, I started with biopolitics-related media theory to develop narratives, then fact-checked with scientists. Based on their feedback, I adjust some parts of my plan and iterate them each time when there are
new inputs.
The project is about making visible, it explores the recursive translation and simulation of data into flesh and flesh into data, in a way, it continues my research of technology as desire. Since it focuses on the body and data, I take an embodied perspective from the viewpoint of amorphous cellular existence. In the video, we will think with interstitial fluid about tissue staining and digital segmentation, memories and algorithms, with leaking neuron activity hitting in from time to time. This project is still in progress, but I am very much looking forward :)
