Mathias Klang | writing notes

The algorithm is a bad guide

Algorithms are flawed. And yet they seem to be the best technology companies have to offer. How many products claim to “learn from your behavior”? But what happens when I am the weaker part in this information exchange? There is no way I can know what gems are hidden in the database. So once again the products recommended to me are repetitive or shallow.

So it was great to stumble upon Susanna Leijonhufvud’s Liquid Streaming, a thesis on Spotify and the ways in which streaming music, selected by algorithm not only learns from our experiences, but more interestingly, acts to train us into being musical cyborgs (a la Haraway)

Starting from the human, the human subject can indeed start to act on the service by asking for some particular music. But then, as this music, this particular track, may be a part of a compilation such as an album or a playlist, the smart algorithms of the service, e.g. the machine, will start to generate suggestions of music back to the human subject. Naturally, the human subject can be in charge of the music that is presented to her by, for instance, skipping a tune, while listening on a pre-set playlist or a radio function. Still, the option in the first place is presented through a filtering that the machine has made, a filtering that is originally generated from previously streamed music or analysis of big data, e.g. other networked subject’s streamed music. Added to this description; if an input derives from the subject’s autonomous system, then the analogy of an actor-network is present on yet other layers. The actor-network of the musical cyborg work both within the subject itself, as the subject is not consistent with an identity as an entity, as well as between the subject and the smart musical cicerones.
Leijonhufvud (2018) Liquid Streaming p. 274

We often forget this feedback loop. Since we are trained by the algorithms the level of serendipity and growth is relatively low and we tend to be stuck in a seemingly narrow spiral – especially considering we are supposed to have access to an almost infinite amount of music.

As a newish Spotify user who is musically ignorant, I often find the algorithm to be laughably unhelpful since it does little to expand my horizons and as such is less of a cicerone (knowledgable guide) and more of a frustrated and frustrating gatekeeper.

It would be nice not to have the things I already know recommended to me ad infinitum, but rather show me things I have not seen or heard. Sure I may hate them but at least I may have the chance of expanding my repertoire.

Susanna Leijonhufvud (2018) Liquid Streaming: The Spotify Way To Music, Doctoral Thesis, Luleå University of Technology, (Fulltext here http://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1171660&dswid=-2263

Posted on May 1, 2019 by Mathias Klang

CFP Digital Ethics

Its time for another symposium on digital ethics, this will be the 9th year running. Here is the call for papers

We are looking for papers on digital ethics. Topics might include but are not limited to privacy, hate speech, fake news, platform ethics, AI/robotics/algorithms, predictive analytics, native advertising online, influencer endorsements, predictive analytics, VR, intellectual property, hacking, scamming, surveillance, information mining, data protection, shifting norms in journalism and advertising, transparency, digital citizenship, or anything else relating to ethical questions raised by digital technology. This is an interdisciplinary symposium, we welcome all backgrounds and approaches to research.

Researchers can either submit a proposal as a team (consisting of one junior and one senior scholar) or individually. In the latter case, organizers will match submitters up with a partner based on compatibility of the proposal. Five teams will be selected to present completed research at the symposium and critique each others’ work during five 75-minute sessions. After further review, the articles will be eligible for inclusion in a special issue of the Journal of Media Ethics.

Important dates:

Abstracts should propose original research that has not been presented or published elsewhere. The abstract should be between 500 and 1,000 words in length (not including references) and should include a discussion of the methodology used. Please also submit a current C.V. of all authors with the abstract. Abstracts are due on May 20, notifications will be sent out by June 5. Completed papers will be due by October 15.

For more information check out the Call for Abstracts: 9th Symposium on Digital Ethics

Posted on April 22, 2019 by Mathias Klang

Privacy Summer Course

This course will explore the effects of surveillance technologies from the everyday devices to the most sophisticated. It will analyze the effects of technology on society, culture and law. Students will gain insights into the impact of surveillance and technological empowerment on communication. Through the study, analysis and application of privacy & surveillance theory the participant will develop a firmer understanding of the role of surveillance on society and its impact on privacy.

In order to cover the topic in five weeks the course will cover one topic each week.

Privacy & Surveillance Theory,
Privacy, Surveillance & The Body
Privacy, Surveillance & The Home,
Privacy, Surveillance & The City,
Privacy, Surveillance & Digital Technology

Check out the syllabus here.

Posted on April 17, 2019 by Mathias Klang

Arrogance of Simplicity

This is from an article arguing against our fetishization of counting calories and BMI but the conclusion contains an important truth that should be applied more broadly:

Humans come in many shapes and sizes. Some people can truly eat whatever they desire and not gain a pound; others chew on leaves and remain portly. The lengths we go to calorie count isn’t a sign of health; it’s orthorexia, which creates cortisol, another factor in weight gain.

Don’t be fooled by the arrogance of simplicity.

Derek Beres Do Caliries Even Count

It’s all too easy to fall for simplistic solutions and slogans. It’s also a reminder to question established truths.

Posted on April 17, 2019 by Mathias Klang

NYC After Break

Spring break is over and the commute has returned. My tired body not used to the rigor of early mornings (how quickly we forget), made worse by the increased darkness of daylight savings. But each trip is its own reward, and this one didn’t disappoint.

One the subway a heavily tattooed man helped a blind man who was losing his balance, despite a grey sky, children celebrated spring by running and smiling, even the panhandler gave me a heartier smile as he shook the coins in his hand.

A rat ran along the side of the street and around the corner. At the site of a former launderette a crowd gathered to watch the removal of debris from the recent fire that had ravished the building.

A man with a black cowboy hat with a large gold brooch on the front was leaning against a wall while making a phone call. His snakeskin boots matching his brown snakeskin belt, which had a buckle that matched the gold on his hat.

There is always something happening here. It’s good to be back.

Posted on March 25, 2019 by Mathias Klang

Post Panopticon & Me

Teaching privacy and surveillance is a great reason to return to the theories that underpin everything, and I do enjoy introducing students to the history, function, and metaphor of the panopticon. While making myself rethink how it actually works.

The basic panopticon is nicely summarized by Jespersen et al in their 5-point list about the character of the panopticon in Surveillance, Persuasion, and Panopticon

The observer is not visible from the position of the observed;
The observed subject is kept conscious of being visible (which together with the principle immediately above in some cases makes it possible to omit the actual surveillance);
Surveillance is made simple and straightforward. This means that most surveillance functions can be automated;
Surveillance is depersonalized, because the observer’s identity is
unimportant. The resulting anonymous character of power actually gives Panopticism a democratic dimension, since anybody can in principle perform the observation required;
Panoptic surveillance can be very useful for research on human behaviour, since it due to its practice of observing people allows systematic collection of data on human life.

So last week I focused on privacy and surveillance in situations of “invisible” panopticons. Invisible panopticons could still be covered by point 2 above. In the panopticon we internalize the rules for fear of being watched, and ultimately punished for transgression. But I was trying to explain why there are situations of of self-surveillance where we could easily “misbehave” and nobody would punish us. A misbehavior that nobody cares about aside from maybe myself. If I binge cookies for dinner, drink wine for breakfast, watch trash tv, ignore my work etc nobody cares (unless its extreme) but I may punish myself. Where is the panopticon/power that controls my behavior.

In this case the panopticon (if we can claim there is one) is… my self image? We really have to contort Foucault’s ideas to make this fit under the panopticon. As he says in Discipline and Punish:

the Panopticon must not be understood as a dream building: it is the diagram of a mechanism of power reduced to its ideal form; its functioning, abstracted from any obstacle, resistance or friction, must be represented as a pure architectural and optical system: it is in fact a figure of political technology that may and must be detached from any specific use.

The power over ourselves in settings where there may be no real social harm if we were found out, is more about the conditioning and identities with which we conform. And our ability to act beyond them, to break free of the constraints of power represents the scope of agency we have.

To behave outside the norms that reside within me requires that I am aware of those norms and that I am comfortable to break those norms. That I recognize that there may be other actions I could be taking, and that I am comfortable enough to take them. So the way in which Butler argues that we are not determined by norms. We are determined by the repeated performance of norms. This is as Butler agues in the conclusion of Gender Troubles “…‘agency’, then, is to be located within the possibility of a variation on that repetition.”

Therefore I am being surveilled by the idea of me. How that me would behave in any given situation is limited by my ability to see myself behave.

Posted on February 12, 2019 by Mathias Klang

Function of University

College Council meetings are rarely the source for inspiration but it’s always a good idea to keep alert because suddenly there are glimmers of light even in administration. Yesterday in came in the form of this quote from W. E. B. Dubois:

The function of the university is not simply to teach breadwinning, or to furnish teachers for the public schools, or to be a centre of polite society; it is, above all, to be the organ of that fine adjustment between real life and the growing knowledge of life, an adjustment which forms the secret of civilization.

We are not here to be the centre of polite society…. Let that sink in.

Posted on January 25, 2019 by Mathias Klang

WTF? The great question of our age

How media has changed and what it means, by David Roberts

The internet changed all that. There are no longer supply constraints — it is trivially cheap and easy to publish something on the web — and there are virtually no constraints left on the supply of information. Libraries are online. Government records are online. Every public figure’s every move is blogged or tweeted.

Two things follow. First, with supply constraints gone, there is no reason to confine web journalism to the length and formal constraints of journalism developed for paper. Any story can be as long as it needs to be, whether it’s 200 words or 2,000. Not every journalist must choose between the view-from-nowhere voice of the objective journalist and stale aphorisms of major newspaper editorial pages. There is room for a greater variety of length, form, tone, voice, and subject on the web.

And second, there’s more need for explanation. Because they were supply constrained, newspapers and newspaper journalists focused on what was new, what just happened, the incremental development. But lots of times, readers had no way of making sense of those developments or contextualizing them. They were getting the leaves, but they’d never gotten the trunk.

Especially as information and incremental developments explode in quantity, there is increasing public hunger for understanding — not so much what happened, but what it means.

The great question of our age is simply, WTF? WTF isn’t asking after what happened. It’s easy to find out what happened these days. Rather, it’s pointing at what happened and asking, well … WTF?

What’s the deal with that? How does it work? How good or bad is it, really? How does it connect with these other things? What can we learn from its history?

People want to know how the world works. They want to know why the things that are happening are happening. They don’t stop wanting to learn when they get out of school.

So journalism is inevitably shifting. These days, it is less about producing new information than it is about gathering information already on the record, evaluating it, and explaining and contextualizing it for an audience, perhaps with some analysis and argumentation for good measure.

Don’t get me wrong: There’s still plenty of information to be dug up. Investigative journalism still very much exists, though it is under-funded everywhere. I look on it with great admiration and some awe, but it’s not what I do. And though many are loathe to admit it, it’s not what most US journalists do these days

Posted on December 29, 2018 by Mathias Klang

Global Surveillance

Again, if the Chinese surveillance state is more far-reaching than what we know to exist in the West—at the moment—this is largely because China has already done what the West is currently proposing to do.

Brian Hioe – Technoleviathan: China, Silicon Valley, and the rise of the global surveillance state.

Posted on November 29, 2018 by Mathias Klang