e-flux Conversations has been closed to new contributions and will remain online as an archive. Check out our new platform for short-form writing, e-flux Notes.

e-flux conversations

Tyler Coburn: Live Coverage of "SECRET" tonight at e-flux

Shhhhh …

Tonight, Tyler Coburn (@tylercoburn) responds in real time to a roundtable between Hito Steyerl, Trevor Paglen, and Matteo Pasquinelli, moderated by e-flux journal editor Stephen Squibb. The conversation is organized as a special preview of the April issue of e-flux journal:

It’s not about the data or even access to the data. It’s about getting information from the truckloads of data … Developers, please help! We’re drowning (not waving) in a sea of data—with data, data everywhere, but not a drop of information.

Analysts are choking on intercepted communication. They need to unscramble, filter, decrypt, refine, and process “truckloads of data.” The focus moves from acquisition to discerning, from scarcity to overabundance, from adding on to filtering, from research to pattern recognition. This problem is not restricted to secret services. Even WikiLeaks Julian Assange states: “We are drowning in material.”

– Hito Steyerl, “A Sea of Data: Apophenia and Pattern (Mis-)Recognition,” forthcoming in e-flux journal no. 72 (password: secret)

To follow the discussion on live stream, tune in here

2 Likes

Good evening, e-locuters. Things are about to begin here at 311 E. Broadway, and contrary to the event title, I’ll be publicly responding throughout. “SECRET” draws on Hito Steyerl’s essay in the forthcoming e-flux journal, in which the writer and filmmaker (with her characteristic virtuosity and “secret ninja technique”) spirals out from a central problem of the information age: extracting meaningful signals from “a sea of data,” or noise. This task, as noted in the above post, beleaguers NSA contractors, hacktivists, and prosumers alike. Moreover, it gives rise to various coping mechanisms, or apophenia: “the perception of patterns within random data.” These, Steyerl notes, are far more consequential than seeing a duck-rabbit in a passing cloud. NSA’s SKYNET, for example, has treated as factual the patterns “extracted by assessing correlations and probabilities,” causing the potential misclassification of as many as 99,000 Pakistanis. With such importance vested on what Susan Schuppli calls “deadly algorithms,” the question inevitably arises: “Who is ‘signal, and who disposable ‘noise?’” Undoubtedly, Pasquinelli and Paglen will have some insights.

1 Like

Pasquinelli and Steyerl prepping

The evening takes form as three, fifteen-minute talks, followed by general discussion. Steyerl is up first, building both on her essay and a talk she recently gave at the Whitney for “Surviving Total Surveillance,” in conjunction with Laura Poitras’s current exhibition.

A paragon of apophenia is IBM’s Watson, originally developed to successfully compete on Jeopardy! (it won against two of the quiz show’s greatest champions in 2011). This impressive computer no longer dallies in mere trivia, but has gained the ability to “separate terrorists from refugees” by culling data from various sources, including the dark web. The combination of these datasets is equivalent to a score for each person in question—just one of the many scoring metrics (academic, credit, “social sincerity,” etc.) by which we increasingly find our informational identities determined. She summarizes: “score-ification is one form of pattern.” For more on the topic, see this excellent article by Frank Pasquale.

“Score-ificiation,” in short, exemplifies apophenia’s “magical thinking,” its meaning-making hallucinations yielding real political effects. Prophetically, Steyerl claims that “the state that’s emerging is a deep algorithmic state, in which there is no due process […] in which legislation is for a large part secret and proprietary.”

Where do you weigh in? The discussion of the biopolitical aspect of informatics is already in the discourse (see Orit Halpern’s Beautiful Data), but might we be on the horizon of informatic governmentality?

Paglen’s image game is on point.

1 Like

Trevor Paglen follows, voicing his suspicion that Steyerl invited him on the panel to discuss his current project on “machine vision.” As the project is in-process, Paglen politely dodges the subject, though does parse some of the essential differences between humans and machines. To put it bluntly, one of these things is made of meat and not the other.

Of course, there are other differences; for one, Paglen notes that the images humans make “seem to point to something out there in the world,” which we call “representation.” These images are never without frames, requiring us to consider the ways “representation has been hijacked” in the service of power.

Paglen moves onto “seeing machines,” crediting Harun Farocki with identifying “operational images,” integrated into machinic systems, decades before their ubiquitous presence in our lives (read Paglen discussing this topic at length in e-flux journal #59). “The majority of images in the world are made by machines, for machines," he remarks. “What does this mean for visual culture,” when human eyes have become the exception to the rule?

Obviously, Paglen is not the first to raise these topics, but with the intensification of machinic perception, they become all the more urgent. For a significant precedent, consider James Bridle’s writing on The New Aesthetic: “Each image is a link, hardcoded or imaginative, to other aspects of a far greater system, just as every web page and every essay, and every line of text written or quoted therein, is a link to other words, thoughts and ideas. Again, in this the New Aesthetic reproduces the structure and disposition of the network itself, as a form of critique.”

Pasquinelli showing a slide about CompStat, a police tool first created, in 1995, for the NYPD.

Paquinelli concludes, working from Foucault’s 1978 lectures at the Collège de France, later collected in the book Security, Territory, Population. “Foucault defined statistics—what we today call ‘data’—as one of the quintessential forms of power,” he remarks. Historically, this was tantamount to the “knowledge of the state,” which needed to be kept secret for the preservation of power. Pasquinelli doesn’t see this secrecy as very different than corporate secretism, from the informational to the infrastructural levels. But he does note that this alignment of statistics and state primarily belongs to an earlier paradigm of power.

The paradigm we’re currently living, he continues, is exemplified by such programs as the aforementioned SKYNET, which doesn’t record the content of conversations, but the metadata. It’s thus a system that “tracks not what people are saying, but doing.” This plays into a “new computational norm,” in which power is figured by algorithms—no longer by human agency.

A projection of animals and figures onto the heavens.

A nebula captured by the Nasa Hubble Telescope, run through Google’s DeepDream robot. The machines, not the humans, are now doing the projecting.

We’re in the Q&A. One audience member notes a theme that runs throughout the different presentations: “the topologies, geographies, the schema.” She’s also struck by the tendency of different interests—from The Intercept to BuzzFeed—to process the “sea of data” through geometric visualizations that ostensibly organize and make sense of this material.

For Pasquinelli, these are human interfaces that we cannot do without, particularly as we enter a time of “incredible abstract complexity.” “Visuality,” he notes, “is just an interface for complexity.”

Steyerl concedes the danger of infographics, in that they can easily come to be seen not as visualizations of data, but their reality. The question of how data is translated into visuals thus becomes paramount.

The audience member also notes that the speakers are addressing a particularly visual audience, implying that we’re predisposed to aesthetic solutions like data visualization. And indeed, the panel at large has given precedence to vision, seeing, and perception over other forms of machinic sensation. But if producing “counter-images,” as Steyerl remarks, won’t do much against the onslaught of machine-made images, might other artistic media provide more effective means of response? What, in short, are the aesthetic strategies that should be taken up within the realm of machinic sensation?