ABSTRACT
The following chapter argues that theorizing algorithmic cultures as being based upon functional, pure and clean networks, hardware and data systems, might conceal its more dirty, dark, dysfunctional and pathological aspects. These negative aspects are too often downplayed, but certainly belong to the cultures in which algorithms are operatively taking part. An inquiry into algorithms that focuses on these cultural aspects should include not only man-made realms, but also the important perspectives of non-human agency, which, while indeed human-made, has gained an effectiveness that far exceeds human culture, communication and aesthetics; self-organized technological processes are an example. Extending the notion of culture towards ecosystem thinking also leads the inquiry to the pathological side of algorithmic culture. Furthermore, an ecosystem is usually considered more living, complicated, sensitive to changes and stimuli than a culture. Its many agents operate in between control and non-control. Algorithmic ecosystems, then, consist of humans and nonhumans, respectively non-machines and machines, constantly exchanging signals not only with one other, but also with their environments and their objects, processes, materials and bodies. Such an open-ended approach resonates strongly with MacKenzie’s concept of agencement as “combinations of human beings, material objects, technical systems, texts, algorithms and so on” (2009, 4). Along these lines, the first of overall four sections will explain the term algorhythm, a neologism that not only enables a specific and at the same time far-reaching understanding of algorithmic culture, namely its rhythmicity, materiality and physicality, but also enables a time-critical approach into its pathologies. The second section will offer short historical accounts of algorithmically-induced resource mismanagement, network breakdowns and other manifestations of unintended bad programming and design, beginning with the 1960s and its first attempts to schedule algorithms in shared systems. It will furthermore explore early issues of the ARPAnet and briefly examine the genealogy of computer viruses. The third section will describe more in detail the socalled AT&T Crash of January 1990, when the long-distance American telephone network was defective for almost a full day and caused considerable economic damage. The final section will develop the notion of neoliberal
pathogenesis as a symptom of our current neoliberal, free market society, which, from a media theoretical and historical perspective, is deeply rooted in the above-mentioned context of early distributed networks and computing.
