Knowledge Machines between Freedom and Control
organized by IMA Institut for Media Archeology in cooperation with Theo Röhle (author from “Der Google-Komplex. Über Macht im Zeitalter des Internets”)
DO 06 10 10:00 – open end (closed)
FR 07 10 09:30 – 18:00 Presentations (public, no registration needed, free entree)
09:30 – 11:00 Index: Introductiory speech Katja Mayer, worklab presentations, Discussion
11:30 – 13:00 Algorithm: Introductiory speech Malte Ziewitz, worklab presentations, Discussion
14:00 – 15:30 Profil: Introductiory speech Astrid Mager, Worklab presentations, Discussion
16:00 – 18:00 Panel discussion moderated by Sonja Bettel
with Konrad Becker (World-Information.Org), Wolfgang Sander-Beuermann (SuMa-eV), Astrid Mager, Katja Mayer, Theo Röhle, Malte Ziewitz, u.a.,
Participants: Seppo Gründler, Seda Gürses, Joris van Hoboken, Jogi Hofmüller, Reni Hofmüller, Astrid Mager, Katja Mayer, Hannes Carl Meyer, Nicole Pruckermayr, Ushi Reiter, Theo Röhle, Elisabeth Schimana, Martin Schitter, Malte Ziewitz, IOhannes Zmölnig
Kulturfabrik Hainburg
Kulturplatz 1 bzw. Donaulände 33
2410 Hainburg/Donau
The symposium is divided into a one-day worklab, in which the participants will examine the three topics index, algorithm, and profile – and a public presentation day on which the groups will present their results to a broad audience; individual presentations are also possible.
Today search engines constitute the most important form of access to information on the Internet. In the hands of users they are a powerful tool for locating information. Thanks to search engines we can browse even the most remote corners of the World Wide Web, without them we would get lost in the flood of available information. The free exchange of information and a democratic basis for discussion on the Web would be unthinkable without this access. And yet search engines are not neutral tools: embedded in economic, political, and technical conditions, they are machines that produce certain kinds of knowledge. This is why we must ask ourselves who makes search possibilities available, how is this access to information influenced, and which demands have an impact on technology development. The aim of the symposium is to take inventory of the current Internet search cultures. Together scientists, artists, and engineers will explore the paths online searches have taken in recent years and identify which new developments can be coaxed out of this new charged realm between the poles of freedom and control.
The initiative for this symposium arose out of two art projects: the insight-Tower a World Machine by IMA Institute of Media Archeology (Seppo Gründler, Nicole Pruckermayr, Elisabeth Schimana, Martin Schitter) that was commissioned by the Lower Austrian state exhibition 2011, and Modells ¬ the perfect profile, a LED Installation by Nicole Pruckermayr and Elisabeth Schimana that was commissioned by Art in Public Space, Lower Austria. The symposium was conceived in collaboration with Theo Röhle, author of the book Der Google Komplex. Über Macht im Zeitalter des Internets [The Google Complex. On power in the Internet age]. His research findings have influenced these projects strongly, producing the need to work through and explore in detail the continuously arising questions in conjunction with scientists, engineers, and other artists.
The topics:
1. Index
The first step in making the World Wide Web searchable is to compile and order an enormous amount of data so that it can be quickly and efficiently processed. Nowadays, this basic search infrastructure the index can only be built at an immense financial expense. The main index of the Internet belongs to Google, a private company with a monopoly and a large technological lead over its competitors. What consequences does this have on information access? Is it enough for a player to make a main index publically available? Could access to this index be controlled differently, e.g. by reliable programming interfaces, long-term contracts with research facilities or the like? Or do we need to start thinking about alternative indices? If there is to be something “beyond” Google in the near future. How would an alternative of this kind be set up – government-subsidized, peer-to-peer, specified indices for individual interest groups?
2. Algorithm
Search engines reduce a data universe of 2 trillion websites to the comprehensible number of ten results. This reduction is made possible by a multitude of algorithms selecting, sorting, ranking, and comparing behind the scenes. The advantage of these processes is obvious: without them the Web would be simply unnavigable. Just weeding out the spam would demand more time and cognitive capacity than users have at their disposal. Search algorithms allow self-determined action on the Internet but also contain aspects of control over this action. Even if we recognize some of these search algorithms and can guess at others, they still operate, for the most part, behind the scenes. It is therefore very important that we give thought to this prevailing monoculture of algorithms. What kinds of knowledge are being produced, which ones get lost? Is there enough transparency or should more criteria be revealed? Which experimental approaches are conceivable – can other regimes of visibility be created through alternative algorithms? Who is responsible for these developments?
3. Profile
More and more user data pours into the search geographic and temporal information is considered, individual search histories are analyzed, and if a page gets a lot of hits, it appears higher up in the list of results. Individual and group-based profiles make sure the results are adjusted to personal preferences. Access to knowledge is more and more effortless; the search engine seems to know what we want to know. These processes relieve the user, the Web adapts in real time to their demands and needs. Users are faced with fewer sorting and selecting decisions. What does this mean in terms of access to information a tendency toward focusing and concentration or rather toward the breaking up into hermetic spheres of information? From a company perspective profiles are interesting above all for commercial purposes they are useful in calculating, evaluating, and optimizing the probability of individual purchase decisions. Which correlations end up catching our attention through these “profiling machines” (Greg Elmer) and which ones get lost along the way? Can there be data mining beyond economic rationalities? What would a self-determined handling of profiles look like?