Gegen Google & Co: Track me not

Das ist durchaus ein zweischneidiges Schwert. Wenn ich einen Blick auf die aktuellen "Queries" werfe, erscheint dort der übliche Unsinn, der in Suchmaschinen abgefragt wird (Britney Spears u.ä.), aber eben auch Begriffe wie "Bit Torrent", "North Korea", "Syria" und "Bone Cancer". Will ich, dass Google meiner IP diese Mischung zuordnet? Man kann natürlich argumentieren, dass durch diese seltsamen Mischungen jede Erstellung eines Profils unmöglich gemacht wird - darauf setzen würde ich nicht.

Die Sache ist komplizierter als ein bloßer Verweis auf "Bit Torrent" etc. Hier weitere Infos
http://www.nyu.edu/projects/nissenbaum/papers/HoweNissTMN.2.8d.pdf

Ich habe Frau Professorin Helen Nissenbaum von der New York University/Faculty of Law mal um Auskunft gebeten. Hier ist die Antwort der Erfinderin des Tools

Dear Prof. Dr. Hoeren:
I wanted to check with Daniel Howe, the technical expert on the project, before getting back to you. This is actually an interesting question both technically and philosophically. We've addressed the philosophical part of it in a paper which will be appearing soon but in the meantime, let me try to give abbreviated answers to both. 
On the technical side:  a search query to Google is approximately 5k. For the average user, making 10 actual searches per day and using the default settings on TMN, their total usage would amount to 180-200 queries per day, therefore, on the order of 1mb per day.  By comparison, the average song is approximately 5mb, and a compressed movie is 1GB, which is a factor of 1000 over TMN and a DVD is approximately 6GB.  
Thus, TMN does use network and server resources but it's use is relatively light, by comparison with what is now considered quite normal usage of the shared resource of the Internet and Web.  
But there are still philosophical questions to consider. Quite apart from TMN, in general, there is a problem with shared resources that if everyone, simultaneously draws on them they may not be able to bear the weight: for example,  if the whole of China, at the same time, sent a batch of "genuine" queries to Google, or downloaded a song, movie or DVD, they could bring down the Net or the server in question (unless there were careful timesharing capacities built in that could handle such congestion.)  I suspect that the vitality of these resources depends to some extent on the fact that we do not, as a matter of fact, act in these ways. Further, because memory is inexpensive, providers of services keep adding to their stock of memory as the demand for bandwidth radically increases with increasing interest in large files and increasingly active usage.  (I'm not a network engineer, so all of this is very schematic!) 
That said, TMN usage is very small, almost imperceptible in the big scheme of things but were it to increase significantly (well before it would big enough to threaten service), it would signal to Google, and to all search engines, generally, that people believe they should not be be personally monitored in the practice of Web search. And it is our hope that when this message is acknowledged, these companies will offer a more credible privacy policy. Thus, in a sense, the ultimate success of TMN will be to make ourselves unnecessary, and that will be the case when our message comes across loud and clear.
However there is one remaining philosophical aspect to this argument. One reason people might criticize TMN when they wouldn't criticize the down and uploading of far far bigger files, or they don't criticize trivial searches for material of little social value, is that they somehow see TMN as illegitimate, as a "waste" in the first place. We really want to highlight and question this assumption for who determines what is and isn't legitimate use? Is it merely what one or another organization, corporation, government, etc. wants, or is there some more democratic approach to determining this? We've argued that in circumstances where privacy is not adequately protected, then the use of bandwidth and server resources in order to protect privacy is quite legitimate.  
I suspect there is more that can and should be said about these issues, but I'll stop here and hope it goes part way to answering your and your colleagues' concerns.
Best wishes,
Helen Nissenbaum

Bitte entschuldigen Sie, Prof. Hoeren, ich wollte Ihnen nicht Ihren Titel abschneiden! ;)

Das freut mich! :-) Ich bin gespannt auf die Antwort aus New York!

"die zufälligen Suchanfragen scheinen allein englischsprachig zu sein"

Die Begriffe werden doch local in der .../TrackMeNot/tmn_seeds.txt gespeichert und sollten wohl auch leicht zu editieren sein.

Man könnte zB eine Liste mit den Top100 deutschen Google Hits schreiben. Das wäre wohl auch besser so, weil sonst google einfach diese Begriffe in der jetzigen Datei aussortieren könnte, um die Profilbildung nicht zu verwässern.