Hi, I am still thinking of a good way to frame the questions for the coming week, but wanted to lay a few points out in the meantime.
First, a quick recap of the Edge Tools event. For convenience I interpret the term as applications of digital methodologies (data mining, storytelling, etc) to social sciences research and beyond, in “an increasingly complex and connected world” characterized by big data. The topics addressed included:
- monitoring social media data for marketing
- applying multiple data sources and flexible human organization to military operations
- locating potential national security threats by data mining children’s stories, collected using a mobile survey platform
- analyzing diverse forms of information ranging from social media to humanitarian information in order to track the activities of military forces in Syria
- crowdfunding a successful game through data analysis of successful precedents and storytelling
That all speakers talked about either military or marketing purposes speak to the general orientation of this event. Not surprising since the military and the economic sector have an important role in the history of computation, to say the least. Underlying all of these examples is the current technological landscape where the world becomes data, both because everything seems to be represented as data and tools are developed to be able to deal with the data. What flows alongside this current is the belief that this change offers the potential for a more profound, direct and wider understanding and/or interaction with the world.
Which is exactly what surveillance is about. The extensive communications monitoring by the NSA is an attempt to better handle the soaring amount of information, much as search engines allow one to navigate the web. Bilton describes how the availability of data and tools for gathering it is a given nowadays, quoting Wizner’s claim that “tracking technologies have outpaced democratic controls.” In addition to technical possibility, the industrial structure also facilitates surveillance. Wu’s article provides historic examples of the cooperation between the government and corporate monopolies, which was also repeated in the NSA case. Both articles point to the idea that the current situation is making it very easy to monitor people, be it the technological development or the industrial structure. And as Grimmelman lays out using examples of Google search, the design, use and regulation of a system, by companies, users and governments, are not neutral; all activities have political implications. When the design of the systems we use move more and more towards keeping our data on someone else’s hard drive (the cloud), there is yet another tradeoff between convenience of use and risk of surveillance.
Achim!
Thanks for the analysis of Edge Tools (I’m obviously still peddling this term at all social gatherings). I do think that this presentation of tools (the SSRC panel specifically) faces the same danger that abstraction to data tends to face — by isolating the tools (as one isolates each piece of “data”) and putting them in a parallel structure, the organizers level these tools and procedures to a baseline “objective” or intent that feels suspicious. It’s hard to see good in tools when the mission to find the similarities in these tools essentializes them in troubling ways. I do think that the tools taken alone and along their own aims might be a fairer way to consider the methods presented. There are certainly tradeoffs happening, but cynicism strikes me as too easy. As with social media, we sometimes want things to be made easier for us, “safer” for us (which is sometimes actual safety). I wonder how much of our uneasiness about the ramifications of military surveillance & corporate tracking vs self-documentation/archiving has to do with the not-knowing enough to be able to distinguish which data have integrity and which do not. Does part of our general anxiety about these tools come from the gloss of a sociological presentation, a surface understanding we get from seeing things in ten minute clips and sound bites?
Excited for more conversation.
-Jojo
Thank you, Achim, for beginning the discussion this week. I feel that the readings this week which included videos, articles, and for some of us, even a museum exhibit, were quite broad in terms of covering the different issues of surveillance. From an artistic perspective, Poitras’ work forces the viewer to grapple with the idea of the unseen watching the seen, and in turn, being watched by other unseen entities. These sort of layers of surveillance cast doubt on the possibility of ever protecting one’s privacy.
In reading the article by Wu, and the description of telephone operators, it occurred to me that this idea of privacy is somewhat newer concern that has only been brought about by the persistence of our personal records in the form of electronic data. That is to say, that it is not so much that our information could be available to others at any given time, but that we may at some point be held accountable for actions taken at any point of time in the past and that it is traceable. While reading the article by Bilton, it occurred to me that we have recently seen a corporate entity resist the surveillance of the federal government in the recent case of Apple vs. FBI on phone hacking. Of course there are likely numerous cases when sharing of data occurred between corporate and government entities, I would almost speculate there will be more cases like the former in years to come.
The presentation by Donovan was informative in the way it attempted to unveil the mystery of the internet and data sharing. I doubt that it is just teenagers who share a perception of cyberspace as somewhat mythical, and I wish that it were possible to educate people about online best practices for safety and security. While watching this presentation, however, I also felt tempted by the idea that such information, particularly about the storage practices of massive amounts of personal data, is better left unseen and unheard of. I see this as similar to the dilemma that the Snowden case brought upon the general public: to what extent should we be responsible for educating ourselves and others about surveillance if doing so increases the chances of risking our own privacy and that of others?
Achim,
Really cogent connections between the Edge Tools talk and the variety of materials we had to engage with this week. Thanks for reminding us about the integral role that the military and economic sectors have played in the history of the internet and communications technologies, which certainly throws some cold water on my initial shocked and impassioned response to the talk — none of this is new. I am struck by this point that you made: “What flows alongside this current is the belief that this change [world as data] offers the potential for a more profound, direct and wider understanding and/or interaction with the world.” What was of course unfortunate about the talk is that by staying within those realms of commerce and the military, it did not explore offer many more profound, direct, and wider understandings. I have been thinking a great deal lately about how computer science education might be taught from more artistic, civic, community or activist perspectives, and the kinds of ideologies and material supports needed to sustain such pedagogical visions. What kinds of talks might have been given about how data mining and analysis serves a social justice purpose? The one example I can think of is Ushahidi, which has been successfully used in natural disasters and for organizing (https://www.ushahidi.com/) — but as Tracy mentioned after the talk last week, if your fellow organizers use a tool to track you, then the authorities can use the same tool. Is a social justice approach to use of data an oxymoron? The Deep Lab folks are attempting to answer that question in their work, but I haven’t had enough time to truly explore that site yet, so I don’t know what their approach to it is yet…
I was also struck by this general idea from Citizenfour: Metadata builds a perception which can tell a story that isn’t necessarily true. How do we keep that point always in mind, so we don’t fetishize big data analyses? Even as I was deeply troubled by the ways that the recorded stories of young people were being mined for evidence of radicalism, and certain that racial profiling would result from an application of information gleaned from that study, I was captivated by the methodology used. The stories behind the algorithms are all constructed by humans, but when it seems like algorithms draw our conclusions, the allure of the positivist is at its strongest.
SV
Thanks for the comments, Jojo and Teresa. A few quick notes before the conversation tomorrow:
Donovan’s presentation employed concepts such as dataveillance, smart city and proprietary ecology in order to address the corporate and governmental practices of gathering and utilizing data on people, and how the production and aggregation of that data is disconnected from the actual users. In his words: “who gets to produce information, where and for what purposes?”
Teresa, like the Apple vs. FBI case, I do hope that more companies will stand up against invasive governmental demands, and I think they should be urged to. But this is probably not most effective if legal requirements demand that companies comply; also, the proprietary ownership of user information, which is one of the technical conditions that facilitate such surveillance, remains unchallenged. As illustrated in the iPhone location data issue, where the company rolled out a security update that removed the tracking data out of the user’s sight instead of doing away with the actual practice, companies also have their way of being invasive. Donovan’s claim that there are no private solutions to public problems seems like a good way to frame this.
I was left a bit unclear when you talked about the threats of issue-raising at the end. My reading of your last paragraph is that you are concerned that as more people learn how everyone’s data is everywhere, and how it can be used to track them, they might be encouraged to engage in surveillance themselves? In that case, I would argue that education about this matter has an effect of leveling the ground, since the power relationship among individual users, companies and governments is not equal and neither are the collection and usage of data in between them. Also, I feel that knowledgeability about surveillance can do more good than harm, both in terms of individual practices like encrypting your email and in terms of formulating thoughts and opinions about technical issues (that are not just a given technical condition per se, but rather a social situation).
Zittrain’s main notion is generativity, which is “a system’s capacity to produce unanticipated change through unfiltered contributions from broad and varied audiences.” A generative architecture is one that is leveraging, adaptable, easy to master, accessible, and transferable—one that facilitates change. He argues that the “information technology ecosystem functions best with generative technology at its core.” This has its vulnerability, as it inherently is a model where things can get started without a comprehensive blueprint, it also “facilitates the tools and relationships through which people can meet the problems when first-round success causes them to materialize.”
I think that on the other hand of the increased world representation as data, there exists the obfuscation of the mechanism required to handle this data. The computational power and scale needed for such operation is one of the factors contributing to the fact that these data and algorithms are developed by private companies. Only part of those are open sourced, and even open sourced tools and datasets are maybe too big to grasp by an individual in a comprehensive way. So while methods like data visualization and mining seems to make direct interaction and perception possible, it also introduces an even bigger level of abstraction—where large datasets are converted to numbers and diverse features compressed into a couple axes in order to help probabilistic decision-making. In Zittrain’s hourglass illustration of different layers of technology, it is as if the middle and lower part of the hourglass kept getting bigger, darker, and ruled by quantum physics. In that sense, making things more generative will be a big challenge.
Oops missed your comment, Sara.
Your last point is very important when thinking about recent developments in CS, which in some way feels like a prejudice-machine building contest. To quote a tweet from my former teacher: “a cool thing to remember is that whenever someone says ‘A.I.’ what they’re really talking about is ‘a computer program someone wrote.'” (https://twitter.com/aparrish/status/713703634635268096)
The two keynotes from the FACETS conference (http://www.facets-con.com/#schedule) were relevant to data tools and social justice. The videos should hopefully be up soon: http://livestream.com/internetsociety/facetscon
Data & Society (http://www.datasociety.net/) is also a great resource/institution on the subject.
Talk more tomorrow!
Great discussion in class tonight, everyone. Here’s that link to the Reddit thread about the person who couldn’t stop trolling so his wife split with him: https://www.reddit.com/r/relationships/comments/2cpgy4/update_i_32f_just_discovered_my_husband_34_of_six/
Another relevant material that recently came out, this one with regards to activism:
(I really like the drawing of internet in the linked page)
“Tactical Tech is excited to share the outcomes of two research projects looking at how human rights defenders learn about and adopt digital security practices.
Working in digital security training for over a decade now, we have now stepped back to document the mechanics of digital security training and the contexts in which human rights defenders encounter it.”
https://tacticaltech.org/news/new-published-research-digital-security-human-rights