Review: Terms of Service: Understanding our Role in the World of Big Data (Al Jazeera America)

July 2, 2015

Late last year, Al Jazeera America released Terms of Service: Understanding our Role in the World of Big Data, an feature published as an almost fifty-page "graphic" or "comic novella," as the network calls it. This caught my attention not in the least both because of the format and because of an interest in the topic, but also because of the interesting way that Al Jazeera released it: simultaneously as a "web comic" (comic readable via a website), as a free title on iBooks and Google Play, and as a downloadable ebook and PDF (not in a comic archive format, however). This is of course a model easier to follow for a network giving away an article than for the mainstream comics publishers that want to both sell but also retain rights management for their digital comics, but I still thought the model was interesting. And it's pretty hard to argue with free.

As a matter of fact, that it's pretty hard to argue with free -- or society increasingly pushes us not to argue with free -- is one of the first and strongest points that reporter Michael Keller and cartoonist Josh Neufeld make in the comic. Early on they talk with University of Colorado professor Scott Peppet, whose "unravelling theory" essentially looks at how, when everyone is sharing on social media, the people who aren't sharing are the ones considered "weird" or outside societal norms. Even as we might all generally agree that sharing too much about oneself online is "bad," there's a cumulative effect of people participating in social media because other people in their social circles already do. And the fee for entry is, of course, deceptively low -- free, monetarily, though the cost is less clear in terms of the value of personal data shared online.

The concrete costs of "big data" -- "big data" in terms of all the information that social media sites, FourSquare, Fitbit, insurance companies, etc. have about a person, and "costs" in the sense of "how this can all be problematic" -- are difficult even for the authors to state, because a lot of what the authors cover here are concerns about what might happen, not necessarily what is happening. They offer one example where pedestrian deaths could perhaps be eliminated if all cars were outfitted with sensors and all pedestrians were outfitted with sensors. The big question here is what kind of society do we want to have, and in a more specific sense how much privacy is it worth giving up to prevent loss of life.

These are not "today problems," as the authors describe, but rather "tomorrow problems" -- decisions our society could one day have to make based on the trends of today (at another point, a character thinks about how the retinal scan he allows today might come to haunt him twenty years hence). I think the vagueness of the threat is one of the reasons, as the authors point out, it's hard to get data laws through Congress and instead the US has a "patchwork" of state laws. Interestingly, one law that did pass was a data breach law that required businesses to tell consumers if their information was hacked. Unfortunately or not, this seems rather routine to us these days, and I was frankly surprised to learn that it had been codified into law -- and in the reverse, to understand that if it hadn't been codified into law, businesses would have no responsibility to let customers know about breaches.

Also interesting is the authors' idea that all this big data might lead to changes in the way insurance works. They focus specifically on the Progressive Snapshot program, in which drivers plug a device into their car's data port and their insurance rate can go up or down depending on what kind of driver they are. When presumably a minority of people have some kind of "car tracker," then it only affects their individual rates. If everyone had a "car tracker," however, and all insurance companies could accurately determine an individual driver's risk, then the concept of shared risk goes away. The even greater problem is should that kind of data be misused or misinterpreted; elsewhere the authors offer the hypothetical example of a person who can't get a bank loan because their Fitbit shows they haven't been exercising or car shows they've been driving erratically, because the bank imagines a correlation between this other data and financial trustworthiness.

This is complicated, to be sure, and Keller and Neufeld's comic is concise and funny and works well to make these issues clear. At the same time, I'd like to believe that it is Keller (if we imagine him as "the writer") making these issues clear that makes them clear, and not just the fact that they're in comic format. I bristled a little at the editors' statement in the prose introduction that "we believe many folks want to learn more about these issues but are turned off by often dense and jargon-laden coverage. So we made a comic," as if a comic could not also be dense and jargon-laden. They're conflating medium with message here; it's the message that's clear and the comic that provides the vehicle for presenting the message, not that a comic is by definition simplistic.

Indeed I think what benefits the authors here is not the comic format, but the ability to present words and pictures simultaneously. In this way Terms of Service reminded me of An Inconvenient Truth (and Al Gore cameos in the novella to boot) or a Morgan Spurlock documentary, in that there's words, but also graphics, and those graphics help to explicate the complex issues -- for example, Gore's "penumbra of fear," right at the beginning, and how big data creates a "constellation" of ourselves that may or may not be accurate. Comics allow for that, but a comic would equally allow for density, too, if that's the way the authors had chosen to go. (Further, I did wonder, when the authors became literal, minimalistic "talking heads" on the page, whether the comics format was truly necessary for this project or if podcast or documentary might have worked better; other times the comics format was more on point.)

What the comics format does perhaps do is allow the authors many different avenues for perspective. At its most humble, Terms of Service depicts reporter and cartoonist interviewing various people who work with or know about big data; generally, reality seems to apply here, in that they video chat with people whom they presumably video chatted with in their original research. Secondarily, there are the images not in the "present"; that is, someone is talking about cars and we see an image of a car with their narrative "voiceover." But then at other points the authors step outside their "characters," as in when Keller speaks gibberish to Neufeld because he's thinking about encrypting his communications, or the unremarked-on moment where Keller is inconvenienced typing his 27-character password to reach his own files. In those instances, the authors are making points through the characters whereas elsewhere they make their points as the characters, and I think comics better allows for the quick switchover between reality and artifice than documentary does.

It's with no ignorance of the irony that I'm posting this review on a blog that utilizes a Facebook page and a Twitter feed and that exhorts users to share posts in those venues to help bring new readers to the site. Obviously I recognize big data is a largely unregulated frontier with great unforeseen implications, even as I myself benefit from and use these networks that take your personal data in exchange for the ability to play, y'know, FarmVille. But I appreciate, at least, that Keller and Neufeld seem to understand these paradoxes, too. Throughout the comic, Keller posits himself as the Luddite, constantly warning Neufeld against the dangers of over-sharing. But on the last page, even after everything they've learned, Keller checks in on FourSquare to get a discount on a burger, and then Neufeld calls him on it, Keller just shrugs. Clearly the authors aren't suggesting big data is all bad; it's just that it's hard to know how far to let it go, and hard to see when it's already gone too far.


Post a Comment

To post a comment, you may need to temporarily allow "cross-site tracking" in your browser of choice.