I had also signed up to write about Equality Lab’s “Digital Security in the Age of Trump,” but it looks like that page changed to something else—and it’s also functioning very wonkily.

Engine Failure

This is essentially a Q&A in Logic Magazine, with Safiya Umoja Noble (assistant professor in the Annenberg School of Communication at USC) and Sarah T. Roberts (assistant professor in the UCLA Graduate School of Education and Information Studies). The discussion centers around these questions:

The internet is dominated by a handful of big companies. How do they organize information to maximize profit? What are the consequences for the public sphere? And what would a better model look like?

Noble’s responses take up most of the Q&A, with Roberts’ coming in briefly at the very end. What’s particularly interesting about Noble’s work is that she’s working on “a non-commercial search engine that makes its biases visible.” Noble goes into the problems with how Google, as a search engine, is built and functions—which is essentially how citation analysis works for papers in journals and publishing houses. The more that you are cited or linked to by someone else, the higher you will rank on Google. However, Noble’s point is that citation analysis doesn’t give you context for why someone decided to link to your work. The ranking system, then, is flawed because it can’t distinguish between whether someone is linking to your work because they are agreeing or disagreeing with you.

These flaws have been problematic and have broader social stakes. Noble bring ups a few examples:

• The porn industry was able to dominate much of the information landscape by having thousands of sites that link to each other and by buying a number of keywords. That’s why, for years, if you Googled terms like “black girls,” “asian girls,” and “latina girls,” without even adding the words “sex” or “porn,” you would get search results of pornography.
• Dylann Roof, the person responsible for the Charleston church shooting, noted that Googling the phrase “black on white crime,” after the Trayvon Martin shooting, played a significant role in forming his white supremacist views. The search results he got, of course, led to many white supremacist websites that used the phrase, but they didn’t return any results that could have provided context on the white supremacist movement.
• Immediately following the 2016 United States presidential election, when you did a search of election news on Google, it’d return results about Trump winning the popular vote—which was obviously false.
• What goes viral on Facebook can be very lucrative for Facebook, even if it’s highly problematic or controversial content. This includes fake news stories posted by Macedonian teenagers and videos of police murders of unarmed African Americans.

To go about improving on and beginning to resolve these issues, Noble believes that the following would help:

• The tech industry hiring people with advanced degrees in ethnic studies and women’s studies and sociology.
• Maximizing profit not being the only “value system” in our society. In Noble’s words, “Engineers may not be malicious, of course. But I don’t think they have the requisite education in the humanities and social sciences to incorporate other frameworks into their work. And we see the outcomes of that.”

When asked whether possible solutions could be trying to reform these companies from within, thinking about regulation, nationalization, and building alternatives, both Noble and Roberts said yes to all of those things.

Discussion Questions:

1. How would we get buy-in to something like the new type of search engine that Noble is working on and proposing? It certainly sounds intriguing and like a great idea—given all the problems that Google presents, which she outlined—but the reality, still, is that Google is widely recognized and used by everyone worldwide.
2. In terms of challenging the model of the internet being controlled by a handful of big companies that care more about profits than the larger ramifications on society, Roberts mentioned international examples where countries actually had success in regulating tech companies:

• • In 2000, the French government threatened to not let Yahoo operate in France if the platform continues to allow people to sell Nazi memorabilia on its auction site.
  • In Turkey, Facebook already has to operate under the condition that any material that relates to the Kurdistan Workers’ Party must be taken down.

Do you think the United States would get to the stage of being just as aggressive in regulating tech companies? Why or why not?

Pedagogy and the Logic of Platforms

In this piece, Chris Gilliard discusses whether and how education can “save the web.” Gilliard goes into a few key concepts:

• Surveillance capitalism: a “form of information capitalism [that] aims to predict and modify human behavior as a means to produce revenue and market control.” This has resulted in Web2.0, which consists of personalization, clickbait, and filter bubbles—“the only web most students know.” Gilliard believes the web is “broken” specifically because of surveillance capitalism, although he admits that “broken” isn’t the most accurate term. He writes, “A web based on surveillance, personalization, and monetization works perfectly well for particular constituencies, but it doesn’t work quite as well for persons of color, lower-income students, and people who have been walled off from information or opportunities because of the ways they are categorized according to opaque algorithms.”
• Digital redlining: “the creation and maintenance of technological policies, practices, pedagogy, and investment decisions that enforce class boundaries and discriminate against specific groups.” Some examples of digital redlining that Gilliard brings up:
  • The process by which different schools get differential journal access. Here, digital redlining dictates how quality information gets locked by paywalls that prevent students from accessing that information.
  • The level of surveillance, in the form of analytics that predict grades or programs that suggest majors to students.
  • The degree that students who perform Google searches get certain information based on the type of machine they are using or get served ads for high-interest loans based on their digital profile. (Gilliard notes that the latter is a practice Google now bans.) Interestingly, this particular example slightly relates back to “Engine Failure.”
• Consent: Gilliard writes that “if higher education is to ‘save the web,’ we need to let students envision that something else is possible, and we need to enact those practices in classrooms.” In order for that to happen, he says that “consent” needs to mean more than “click here if you agree to these terms.”

Discussion Questions:

1. Gilliard ends his piece with this:
   

   Using higher education to “save the web” means leveraging the classroom to make visible the effects of surveillance capitalism. It means more clearly defining and empowering the notion of consent. Most of all, it means envisioning, with students, new ways to exist online.

   How would you go about doing these things in your own classroom? Is it even possible to use higher education to “save the web?”

2. What would the web look like if surveillance capitalism, information asymmetry, and digital redlining were not at the root of most of what students do online? This is a question Gilliard mentioned that we don’t know the answer to, but what do you think the web would look like if that were the case? Would it be the most ideal version of the web that we all wish we had now? Why or why not?

Education Technology has spread into every field and discipline, with numerous applications in each. This creates a seeming invincible ubiquity, untangling the interconnected lines of data-tracing is nearly impossible. Meinke’s “Student Data Harvested by Education Publishers: They haz more than u think,” discusses the overlapping strategies used by social media platforms and educational platforms to collect user information and process it to ostensibly create better interactions for all involved.

Many of the articles in this week’s readings discuss Cambridge Analytics’ use of Facebook data to facilitate getting targeted outreach and content to users for political ends. This revelation had little true impact on Facebook: their stock value temporarily dipped, and Zuckerbot had to go before Congress. Otherwise, not much happened, people still regularly went on Facebook to add their own personal content and click preferences to the information they explicitly volunteered to the create their user profile. “But education is different,” says Meinke: we take classes to earn degrees and are entitled to a sense of safety and guardianship from our institutions. I agree that education is different but for very different reasons, unlike with social media- which has no proven role in education and is ostensibly “free” in the market sense of the word- interactive course content comes with a hefty price tag. Nowhere is it understood that the students’ data will be held, analyzed and disseminated in perpetuity in payment for access to said content/platform. Indeed, the price of textbooks (e-book or dead tree) has gone up by triple digit percentages in the last decade, even adjusting for dollar values. Students have paid enough upfront for their privacy to have it remain intact, and to have access to their course content beyond the semester. Program learning outcomes are based on the idea that students will enlarge upon skills learned in previous courses.

The rationale for granting access to learning management system data to publishers is posited as the facilitation of platform use by both faculty and student. However, in the nascence of online textbook augmentation students entered a long activation key and would create their profile by selecting the school and course in which they were enrolled. There was nothing wrong with this model, however, the trend towards single sign on (SSO) has created a myth that people cannot handle having multiple logins and also that unique logins will deter platform uses. We can all agree that having to create a new account instead of using an existing Facebook or Google account will probably give some users pause enough to reconsider their buying or comment-posting choices- but there is a critical distinction between the user of mass-published educational content and just about any other instance of use that online learning is frequently compare to: the end-user has already paid to access the platform. The enticement and marketing ships have long sailed. Also, the pursuit of credentialed education should not be trivialized as being tantamount to posting comments or other casual engagements.

I would add to some of the responses in the Davis and Bulger interview with a recommendation that social media not be used in a classroom setting.  “One of the things that excites faculty the most and what they want to learn more about is different ways to use social media in the classroom. An assumption I encounter frequently is that as students are already using social media, social media will be easy to adopt and adapt for learning both for the professor and the student….” We may argue that there is no such thing as true privacy online, but the latent, abstract and remote impact of Cambridge Analytics knowing bits of your professional, social and academic life versus the immediate, and potentially devastating impact of your teacher or boss seeing these overlaps are incomparable.

For the class:

Do you use social media as part of your course activities? If so how?

Can open educational resources combat the integration and slick marketing of the big publishers?

In the wake of Cambridge Analytics, repeated data-breaches across industries, and the lackadaisical responses they’ve garnered, along with the high usage and access provided by freemium platforms, is their room for social action to preserve privacy or is it a matter of personal choice?

In her 2013 article, Noble describes a Google search that she conducted in 2011 using the keywords “Black girls.” She observed that both the advertisements and top 10 search results were dominated by hypersexualized and pornographic portrayals of Black women. Noble argues that these results demonstrate how the Internet is far from a democratic utopia. Instead, search engine results reproduce and amplify misrepresentations of race and gender and reinforce the social, political, and economic disenfranchisement of women and people of color. Noble’s research demonstrates the inaccuracy of users’ perceptions of search engine results as being neutral or objective and reveals how the organization of results is heavily influenced by commercial interests.

In their 2017 article, Noble and Roberts critique higher education’s reliance on learning technologies which also monitor students by collecting their data. These “black-box technologies” are not transparent about their data collection practices and often use this data to improve their services and attract clients. The authors argue that IT professionals and technology scholars need to work together to think through the social implications of using these learning technologies before adopting them.

Discussion Questions:

To what extent did Noble’s 2013 article change your views about search engines? Have you observed biases or misrepresentations in your search engine results?

Noble and Roberts ask: “Will the future of knowledge reside with powerful information systems, unknowable algorithms, and privatized islands of data? If so, at what cost? Further, what role do we, as information technologists and educators, play in identifying and discussing these nuances with our students, staff, faculty, and campus administrators?”

In interviews with faculty, Head et al. (2020) observed that faculty rarely discussed these issues with their students. What would you say to your students about search engine algorithms and black-box technologies? How could you integrate an assignment related to these issues into your course?