AFA DIALOGUE

The AFA Dialogue has been created to air concerns of all faculty. The AFA Update is the factual voice of AFA, while the AFA Dialogue encourages conversation and publishes personal opinions about workplace issues and political concerns. We invite any faculty member to submit letters, articles, or opinion pieces. The opinions contained herein are solely those of the writer, and AFA neither condones nor condemns these opinions. AFA reserves editorial prerogatives.

Submit comments, letters, and/or articles via email to afa@santarosa.edu or via fax to (707) 524-1762.

AFA members who submit original articles of 500 words or more that are published in an issue of the AFA Dialogue will be awarded a Stipend of up to $50.

Canvas Is Harvesting Our Data to Build Robot Teachers

Now that I have your attention, I will admit that it’s more fair to say that maybe Canvas is harvesting your data to build robot teachers. I would even say, probably. Instructure, Canvas’s corporate parent, is private, so it’s difficult to get a clear look inside their operations. But the possible advent of robot teachers, built out of our data, is important, yes? So let’s review the evidence.

In an earnings call with investors last spring, Dan Goldsmith, CEO of Instructure, signaled a change in the company’s business plan, announcing the “DIG” initiative, “focused on analytics, data science and artificial intelligence.”¹ Goldsmith boasted that Instructure’s “data assets” represent “the most comprehensive database on the educational experience in the globe” and announced plans to “monetize” the machine learning and AI capabilities of DIG, using those “data assets.” Goldsmith also let investors know that the DIG initiative will apply to Instructure’s new “Bridge” product for the corporate market. Instructure is touting Bridge as a Canvas-like product. It clearly views Bridge as, well, a bridge between the academic and corporate realms.

An observation, and clarification, before we move on. By “data assets” Goldsmith means the enormous volume of data Canvas has gathered by tracking our use and our students’ use of Canvas. In short, Canvas/Instructure surveils us and our students, and declares the results of its surveillance of us to be Instructure’s own property. As Goldsmith makes clear, that raw material then becomes capital for Instructure’s development of new products, and pursuit of further market share and revenue.

Most of us will recognize this business model as the same one employed by Google, Facebook, Amazon, Microsoft, and other digital tech enterprises of the sort that Shoshana Zuboff dubs “surveillance capitalists,” in her recent book, The Age of Surveillance Capitalism. Zuboff writes that the moment in 2003 when Google realized that it could simply declare that the surplus data it had gathered by surveilling users of Google Search was a new “asset,” which Google could then “monetize,” was the founding step of surveillance capitalism. It really looks like Canvas has taken that step, wouldn’t you say?

Zuboff shows that what surveillance capitalists do next with their “assets”/our data is to process the data, using machine learning and AI technologies, to generate extremely accurate predictions of our future behavior. As Eric Schmidt of Google has claimed, “We can more or less know what you're thinking about, which means that "we can suggest what you should do next.” The companies then sell those predictions, mostly but not exclusively to advertisers, in a market whose value has grown, between 2003 to 2019, from zero to the hundreds of billions of dollars annually. The DIG initiative may be new for Instructure, but by now it looks very familiar, following a path well-beaten by some of the world’s wealthiest and most powerful corporations.

There is more evidence that Instructure has committed itself to surveillance capitalism along the lines set forth by Google. In a July 2019 interview Goldsmith touted Instructure’s plans for developing “personalized learning services,” which will be built on Instructure’s growing abilities to understand “individuals[students], their paths, their passions, and what their interests are.”²  Think Alexa, Siri, or Ok Google.

In the bad, old, slow, dirty days, before we all learned that tech capitalists were paving the golden digital road to our shining future, students seeking education had no choice but to look to human teachers for “understanding of their paths, their passions, and what their interests are.” But human teachers have many flaws. Limiting our consideration of those flaws to Instructure’s business interests, and their congruence with the Chancellor’s interest in making the CCC system more “efficient” in producing certificates and degrees, teachers have the grave flaw of being extremely expensive. We are irremediably inefficient economically. A teacher can attend seriously to the interests and passions of only a few students, at most, at a time. Once the new DIG initiative has finished digging our data out of us, Canvas should be able to “know what students are thinking about,” and “suggest what they should do next,” by the tens of thousands at a time, for cheap.

Do you think, as I do, that there is an awful lot baked into that “should”?

It sure looks like Instructure is selling the California Community College something in the way of automated teaching technologies—robot teachers, but not like the Terminator, or Gort, from the Day the Earth Stood Still. More like Alexa, or Siri. Or Hal, from 2001. (Canvas to tens of thousands of students, anxious about their human future: “I honestly think you should sit down calmly, take a stress pill, and think this over.”) We don’t have a say in whether the system is buying. We’re not the customer. Canvas is “free” to us. We’re not the product any more, either. We’re the source of raw material for building our own replacements. The Chancellor’s office isn’t even asking us to go along. It’s telling us that we will go along. Or perhaps, assuming that we will go along.

After all, what could possibly go wrong?