I was giving a demo of Google My Maps to our group yesterday (for our SurplusBanff map) and while we were zoomed in on the Banff Centre we noticed a POI for “Banff Wine Store” located in the forest behind our residence hall. Either there are parts of the campus that we haven’t explored yet, or Google has located that point incorrectly. Clicking on the icon tells us that the true address for the Banff Wine Store is 302 Caribou street downtown.
Mike asked if it was possible to tell Google to fix that error. I said “Probably, but why would I want to help Google make money using my free labor? If the error was on OpenStreetMap, that would be different.” Sure, I was exaggerating my political stance towards Google (after all, I was teaching the group about My Maps instead of something else, right?) but it gave us a good opportunity to talk about Google Maps vs OpenStreetMap, and I gave a quick demo on editing OSM for good measure. Note how OpenStreetMap has much better coverage of the Banff Centre, including most of the paths and major buildings, while Google has close to nothing.
The question of Google (and others) profiting from (or “exploiting”?) volunteered labor has been coming up in my own research, and we’ve danced around the topic at this residency without really taking it head on. The issue here is surplus network labor, which is just a specific form of the broad material and social surpluses that our group has been grappling with. This digital labor is the collective cognitive surplus of a billion residents of the global North with too much time on their hands. With the advent of the internet, Clay Shirky argues, billions of man-hours that would previously have been wasted watching TV commercials can instead produce massive projects like Wikipedia using nothing but volunteered labor. But… most of the hours spent online are not so organized. There’s value in the chaos of surplus information and patterns of activity online, but teasing it out is beyond the capabilities of most. Google is on of the few entities in a position to make sense (and money) of it all.
Today I came across an article by Matteo Pasquinelli that is getting at that point exactly: Are We Renting our Collective Intelligence to Google? Pasquinelli introduces the terms “cognitive capitalism” and “network surplus value” to describe how our collective linking and clicking online is converted, via PageRank and AdSense, into Google’s core product. The free services provided by Google (such as Gmail, Google Docs, Maps, even search itself) are only possible because we have paid for them with our collective intelligence.
In his conclusion Pasquinelli asks some good questions about possible non-monopolistic alternatives to PageRank, from trust-based systems to a voluntary hand-indexing of the web, but I’ve been thinking about what could be done on an even smaller scale, to at least shift the results that PageRank returns, if not bypass it entirely. My impression is that most people (with the exception of those rabidly link-sharing, Search-Engine-Optimizing webmasters out there) don’t really think about how Google’s search works, and how their actions are an integral part of Google’s rankings. Every time you click on one of Google’s search results (in my case it’s probably hundreds of times a day) you are validating Google’s algorithms. You are agreeing with Google’s opinions.
A recent example: right now if you search for “Michelle Obama” on Google’s image search, the first result is a racist photoshop job which has prompted Google to buy ads on its own site to explain and apologize. The weird logic of the ad buy and the question of whether or not Google should take down that search result are topics for another time. What’s amazing to me is the extent to which Google’s generic apology underplays the role of PageRank–of actual collective human actions. “If you are offended by what you saw,” Google seems to be saying, “blame the computers! Don’t blame us… nor the racists… and certainly not all the mouse clicks and web links made by average people just like you!” There’s no way of telling whether the offensive Michelle Obama image is #1 because of racists linking to it, or if it’s staying #1 because of everybody else still clicking on it out of curiosity or outrage. I’d rather have Google’s apology page say something like “if you don’t like it, don’t click on it!”
In no way am I trying to shift the blame away from the people who created and distributed that image in the first place. My point is that we make the web through our actions, in both subtle and explicit ways. Of course, it’s rarely as explicit as in this case, but the fundamental question remains: how do we make the net we want? Instead of PageRank looking for what is “relevant”, how about searching for what is “valuable?” How do we use our cognitive surplus more wisely to make that a reality?