It’s become a common refrain in my ongoing debates with Brian Greenberg: he says something completely cockamamie (IMHO) about George Bush or the financial industry, and I reply, “you and I live on different planets.” It’s been my theory for a while that despite our both being raised as Northeastern middle-class Jews, he and I just have completely different frameworks for looking at the world, and how we filter information about it.
Then, twice today, Eli Pariser jumped in with some reasons why this might be true, first from the following TED video, then in an hour-long interview on the Diane Rehm show.
Shorter version of Pariser: Google, Facebook, Amazon, Netflix, and most other gatekeepers on the web are personalizing the results they give you for your searches. As a result, you don’t know what’s being excluded from these results, so you increasingly live in a filtered bubble of information that’s been selected for you, based on not only your login cookies, but other data such as what wifi hotspots you’re nearby, and what computer you’re using. As a result, when Brian looks up information about Goldman Sachs, he’s going to get results from the Wall Street Journal and Bloomberg; I’m more likely to see Matt Taibbi and Mother Jones.
I’m still chewing over this idea, but my initial impression is to reject it. The people who are most going to be affected by this kind of personalization are also the people who are unlikely to go seeking alternate news sources in the first place. If I want to read a WSJ article, I’m not going to be dissuaded by not seeing it in the first four links on Google. But that presupposes that I know about the Journal, and I know when I want to look for their information (and when I want to avoid it as a self-contained and self-satisfied echo chamber).
The idea that we all benefit from having a randomized percentage of our news feed is also nothing new: David Brin suggested it in 1990’s Earth, which talked about a personalized news filter that we still don’t have 20 years later. The difference between Brin’s solution (which predated the common use of the Internet), and what Google is doing, is precisely what Pariser is calling for: transparency in what’s being filtered out. There’s no way to tell Google, “I place a high value on seeing topics that are outside some of my boundaries—but really, I don’t ever need to know about sports. Oh, and please show me a cogent Republican argument if they ever happen to come up with one.”
But this brings me back to something I wrote about years ago, and which surprisingly (and amusingly) I can’t find in Google. Head over to Google News, and what do you see? My page, considering only the headlines, is currently being built from the following sources: ABC, ABC (Australia), ArsTechnica, the BBC, Bloomberg, the Boston Herald, the Chicago Tribune, the Christian Science Monitor, CNN International, Computerworld, the CBC (Canada), CTV (Canada), the Daily Mail (UK), Fox News, the Guardian (UK), the Hindu (India), Hollywood.com, InformationWeek, iVillage, the Kansas City Star, the Los Angeles Times, MediaPost, the Nation (Pakistan), the New York Times, NPR, PCWorld, the Register (UK), Reuters, Reuters Africa, the Seattle Post Intelligencer, the Sydney Morning Herald, the Telegraph (UK), TMZ, USA Today, WHDH (NBC Boston), and ZDNet UK.
That list would be twice as long if I checked the sidebars, or if I did my usual thing and opened Google News in tabs. Compare that list to the access to media I had twenty years ago: the Philadelphia Inquirer, NPR, CNN, and perhaps an early version of ClariNet. (Granted, 20 years ago September I started grad school at the Annenberg School for Communication, so at that point my media buffet broadened quite a bit more.)
It’s my theory that even with constricted filter bubbles, the greater constriction is imposed by most people’s apathy, the same dynamic that leads to more interest in Donald Trump than a discussion of actual spending policy. Regardless of what might be filtered out of the Internet tidal wave of information, there’s still a much better chance of accidentally lodging an interesting headline in the brain of someone who doesn’t give a damn, which might just lead to a painless and cost-free click to find out more.
What Pariser is concerned about, in my opinion, is that the Internet has near-infinite potential to make all of us much more informed of the world around us, and that content filtering might reduce from this theoretical maximum. I agree with him, and stand behind the idea of making these filters more transparent. But what’s keeping us from realizing this potential is simple human nature, which isn’t going to be fixed by any algorithm.
meh you and brian are just both wrong. :P
Pariser has an interesting point about algorithms, and no – I didn’t realize that Google search results were so different for different people.
But where his argument falls down is when he (implicitly) assumes a lack of intellectual curiosity on behalf of the searcher. To use his example, if I’m Googling “Egypt,” then I’m probably looking for the Wiki page, so I can see Egypt’s population or land area or whatever. If I want to know about Egypt’s revolution, I’m probably Googling “Egypt Revolution” or something like that. In other words, I don’t depend on the algorithms to tell me what I’m interested in – I actively seek out what I’m interested in. This, incidentally, is not something borne of the Internet. Back when I used to read the New York Times or the Wall Street Journal on the way to work, there were certain articles (or even entire sections) that I skipped over because the headlines didn’t interest me. It was just as likely that 1992 Brian would read a totally different New York Times than 1992 Jeff as it is that 2011 Brian will have a totally different newsfeed than 2011 Jeff.
What’s different about the Internet-enabled world is this: in 2011, Jeff IS one of Brian’s information sources (and vice versa). Rolling Stone magazine is not in my newsfeed (particularly not in the financial section of my newsfeed). And yet, I can confidently say that I’ve read every major article Matt Taibbi has ever written about Goldman Sachs for Rolling Stone. Why? Because, while it doesn’t appear in my newsfeed directly, someone tells me about it. Not only that, they usually tell me about it via expressing their opinion on what he wrote (sometimes agreeing with me, sometimes disagreeing with me). Either way, I get to read the article. And I get to click through to the related articles, read comments, create my own Google searches, etc..
To sum up: algorithms can help and, as Eli Pariser has just taught me, they can even hinder. But nothing replaces an active curiosity and a recognition that the source of information matters. I don’t think it ever will…