Monday, February 13, 2012

Why I Don't Buy the Filter Bubble Arg

Update (3/3/14): in case anyone comes across this, I'd recommend reading nina de jesus' post "The Filter Bubble Is a Misguided, Privileged Notion" over this. It's an excellent piece which touches on deeper and more troubling problems with the filter bubble as a concept.


For those unfamiliar, the Filter Bubble Argument goes like so: personalized search/social media creates an echo chamber wherein, every time someone seeks information, they only receive answers that they agree with or that come from within their peer group. In other words, by adapting results to what people want to see, search engines necessarily hide the uncomfortable/disagreeable/different/Other.
I'm sorry, but I've always thought this arg is weak. Below is a diatribe to that effect.
A) And this is new...how?
The Bubble is hardly unique to 21st century web services. People have always sought out ideas they agree with & avoided those they find uncomfortable. Furthermore, modern transportation & communication technologies have made it easier than ever to contact people outsides one's peer group, whether that group is a nationality, political party, or profession. Compare today to the mid-17th century; when were people more insulated?
B) Right Issue, Wrong Solution
Changing algorithms or avoiding personalized services isn't the way to defeat the Bubble; education is. The Bubble is an information literacy issue and the best way to address it is via information literacy education. Seeking out adverse opinions, problematic facts, and antithetical arguments is perhaps the most vital step in framing your own opinion or writing anything that could genuinely be called research. Making debate a required course in high school would go further to remedying this problem than mandating that Google give everyone the same results.
C) Alternatives exist and you don't use them
It's frustrating that numerous alternatives exist and even librarians either aren't aware of them or simply don't care enough to employ them. Don't like Google? Hey, there's this excellent search engine called DuckDuckGo which is incredibly forthright about privacy and lack of personalization. Blekko is another option. Lastly, you can always search Google from an unauthenticated browser window, which severely constrains personalization. Yet I have yet to hear anyone proffering the Bubble Arg mention them, much less advocate for them. To me, this sounds like people who complain about Microsoft Windows' monopoly whilst ignoring GNU/Linux. There are options. The more people use them, the more viable they become.
D) The Bubble Is Good
Disclosure: I use Google far more than DDG or Blekko and, more often than not, I like how it personalizes results. Often, I'll perform a "known item" search and Google will promote precisely the link I'm looking for simply because it knows I've visited the site before. That's desirable behavior. The alternative of "same search results for everyone all of the time" is clumsy most of the time.

To be clear: this is an important discussion to be having. I'm glad people are drawing attention to search & social filters. But I'm not persuaded in the least.

Wednesday, February 1, 2012

Computer Science & Librarianship

I often regret not studying computer science in college. I was at a school with an excellent program, I'm mathematically inclined, & I already had a little experience with BASIC so it would have made a lot of sense. Instead, I've ended up a techie librarian with a humanities degree who spends a lot of time working on the web, but has little formal training. This situation doesn't seem all that abnormal; web design courses were well-attended in my MLIS program & I constantly hear LIS student bemoan a lack of programming courses. Librarians are well aware that it's a "program or be programmed world" & we're working hard to become programmers. The number of librarians I see participating in #codeyear is one indicator. But also there are a number of open source library software packages, from the VuFind discovery layer to the LibStats reference tracking program, written by librarians.
So we're taking things into our own hands & I'd like to contribute more to that. On the one hand, being one of those rare librarians with a CS background would set me up to be a valuable professional. But on the other hand, would I have ended up a librarian if I majored in CS? Most likely not. I probably would've landed a real, salaried job out of college, instead of the meager hourly position I ended up in. My dismal post-collegiate years instilled a political will in me too, a will that drove me to librarianship and public service. In comparison, the start-ups run by young, talented CS students appear banal: more music recommendation algorithms, personalized finance software, social anything, web apps that replace desktop apps but aren't unique in the least. These are doubtless exciting projects from the inside, but I joined librarianship largely for the ideals & not the glory. In the end, I think I would prefer being at a job I believe in.
As an addendum, I've found quite a lot of coder librarians I look up to. They're exactly where I want to be. & you know what? They mostly have humanities backgrounds. Philosophy & English are startlingly common degrees for web services librarians. So maybe I made the right choice after all.