The Filter Bubble: More Program . . . or be Programmed
Wednesday, May 18th, 2011Great TED talk by Moveon.org founder Eli Pariser on how the personalization reflex/mandate/standard of content on the web keeps us in a bubble. Couple great slides/moments:
- an examination of his Facebook feed which, over time, as FB’s algorithm’s started to weed things that appeared not to be what he wanted, started suppressing the display of statuses from conservative friends.

- showing how two people Googling “Egypt” during the uprising get vastly different results based on what the Google algorithm concludes each person wants. One person got news links about the uprising, the other got travel information!

- a fun, but spot-on mock-up of what useful personalization choices should look like:

The Google search screen is possibly the most unsettling. From the beginning, much of Google’s algorithm has been focused on following the herd. In addition to attempting to understand what a page was about, Google cleverly added in the power of “this is what other people think this page about” (because they clicked on the result with a similar query) and “this is where influential linkers (content providers) think you should go.” So already, you’re getting knowledge that is pre-validated (or pre-sanitized). But now, by personalizing the results, you’re narrowing the results even further: this is what people who think, spend, and live like you believe is the best result.
The fun of the internet is to see things that surprise you, connect laterally to your existing thinking, create serendipity, or just put things in front of you that you should know as a member of a community and a polis. The “service” of personalizing – while valuable – reduces that function, makes the user work to be exposed to new things. This line from Eric Schmidt, suggests that there is an end state goal for the algorithm to manage everything you see: “it will be very hard for people to watch or consume something that has not in some sense been tailored for them.” Oi, oi, oi.
Parisi ends with a rousing, but almost disturbing plea to programmers to stop programming us:
we need the new gatekeepers to code that kind of responsibility into the code that they’re writing … I know there are a lot of people here from Facebook and Google . . . wq really need you to make sure that these algorithms have encoded in them a sense of the public life, a sense of public responsibility, we need them to be transparent enough that we can see what the rules are that determine what gets through our filters . . . we need you to give us some control so we can decide what gets through and what doesn’t.
