Discovering new things on the Internet is not random. We live firmly in the age of the feed: Facebook is moderated (and occasionally tinkered with) by the invisible hands of an algorithm, our Google search results are skewed to our interests and social circles, and even Twitter, that last bastion of personal choice, has begun experimenting with injecting users’ feeds with “popular” content.
To combat the creep of personalization, Sean Schuster-Craig, an artist and musician who, under the alias Jib Kidder, makes the kind of damaged YouTube collage videos that only a true mastery of the medium can engender, often advocates a process for finding genuinely random video on YouTube.
It requires the intentional dismantling of YouTube’s tailored search systems, which he calls “unFacebooking the thing.” Schuster-Craig enters arbitrary dates into the YouTube search bar and removes the “Relevance” filter from search results, effectively summoning the long, long tail of YouTube’s untapped content. The videos that turn up—never really intended for public consumption, with views in the single digits—are real as fuck.
Facebook, on the other hand, will not let you unFacebook Facebook. It is impossible to discover something in its feeds that isn’t algorithmically tailored to your eyeball.
“The laws of Facebook have one intent, which is to compel us to use Facebook," Schuster-Craig explains. "It believes the best way to do this is to assume it can tell what we want to see based on what we have seen. This is the worst way to predict the weather. If this mechanism isn't just used to predict the weather, but actually is the weather, then there is no weather. And so Facebook is a weatherless world.”
As we navigate the social web, our every peccadillo is monitored and mirrored back to us through sidebar advertisements and canny clickbait, scraped clean of individual tastes and served back to us as goods and ideas for mass consumption.
On the Internet in 2014, we’re all reading the same articles, cooing over the same skittish goats, tumbling down the same clickhole. The internet activist and critic Eli Pariser calls this weatherless world a filter bubble: a state of ideological limbo, where users are isolated from divergent viewpoints by the monolithically misguided helpfulness of the social web’s content algorithms.
In brief, our feeds are fed to us. A “feed,” explains Jarno Koponen, the Finnish designer behind Random, an iOS app that aims to break the filter bubble with a combination of machine learning and irrationally-served content, “is a hierarchical, linear and inflexible way of presenting information. It allows only a limited set of interactions.”
Random, on the other hand, serves up content—long-form articles, videos, news, blog posts—through a custom adaptive interface that captures your interests and preferences as you browse, building what Koponen calls a “unique anonymous choice profile.”
What does that mean? Allow me to illustrate: I just opened Random. I’m presented with a cluster of topics, including “success,” “torture,” “robotics,” and “deep space.” Fueled by my natural curiosity for the subject, I choose “robotics,” and Random serves me an article from Cornell University about Robo Brain, a repository of knowledge culled from the Internet to teach to robots, which (obviously) I read in its entirety.
Now Random knows I find this subject interesting. When I return to the cluster of topics, it’s more my speed. The topics now include “drones,” “robotic grasp,” “interaction,” “haptics,” and a few wildcards, like “performance” and “smiley-face.” The more I browse, the more customized my experience becomes.
But although the choice profile gets to know me, Random never turns into a filter bubble, because it perpetually injects the irrational into my experience—those wildcards in the cluster—in a cocktail of relevancy and serendipity. Koponen explains, “the system needs to learn and understand your interests in order to know what is unexpected and surprising for you.” Call it randomizing.
The unexpected, in this context, can become meaningful; seemingly unrelated ideas, daisy-chained to one another in the curiosity-driven dérive of browsing, can manifest previously unseen connections. Perhaps “robotics” and “performance,” thrown together, will spark a connection in my mind.
“Different things appear irrational or random to different people," as Koponen puts it. "One could say that all the things in the known universe are somehow connected and you can link everything in a rationally comprehensible manner. In Random, randomness becomes a situated phenomenon. It exists as an experience and as an impression that is tied to a certain individual.”
To discover something interesting in the early days of the web took patience. We browsed without guidance, searching for the unexpected. It was often unwieldy, or boring, but the discoveries were ours—to be treasured like thrift shop scores or a gem dug from the B-movie rack at the video store. To experience such aleatoric discovery now takes pointed unFacebooking, as Schuster-Craig puts it, or disavowing the feeds entirely in favor of something like Random.
These tools may be imperfect: unFacebooking is limited, a loophole not long for this world, while Random remains a controlled experience, a kind of anti-filter bubble filter bubble in which randomness itself is curated by machine. In our increasingly ad-driven social web, we may not ever truly pop the filter bubble, but if we try, we can peek across its boundaries just long enough to rekindle our appreciation for true discovery.