18 PHI KAPPA PHI FORUM
That’s the power of a Facebook algorithm. The social media
service uses more than 150,000 different factors — what it
calls its EdgeRank — to try to anticipate what you’d like to
see, what you’re most likely to pay attention to, and what
you have less interest in. The impact of Facebook filtering
news and information, as well as the impact of algorithms
on public information as a whole, has raised considerable
concern among academics, politicians, and policymakers.
But the hype is not the reality.
The concerns about Facebook and news consumption
generally fall along three domains. You’ve likely been
told that Facebook paves the way for a “filter bubble,”
or an “echo chamber.” 1 This argument suggests that since
Facebook is trying to optimize for content you see, you’ll
never see opposing views. Similarly, others worry that
Facebook’s algorithm is being manipulated to censor or
suppress particular content, such as conservative news
sources. There’s also the fear that Facebook’s ability to
manipulate the news feed might have a significant impact
on public behavior. The truth to all of this is somewhere in
the middle between utopianism and dystopianism — and
depends on whether you understand how algorithms work
and how researchers use data.
First, it’s important to understand what an algorithm
actually is. Put simply, an algorithm is a sequence or
series of steps; you might think of a recipe as a basic
algorithm. An algorithm might weigh particular steps
on this sequence more heavily than others — perhaps
more emphasis on shares rather than likes. At the end of
the algorithm is your Facebook newsfeed, Google search
results, or Netflix selection.
Algorithms aren’t just computational or machine
intelligence. Algorithms are made by people and make
judgments based on criteria people decided were important.
The algorithm is nothing without the human who
programmed it. 2 And while each result may be different
based on input, the secret genies who shape the Google
search algorithm are doing so extremely consciously — an
algorithm only renders its machine judgment thanks to the
decisions already made by people.
The scary thing, you might think, is that what powers the
Google search algorithm is secret, and you’ll never know
— for competitive reasons — what goes into producing
a Google search result. This is why the algorithms are
sometimes called black boxes3, an input is put into the
algorithm, and we don’t have access to understanding where
Facebook could show you every single activity of every single one of your friends and
FACEBOOK, FILTERING AND
all the activity on pages you’ve liked. Instead, most people are only seeing less than a
tenth of the posts they could see each day in their news feed.