“Birds of a feather flock together.”
Such is the cliché that often describes those with like minds.
Interestingly enough, this same cliché can be used to describe the algorithms
applied by major internet websites to tailor the information presented to their
users.
The first example, mentioned by Eli Pariser is the omission
of choice posts on his Facebook newsfeed. He claims that Zuckerberg has created
an algorithm that detects what types of links its users frequently click on.
Using this algorithm, Facebook will then filter out links or posts that are not
related to what has been calculated to be a user’s “preference.” Specifically,
Pariser claims that based on his previous actions on Facebook, his preference
has been calculated to be that of a democrat, therefore omitted any posts that
may relate to the Republican or Tea Party.
Similarly, Google has established a system that tailors the
results of a particular search to the predicted audience. Essentially, this
means that rather than having a uniform set of results for a given search,
results will vary from person to person. The algorithm that calculates these
preferences is based on websites and previous searches recorded for each
Internet user. The reasoning behind such a system is meant to make a person’s search
process more efficient, and therefore help each user find what they are looking
for with less of a struggle. However, by restricting a person’s search results
to information to affirm their current preference/opinion, a person’s exposure
to alternate (inevitably contrasting, but not always negative) opinions.
When an individual is not exposed to various points of view,
it creates conditions that can potentially cause this individual to be
close-minded in regards to any idea
that contrasts their beliefs. We must realize that this is an issue because in
order to effectively counter ideas that contrast with your beliefs, you must be
able to understand contrasting points
of view. It is impossible to make well rounded, rational, and informed opinions
if these filtering algorithms continue to hinder those who are not aware of
their presence.
Conversely, Thompson believes that this “homophily” that is
enforced by these algorithms can help individuals strengthen their worldviews.
He states that although being exposed to contrasting opinions may in the long
run be beneficial, it is the volume of a person’s interactions with both
similar and different opinions that are fundamentally important. This basically
states that individuals will interact with those who share their point of view
on a frequent basis, whereas, it is likely that they will interact with those
they disagree with as often. Ergo, even though many of the same ideas are
shared between peers with similar beliefs, it is inevitable that from time to time,
a new idea will be shared between peers. This interaction over time is more
cumulative than the amount of new ideas shared between dissenting peers,
because as previously stated, the volume of exchanged ideas is higher,
therefore there is a higher probability of new ideas being shared between
similar minded peers.
In my opinion, homophily seems like a recipe for
restriction. Restriction of thought, restriction of exchange, and restriction
of change. We must be informed citizens of the world, and it is impossible to
do this if computer generated algorithms are telling deciding our preferences,
and therefore the information that it deems valuable for us to know. We should
be able to chose what we value as important, and be able to find information to
contrast our values just as easily as those that support them.
No comments:
Post a Comment