It seems that, in general, this article is meant to notify it's reader of a present or emerging problem. Though it doesn't present any alternatives or solutions (annoying) it does make a token effort to analyze the positives of the situation. It's clear otherwise that the author views the negative effects as far exceeding the positive ones, at least in my opinion.

Author seems to be playing devil's advocate; playing the Believing/Doubting game.

Strength in numbers, mob mentality. Outside validation, whether justified or not, encourages us to hold tighter to our own beliefs and question ourselves less.

Grade A example of GroupThink right there.

I'm wondering why they didn't test to see what would be the result of creating an evenly mixed group. I'd be interested to see if it ended up creating more polarized opinions, or made people more open to other points of view. It certainly would have helped to solidify the point the author is trying to make. As it is it seems as if the only message being delivered here is that "the internet can make it easier for people to create false dichotomies". A counterpoint showing the advantage of mixed groups would have certainly strengthened their position.

Funny word, here :P

Preparing myself for the Ethos/Logos argument here.

People like to have their opinions validated. Taking advantage of their bias is natural, at that point. Especially for companies who's income relies entirely on the amount of time a user spends clicking around on their homepage.

Reminds me of the Cambridge Analytica Facebook scandal.