They came to this conclusion using disparate impact theory (DIT).
A standard algorithm makes suggestions based on the user's engagement. If people engage with conservative content, then the algorithm will suggest more conservative content. This is what happens in a free market; people choose what they want and the market provides it for them.
But what if the people choose "wrong" according to the leftist elites? Then they deploy Disparate Impact Theory. They look at the end results, and if those results are not in keeping with leftist desires, then they claim bias. It doesn't matter if the results are skewed for perfectly good reasons. All that matters is there's a disparity that doesn't favor leftism. And according to DIT, that disparity justifies intervention.
So leftists see that the algorithm is suggesting more conservative content (because that's what people want) and immediately declare the algorithm to be biased in favor of conservatism. Then they demand that the algorithm be changed to "correct" this disparity. And how does the algorithm accomplish this? By turning biased against conservative content, in direct opposition to the organic preferences of the users.
Disparate Impact Theory is the most pernicious and insidious tool in the leftist handbook. It is at the heart of all leftist policy. It needs to be destroyed.
They came to this conclusion using disparate impact theory (DIT).
A standard algorithm makes suggestions based on the user's engagement. If people engage with conservative content, then the algorithm will suggest more conservative content. This is what happens in a free market; people choose what they want and the market provides it for them.
But what if the people choose "wrong" according to the leftist elites? Then they deploy Disparate Impact Theory. They look at the end results, and if those results are not in keeping with leftist desires, then they claim bias. It doesn't matter if the results are skewed for perfectly good reasons. All that matters is there's a disparity that doesn't favor leftism. And according to DIT, that disparity justifies intervention.
So leftists see that the algorithm is suggesting more conservative content (because that's what people want) and immediately declare the algorithm to be biased in favor of conservatism. Then they demand that the algorithm be changed to "correct" this disparity. And how does the algorithm accomplish this? By turning biased against conservative content, in direct opposition to the organic preferences of the users.
Disparate Impact Theory is the most pernicious and insidious tool in the leftist handbook. It is at the heart of all leftist policy. It needs to be destroyed.