Just how in sync with your wants and needs is your Netflix account? If it takes you 20 minutes to find a mediocre movie, does it really know what you want?
Pop culture commentator Thomas Flight breaks down the complicated relationship with our viewing habits and just how closely Netflix’s machine learning is getting to know you. In roughly seven minutes, he addresses its custom artwork, which is meant to draw you into clicking, and how AI is used to personalize the art for each user. If you’ve watched several movies with a certain actor in it, the art will be tweaked to reflect that, because the machine learning thinks that will please you. Netflix has been experimenting with this for a few years, and Flight cites a December article from Netflix about its design.
Netflix is also tracking all your usage metrics to find “quality engagement,” a term for when a viewer finally settles into something they like, though it’s not clear if that applies to a movie you start watching then quickly lose interest in, but for some cruel reason leave playing in the background while you do something else, like I did with The Polka King.
Flight remarks that his homepage looks different than his wife’s and brother’s; I tested out this out with mine and my boyfriend’s and found the same subtle gendered shift: His homepage had more comedy and action with blues and reds in the thumbnails, while I was served more episodic fare with oranges and pinks. He posits that Netflix even knows just how much Netflix you should be watching.
This personalized approach, he says, is mostly good—at least for Netflix. But, he asks, what will happen when news becomes customized, too? Or information databases like Wikipedia? That might already be happening, but more pressing: Does Netflix count my Polka King viewing as a quality engagement?
