Who Wrote That Recommendation?

Open almost any app on your phone and something is deciding what you see next. Netflix chooses the poster art. TikTok chooses the video. YouTube chooses the next autoplay. These recommendation algorithms are not neutral: they are trained to maximize the time you spend inside the app, because that time is what the app sells to advertisers.

In practice, this means an algorithm learns which images, phrases, and outrage-levels keep your finger on the screen. If political anger keeps you scrolling, you will see more of it. If a certain celebrity keeps you tapping, you will see more of them. The content you consume, then, is a mirror not of the world but of your own attention — amplified.

Media scholars argue that this creates two problems. First, it narrows understanding: we increasingly meet only ideas that agree with us. Second, it hides authorship: no human editor 'picked' the feed, yet real choices — about what to promote and what to suppress — were made. Reading in the 21st century, they warn, requires a new kind of literacy: not just of the words on the page, but of the invisible hand that put them in front of us.