Algorithms Replaced Critics, Journalists Followed
Universal Pictures released an animated film called Forgotten Island this year. It scored 95 percent on Rotten Tomatoes—yet entertainment media outlets began recommending it as essential viewing for fans of KPop Demon Hunters, an unrelated animated property from two years earlier, despite sharing no director, writer, studio, genre framework, or thematic architecture.
They share only an audience demographic and a position in the same streaming recommendation algorithm. This is how discovery has been industrialized. When a platform's algorithmic engine groups two films together based on user behavior (viewers who watched one also clicked on the other), entertainment journalists treat that proximity as artistic kinship.
The algorithm becomes invisible. The recommendation becomes reviewable, defensible, a matter of critical judgment rather than data adjacency. The reader learns to accept that if Netflix or Amazon positioned these titles near each other, some meaningful artistic through-line must connect them.
The pattern repeats wherever recommendation systems have become the primary discovery tool. Spotify playlists curate songs by algorithmic similarity, not by harmonic or lyrical logic. Listeners describe them as cohesive visions—a lo-fi hip-hop track grouped next to ambient drone because both have similar listener retention rates creates the impression of a shared aesthetic vision. Music journalists write about the "evolution" of a sound that is actually the evolution of a data model. The curator has been replaced by the database.
The mechanism is simple. Algorithms measure behavior, not intent, and two films watched by the same people are mathematically adjacent. That adjacency is not narrative or thematic. It is visible and quantifiable, which makes it easier to defend than a genuine editorial judgment would be. A critic recommending a film because she felt something shifting in its structure risks being wrong. A journalist recommending a film because the algorithm already placed them together has outsourced the risk to mathematics. You receive a job candidate because an internal system flagged her as similar to your last three hires, based on resume data that shares no common purpose, or you're offered a book by an algorithm trained on what you've already finished rather than what you're ready for next. Adjacency has replaced curation so completely that we've stopped noticing the exchange.
