"Did I choose this, or did the algorithm?" How recommendation systems erode personal taste

Recommendation algorithms have quietly taken control of what once seemed the most personal thing about us: our preferences in music, film, fashion and books. More and more people find it hard to answer a simple question: "What do I actually like, as opposed to what an algorithm pushed on me?" The Guardian examines the scale of the phenomenon and meets those who decided to push back.

Personal taste as a casualty of technology

Two decades ago, your list of favorite bands, directors or authors worked like a cultural passport. People built their taste slowly, through friends' recommendations, chance finds in shops and magazine reviews. The process was alive, chaotic and deeply personal. Discovering a new band or writer took effort, and that effort is what made the discovery truly your own rather than something suggested to you.

Today the algorithms of Spotify, Netflix, TikTok and Instagram do that work for us. They analyze billions of behavioral patterns with a single goal: holding your attention for as long as possible. As a result, we consume what is statistically similar to what we have already watched or listened to, and gradually lose the ability to be surprised, to search, and to make a genuine choice that is not prompted from outside.

The paradox of infinite choice

The Guardian's authors gathered accounts from people who asked themselves: "Did I choose this, or did the algorithm choose for me?" Many admit they can no longer draw that line. Researchers of media consumption note a characteristic paradox: with unlimited choice, people now listen to, watch and read less varied content than in the era of physical stores with limited selection.

• Spotify offers 30 personalized tracks every week, and most users listen only to those, without looking for anything beyond them
• Netflix hides little-known films beyond the first two screens of scrolling
• TikTok can lock a user into a stable "aesthetic bubble" within a few days
• Instagram's algorithms standardize fashion: trending looks in different countries become indistinguishable from one another
• Amazon's book recommendations are built on sales data, not literary merit

This is how "taste homogenization" emerges: people in different cities and countries discover the same albums, wear the same capsule wardrobes and read the same bestsellers, not because they decided to, but because the algorithm led them to the same point.

Those who resist

The article describes people who have deliberately given up recommendation systems, and the journalists call them "style rebels". Some went back to physical media: vinyl records and paper books chosen blind, going only by the cover or the blurb. Others set themselves a rule: no reading reviews before watching a film, and no studying the charts before buying an album. Still others agreed to swap recommendations with friends strictly in person, ignoring digital platforms entirely.

"I wanted to feel again that my taste is my taste, not a reflection of

what the algorithm decided to show me," says one of the people featured.

This is neither Luddism nor romantic nostalgia for an analog past. It is conscious resistance to a system that monetizes individuality and turns cultural choice into predictable, impersonal statistics.

What it means

Algorithmic filtering solved a real problem: information overload in a world with millions of new releases every year. But the side effect was unexpected: we quietly delegated the shaping of our own identity to machines. The answer to "who am I" now comes partly from a recommendation system. Recognizing that is the first and perhaps most important step toward reclaiming authorship of your own taste.

*Meta has been designated an extremist organization and is banned in Russia.