Why Algorithms Won’t Always Control How We Discover Great Art

On seeking a better discovery system for creators and consumers

“Work” (1863) by Ford Madox Brown. Oil on canvas. Public domain.

The paradox of the algorithm is that it is designed to be highly personal: It notes the user’s preferences and recommends things that are likely to appeal to that specific user. And yet it’s one of the quickest ways to spread sameness.

I think a 2021 on Netflix’s recommendation system is a great demonstration of why this happens. The paper’s author, communication strategist , aimed to investigate how Netflix’s recommendation systems worked and how it might impact personal taste.

Pajkovic designed a set of fake accounts with different personalities like “The Culture Snob,” “The Die-Hard Sports Fan,” and “The Hopeless Romantic.” These personalities strictly watched shows that corresponded to their profile: The snob only watched arthouse films and obscure foreign directors, while avoiding reality shows; the sports fan only watched shows that featured strenuous physical activity and ignored romantic comedies, and the romantic exclusively watched high drama shows that vary between high and lowbrow. Expectedly, the homepage for each account showed identical shows in the beginning. But that changed as each account started watching shows that corresponded to their “personal…