Jarrett Fuller

04/08/2024

Recommending Recommendations

Our content is perpetually served to us through algorithms. I get a lot of my news from Apple News which learns what I like to read and sends me more of it. I listen to music on Spotify that makes playlists for me each day based on what I’m listening to. I watch movies and television shows on all the platforms, each of which offers its own “You Might Also Like” section. The videos I watch on YouTube recommend more videos like the ones I’ve been watching. Despite being plugged into these platforms for a decade or more1, I rarely find the algorithms’ recommendations offer me media that truly enriches to my life. In the short term, sure, it might entertain me for an hour and half. But when I look at my favorite media of the last few years: the books that mean the most to me, the bands I return to again and again, the movies and television I now count among my favorites, none of them came from the algorithms.

So where’d they come from? The short answer, of course, is people2. Media is a topic among my friends, family, and colleagues and recommendations from them have pointed me bands now I love, television I’ve watched multiple times, and movies I can’t get out of my head. But other recommendations come from my own research, a trail of content consumption where one thing leads to the next. I’ve found my more invested in consuming something I’ve read about before, whether that’s a review about the thing or just an essay that happens to reference a particular cultural touchstone. So much of what I consume comes from these lists I make for myself, picking up crumbs of references to return to later.

When I was a teenager — who took discovering new music and films very seriously and didn’t have people around to recommend things to me — the way I found new media was from looking at the influences and recommendations from the artists I already loved. Who did the bands I like listen to? What films were my favorite filmmakers inspired by? What were my favorite writers’ favorite books? This is how I got into Andrei Tarkovsky’s films or R.E.M. or Annie Dillard.

I was thinking about all this last night watching yet another edition to The Criterion Collection’s Closet Picks YouTube series. In each video, never more than ten minutes long, a filmmaker, actor, musician, or writer stands inside a closet filled with The Criterion Collection’s inventory. A totebag in hand, they go through the Blu-rays and box sets, pulling out the films that mean the most to them. I watch all of these. Sometimes I just like hearing someone talk about a movie I’d seen before in a new way and other times, I’m introduced to new films or given an entry point into something I’ve always wanted to watch. While watching Ben Gibbard’s picks I found myself adding them all to my Criterion Channel watchlist — I’d always wanted to watch Kelly Reichardt’s films, for example, but never had a way in: his description of Old Joy gave me what I needed to finally start watching. I’ll probably watch it tonight. In many ways, this is me discovering new content the way I always have.

I feel the same way about Amoeba Music’s What’s In My Bag YouTube series, which I’ve watched off and on for the better part of a decade. It’s the same concept: a musician walks around Amoeba Records in LA and talks about the records that influenced them. I’ve been listening to Megafaun for ten years because John Hodgman talked about them in a video. I got obsessed with Roberta Flack after Matt Berninger of The National talked about her, helping me both understand her music and his music better.

This is the power, I think, of recommendations like this: they are contextual. They expose the links, the influences, the personal stories behind the media. The algorithms, right now, can’t do that — they are just a stream of new things to consume. But when the recommendations come from somewhere, from someone, we have a specific entry point that gives us a deeper connection to both the media and the person. In his book, How To Talk About Books You Haven’t Read, Pierre Bayard writes that its the connections between books that matter just as much as the individual books themselves: how do they relate to each other, what influenced what, who was reading who. I think this is why the recommendations from Criterion or Amoeba stick for me in a way the algorithms don’t: they are creating a landscape of influences, a network of inspirations.

Or perhaps this is just a longwinded way for me to recommend two YouTube channels that have meant a lot to me and I want to share them with you.


  1. I’ve used Netflix for close to fifteen years. I’ve paid for Amazon Prime for over ten; Hulu, HBO, and Spotify all around a decade, give or take. 

  2. Another answer is brands. A record label, a movie studio, a television network build up brand reputation that may or may not make me more inclined to sample something new. I’m more likely to try a new show from HBO or FX than Amazon. I’ll watch a new A24 film no matter what but I’m skeptical of anything Netflix produced.