![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Well, I did some things. I did, in fact, do some laundry, put said laundry away, clean the bathrooms, clean the basement, take out the trash and whatnot. I then ordered fried chicken from Ezell’s at 8:15 pm because I could, and I THEN made a pot of jello vanilla pudding using 1/3 milk, 1/3 heavy cream and 1/3 oat milk, because it’s what I had on hand. I am now eating said pudding.
I’ve managed to clean most of the food out of the fridge by either eating it or tossing it if it was already bad. It will be nice and empty for the family’s return tomorrow, ready for a Whole Foods run. It’s about 9:15 pm, still time for me to read some more tonight.
I’ve been reading a book called “Computing Taste” by Nick Seaver, about the people who design music recommendation algorithms. Since I work at a major music streaming company, I’ll leave you to guess which, I’m super interested in how this is perceived by people outside of tech. Being from the Midwest, and having a number of facebook friends from my childhood still, I see a real dichotomy between how the people I know in Seattle view/understand tech, and the people from my childhood that aren’t in this world. And then the people in academia, of whom I also know quite a few, are shades of in between depending on their field and age. I’m fascinated and dismayed by how little most people seem to understand about the way the internet works, how companies use their data, what “free” means when it comes to ad-served content, etc.
I myself don’t know that much about how it works, honestly, when it comes to music recommendation. I’m not a machine learning scientist. I know enough to know what is possible and what isn’t, and I have a vague sense of how it all works, and I guess this is more than a lot of people but I certainly can’t claim to be an expert. Before this book I read a book called “Spotify Teardown” which was fascinating and weird — part performance art, part academic experiment, equal parts bumbling and insightful and naive. It was truly bizarre. Instead of talking to anyone at Spotify, much of the book consisted of first, plumbing the depths of Swedish media in the aughts to reveal the early Spotify backstory (fascinating). Next, it engaged in a number of experiments to game the Spotify system by putting content on Spotify and observing its trajectory through the system, then testing its performance using bot-facilitated playback. It was kind of endearing, kind of ridiculous. Some of the book was very insightful and other parts read like willful ignorance.
At any rate, my point is that “Computing Taste” is much more academc and rigorous than “Spotify Teardown”. The author compares the notion of using algorithmic recommendations to hook a listener to anthropological literature about animal trapping and it’s an interesting lens to consider this through, this notion of keeping people on one’s platform as some sort of most dangerous game of hooking people with their own taste.
I think it is super interesting that music streaming companies are not really interested so much in selling “music” as they are a perfect algorithm. It’s a weird sort of problem, in the world of retail, because they all have the same catalog. With very few exceptions, the tracks are the same, and most listeners do not care that much about high-fidelity sound. Exclusive content isn’t an especially good way to win the game. They ostensibly cater to music fans but really the people they care most about are the mildly opinionated people, the ones that aren’t that picky about their music and are willing to have recommendations fed to them as long as they are just good enough. One thing Seaver really gets right is the difference between the people working for these companies and their listeners. The people working there are not the users. They are the most geeky of music geeks, for the most part. There are some folks that are indifferent but many of the employees are deeply passionate about music, and this was true at the last music streaming service I worked for as well, even tho that one wasn’t the primary business of the overall company. They are the super users, and they are the people least likely to use a service like this in reality. But they have this shared vision of making it great for others. Of delivering the perfect music at the perfect time. Why?
My job is more concerned with information, the facts around the music, the tying together of metadata across verticals and across formats and across different manifestations. Ensuring that we understand that Johnny Rotten and John Lydon are in fact the same person. There’s a recommendation value in there for sure, because music doesn’t exist in a vacuum, but that’s not what interests me about it. It’s representing the complexity of life, its context and its relationships in a way that is extensible, for me.
Music genres are incredibly fascinating to me. All genres, really, but music genres in particular are super weird. Think about movie genres, or book genres. As a rule, they describe the aboutness of the book. The subject matter or theme, typically — Fiction, Nonfiction. Mystery, Horror, Romance. History, Science, Philosophy. You can quality it by place or time period or both - Regency Romance; Gothic Horror. Some of it is more specific than others. Cozy Animal Romance is a lot more hyper focused than Epic Fantasy, but each has a dedicated following and set of representative authors.
With music, so much is packed into a single word or short phrase that has little to nothing to do with the sounds you are hearing, and absolutely nothing to do with the subject matter. It’s completely different from other media in that respect. A rock song can be about anything - suicide, love, death, drinking at a bar, all of the above. The genre would tell you nothing about that. Genre is really about the sound and the feel. Certain themes tend to come along with certain genres from time to time, but not usually. I would say country music is a little more focused on themes than sound than some others. Musical taste says so much about people, or so we think, and yet I feel it says a great deal less about us than our taste in books might. Or am I wrong about that?
I’m off on a bit of a tangent. Anyway, I’m enjoying the book is my main point, and I also think it’s super interesting how people choose to write about and represent tech, especially from an academic standpoint. I have a couple of friends who are writing books in the field and I’ve offered to both of them to make introductions at my place of employment; is this wrong? I don’t know, probably. But I feel really passionately about the need for journalism and academia to study these places, particularly as algorithms are both proprietary and increasingly obtuse and unexplainable. Government is not going to regulate this shit. We need people who are investigating and writing about it and thinking about it in philosophical ways, outside the business world.
So, back to regular life. Round up of past few days: horrifically unproductive, but for once in my life actually had time to think. How do I build this in in future? How do I ensure space for myself? I don’t know, I excel at not considering my own needs, but let this be a reminder that it’s worth it to try.
I’ve managed to clean most of the food out of the fridge by either eating it or tossing it if it was already bad. It will be nice and empty for the family’s return tomorrow, ready for a Whole Foods run. It’s about 9:15 pm, still time for me to read some more tonight.
I’ve been reading a book called “Computing Taste” by Nick Seaver, about the people who design music recommendation algorithms. Since I work at a major music streaming company, I’ll leave you to guess which, I’m super interested in how this is perceived by people outside of tech. Being from the Midwest, and having a number of facebook friends from my childhood still, I see a real dichotomy between how the people I know in Seattle view/understand tech, and the people from my childhood that aren’t in this world. And then the people in academia, of whom I also know quite a few, are shades of in between depending on their field and age. I’m fascinated and dismayed by how little most people seem to understand about the way the internet works, how companies use their data, what “free” means when it comes to ad-served content, etc.
I myself don’t know that much about how it works, honestly, when it comes to music recommendation. I’m not a machine learning scientist. I know enough to know what is possible and what isn’t, and I have a vague sense of how it all works, and I guess this is more than a lot of people but I certainly can’t claim to be an expert. Before this book I read a book called “Spotify Teardown” which was fascinating and weird — part performance art, part academic experiment, equal parts bumbling and insightful and naive. It was truly bizarre. Instead of talking to anyone at Spotify, much of the book consisted of first, plumbing the depths of Swedish media in the aughts to reveal the early Spotify backstory (fascinating). Next, it engaged in a number of experiments to game the Spotify system by putting content on Spotify and observing its trajectory through the system, then testing its performance using bot-facilitated playback. It was kind of endearing, kind of ridiculous. Some of the book was very insightful and other parts read like willful ignorance.
At any rate, my point is that “Computing Taste” is much more academc and rigorous than “Spotify Teardown”. The author compares the notion of using algorithmic recommendations to hook a listener to anthropological literature about animal trapping and it’s an interesting lens to consider this through, this notion of keeping people on one’s platform as some sort of most dangerous game of hooking people with their own taste.
I think it is super interesting that music streaming companies are not really interested so much in selling “music” as they are a perfect algorithm. It’s a weird sort of problem, in the world of retail, because they all have the same catalog. With very few exceptions, the tracks are the same, and most listeners do not care that much about high-fidelity sound. Exclusive content isn’t an especially good way to win the game. They ostensibly cater to music fans but really the people they care most about are the mildly opinionated people, the ones that aren’t that picky about their music and are willing to have recommendations fed to them as long as they are just good enough. One thing Seaver really gets right is the difference between the people working for these companies and their listeners. The people working there are not the users. They are the most geeky of music geeks, for the most part. There are some folks that are indifferent but many of the employees are deeply passionate about music, and this was true at the last music streaming service I worked for as well, even tho that one wasn’t the primary business of the overall company. They are the super users, and they are the people least likely to use a service like this in reality. But they have this shared vision of making it great for others. Of delivering the perfect music at the perfect time. Why?
My job is more concerned with information, the facts around the music, the tying together of metadata across verticals and across formats and across different manifestations. Ensuring that we understand that Johnny Rotten and John Lydon are in fact the same person. There’s a recommendation value in there for sure, because music doesn’t exist in a vacuum, but that’s not what interests me about it. It’s representing the complexity of life, its context and its relationships in a way that is extensible, for me.
Music genres are incredibly fascinating to me. All genres, really, but music genres in particular are super weird. Think about movie genres, or book genres. As a rule, they describe the aboutness of the book. The subject matter or theme, typically — Fiction, Nonfiction. Mystery, Horror, Romance. History, Science, Philosophy. You can quality it by place or time period or both - Regency Romance; Gothic Horror. Some of it is more specific than others. Cozy Animal Romance is a lot more hyper focused than Epic Fantasy, but each has a dedicated following and set of representative authors.
With music, so much is packed into a single word or short phrase that has little to nothing to do with the sounds you are hearing, and absolutely nothing to do with the subject matter. It’s completely different from other media in that respect. A rock song can be about anything - suicide, love, death, drinking at a bar, all of the above. The genre would tell you nothing about that. Genre is really about the sound and the feel. Certain themes tend to come along with certain genres from time to time, but not usually. I would say country music is a little more focused on themes than sound than some others. Musical taste says so much about people, or so we think, and yet I feel it says a great deal less about us than our taste in books might. Or am I wrong about that?
I’m off on a bit of a tangent. Anyway, I’m enjoying the book is my main point, and I also think it’s super interesting how people choose to write about and represent tech, especially from an academic standpoint. I have a couple of friends who are writing books in the field and I’ve offered to both of them to make introductions at my place of employment; is this wrong? I don’t know, probably. But I feel really passionately about the need for journalism and academia to study these places, particularly as algorithms are both proprietary and increasingly obtuse and unexplainable. Government is not going to regulate this shit. We need people who are investigating and writing about it and thinking about it in philosophical ways, outside the business world.
So, back to regular life. Round up of past few days: horrifically unproductive, but for once in my life actually had time to think. How do I build this in in future? How do I ensure space for myself? I don’t know, I excel at not considering my own needs, but let this be a reminder that it’s worth it to try.
