This is a really thoughtful article and also in a way, it’s reassuring that the algorithm doesn’t ACTUALLY know me, it’s just categorized me into demographics. I’ve always been freaked out by the algorithm knowing too much about me, but when you break it down like this and what it’s actually motivated by, it’s a little more reassuring.
Here from ironically, the fyp. I’m always super engaged by your videos but this is the First time reading your Substack. I really appreciate how digestible this is. No bells n whistles just good, clear and insightful educational content. There’s so much other overwhelming media, this was a breath of fresh air. Will be keeping up w future substacks (-:
This is an interesting concept to digest and actually I think it makes so much more sense now how easily AAVE has spread and been diluted on social media. It's not just the simple fact that people are being exposed to certain words and phrases more, but that they feel it was intentionally placed before them so they've gained permission to use it. Cool stuff.
It does make me want to explore the wider implications on black culture as a whole and see where else this dilution occurs.
this but you put in such a well thought out way that the concept was sharpened and clarified in my mind. I watch your tiktoks cause i’m a writer and an english teacher; I think linguistics is interesting and fun, and it was probably the algorithm that led me here in the first place, but i gotta say your sub stack posts are very well written and very insightful.
u are literally so cool god bless the day i subscribed. your articles make my day fr and im studying linguistics rn partly because of how interested i became in yours and similar creators' videos
Isn't the core point a truism that people, for the most part, do already understand? That "my" algorithm isn't about me but a string of numbers that represent interests similar to mine is something widely understood, isn't it?
I absolutely love your work and I do appreciate how you've explained your argument here. I'm just curious whether this egotistical assumption that the FYP is "for me" is really one that is made very often.
I couldn't help but think about the people who comment, for instance, in a recipe video that uses a lot of cheese, that they're lactose intolerant. As if it were the creator's fault or mistake for not having that in mind when posting that recipe. I wonder how many of those people truly forget that those videos are not specifically for them.
I took a philosophy of mind class recently that delved into the token manipulation systems some AI's are founded on. It's really interesting stuff. The level of abstraction these models use basically gets to a level of perfect, digital games like chess (although how accurate this is to the algorithm's ran by social media companies today, I do not know). By this logic, it makes sense that the algorithm isn't getting the full picture on what the user themselves is like.
Still, I never realized just how depersonalized content creation can be as it turns to a way of maximizing "engagement" over "views" or "subscribers". I'm used to the latter with regards to YouTube, but short-form content seems to take things a step further in a way I never fully digested until AFTER I left it. It seems so obvious now, but it's really hard to tell just how warped the algorithm's idea of you is when you're constantly being bombarded with more information. I really feel like I'm getting the bigger picture about social media when I read these pieces. Bless Substack, and bless you, Adam.
"[The algorithm] is working with the cave-shadows it can encode out of both the creator’s message and your personality (and selectively reinterpreting them to maximize profit)"
In your model describing the process of encoding and decoding including the algorithm, shouldn't there be a visual channel as well? I don't know how relevant it is linguistically, but since we are now mainly consuming audiovisual media on Instagram/Tiktok I thought it would be worth mentioning.
This is a really thoughtful article and also in a way, it’s reassuring that the algorithm doesn’t ACTUALLY know me, it’s just categorized me into demographics. I’ve always been freaked out by the algorithm knowing too much about me, but when you break it down like this and what it’s actually motivated by, it’s a little more reassuring.
Here from ironically, the fyp. I’m always super engaged by your videos but this is the First time reading your Substack. I really appreciate how digestible this is. No bells n whistles just good, clear and insightful educational content. There’s so much other overwhelming media, this was a breath of fresh air. Will be keeping up w future substacks (-:
Here's an interesting piece of research on how algorithmic recommendar systems and captology can be understood through the anthropology of trapping:
https://www.semanticscholar.org/paper/Captivating-algorithms%3A-Recommender-systems-as-Seaver/174ebfffe983ee2b14dc3ade9dbd4c03e5b4068c
Definitely looking forward to checking out Algospeak, though, because the way things are framed in this essay make a ton of sense as well!
This is an interesting concept to digest and actually I think it makes so much more sense now how easily AAVE has spread and been diluted on social media. It's not just the simple fact that people are being exposed to certain words and phrases more, but that they feel it was intentionally placed before them so they've gained permission to use it. Cool stuff.
It does make me want to explore the wider implications on black culture as a whole and see where else this dilution occurs.
I feel like I already vaguely knew all of
this but you put in such a well thought out way that the concept was sharpened and clarified in my mind. I watch your tiktoks cause i’m a writer and an english teacher; I think linguistics is interesting and fun, and it was probably the algorithm that led me here in the first place, but i gotta say your sub stack posts are very well written and very insightful.
u are literally so cool god bless the day i subscribed. your articles make my day fr and im studying linguistics rn partly because of how interested i became in yours and similar creators' videos
Isn't the core point a truism that people, for the most part, do already understand? That "my" algorithm isn't about me but a string of numbers that represent interests similar to mine is something widely understood, isn't it?
I absolutely love your work and I do appreciate how you've explained your argument here. I'm just curious whether this egotistical assumption that the FYP is "for me" is really one that is made very often.
I couldn't help but think about the people who comment, for instance, in a recipe video that uses a lot of cheese, that they're lactose intolerant. As if it were the creator's fault or mistake for not having that in mind when posting that recipe. I wonder how many of those people truly forget that those videos are not specifically for them.
Fantastic use of Plato's cave! It captures the message in such a succinct manner.
Oh Lord, not the Allegory of the Cave. That really makes me rethink so many things
Much needed.
I took a philosophy of mind class recently that delved into the token manipulation systems some AI's are founded on. It's really interesting stuff. The level of abstraction these models use basically gets to a level of perfect, digital games like chess (although how accurate this is to the algorithm's ran by social media companies today, I do not know). By this logic, it makes sense that the algorithm isn't getting the full picture on what the user themselves is like.
Still, I never realized just how depersonalized content creation can be as it turns to a way of maximizing "engagement" over "views" or "subscribers". I'm used to the latter with regards to YouTube, but short-form content seems to take things a step further in a way I never fully digested until AFTER I left it. It seems so obvious now, but it's really hard to tell just how warped the algorithm's idea of you is when you're constantly being bombarded with more information. I really feel like I'm getting the bigger picture about social media when I read these pieces. Bless Substack, and bless you, Adam.
"[The algorithm] is working with the cave-shadows it can encode out of both the creator’s message and your personality (and selectively reinterpreting them to maximize profit)"
Absolute killer line.
In your model describing the process of encoding and decoding including the algorithm, shouldn't there be a visual channel as well? I don't know how relevant it is linguistically, but since we are now mainly consuming audiovisual media on Instagram/Tiktok I thought it would be worth mentioning.