“Indeed, are we each not essentially the sum of what we like, what we don’t like, what we find interesting, what makes us sad, what makes us happy? And are the decisions (likes, saves, views) we make when uninhibited not a near-perfect manifestation of these things?”
I've given this a fair bit of thought, not for the purposes of predicting human behavior so as to maximize engagement for profit, but for the purposes of analyzing what makes an artist's style distinctive, in my search for strategies to help me refine my own. The way you indirectly define identity here is, interestingly, very similar to how I've come to conceptualize style. The two concepts are closely related, of course, as evidenced by the existence of phrases such as (an artist's) “creative identity.” Whether or not we identify as artists is beside the point; indeed, it's uncanny how any red-blooded human's fingerprints end up being all over everything they've ever engaged with in an ”uninhibited” way, from the observable tendencies shot through their preferred creative aesthetic to the quirks of their handwriting to the idiosyncratic trail of digital breadcrumbs they leave behind on social media. People's behavior is unique to them in all of those realms, and in this regard, there's no question that ”we dramatically underestimate the substance of a meme.” (Well put, haha.)
That being said, I've also come to think of social media as being an avatar factory... one that has us making avatars of ourselves, for better or worse. You know the old chestnut, “If something's free, you're the product”? Let's run with that for a moment, since your post's title explicitly does just that. If ”I'm the product,” what kind of product am I?
The answer seems obvious: “I,” as the product, am a large collection of data points sold to advertisers who'll pay untold sums to maximize the time my eyeballs will spend in contact with an ad for their product or service. In other words, the data set that is “me” is one that enables profit, namely, transactions from which I get no kickbacks (aside from the entertainment value I get from shit-posting, and my option to talk to friends while ads float down my feed).
Getting back to objective facts: I am, and you are, a human being, not a data set. So I ask you again: What kind of product are we?
The machine does not, and is not designed to, recognize my humanity, nor can it *understand* my behavior (though it may well exhibit a stunning capacity to predict it, as an extremely fine-tuned algorithm designed to process unimaginable amounts of data). It exists only to document, in excruciating detail, the history of my online existence. Strictly speaking, there's no ghost in the machine capable of understanding identity, only an algorithm finely coded and calibrated to predict behavior. (Understanding involves processes such as perception, interpretation, inference, assumption, and others which necessarily involve human elements that are subjectivity and acts of will, i.e., decision-making.)
Your post's title tells me you've come to view yourself as the people behind social media platforms do, internalizing the notion that someone's identity can be aptly summarized based on what the machine ”makes of you,” and that this is a desirable outcome, since it enables bulletproof conclusions about a person's essential nature.
For your own and humanity's sake, I would suggest you revisit that conclusion.
Hyper-rationality can be valuable. It's also a double-edged sword, and a horribly dehumanizing one at that on the wrong side of the blade.
You'll probably disagree with a few of my points, 02Tenon, but be sure that I really enjoy reading your thoughts, and would love to read new posts more often. Cheers!
I suppose the way I'm describing the true self, the true identity, is who are you actually at your core. And in an environment where you can be uninhibited and have no real incentives to do anything, that core self is bound to manifest in various ways - and I would definitely agree that that would include someone's artistic style, to the extent that they did not start their art piece with a goal in mind.
" The answer seems obvious: “I,” as the product, am a large collection of data points sold to advertisers who'll pay untold sums to maximize the time my eyeballs will spend in contact with an ad for their product or service. In other words, the data set that is “me” is one that enables profit, namely, transactions from which I get no kickbacks (aside from the entertainment value I get from shit-posting, and my option to talk to friends while ads float down my feed).
"The machine does not, and is not designed to, recognize my humanity, nor can it *understand* my behavior (though it may well exhibit a stunning capacity to predict it, as an extremely fine-tuned algorithm designed to process unimaginable amounts of data). It exists only to document, in excruciating detail, the history of my online existence. Strictly speaking, there's no ghost in the machine capable of understanding identity, only an algorithm finely coded and calibrated to predict behavior. (Understanding involves processes such as perception, interpretation, inference, assumption, and others which necessarily involve human elements that are subjectivity and acts of will, i.e., decision-making.)"
I think we may have different definitions of understand. I more mean it in terms of 'I know what this thing is and how it works, even if I don't know why'. And I get what you're saying. But at the same time, if someone claimed they knew everything about their spouse, what they like, what they dislike, what assumptions they hold etc. and they were able to prove that by accurately predicting how they would respond to things, I think we would say 'wow that person really knows their S/O'. So at a certain point I think we generally accept, in other domains, that accurately being able to predict something is a great test of whether you truly understand it.
I think it freaks people out to contemplate the possibility that a machine knows us more than we know ourselves or those we are close to. But that doesn't make it false. I think there are romantic and whimsical elements of human nature. But we viewed rain, for instance, the same way [romantic and whimsical] before we understood how weather patterns work. Point being, I don't think disliking the dystopian implication of an algorithm knowing you more than you do necessarily makes it wrong, even if it is distasteful.
My view on this is that it is as simple as 'if you can accurately predict how someone will respond to something, you understand them'. Understanding is fundamentally about knowledge, and I don't think we can designate humanity and identity as being so rarefied that it cannot possibly ever be understood, even if we ourselves lack the ability to perfectly understand people (which I would argue is more the product of things like bounded rationality and that we simply never have the opportunity to have someone's mind laid out in front of us to see as a complete object, because the mind is an inherently nebulous and un-observable thing).
This reminds me a lot of how, I'm told, Renaissance got their investment algorithms so good. Really they just stopped caring about causality. They didn't care why something was going to happen, just that it was. So they were able to predict it. To blend that with this discussion, perhaps the algorithm doesn't know why you gravitate to certain types of content (perhaps it's because of something you experienced as a child - and frankly I don't think we even understand ourselves at that level, even if we think we do), but it does know what sentiments you like, what makes you happy, what makes you sad. So while I think you make a fair point in saying that that isn't the same thing as truly knowing someone's core, I do think it is a uniquely perfect snapshot of what lies immediately downstream from that core. But I do think some of your skepticism is rooted in an perception that conceding this point is the same as bowing to a tech dystopia. And I get why someone would hate that idea, but that doesn't make it false.
Weirdly, this discussion reminds me of what I've heard about our understanding of things like glass, electricity, and photosynthesis. In reality, we still don't know fundamental things about them. If my recollection is correct, our knowledge gap on those things really comes down to understanding how they work at their core; why they work, largely meaning we don't understand the quantum properties they abide by. And I think that's a good analogy to the algo's understanding of identity. True, we don't understand why the quantum physics of photosynthesis can work at room temperature yet our quantum computers need to be super cooled. But we still know photosynthesis works at room temperature; we understand what is immediately downstream from the core truth of the matter. And at the end of the day we understand plants so damn well that we are able to manipulate their genes and grow them at an immense scale. So perhaps the point is that algorithms are the best possible understanding of who we are that is in existence, but it is still one step removed from having an advanced understanding of who we are at our core; they are just unique in that they are the only entity who can even see our uninhibited core manifest at such scale (e.g. hours of scrolling).
re: "I think it freaks people out to contemplate the possibility that a machine knows us more than we know ourselves or those we are close to."
Yes, it does! – not because I'm afraid of the machine's knowledge of my most-secret secrets, but because I'm leery of the people running the damned machine, precisely because they've fine-tuned it so well to serve their goals and agenda(s).
re: "some of your skepticism is rooted in an perception that conceding this point is the same as bowing to a tech dystopia. And I get why someone would hate that idea, but that doesn't make it false."
I actually agree that "the algo understands me better than I do" may not be objectively false. However, my skepticism isn't (only...) to do with refusing to surrender to a tech dystopia. Admittedly, the idea of being subjected to the machine alternately makes me queasy and furious. But I'd counter that my skepticism is rooted in the idea that collective misattribution of *sentience* to the machine would not only misguided, but phenomenally dangerous. (To get my gist here, just replace a fallacious, politically-motivated "the Devil made me do it" with a fallacious, politically-motivated "the machine wrought X havoc of its own free will." I guess we're still talking about an unwillingness to bow to a tech dystopia, but the point I'm making turns on the eventuality of people's flawed belief in machine sentience, and not on reasonable assertions around the machine's ability to extrapolate predictions around human behavior from mounds of data.
I have to run, but this has been fun. And I'd love to know more about Renaissance investment algorithms, if you care to post a link.
Oh yeah I don’t mean the algo understands us in that it is sentient. Im defining it much more simply than that. Just meaning ‘in possession of knowledge’. Like that profile exists somewhere. Can’t imagine it’s directed to actually make written profiles of people.
The Renaissance book is called ‘The Man Who Solved the Market’. Super interesting.
“Indeed, are we each not essentially the sum of what we like, what we don’t like, what we find interesting, what makes us sad, what makes us happy? And are the decisions (likes, saves, views) we make when uninhibited not a near-perfect manifestation of these things?”
I've given this a fair bit of thought, not for the purposes of predicting human behavior so as to maximize engagement for profit, but for the purposes of analyzing what makes an artist's style distinctive, in my search for strategies to help me refine my own. The way you indirectly define identity here is, interestingly, very similar to how I've come to conceptualize style. The two concepts are closely related, of course, as evidenced by the existence of phrases such as (an artist's) “creative identity.” Whether or not we identify as artists is beside the point; indeed, it's uncanny how any red-blooded human's fingerprints end up being all over everything they've ever engaged with in an ”uninhibited” way, from the observable tendencies shot through their preferred creative aesthetic to the quirks of their handwriting to the idiosyncratic trail of digital breadcrumbs they leave behind on social media. People's behavior is unique to them in all of those realms, and in this regard, there's no question that ”we dramatically underestimate the substance of a meme.” (Well put, haha.)
That being said, I've also come to think of social media as being an avatar factory... one that has us making avatars of ourselves, for better or worse. You know the old chestnut, “If something's free, you're the product”? Let's run with that for a moment, since your post's title explicitly does just that. If ”I'm the product,” what kind of product am I?
The answer seems obvious: “I,” as the product, am a large collection of data points sold to advertisers who'll pay untold sums to maximize the time my eyeballs will spend in contact with an ad for their product or service. In other words, the data set that is “me” is one that enables profit, namely, transactions from which I get no kickbacks (aside from the entertainment value I get from shit-posting, and my option to talk to friends while ads float down my feed).
Getting back to objective facts: I am, and you are, a human being, not a data set. So I ask you again: What kind of product are we?
The machine does not, and is not designed to, recognize my humanity, nor can it *understand* my behavior (though it may well exhibit a stunning capacity to predict it, as an extremely fine-tuned algorithm designed to process unimaginable amounts of data). It exists only to document, in excruciating detail, the history of my online existence. Strictly speaking, there's no ghost in the machine capable of understanding identity, only an algorithm finely coded and calibrated to predict behavior. (Understanding involves processes such as perception, interpretation, inference, assumption, and others which necessarily involve human elements that are subjectivity and acts of will, i.e., decision-making.)
Your post's title tells me you've come to view yourself as the people behind social media platforms do, internalizing the notion that someone's identity can be aptly summarized based on what the machine ”makes of you,” and that this is a desirable outcome, since it enables bulletproof conclusions about a person's essential nature.
For your own and humanity's sake, I would suggest you revisit that conclusion.
Hyper-rationality can be valuable. It's also a double-edged sword, and a horribly dehumanizing one at that on the wrong side of the blade.
You'll probably disagree with a few of my points, 02Tenon, but be sure that I really enjoy reading your thoughts, and would love to read new posts more often. Cheers!
I suppose the way I'm describing the true self, the true identity, is who are you actually at your core. And in an environment where you can be uninhibited and have no real incentives to do anything, that core self is bound to manifest in various ways - and I would definitely agree that that would include someone's artistic style, to the extent that they did not start their art piece with a goal in mind.
" The answer seems obvious: “I,” as the product, am a large collection of data points sold to advertisers who'll pay untold sums to maximize the time my eyeballs will spend in contact with an ad for their product or service. In other words, the data set that is “me” is one that enables profit, namely, transactions from which I get no kickbacks (aside from the entertainment value I get from shit-posting, and my option to talk to friends while ads float down my feed).
"The machine does not, and is not designed to, recognize my humanity, nor can it *understand* my behavior (though it may well exhibit a stunning capacity to predict it, as an extremely fine-tuned algorithm designed to process unimaginable amounts of data). It exists only to document, in excruciating detail, the history of my online existence. Strictly speaking, there's no ghost in the machine capable of understanding identity, only an algorithm finely coded and calibrated to predict behavior. (Understanding involves processes such as perception, interpretation, inference, assumption, and others which necessarily involve human elements that are subjectivity and acts of will, i.e., decision-making.)"
I think we may have different definitions of understand. I more mean it in terms of 'I know what this thing is and how it works, even if I don't know why'. And I get what you're saying. But at the same time, if someone claimed they knew everything about their spouse, what they like, what they dislike, what assumptions they hold etc. and they were able to prove that by accurately predicting how they would respond to things, I think we would say 'wow that person really knows their S/O'. So at a certain point I think we generally accept, in other domains, that accurately being able to predict something is a great test of whether you truly understand it.
I think it freaks people out to contemplate the possibility that a machine knows us more than we know ourselves or those we are close to. But that doesn't make it false. I think there are romantic and whimsical elements of human nature. But we viewed rain, for instance, the same way [romantic and whimsical] before we understood how weather patterns work. Point being, I don't think disliking the dystopian implication of an algorithm knowing you more than you do necessarily makes it wrong, even if it is distasteful.
My view on this is that it is as simple as 'if you can accurately predict how someone will respond to something, you understand them'. Understanding is fundamentally about knowledge, and I don't think we can designate humanity and identity as being so rarefied that it cannot possibly ever be understood, even if we ourselves lack the ability to perfectly understand people (which I would argue is more the product of things like bounded rationality and that we simply never have the opportunity to have someone's mind laid out in front of us to see as a complete object, because the mind is an inherently nebulous and un-observable thing).
This reminds me a lot of how, I'm told, Renaissance got their investment algorithms so good. Really they just stopped caring about causality. They didn't care why something was going to happen, just that it was. So they were able to predict it. To blend that with this discussion, perhaps the algorithm doesn't know why you gravitate to certain types of content (perhaps it's because of something you experienced as a child - and frankly I don't think we even understand ourselves at that level, even if we think we do), but it does know what sentiments you like, what makes you happy, what makes you sad. So while I think you make a fair point in saying that that isn't the same thing as truly knowing someone's core, I do think it is a uniquely perfect snapshot of what lies immediately downstream from that core. But I do think some of your skepticism is rooted in an perception that conceding this point is the same as bowing to a tech dystopia. And I get why someone would hate that idea, but that doesn't make it false.
Weirdly, this discussion reminds me of what I've heard about our understanding of things like glass, electricity, and photosynthesis. In reality, we still don't know fundamental things about them. If my recollection is correct, our knowledge gap on those things really comes down to understanding how they work at their core; why they work, largely meaning we don't understand the quantum properties they abide by. And I think that's a good analogy to the algo's understanding of identity. True, we don't understand why the quantum physics of photosynthesis can work at room temperature yet our quantum computers need to be super cooled. But we still know photosynthesis works at room temperature; we understand what is immediately downstream from the core truth of the matter. And at the end of the day we understand plants so damn well that we are able to manipulate their genes and grow them at an immense scale. So perhaps the point is that algorithms are the best possible understanding of who we are that is in existence, but it is still one step removed from having an advanced understanding of who we are at our core; they are just unique in that they are the only entity who can even see our uninhibited core manifest at such scale (e.g. hours of scrolling).
Thanks for the read!
Thanks so much for the thoughtful response! :)
A couple quick thoughts,
re: "I think it freaks people out to contemplate the possibility that a machine knows us more than we know ourselves or those we are close to."
Yes, it does! – not because I'm afraid of the machine's knowledge of my most-secret secrets, but because I'm leery of the people running the damned machine, precisely because they've fine-tuned it so well to serve their goals and agenda(s).
re: "some of your skepticism is rooted in an perception that conceding this point is the same as bowing to a tech dystopia. And I get why someone would hate that idea, but that doesn't make it false."
I actually agree that "the algo understands me better than I do" may not be objectively false. However, my skepticism isn't (only...) to do with refusing to surrender to a tech dystopia. Admittedly, the idea of being subjected to the machine alternately makes me queasy and furious. But I'd counter that my skepticism is rooted in the idea that collective misattribution of *sentience* to the machine would not only misguided, but phenomenally dangerous. (To get my gist here, just replace a fallacious, politically-motivated "the Devil made me do it" with a fallacious, politically-motivated "the machine wrought X havoc of its own free will." I guess we're still talking about an unwillingness to bow to a tech dystopia, but the point I'm making turns on the eventuality of people's flawed belief in machine sentience, and not on reasonable assertions around the machine's ability to extrapolate predictions around human behavior from mounds of data.
I have to run, but this has been fun. And I'd love to know more about Renaissance investment algorithms, if you care to post a link.
Oh yeah I don’t mean the algo understands us in that it is sentient. Im defining it much more simply than that. Just meaning ‘in possession of knowledge’. Like that profile exists somewhere. Can’t imagine it’s directed to actually make written profiles of people.
The Renaissance book is called ‘The Man Who Solved the Market’. Super interesting.
Will check it out. Thanks!