Many argue that the internet has destroyed more jobs than it’s created, and that as such our information economy has obliterated the middle class. The common “solution” proposed for this problem is to create systems that pay users for their data. If you join Twitter or Facebook (Meta) then you should somehow be compensated since it’s your data that make these companies successful. This is the argument Jaron Lanier makes in his book Who Owns the Future. I will argue that compensating users for data is a non-solution because of the way networks convert user data into product features. I will argue that rather than compensate users for their data it makes more sense to change the network topology such that more people can create lucrative enterprises under the current model.
Check out the video version: https://www.youtube.com/@nontrivialpodcast
Many argue that the internet has destroyed more jobs than it’s created, and that as such our information economy has obliterated the middle class. The common “solution” proposed for this problem is to create systems that pay users for their data. If you join Twitter or Facebook (Meta) then you should somehow be compensated since it’s your data that make these companies successful. This is the argument Jaron Lanier makes in his book Who Owns the Future. I will argue that compensating users for data is a non-solution because of the way networks convert user data into product features. I will argue that rather than compensate users for their data it makes more sense to change the network topology such that more people can create lucrative enterprises under the current model.
Check out the video version: https://www.youtube.com/@nontrivialpodcast
Many argue that the internet has destroyed more jobs than it's created. And that as such, our information economy has obliterated the middle class. The common solution proposed to this problem is to create systems that pay users for their data. So if you join Twitter or Facebook, now meta, then you should somehow be compensated since it's your data that make these companies successful. This is the argument that Darren Laer makes in his book who owns the future. I will argue that compensating users for data is a non solution because of the way networks convert user data into product features. I will argue that rather than compensate users for their data, it makes more sense to change the network topology so that more people can create lucrative enterprises under the current model. Let's get started. OK. So we're all used to using networks when I, when I say networks, I mean things like these services or these products that we use like Twitter or Facebook, which is now called meta uh even Google search engines, right? We all log on to these uh you know, massively connected as in millions, hundreds of millions, sometimes billions of people networks that provide a value to what we are looking for or who we're connecting with or somehow provide a value back to the user. Ok. So if I'm gonna, if I'm Facebook slash meta, uh, or if I'm, if I'm logging on to Facebook slash meta, then there's obviously a lot of value to that network because I can connect maybe with, you know, former classmates or friends and family. Uh, you know, I can put myself out there. Maybe I have a brand that I'm trying to get, you know, there's a lot of value that I can get, um, out of using a service like Facebook slash meta. Of course, Twitter, the same thing. Uh, you know, you could also think about it in terms of maybe even news. I mean, maybe you don't really trust the news these days, maybe you think it's too biased. Maybe you want to go to something like a Twitter because at least it's a lot of people just talking about what's going on. You know, maybe you think it's less of an echo chamber. I don't know, but there's a lot of value potentially to joining these networks. OK. That's what I mean, when I say network, you know, Facebook slash Metas, the Twitters, you know, the Google, the ebays, these companies are, are, you know, essentially in place to connect millions of people together and they do that by, you know, kind of being this middle man that says you know, we're just gonna connect you and then you guys kind of do what you want. We're gonna have some kind of minimal regulation presumably over the content and what happens. And uh and we're gonna try to take advantage of essentially these network effects and that's what I want to start talking about in this episode of the network effects. OK. So the way we're gonna, you know, the way I'm gonna do this is, is let's talk about networks, let's understand the properties of networks. And then I'm gonna talk uh talk about some of the challenges that that actually sets up for individuals who are, you know, just in a society participating and trying to get the value of these networks, you know, and that kind of leads into some of this wealth disparity that we see in today's society. In other words, you know, a lot of the blame of the kind of obliteration of the middle class if you will, is on these tech companies that exist in the information economy because they leverage these network effects, they concentrate, you know, the wealth towards the center and then people are using these free networks but not getting compensated themselves. And so, you know, that, that this is evidenced essentially by the fact that, you know, if, if, if the internet was, you know, creating all these jobs, why is there this kind of dearth in the middle class? Why is there seem to be this concentration of wealth and not, and uh and, and not a lot of people making money in the middle uh as the argument goes, and then we'll take a look at some of the underlying mechanisms of why those things are happening. And then I'll wrap that up with uh you know, the, the Jaron Laer solution from his book, uh who owns the future. And then my solution that I think, uh you know, makes a little more sense based on respecting the properties of complex networks. So let's start talking about networks in general. I think it's important that people understand how these things work. So again, when I say network, I'm uh I'm just simply talking about these uh these organizations, these companies that are providing uh you know, basically like a social network, right? Connecting people together uh allowing them to maybe make transactions if it was ebay, allowing them to search the internet, if it was a Google search, allowing them to uh connect with um you know, all kinds of people if it's Twitter, uh you know, Facebook slash meta. So, so these are networks, these are a bunch of think of them as a bunch of nodes connected together and a lot of those nodes um when they make these connections will end up uh well, let's step back a little bit. So, so if you have a bunch of nodes that are connecting uh the more nodes that would come into the network, the more valuable that network is going to be. And this is called MET CAF S law. So this is kind of just a general law, general property or rule uh about networks in general. So the more in the case of a social network, people that are being added uh to the to the network, the more value you can potentially get out of it. And of course, that makes sense. So if I'm gonna be using, you know, a bunch of um you know, machines that are connected via the internet and I can search those machines for information, presumably, it's a lot more valuable, the more machines that are there because there's more information that can be, you know, weighed and compared and then brought back to me in my search. In other words, my search can be more relevant, right? Um And we know that, you know, there's these page rank algorithms that Google uses to kind of make all that happen. So the more knows that join that network that's more valuable to me as a user and that's more valuable to the owners of these companies like Google. Um because the more value they can add, they can of course sell that to advertisers, they get a bigger share of the market and on and on. Um if I'm, if I'm jumping onto Twitter and I know you got a lot more people on Twitter that are saying a lot more things presumably that could be more valuable to me. I mean, if I'm trying to put my brand out there and I'm, I want to be heard, maybe it's not even a brand, maybe I'm just like tweeting and II I want to assume that people are, are reading my tweets, you know, I want to be heard, so to speak. Then the bigger that network is more valuable potentially, that is to me, of course, if I am my own company or I'm a, I'm advertising or I'm trying to take a specific maybe agenda approach to Twitter, you know, that's more valuable to me knowing that a lot of people are connected. OK. So, so this Met CAFS Law is this general property of these networks that have a lot of nodes that are connecting. And so of course, we can consider these complex networks as I've discussed on other episodes. Um That's gonna have more value, the more people that are connected, the more nodes that are in that network as a general property. So let's understand that I'll, I'll use that, I'll carry that um property throughout this episode as we start to think about the challenges that, that, that, that uh a arise from that some of the more core core patterns that emerge from that. And then again, the the solutions that we get into for this kind of disappearance of the middle class and what have you um another property of these complex networks is the formation of hubs. So when you have many nodes coming together in the network, they're not just all evenly dispersed. OK. Imagine throwing a bunch of rice grains up into the air and then all those rice grains fall down. Would you expect to see all those rice grains kind of uh uh equidistant from each other? In other words, they're all evenly spread out. Well, now you kind of get these clumps, right? You get spread out in some areas and be clumps in others. Well, we see this in complex networks. So if you think of the nodes being connected by lines, when I say node, just think of like a little ball, right? Those little balls coming into the network, that's an agent, that's a human, that's a machine, whatever it is that comprises the network, those are connecting to each other with lines. Well, some of those nodes, right, some of those individual pieces or people will have a lot more lines connecting to them than you know, the the average or the or the the you know what the normal person coming in. So in a social network, this would make sense, right? We have influencers. So you're gonna have people on Twitter, for example, who will have, you know, a million followers or whatever, you know, or you know, 5 600,000 followers. And you consider those kind of influential people because they have a massive audience, anything they tweet anything they say is gonna be read by, you know, many, many, many people, potentially millions. And so there's a lot of weight to what they do because there's that concentration of information towards those hubs. And so all complex networks typically have these hubs that form and they form naturally, that's an expected property of complex networks. OK. So, so we've got met CAF S law, which is saying that, you know, in, in, in a kind of an obvious sense, right? The more nodes or the more people that are joining a network, the more valuable it is that's more valuable to the owners of the networks. OK. So if you're running a server, if you're the, you know CEO of Twitter or the original founders of, you know, Instagram or whatever, that's gonna be very, very valuable to you as the owner, to the, to the creator of these servers. Um the fact that the value goes up and, and, and under met CAF Law specifically, you know, the value is increasing as the square of the number of nodes. And so you can kind of think of this essentially as an exponential increase in the value of the network. So it's not just like, you know, one user, two years or three years, three user is, you know, one value, two value, three value, right? You're 2468 10, you're, you're, you're going up in an exponential fashion. So it it grows as the square So it's a very rapid growth. OK? It's a very rapid growth, the more users I get one more user that's already a lot more value. Um And then the other property is the hub formation. This this the fact that as more nodes are coming into the network, you don't expect them to all be kind of evenly, you know, dispersed. It's not like, you know, a, an, an ideal gas law from chemistry or something where, you know, everything's like a billiard ball and it's kind of its own thing and it's discrete and they all just interact in this equilibrium. You've got this concentrations, these concentrations that happen, these major hubs that have the overwhelming majority of links attached to them and then the overwhelming majority of other nodes in that network, I just kind of have one or two links or maybe 100 links or maybe 1000 links but not a million. OK. So, so concentrations of hubs. So those are the two properties that I want, I wanna take into this conversation, you got the Met Law and you got the concentration of hubs. So what challenges do these now present? Right. Um You, so, so I've kind of just taken that unbiased look at, at networks in general like we know this about networks, these are how they work, right? Well, when we start to think about society and you know, the economy, the fact that we are so many of us if not all of us are jumping onto these networks, this presents some challenges. So, with respect to Metcalf's law, the idea that the value increases, the more people that are on it, that sounds like, you know, would not only be a positive thing. Well, you got to think that a lot of these networks are almost more like a public utility. Ok. And, and by public utility, I really mean, well, let's think about what the definition of public utility is, right? It's really an organization that maintains the infrastructure for our public service, right? It's uh you know, public utilities are subject to forms of public control and regulation ranging from local community based groups to statewide government monopolies. But the point is that they are meant to supply goods and services that are considered essential, right? Water, gas, electricity, telephone and other communication systems represent much of the public util uh utility market. Ok. So goods and services that are considered essential. So when you think of something like a Twitter, you know, a Facebook slash meta might not think, well, it's not essential, right? I can jump off that network at any time, you know, and, and what would be the big deal. Um But that's not necessarily true because there could be a downside to me leaving that network right now. One is, you know, I'm connected to family and friends and maybe that's my primary mode of connection. And if I leave that network, that net that, that connection gets severed. So that's a downside. But you might be like, yeah, ok. But I mean, it's not like it's the only way to connect. But, you know, maybe that's the way I'm putting a brand out there. Maybe that's the way I primarily communicate. Maybe that's, you know, maybe I want, if I'm applying to jobs, maybe I want people to go back to the Facebook page or the Twitter page and kind of see what I've built there, the things that I talk about the links that I'm providing, you know, maybe it's almost, um I mean, it could be a number of things, but basically, your identity is essentially tied up into these social networks. Now, if I'm running a business and I'm, and I'm, you know, either directly or indirectly kind of promoting that business through the social network, then that, that ends up having a lot of value and you can go on and on and really think of examples. And if, if you think of today's social networks, especially ones like Facebook meta that have, you know, billions of users, these things are almost like a public utility. They uh people depend on them, they depend on them to communicate, to connect with people to know what's going on in the world. You know, if, if I am not on this network, you know, there might be other organizations that I need to use or want to use and they say, well, go ahead and check our Facebook page or connect with us on, you know, Twitter or, you know, I'm only selling this product on ebay. So you need to go do the transaction there or on Amazon, whatever it is, if you pull away from these, you know, Amazon would be another good example, right? Millions and millions of all these products, you know, these buyers, buyers and sellers are kind of being uh brought together right on Amazon and, and you're getting things shipped to you that you wouldn't have access to otherwise and maybe it's getting, you know, shipped to remote areas, you know, you don't have the ability to go into a store, whatever you can go on and on. I mean, these things are becoming almost like public utilities because we really, really depend on them, you know, think about COVID the last couple of years. I mean, how, you know, Amazon did quite well, I think, right. And, and it makes sense because, you know, you could just go find what you want online and get it delivered to you. So there's a, there's a real dependence there. And so when we talk about the Met Caf Law property, the fact that networks um have a lot of value and the more people that are on it, the more valuable it gets there is a downside to that because you can think of that value as a dependency, right? As an individual. I am signing on to a super valuable network where there is a social expectation that I'm going to be on there, go to Amazon to find that product, go to ebay to make that transaction, go on Twitter to see what's going on, connect with us on Facebook slash meta. In order to yada, yada yada, that's a real social dependency if you will. And if I pull away from that, you've kind of got the FOMO, right? FOMO, the fear of missing out that that's already a bit of kind of anxiety for people. But you've got even deeper than that, this idea that if I walk away from this network, um I might really be missing out on opportunities, maybe jobs are getting posted on those networks, whatever, right? There's all kinds of examples. So, so Metcalf's law brings a lot of value to it, but it also creates dependency. So keep that kind of um challenge in mind at a social level, these things are not simply neutral. OK? And um yeah, so, so, so there's that and, and then I want to talk about the challenge also of this middle class, potentially disappearing. Now, it might seem weird, you know, why are you talking about the middle class disappearing along with the information economy? I mean, is that really the internet's fault? So Jared LA in his book, who owns the future talks about this, right? If you look at these serve. So when I, when I say a company like um you know, like a Facebook meow or a Twitter, these organizations are, what they're doing is, is these are companies that are trying to take advantage of these network effects. The network effects are things like the Met CAF S Law um or things like the concentration of, of information to, to a very few hubs. These are very scalable things. So if I'm gonna start a company and I want that service to be provided through software, we already know software. And I've talked about that before is, isn't, you know, because it's an informational thing, not a physical thing, it's very, very scalable, right? Remember the difference when we talk, you know, you know the Wework episode, right? If you have an under understand uh an an underlying physicality like a real estate company and you pretend to be a software company, that's gonna be a really big problem because you're not actually a software company or a real estate company, the physicality of real estate can only grow so fast. But if you're truly a software company, that software company can grow very, very fast because it's just information, information can be copied, it can be replicated. Um you can make, you know, a million copies of something for the same price as making one copy of something for the most part. I mean, not getting into how you obviously manage that infrastructure, but generally speaking, software grows, grow uh grows very rapidly. Well, in addition to that, if my software product happens to be uh something that provides a service over a very large network like this. So millions hundreds of millions of people can join this thing and they can interact and they can do transactions, do information, the value keeps growing and growing and growing. Uh you know, this, this is extremely scalable. And so as a company, you know, and you see this in Silicon Valley a lot, right? They can scale very rapidly, they can generate a ton of wealth in a very short amount of time. That's why typically companies like this uh attract a lot of VC dollars, right? A lot of venture capitalist dollars because they're so scalable because they can concentrate a lot of wealth very, very rapidly. Ok. Well, in that network, one of the reasons so many people jump onto it is because it's free, right? I don't have to pay anything to join Twitter, right? In terms of actual money, right? I don't have to subscribe to that. Um You know, I can join that for free. I can join uh you know, Facebook slash meta for free. I can go use Google for free. I mean, other than, you know, uh paying your internet service provider for the internet connection, all the actual service that's provided by these by these companies are free. So it's easy. There's, it's a kind of not a lot of friction to go ahead and join them. But if I'm jumping onto these free networks, that's not all necessarily positive because I'm not getting compensated for my use of the network. Now, you might say, well, why would you get compensated for the use of the, I mean, why, you know, what are you talking about? Right. It's just free, like, go ahead and use it. But start to think and, and Jaren Lanier talks about this, how the value of these uh companies are actually gene, how, how are they generating their money? Right. Well, what they're doing is they're taking your data and they're turning that into product features that in turn essentially kind of get sold back to you or get brought back into the product and, and continue to improve the product and that increases the value and that invites even more people to use it or to, to the the same people to use it for a longer period of time. They kind of get addicted to it. You get a lot of these addictive kind of features baked into the uh into the products that are generated from user data. So it's not just you using a service and, and, and it's kind of that's it, you know, walk away whenever you want to use it, you just kind of input data and then you know, you just get your search result or you just connect with people or you do your transaction, your data is being captured by these pro these networks and it's being mixed and mangled with other, you know, millions of other pieces of user uh data. And then you have a lot of technologies like machine learning that come in and they find these statistical correlations between all these data points and the the outputs that they want to produce in the product. So for example, a recommendation engine, right, if I go on to Amazon and I am seeing, you know, books being recommended to me or Amazon Prime or something, you know, videos being recommended to me, you have advertisements on Twitter, um annoyingly, right? You're going through Twitter and you have all these advertisements on, you know, Facebook, same thing, whatever these companies generate a very large most of their revenue often is through uh uh the these marketing agencies, it's through ad revenue. OK? Because they're free, they're free for people to join and they need to make that money somehow. So they're taking your data, they're understanding what the market wants and then they're presenting ads to you, um you know, you know, via this mechanism and that's essentially how they make money. OK? So you might say, OK, well, I mean, that still makes sense, right? So what's the issue and what's the connection to the middle class? Well, this means that people are joining these networks and don't necessarily have the ability to create wealth themselves, right? So let's go back to thinking about these networks as almost utilities, we have this dependence. It's not just a fun, you know, you feature uh a product on your phone, you can use. We're actually talking about, you know, major social connections, you know, major networks that people are jumping on. That's where they get their information. That's where they find out where things are. That's how they connect to companies. That's how they do transactions that, that, you know, that's how they make purchases and get things delivered to their home. It's almost more like this public utility. And so it's this massively connected system, but when you're a part of it, it, it, it's, it's also using your data to generate all these features and you're giving that data, you're giving that data, you're giving that data, but you're not getting compensated for it. And if you now look at this, this kind of economy that we're in and we have this extreme concentration of wealth, you know, you've got the Elon Musk and the Jeff Bezos of the world. You know, you got a single individual worth $300 billion and you don't have the middle class that you used to. And you've got, you know, the millennials and what have you, who can't buy homes. It's not like what their parents used to have. You, you really get into this idea that, well, wait a second, the economy isn't set up that well and the old and, and the things that drive the economy are these networks are these software products by and large and those software products have properties about them that concentrate the wealth, right? They have, you know, this the Mark Zuckerberg's can be so rich because of the way that you have this aggregation, this hub formation and networks that concentrate the wealth. In other words, would Mark Zuckerberg have that much money if users were being compensated for the data. OK. That, yeah, so, so think about this argument, right? Like I, if I go on the Facebook slash meta, I, I call it both, right? Because people still think of it as Facebook, right? Facebook slash meta. And I keep giving you my data, giving you my data, giving my data and you turn that into and into advertisements to sell back to me, you're making millions and millions of dollars, you know, why shouldn't I be compensated for it? Especially if I'm in an economy where one you get to have that much money because of these network effects. But two me on this network does not get to have any wealth generation even though I'm giving it to you. So I'm kind of giving giving you, let's say the Silicon Valley entrepreneurs the ability to concentrate their wealth through my data. And we're in an information economy that uses data, but I can't any of that wealth for myself slash I can't get a good job or I, you know, I maybe I can't get into that middle class. There's this huge wealth disparity and on and on. Ok. So that's why the, the, these networks kind of get tied up, at least in Jaron Lanier's book, who owns the future, get tied up into this middle class conversation. Ok. We're in an economy. We're creating these networks, they concentrate the wealth. We, we don't want to walk away from them. Right. Matt cash. There's all this value. I don't wanna, I have this fear of missing out. It's more like a public utility, but I can't in this economy necessarily generate much wealth for myself. Uh Only a few people seem to be able to do that. Ok? Um And so, and so that's the argument that's being made. It, well, that's, that's kind of the situation that's being painted uh specifically by Jaren LA in this book. And that's the connection to these, these networks uh to the disappearance of the middle class. So it seems like a not an ideal situation and, and Jaren calls these siren servers, siren servers or like the Greek mythology where you got the sirens who, you know, attempt the Mariners away from their boats and because they have this beautiful voice, then it leads to their destruction. So this idea that when these companies go create uh these services like the Facebook matters, the Twitter is, it's, it's almost like a siren. We all get attracted to it, right. We got Matt Cash law there, there's a ton of value, we get attracted to it. And now all of a sudden now we're part of this information economy that doesn't really benefit the majority of people. OK? And that's, that's kind of the argument that's being made there. OK. So we talked about the network effects that we can expect, right? We've got net CAF S law, the more nodes, the more people that enter into a network, the higher the value is gonna be right, that value grows as a square of the number of nodes that are in the network. And then we talked about this hub formation. You should expect that networks will concentrate a lot of the kind of links within that network, the connections to very few hubs and they grow really, really big and then most of the other nodes do not. And that's an expectation. We took a look at um you know, talking about some of the challenges that this sets up with respect to the value of a network, they can become so valuable that essentially we as people become dependent on them. It's more like a public utility, it can be very hard if not damaging to, to walk away from the network. And then we took a look at, you know, connecting essentially the disappearance of the middle class to these large tech company networks that get created. Um You know, we are adding data to these services that make them possible right, our data goes into the product features and yet we don't generate, you know, the the wealth that gets generated is by very few people, you know, the original kind of founders of the company, the ones who own the servers and we don't actually get compensated uh for the data input. And so you can kind of have this argument that, you know, the disappearance of the middle class might be connected to the fact that we're in this information economy. But the information economy is largely driven by these products that concentrate the wealth, right? So let's take a look at the underlying mechanism to, you know, that kind of the concentration of wealth or, or more specifically the hub formations that happen within these networks. So this is essentially a rich get richer mechanism, right? We see this every day in life, I mean, if you're an investor and you're just starting out and you have a couple of grand in your bank account you want to play with versus someone who has $100 million in their bank account, you know, who's gonna find it, who's probably going to be more successful in their investment? Well, probably the person with $100 million simply because they have more to play with, they can risk more, they can get into, you know, kind of these diversification mechanisms within their portfolio more effectively because they just have more to play with, right. So rich get richer, you know, if you uh are a new kid in a high school and you walk in and you don't know anybody. Uh You know, what, what, what is the likelihood that you're going to maybe uh attract a bunch of people to your, to, to make a connection with you so that you can grow a group of friends versus someone who already has, you know, who's already, let's say really popular and they have a lot of connections. You know, again, there's all kinds of examples of the rich get richer, right? People that already have a lot of social connections are going to be able to grow their, their network much more readily than somebody who doesn't have any connections. So in social networks, we see this, right, the the influencers, the people that already have a million connections are going to, you know, be able to more rapidly grow their connections and somebody who's just starting out, right? Uh You know, big companies are going to be able to um take market shares, not, not always, right. There are obviously exceptions to this rule, but often can, you know, use their size almost in the sense of, you know, being too big to fail, right? To continue to grow, to continue to expand, to buy up. Um you know, you know, start, you know, even if you have a startup, let's say no, but I'm, I'm a disruptive startup and I'm small, but I'm really different than the big company. Yeah, but you're probably gonna get bought out by the big company, right? And that's how this mechanism works. So the rich get richer and we see this all the time. Well, the technical term for that is preferential attachment and that's, that's essentially the underlying mechanism that's happening here is that, you know, the nodes that already have a lot of connections are going in in the network are going to be able to to more readily create a lot more connections than the new nodes that come in. OK. It's a preferential attachment. I had a tweet not too long ago. Well, it was a while ago but uh I had this video of it was a water fountain and the water was coming down. And if you know how you know water hits the water, it creates bubbles, right? And so I was recording the bubbles that formed as the water came out of this fountain and landed in the water. And I was showing that if you look at the way those bubbles form, they don't all, they're not all Equis existent from each other, but you don't have all these individual bubbles that are just kind of separated out, you have clumps of bubbles at form, you have these hubs, right? And so this is the physical analogy or literally the physical manifestation of this exact same rich get richer or more specifically preferential attachment mechanism at play, right? And that's because you know, you got some initial bubbles at form, they first make the connections and now they're bigger. So you literally have more surface area for which new incoming bubbles can attach to. And so you see the the the the actual physical manifestation of the rich get richer, you know, you're going to have the bigger clumps of bubbles, the bigger groups of of aggregated bubbles attract the new bubbles that come in as opposed to them. Just kind of the new bubbles randomly dispersing and being acquisition from each other, right? So you can go check that video out on Twitter on my account. It's the same thing, right? That's what you should think that that when you have a lot of pieces, a lot of components, a lot of agents coming together and they can interact in the case of bubbles, they just have these, you know, kind of interact, you know, intermolecular interactions between the water molecules, right? They can come together, they're going to do that in a clumpy fashion when you get enough of them coming together. That that's that, that is the pattern that you should expect in these networks. OK? Um And, and that's why I want to kind of lay that mechanism down is that, you know, if you start thinking about the social consequences of things like this, you know, a lot of people might come in and say, well, I want to change that. I want it to be more, you know, I want the bubbles to be Equis from each other. I want equal outcomes for people. I don't want there to be a concentration of wealth would, would essentially be what we're talking about here. In the case of, you know, we're tying it back to the middle class and people, you know, in an individual's ability to, to generate wealth on these very socially important networks. I don't think they should be so concentrated. I want more equal outcomes. You know, I'm gonna talk about some of the, the the problems with that. Not that that's all bad but that there are some challenges. Remember my episode when I was talking about, you know, free speech and the edge of chaos and all that, you know, you've got these networks that are, that are trying to do content moderation and of course, up to a point that makes a lot of sense. But if you overdo it and you intervene on an otherwise natural process, it's going to fragile the system and cause a lot of problems. So we need to respect that networks do this aggregation, these hub formation, the the rich get richer is a real mechanism in networks that should be expected to be there. There is something natural about that because preferential attachment will happen in these networks. So let's keep that in mind. That doesn't mean you can't do anything about it, which I'll talk about in a bit but respect the fact what whatever it is you want to intervene with whatever it is you want to try to kind of engineer, you know, social engineering gets into this problem, right? As soon as you overdo it and, and that line starts pretty early, as soon as you overdo it, you're gonna start to interfere with these otherwise natural mechanisms that occur. And that's when it's gonna fragile the system, that's when it's gonna cause problems and it can lead to catastrophe over the long run. We see this historically, we see this politically in all kinds of different situations, economically, socially, politically, right? We see this in organizations, you have to first and foremost respect the way nature works. And then given that mechanism, say, OK, so what can we do about it as opposed to just running in up against it and say, I'm not gonna, I'm not gonna pay attention to the way nature works. I I just literally going to intervene and do my own thing and that, that's just will always cause catastrophe because things have to be commensurate with the way uh you know, nature happens, the mechanisms are at play. So when we talk about these networks, when we talk about these big companies that set up the Twitters and the Facebook metas, right? You have to respect the fact that preferential attachment is a mechanism that is expected to happen. And if you don't let it, you know it, it can cause a lot of problems down the road. So that's the first mechanism, preferential attachment. Think of those water bubbles that form in the water fountain, they don't form evenly. You get aggregates, you get clumps of water bottles that you know, a few really, really, really big aggregations and then some of the others are dispersed. OK? The other one is the way the internet has kind of been set up to be one way. So the the, you know, the protocol, the html that is used to drive the internet hypertext markup language, right? You, you don't need to get into the specifics about how it all works. But we should understand and Jaren Laer talks about this in his book, who owns the future that this is, you know, essentially been set up to be one way. So if you think about, you know, doing a search on Google, you know, you don't really care where something necessarily came from, you just want the results. It's not like it gets tied back to the original person who necessarily created it. I mean, it's probably the author's name on there maybe, but, but things are getting mixed and matched on the internet, things are getting copied, things are getting replicated, you know, and even if they're referencing that, you know, they, you know, got secondary or third, you know, versions of the sources of the information at the end of the day, there's no attempt of HTML and the internet to specifically go back to the original creator necessarily, whether that information is available or not, the, the, you know, there's nothing that locks that in place, there's nothing that guarantees the information you're looking at is particularly original, right? And, and that's, you know, essentially on purpose. I mean, we want to be able to replicate mix and match. Now, of course, this causes all kinds of concerns, you know, in the music industry, right? So, you know, probably the original case of this was with Napster and Napster was basically allowing, you know, the, the the piracy of, of music, right? You could, you could replicate data. Uh you could, you could download in this peer to peer fashion, uh all the songs you could possibly dream of on the internet and of course you weren't paying for them. And so there was the underlying legalities and the challenges. And so part of that was, well, this is good, this is the internet, you know, get with the times and part of that is bad, you know, you've got the Metallica, you've got the bands that were, you know, very much against this and wanted to say, well, no, this is our music, understandably, right? We are the creators, we don't want people to just kind of replicate this out. So you've got this copying replication one way mechanism that happens or things get copies things, things get uh replicated and they go get turned into all these other things and there's no attempt of the internet to, to kind of guarantee that the original creator of that is going to be, uh you know, compensated or rewarded or at least recognized as the person who, who came out with the thing, the internet kind of doesn't really care. So it's one way, ok, that's the one way mechanism that, that the internet is set up as. Ok. Now, um, and Lanier kind of compares this to mortgages, which is kind of an interesting analogy. So he says if you take and, and mortgage, you know, banks are getting caught up. Well, everybody is getting caught up in the information economy is another example of the way this works is is if I'm a bank, I can sell you a mortgage and that mortgage. Uh you know, you, you, you know, you, you get a chunk of money, you go buy the home and now you have this big debt and you have to pay that off over, you know, 10, 20 whatever years. And so we know that situation, that's how mortgages work, but I as a bank can actually take that mortgage and repackage it and I can do some, you know, kind of fancy financial engineering to repackage that debt repackage that mortgage into this new financial service. I can basically securitize that mortgage, right? And I can sell that to investors, you know, and of course the 2007, 8 2008, 2007, 2008 crash was really centered around these subprime mortgages. Right. And so that's essentially what was happening is we would, you know, they'd sell these, uh, you know, uh, these mortgages to people that maybe shouldn't have been getting into them, arguably. And, uh, and those would get repackaged and kind of subsumed into these new financial instruments which would then get sold as mortgage backed securities to these investors. And they would kind of, you know, we know the story, they had all kind of wrong ratings attached to them. And, and the reason why that happens is because you're, you're, you're folding in something that was originally just this kind of pure idea of a mortgage, which made sense. That transaction makes sense. Give me, you know, borrow, lend me some money, I'll go buy a home, I'll pay you back that money over time and then you re, you know, you copied it, you kind of replicated it and you repackaged it into this financial instrument that gets sold to investors and it can get repackaged again and again and copied and replicated across the network. And so it becomes this entirely different beast that is not really tied back to the original, you know, kind of owner of that debt if you will, right? And yet the original owner can be exposed to new risk because of the way it got repackaged. Ok. So it's getting repackaged. It's getting sold. People are investing that can have an effect on the housing market if that housing market starts to crash. And the original people who took the debt, who took on the loan now they can't repay the home and on and on and on. You kind of know the whole story, right? But, but that's this idea, the danger that you can run into by replicating and copying things across a vast network. The the original person is not getting compensated for all the different ways that that data is getting repackaged and repurposed. And not only am I not getting compensated if it's me, but I'm also maybe getting exposed to all kinds of risk. And, and the important point to realize here is that the the the network, if you think of the bank as kind of the owner of this system, now, that bank is, is almost like an impartial mediator, right? So when they sell the mortgage, they are kind of just writing that as a well, I sold it, that's a sale done. We repackage it, we sell it to the investors. It's now just the investor's problem and the original person who kind of got into that debts problem, but it's not really the bank's problem anymore. They sold the mortgage, they repackaged it, they're out of it. These so-called siren servers as Jaren Laer calls them are, are, are acting in a lot of ways in that same fashion. There are these impartial brokers that facilitate the transaction. But once that transaction is done, they, they, they, they kind of just as long as they keep facilitating it, they, they don't really have skin in the game if you will, they're not really taking on much risk themselves. I mean, imagine ebay or Amazon, you know, how many people are reading the kind of disclaimers or the user agreements to, to the software that they use? right? I mean, they're so long they're technical, they got legal and not a lot of people aren't using are reading those, but that actually exposes the user to a decent amount of risk if something were to go wrong, you know, you've got a buyer, you've got a seller. If, if, if you don't get your product or it's damaged, the buyer and seller can kind of work out that out of themselves, work that out themselves. The, the Amazons or the ebays are only going to be so involved to try to guarantee anything beyond the transaction itself. So they're kind of this impartial broker to that transaction. OK? And, and this has a way of kind of amplifying risk not to the owner of the network, not to the siren server, not to the Amazons or the ebays or the Twitters or the Facebook slash Metas. Rather, it, it tends to amplify the risk just to the users of the network, whether that's the buyer or the seller or whatever the network is doing in the service. OK. The, the the this impartial broker role that these networks take on and this is kind of a messed up situation if you think about it because you have a concentrate, you have scalability and concentration of wealth happening in these networks. And you know, the general rule is if you are uh you know, generating more wealth, you must be taking on more risk, right? You have this relationship between risk and wealth that's essentially supposed to be there. But there's something about these networks that kind of bypasses that a bit. And this is when you kind of get into these problematic situations just as the bank can repackage a mortgage and replicate it and copy it at scale without really taking much risk on themselves directly and and kind of just passing that risk on to the investors and to the the the person that would have kind of originated it. So, so you've got that issue happening of the coping and replication and and and now going back to this one way mechanism, you know, at the heart of this is kind of the one way mechanism that exists in the, in, in these networks and in the way the internet has been set up through HTML, you can copy, you can replicate it. But the the the deep connection back to the to to the original place that these things are, the digital asset was created, you know, who created that, who owns it? You know, if they're taking on more risk, are they getting compensated for that additional risk? That's just not there? Ok. And so to Jaren Lanier's defense, I mean, in, in his argument, you know, the fact that the internet is set up as a one way fashion is kind of seems, seems a bit problematic. So you've got that one way mechanism to these networks and then you've got what I just talked about previously, which is the preferential attachment mechanism. So this is what it is right now, let's talk about some of the potential solutions. So, Jaren la one of those could potentially be changed. Well, they both could be changed. So, so let, let's, let's, I'll just leave it at that. Technically, we could try to make, you know, the internet, maybe two way and maybe the, you know, maybe we can get into some web, you know, web, 3.04 0.05 0.0 whatever, maybe we can create a different web, whether that's a different protocol or a different way of using existing protocols. So that it's not just one way, maybe we can kind of guarantee that the the original creator of the of some digital asset is always recognized and maybe even compensated for. OK. Um You can kind of think of, you know, some of the new stuff with Blockchain technologies. We have NFTS now, these non fungible tokens, right? If I create a piece of artwork. You know, it's all art for some reason right now, if I create a piece of artwork, some goofy little picture and I put it on the internet and I can say, hey, you know, you know, people can go ahead and copy and replicate and mix and match all you want, but it's all, it's got a protocol in place to always guarantee the original creator is known, it can always be tied back to me. And so in that kind of fashion, you know, that might be a way to, to get people to, to generate wealth for themselves. So Jared Lanier's argument is essentially this, he doesn't talk about NFTS or anything. But, but he says that users, the way we change the problem of, of the disparity of the wealth concentration of the disappearance of the middle class is to pay users for their data. And of course, this would just be nano payments, right? Nano payments as he calls it, it'd be very, very kind of micro size. Well, nanos size payments very, very small, you know, might be a fraction of a penny or a few pennies every time you interact with a network. But hey, you're being compensated and the more that you use the network, the more you would be compensated and so on the surface, you know, maybe this makes sense. I mean, it's my data that makes these networks possible, you know, it's, it's these, it's it's the product features are bringing together data from users. And if you take a look at any normally functioning account, uh normal functioning economy, if you take a look at the definitions of wealth and how they work and those mechanisms that play out in the economy, I mean, why should these networks bypass those? I mean, if I'm taking on the risk, it's my initial data that gets replicated and copied, just kind of like these financial instruments, these mortgages, it gets mixed and matched in all these ways that make these companies successful. Even if these little nano payments, whatever I should still be compensated for that data. And the more I use your network, maybe I should be compensated even more. You know, I kind of have this picture in my mind of uh a bunch of people on exercise bikes. Imagine there's an exercise bike in everyone's home, but maybe those exercise bikes can add electricity back to the electric, uh the electricity grid, right? The electrical grid. And so the more that I bike on the bike, the, the, the more I exercise, the more I'm giving back to the electrical grid. So maybe I could get paid for that. Right. So the people who use it more should get paid more, that kind of thing. So you can kind of think of the, you know, Jaren Lanier's idea of, you know, the more you use a network, the more you should get paid for your contribution. And if you do that, then what that starts to do is, is, is essentially siphon some of the wealth away from these concentrated, concentrated hubs in the networks back to the users. And now everybody has a chance to, to uh make money on these networks. I mean, let's step back again. Remember the information economy as it is now is largely driven by these networks by these so-called siren servers, right? The servers that tend to concentrate wealth through the network effects. We've got met CAF S law, we've got hub formation, right? The preferential attachment, we've got these mechanisms at play. They're very scalable. It's how the world works currently, right? Software has eaten the world. And so if we are going to participate in the world, it means we're jumping onto these networks. Well, if that's where the playground is, if that's the, that that's how the economy is functioning, doesn't it make sense for users to get compensated for their contribution to the network? Because it is your data that is going into these product features that is being sold to revenue agencies, uh sorry, marketing agencies. So that, that, that, that these companies can make their revenue off of advertisements and on and on and, and recommendation engines and whatever continues to increase the value of those networks. That's your data that's doing that. So shouldn't you be compensated for it on the surface? That's he that, that does make sense you know, there's not, it's not a blatantly obviously stupid idea. But if we start to pick that apart a little bit, it does become a little bit problematic because I, in one sense it's a little bit goofy. Right. Because, yeah, I'm using this network, you know, but it's not even really like a, so an exercise bike, I'm adding to the electrical grid. I'm just using it. I mean, and who's to say what that value is? And if I'm just running a search, I mean, is that really, I mean, yeah, it's valuable because you can, you can look at the statistical correlations of what I'm searching for and what, you know, what page I ended up on and I, I get that that's valuable to you to, to, to, to a company, you know, but what is that? I'm just using it, right? I'm just using it and, and it's kind of hard, you know, the exercise bike has a physicality to it. It kind of makes sense the information. Sometimes it's hard to get our mind around. Like, is that really valuable? You know, but it is. And so who cares what, you know, if it's value, it's value, I guess you should be compensated for it. But here's my real issue with this as a, as a, as a solution. Uh Well, there's two things here. One is you're, you're taking what is a, essentially a fat tailed or skewed distribution and you're trying to put it into a Gaussian a bell curve. So I'll explain that in a second. OK. This happens all the time, you know, drives me crazy. Anybody who really appreciates how complex systems works that should be driving you crazy when you try to, you, you know, complex systems don't work uh with bell curves, they just don't, it's not what you see in complex systems. OK? You see extremely skewed distributions. So I'll talk about that in a second. You try to force the bell curve that causes all kinds of problems. OK. The other is this notion of value and how you would possibly trace that back to an individual's value. So um let me start with the first one. Actually, I'm gonna talk about this, this idea of the skew distribution. So if you were to take a look at the, uh you know, the number of people in a network and you were to plot, let, let, let's just say you had for, for the sake of argument, 100 people in network, right? Obviously, we're talking millions, but let's just say 100. So you can kind of visualize this in your mind and, and think of it as a histogram or like a bar chart, you know, whatever, just just to keep it that, that basically says, you know, in a histogram, you would say, OK, for each individual, tell me how many links are attached to that person. OK. So you've got 100 people in network, let's say I'm number one. OK. So Sean mcclure is the first person and tell me how many links in that network are attached to that first person. OK. Now remember those bubbles that were forming in the fountain? Maybe if I got there in early or I'm the founder of the company or whatever, I've got a lot of links to me. So the bar on that, on that histogram just think, you know, just think of it kind of like a bar chart, right? That bar would be really high because there'd be a lot of links, right? So the Y axis is number of links and the X axis is just each individual in that network. So Sean mcclure, you know, hundreds and hundreds of links, whatever or I've only got 100 people, so whatever, maybe it's 20 link, 30 links, whatever. And then the next person beside me, which would be the bar next to me in this bar chart, right? Would be, you know, person number two. Now, maybe they've got, you know, 10 links and then the next person and next person. Well, if you, if you look at the shape of those bars, right? You would see this really dramatic slope, right? And that's the concentration of connections in the network, right? Me would, you know, Sea mcla has a ton of links. The next person has quite a few links and then it drops off dramatically and the overwhelming number of people in that network have one, maybe two links, right? Not very many links. And so this is this kind of fat tailed or skew distribution. It's got a very long tail to it. OK. It's very one sided, it's very asymmetric. And this is the, the what you would expect if you were to analyze the networks when we talk about the hub formation, when we would talk about preferential attachment, and you were to measure literally the links and, and, and turn it into this distribution, that's the distribution you would see that's the distribution you should expect. OK. So that distribution is reflective of this uh preferential attachment mechanism that I've been talking about, right? It's very skewed, it's very asymmetric, you've got a very few people with almost all the links and then almost all the people with very few links. And so that skewed distribution, that kind of fat tail distribution is kind of the picture of preferential attachment. It's, it's, it's, it's indicative, it's a, it's a hallmark of the preferential attachment mechanism at work. It's also a picture of the wealth disparity. You could argue, right? Uh If you think about the information economy and it's these networks that are driving it and you have a concentration of links kind of flowing to just a very few people that's like the flow of capital being very concentrated to very few. And then everyone else is not really getting getting much of that, right? Because there or any of that, because the, the, the, the we don't have a two-way mechanism at play where people are getting compensated for the data. It's all one way, it's all kind of flowing to a very few people, right? Lanier has the Instagram example. In his book, you got 11 or however many founders, very few. And then, you know, Instagram gets, you know, bought, you know, by Facebook at the time for a billion, you know, and all that money kind of went to just the founders and, and, and yet you've got this product that was created by millions and millions of people really. And they never got any of that money, they never got compensated for that. So this skew distribution think, you know, picture that in your mind as the, as the wealth disparity as the disappearance of the middle class, right? Because if you had a middle class, then a lot of that wealth would not just be concentrated to a very few people, it wouldn't be that same skew distribution. You would have uh you know, something that was kind of bloated in the middle, more like a bell curve. Now, I now I want to be clear here, we've got two things at play here. The, the, the distribution that we're painting, that was the number of links. OK? That's the topology, right? Topology is just how things are connected. It's kind of like the shape of the network, right? So the topology is being reflected as the skew distribution. And we can think of wealth as being the same thing as the topology, right? Because in these networks, if you have the majority of links being concentrated to just a few hubs, just a few people, and those are the people that are being compensated because they own the company, right? They're the ones that are benefiting lucratively monetarily from the network effects of complex networks. Then that that that network apology is kind of the same thing as the wealth. So when we think so, so when we look at that distribution, you can also think of that as the wealth disparity, the concentration of wealth, you know, the one percenters uh or the lack of middle class, right? Network apology is essentially the same as wealth. Well, when Jaren Laer comes in and says people should be compensated for their data as do others as kind of the common solution, then what you're doing is you're actually separating the notion of wealth from the topology. In other words, you're saying, OK, the network shape is still the way it is. You still have a concentration of links to very few people. You expect to see the skewed distribution. But wealth should not be equated to that wealth should be uh not concentrated in the same fashion because we should have a two-way mechanism, right? E HTML protocol, internet should allow things to happen in both ways with which again to my previous point, if, if you're copying and you're replicating something, you should always know where that came from. So if I'm on a network, don't just take my data and mix and match and copy it and replicate it and turn it into new product features. So that other, you know, so that you can continue to generate wealth from it, you should always have a memory, you should always have a memory that it came from me. And because of that causal chain going all the way back to me, you should be able to pay me even if it's just these little nano payments. So it separates the notion of wealth from topology. You still have link wise, a concentration of links going to the people who own the company, but these companies are taking it upon themselves to compensate their users. And so if you imagine you had another line on top of that chart, you, so you've got the, you know, this, this fat tailed skew distribution, you, you're, you're seeing the the asymmetry, right, the long tail, right? But it's very concentrated on one side. And then on top of that, I do wealth which instead of looking like the same thing because now we're separating wealth from topology, it's going to be more like a bell curve. In other words, the over, you know, like like any bell curve means the overwhelming majority of people are kind of concentrated in the middle and, and have a high value and then it's only those outliers that have the smaller values, right? And so, uh and so in this case, you know, we're, we're kind of plotting different things now. So I probably shouldn't say put the line on top, but just think of a better way to think about this is actually let's have two different charts. OK. So you, you've got the original one that I said, which is the skew distribution. And now let's make another chart that reflects wealth because we're separating wealth from topology. And on wealth, we, you know, we can plot um you know, the, the uh the uh amount of money on the X axis, let's say, so wealth is on the X axis. And so if we increase wealth, it's going to go really high in the middle of the number of people. So the, so the y axis is the number of people, the X axis is wealth. If you increase along the X axis, it's going to bloat to, to the highest number of people kind of in the middle. And then the outliers are going to be, you know, uh at, at the very left and very right side of things. So on the very left side, you've got low wealth, but there's, there's not that many people with really low wealth. And on the very right side of the X axis is the high wealth, but there's not that many people with super high wealth and the overwhelming majority of people, right? Take on that bell curve shape are kind of in the that, that big bloat in the middle, that's the middle class, right? Most people are making decent money, right? That's what that says. OK, so, so we've got two charts, we got the one on the left that histogram that shows the number of links for each people, ee each person in the, in the network, each know in the network. And very few people have the overwhelming number of links. And if, if wealth is the same as topology, then that's also the kind of the same as wealth, you see the concentration of wealth. But if we create a second chart and we separate the wealth from the, from the, from the notion of topology and instead just plot the wealth as a separate thing because now we've got this two way mechanism happening, then you've got more of a bell curve. OK. So, so, so the take home message here, hopefully you can kind of visualize that is in Jaron Lanier's type solution where we create a two way mechanism where people are now getting paid. What you have here is this notion that you're separating wealth from the topology and the we should take on more of a bell curve. Well, my issue here is that, you know, going back to the fact that complex networks need to be you, you have to respect that they are complex. They will have the properties. This asymmetry exists for a reason. And by just separating the notion of wealth and saying, well, now we've got this kind of nice bell curve situation and that's what we should do. We should try to bell curvy things. We should, you know, it's a Gaussian curve. We should try to go things for lack of a better term. We should try to force the bell curve into the situation so that we have this nice big middle class and people are making money that looks kind of nice on the surface. But you should always be very suspicious whenever someone comes into a complex situation and tries to make it a bell curve, you should always, you know, and, and it's, I'm not saying it's 100% guaranteed that it's going to be a problem, but you should be very, very suspicious. It's a red flag. It really is. And that's because complex situations do not follow these nice symmetric bell curves. They don't, they're asymmetric nature operates on asymmetries. You see this all the time, you see multiplicative logarithmic growth of systems, you see them reflected as these more like prototype distributions, these very one sided asymmetric fat tail distributions, those those are reflective of the kind of behavior you expect in complex systems to have. And of course, nature has those behaviors for a reason. So when somebody comes in and says, hey, I have a solution to kind of get, get, get a better equal outcome. In this case, socially, this kind of social engineering you're gonna do. And it's gonna make this nice bell curve in a situation that you know, nature gave us this thing. That's not a very nice bell curve. Let's make it, let's use human ingenuity to kind of belly this situation force the bell curve into an otherwise asymmetric distribution that's going to be problematic, that's gonna be problematic on the surface. You should know that because you're doing something that nature doesn't do. And some people don't like that argument that it wasn't that kind of like a naturalistic fallacy. I mean, you know, just, just, just because something's natural. But what I'm saying here is that you have to understand that things are not natural for random reasons. Things don't survive, you know, I tweeted this yesterday, right? Things don't survive for random reasons, right? You, you don't need to know why something survived. You just need to know that they survive for non random reasons, right? There's, there's, there's a reason that nature quote unquote chooses. I don't mean that in an anthropomorphizing sense. But the, the reason why nature quote unquote chooses to, to have these fat tail distributions, they work. They're what work hub formation, preferential attachment is a powerful mechanism that allows systems to survive and adapt and work and, and, and we don't necessarily need to know what those reasons are. You can try to peel the layers back and say, oh, well, you know, maybe there's hidden dependencies and maybe there's a you can do that, but those dependencies are ultimately hidden. You don't really know. OK. So you have to respect that. So it's definitely a red flag when somebody comes in and says I'm going to make a ghost thing and, and I would argue, you know, it makes sense like look you network topology being separated from the notion of wealth is probably not a great idea, you know, because that is the network, that is how people are connected and the transactions are expected to happen on that network. And so really, there should be this connection between wealth and the topology of the network if you kind of artificially separate it and then you kind of belly or add this Gaussian distribution because it looks nicer and it's a more equal outcome and it's more symmetric. It's easy for our minds to interpret. Shouldn't that be better? Maybe in the short time it would be? But, but you should expect that to essentially fragile the system, it should lead to some pretty bad outcomes. I would argue because it's not natural and, and natural systems have their behaviors for a reason. OK. You know, II I, you know, I've written extensively and I've argued a lot against IQ studies, you know, IQ studies takes, you know, what is literally arguably the most complex phenomenon that we know of the human brain, right? Or you know, the human brain or, or you know, the the way the mind works, intelligence, right? Intelligence emerges at this extremely most complex network that we know of. And then people go in and say, you know, I think I can model intelligence with the GST distribution. You know, I can come in here with linear aggressions and bell curves and very mathematically convenient, easy to work with uh you know, analytical techniques, statistical techniques. And now I'm gonna kind of ray that into intelligence. And yeah, it was you can go, you just look up Sean mcclure and IQ and you're probably gonna get this article. Uh you know, I just, this kind of thing is just problematic because we know, you know, I talk about properties versus reasons, right, properties over reasons. You don't have to, you the causal opacity of something means you're often not going to be able to uncover the reasons. But as a property we can see at the top level that complex systems have certain properties about them. We can look at met CAF S law, we can look at hub formation, we can look at this rich, get richer or this more specifically preferential attachment mechanism that happens in complex networks. And we know that that happens so to come in and say, oh well, it doesn't have to happen where we can separate wealth from topology and we can go put a bell curve in here. That's red flag. It doesn't mean you can't still debate it and there maybe there's still some truth to it, but that is definitely going to be problematic. And I think that a lot of people that offer these solutions, it's just they have a lack of understanding of how complex systems actually function or they or they're trying to uncover reasons more than respect the properties. And you should be really looking at the properties of things because reasons are largely myth, right, causal opacity, we don't get to know the reasons for how things add up in complex systems. That's not an opinion. We know that you just don't, you don't have access to that. So you're pretending to have access to information that you don't have. So anyway, when people try to belly or go signify the situations, try to force the bell curve on things that is arguably complex, which these systems are, these networks are, it's going to be problematic. So that's my first issue just at the top level of Jaron Lanier's solution of trying to come in and saying, let's let's separate the notion of wealth from the topology. Um The, the and, and I'll talk about kind of what my solution is in a second. But the other issue I have is when you create this two way mechanism and say we're going to pay people for their data, you know. Well, it's a little bit goofy. Right. Like, what exact, how are you going to tie it back to value? Like, should I get value just because I'm on your network and I'm using it more than someone else or, or does everybody get kind of paid the same thing? What is the value? What am I doing? I mean, I get it, my data is getting turned into product features that get sold back to many people and you generate millions of dollars off it, but you're not getting millions of dollars off of my data, right? I mean, how much money are you making off of my data? Well, I can tell you right now in line with respecting about in respecting how complex systems works that there's no way to make that causal connection. There is not, there is not, you cannot tell me what my value actually is. So the best it can be is this kind of ephemeral goofy concept of, well, I'm on my network somehow my data gets turned into value. So I'll just add, you know, kind of a nano payment every time you're on it. It's a little bit goofy because you you can't. And so if you take it, so a lot of these product features get created through machine learning technologies. And the way machine learning works is that, you know, it finds these statistical correlations but really what it's doing is it's amalgamating or entangling a bunch of different data and noticing patterns that emerge from that emerge from that. Right. There's that causal opacity. There's a reason that we call A I black box. Today's a I quote unquote A I is driven primarily off machine learning technology and it's a black box and it's a black box. Fundamentally, it's not a black box because, oh, it's kind of tricky to figure out what's going on. No, you don't get to know what's going on period. Full stop. You don't get to know if you knew what was going on, it would no longer be machine learning, it would know by definition, you wouldn't be doing intelligence of any form, even even the narrow form. I'm not saying it's true intelligence of human intelligence, but even the narrow form of A I that we do with machine learning by definition. As soon as you unblock block that unblock box that you're not doing machine learning anymore. If you are, then you're not, you know, it's kind of like in, in quantum mechanics and we're like, well, if it makes sense, like fully makes sense, like it's, it's it becomes completely intuitive to you, then then you're doing the wrong thing. It's not quantum mechanics anymore, right? Because quantum mechanics at the most fundamental level is counterintuitive. It's not like there's something beneath the counter intuition that you could uncover and then all of a sudden it would make sense like no, you don't understand, right? Like Heisenberg, uncertainty principle. You don't get to know the position and momentum of something at the same time. Full stop. There's nothing else to say about it. That's just it, it's a fundamental property of the universe, right? Machine learning is by definition black box. If it wasn't, it wouldn't be complex if it wasn't complex. And it's not actually what machine learning is supposed to be doing. So, so it's not an opinion. It's not like, oh someday we'll figure it out like, no, you're not getting it a fundamental property of the complex systems is the causal opacity. You do not full stop get access to that information. OK? So when your data gets added to a network, there is not this nice causal chain between your edition and then I can see where your addition specifically played a role in the product feature. In other words, I can't tie it to the amount of money it was made. That's absolutely impossible. And anybody that says you can do that is, is, is either lying or deluding themselves or just really what it is is they're not understanding comple complexity, they're not understanding how it works. And today's machine learning does is right on the edge of genuine complexity. It is, you know, there's papers about this, you can go and it's why it's a black box, right? So, so you don't get, you're never gonna have access to that information. There is no causal chain, there is no path you know, I've talked about multiple reliability in all these properties and complex systems right before in, in really all my episodes, right. So, so this idea that you're going to be able to take an individual's data and say, oh, I can see which role it played and I can connect it to the money that we made, you know, this last quarter. And therefore here's your nano payment or here's your set of payments and now you're, you know, correctly, um, you know, correctly compensated and rest assured if you know, someone could say, well, yeah, I know we're never gonna be able to know, but we still know that your data did play a role, right? Any kind of, you know, model interpretation of machine learning would do this, they would say, yeah, we don't know what the role is, but we do know it played a role, right? A ifs gets into this, right. You know, I'm not gonna use zip codes because zip codes could be tied to certain, maybe minorities in the population and you're going to mystery even if I don't really know how those zip codes exactly get used. I know that. Incorporate Ya ya ya. So you can know that it plays a role without knowing what role it plays. So you could argue that well, just knowing that your data plays a role, I can still compensate, you compensate you. But rest assured people are going to get into these debates and these arguments are eventually going to say well, yeah, but isn't my day, I think my data is more valuable because I'm not just like talking on the social network. I'm actually, you know, posting relevant articles or I'm getting into these, these uh you know, really interesting debates and, and then one time one of those debates connected this investor and then this happened and this happened, you know, people are gonna start to, to say, well, you know, isn't mine better, shouldn't I be compensated? It's not gonna work. I, I don't think it's gonna work. Maybe I'm wrong. But at the end of the day, what you're trying to do is you're trying to add a bell curve to something where a bell curve does not belong. And that just always smacks of problems when it comes to complex systems. OK? So I can tell you right now, anybody with a cursory understanding of how complexity works knows that you can't trace back the value to an individual like that. It's not gonna happen. And two even more problematic is when you start saying I'm gonna create a bell curve of wealth and that's gonna be the, so that's very problematic. OK? And I'm not saying it's, it should be completely dismissed out of hand because there's aspects to it that have kind of the right idea, but I think it's the wrong way to go about it. So how would I go about this? To kind of wrap this episode up. Well, I don't think you should separate wealth from topology. I think you should say this is the shape of networks, this is how they form, right? Complex networks happen and behave this way for a reason, even if you don't know the reason, right? You don't have to know the reason. You just have to know that these things don't happen for random reasons. It's really important that people understand that right, properties versus reasons. This is a property of complex systems. So instead of separating the wealth from the topology, respect that that's the topology, but maybe have something in place as a type of minimal regulation that says, let's not let the monopolies get so big. Let's not let the Googles and the Twitters become these monsters where we have this massive social dependency almost like they're a public utility. And if something were to go down or to go wrong, or if I were to leave that network that I would be devastated, that should not be the case. People should be able to um you know, create their own enterprises that can compete with these other enterprises. They shouldn't just always get mopped up, always get bought out necessarily, although the buyouts is still not necessarily a totally bad mechanism. Because if you create something as a startup and it grows to a decent size and then you get bought up and that's your way of making wealth but there is still AAA very, uh, you know, a huge wealth disparity with the current topology. So if you put some level of regulation in place and you could, you could say, well, aren't people doing that? Well, not really. Right. I mean, antitrust is kind of not much of a thing anymore, right? I mean, the, these, you can say, well, under the current, you know, the real definition of monopoly, I mean, Google is not actually a monopoly. I mean, look whatever you wanna call it, right? These are, there's only four or five companies that are super, super massive that are just buying everything else up. And, and so we have an extreme extreme kind of fat tail distribution, a very skew distribution. And you can argue that from a topological standpoint, it might make sense to de fatten the curve a little bit. And so and so to Jaren Lanier's defense, you know, you've got the kind of the right idea where maybe you want a bit more of a belly to it, a bell curve to it, but not to the whole, let's separate wealth from topology and just create the wealth curve by paying people. I'm saying don't compensate people for the data because the whole notion of compensating people, the data is problematic instead if you were to make a change slightly to the topology so that companies could not get so big. So it wasn't as skewed. So you could de fatten or de skew, if you will uh add a little more symmetry to that topology, topology would still equal wealth. Right? Tho those would still be tied together. I'm not separating it like Jaren La, I'm saying stick to asymmetric topo and yes, that's wealth. So don't pay people for their data. But if you don't allow certain companies to get as large as they are via regulation, you know, put a cap on it. Every company can go to 5 billion. No high. I mean, I don't know. Right. I mean, Apple's worth a trillion now. Right. Or at least it was last time I checked a trillion dollars that you got Elon Musk with 300 billion. Those are extreme extreme concentrations of wealth. And I don't think the answer is just, is, is just to say, well, look like that's nature, that's free markets and, and it's supposed to be asymmetric. So just let it go. Right. That's not what it is. Just like free speech. I think free speech has to be respected. I think people have to be able to say what they say, but that doesn't mean every single thing goes like blatantly obvious things. And this idea that we can never draw the line anywhere is a little bit ridiculous. I think there are ways to draw the lines. So you can come in with some type of regulations and put a cap. You, you can get really big, you can be really successful, but you don't need to own the world, right? You don't need to own the world. Remember, we were got valuations and billions and billions of dollars for barely doing anything, right? You when, when it was just a complete free reign on the growth and the scale of these networks and these companies, you're going to run in the problems. So, so you don't just let the asymmetry go to an extreme nature, doesn't do that either, right? There are stressors in the environment that put caps on how rapid and big things can grow. So we should also be respecting that aspect of these asymmetric distributions of these growth processes, these asymmetric preferential attachment growth processes. OK. Now, I don't know exactly know what that is. Maybe it's a government that comes in and, and just like I said before or maybe a company can get to, you know, so many billion and then, you know, you got to stop whatever that means. However, that works, right? You have to allow, you know, other competitors to come in against the incumbent, have more competition, more people can start companies, more people can create value, those will create jobs for people. And then you get some, you know, some some some more kind of uh market competition as opposed to it all being completely concentrated, right? Uh it's not an exact solution because you gotta kind of get into the details of how does this work? How do you know what, who's doing the regulation, you know, what is the cap, but it's the same thing with free speech. So you're respecting the fact that these are how complex systems work, but you're still putting stresses and regulation regulations in place up to a point that helps uh a more healthier asymmetric topology to exist, right? You could argue that ours is so extreme now that it's not a healthy one and it's actually kind of maybe unnatural but with a little bit of proper environmental stresses in place to use the analogy. But but but more specifically regulations in place on, on that, that cap how big a hub can get, you can still respect the asymmetry, you can still respect the preferential growth attachment or preferential attachment mechanisms to the growth process, you know, but you're allowing a slight de fattening of the curve, you you're adding a little bit of bulge if you will to to that topology more towards the middle. So it's not so skewed and so that other people can participate in a number of ways. So anyway, I think that makes sense. So in, in, in, in my approach, don't separate wealth from topology. Keep it there, don't pay people for the data. That's, that's kind of a ill defined process anyways. And, and ultimately, I would argue impossible um allow preferential growth attachment to happen, allow wealth to be associated with a number of links, but kind of de fatten that curve de skew that curve a little bit so that you get a slight more bloating and you can do that. And so you can imagine if you're running a simulation for those that are kind of mathematically inclined and computationally scientifically inclined. And you've got these nodes coming together, you could change maybe the local interaction rules on the individual nodes or more top level. You can just put a constraint right on the optimization. Maybe one of those constraints is, you know, a node can only go up to a point and then not do that. I'm gonna be putting a simulation like this in place on the non-trivial playground that I've talked about before, I'm gonna, it's gonna come a little bit later, but go, you know, pay attention to that, you'll see the simulation run, you can put the constraint in place and then you'll see what that does to the topology, that distribution and it's gonna kind of de skew the distribution a little bit and that might be a way to potentially um uh get us to a situation that's a little more healthy. So I think that respects complex systems and I think that is a better way to do it. So that's all I've got for today. Uh Hopefully, that was a decent episode. Hopefully that all made sense. If you have any questions, let me know uh if you'd like to support nontrivial head on over to Patreon dot com slash nontrivial to listen to the bonus hour, which is what I'm gonna do now. So I'll do a little extra content on this. I'll, I'll talk about this a little more in depth if you want to hear that. So dot com slash nontrivial, a few dollars a month, you can support, uh you know, so, so support the episode and get a bonus hour. If you like what you hear, please consider giving nontrivial a five star rating on Apple podcasts. If you'd like to read the article versions of Nontrivial, subscribe for free at shaw dot substack dot com. You're listening to Nontrivial the podcast that uncovers the patterns that help you understand and navigate our complex world. Until next time, take care.