Stratification of Society as a Function of Average Rate of Information Exchange
I must admit it is with some trepidation that I start this project. This is for many reasons, but as with most things in life, we don’t usually get to pick the time, we only have some say in the place, and the only real choice we have is our actions, and even that is debatable at times. I have been writing this article in my head for several years now, and at some point, I need to commit words to paper, or in this case, bits to databases, about a project I have been working on for some time.
In the many iterations in my head I have bounced back and forth between trying to outline what it is I have been doing, or whether I should start with what I have not been doing. I have decided on the latter, as I worry the former would carry implications that I do not wish to be included.
What I have NOT been doing
I want to be clear that what I have not been doing is working on some preplanned process. As much as my ego would like to claim as such, anyone who has ever known me for more than 5 minutes knows I am not that organized. While it is true to say I have been working on this project since at least 2012, and in many ways studying it’s aspects since 2008, I do not want to give a false impression that 8 or 12 years ago I set out on some journey with a roadmap. I did not. For many reasons certain elements in public discourse caught my attention while attending college on campus in 2008, and those elements have simply gained more and more of my attention as the years have gone by. It’s much more appropriate to say that something shiny caught my attention, and while at times I might be able to turn away, over time, that something shiny turned into an obsession, and eventually became something like a mission, finally evolving into the ideas and plans I want to start laying out in this article.
What I have been doing
What I have been doing is a much more difficult thing to state, as I outlined above, it’s not like I started with any plan, rather than just following impulses and curiosity, but I can sum up what I’ve been doing is trying to figure out how to communicate ‘across the aisle’ so to say, though I would grant in total honesty, that for the first several years, it’s more correct to say that I my goal was to annoy and shame the other side of the aisle in what I viewed as the only proper reaction to the ideas that I encountered, most amazingly, in people that I have known for large portions of my life. I am not proud of my behavior on Facebook in the early years, my only defense being my own ignorance and self-resentment.
But one of the things I decided in my mental rewrites is I do not what this to turn into a narcissistic rehash of my own sins on social media. I have a goal with this writing, and that goal is to outline what I have learned from the past 10 years of social media addiction, describe my own thoughts and opinions on how it works, how it’s evolved, and more importantly, how it’s changed me, as well as how I think it will eventually change all of us. For the more scientifically inclined reading this article, I am even in the processing of learning the data analysis techniques I need to apply machine learning algorithms to my facebook posts for the past decade to better determine and quantify these changes and insights.
So what have I been doing?
I’ve been trying to figure out what has been going on, engaging incessantly with those who disagree with me, for over a decade, and I have it all online, catalogued, and downloadable on social media. In addition to that, I have used Facebook as well as Twitter for pushing my streams of consciousness into the global cloud for retrieval and analysis at a later date. If there’s anything I’ve learned in my life, it’s that the more that I learn, the more ignorant I become, so I won’t claim to to be the first person to apply machine learning algorithms to their own thoughts and streams of consciousness over a 10 year period, but given the age of social media, I feel confident in thinking I will be one of the first.
In the beginning
All good stories have an origin story, and this one is no different. I hope to carve out the time to dig into my own personal thoughts and opinions on the details of social media in other articles, but that is not my purpose here. For now, let’s just say that social media is something that has interested me since I was able to meet and talk to girls in America while I was stationed and participated in some of the worst combat in Iraq. I mean, think about that for a moment. That’s an entirely new thing in human history, one that wasn’t even imaginable as recently as the prior generation. Do you have any idea how many girls will message you back when your profile shows you’re stationsed in Iraq in 2004?
All of them. The answer is all of them.
But I digress, in the beginning, was a young man who loved to engage with people and learn how they think, infatuated by the different opinions that everyone had, the different perspectives, the varying conclusions they carried and how they effected their thinking and actions. I love to learn how people think, and social media, and the internet in general, has been a buffet of such interactions.
Of course, in the beginning I was really horrible at this. Meeting someone at a bar who lives in the same geographical area that you are in is one thing. But being online not only removes the filtering aspect of geographic location, but also, as I would not learn until years after it ruined many relationships, it introduces the problem of inter-human communication through textual medium. Mainly that approximately 93% of human communication is non-verbal, and textual interaction loses even more information in the exchange due to the lack of intonation, tone, cadence, and the other many aspects with which we humans use to convey meaning and intent with our spoken words. Modern society has found itself embroiled in a problem where communication between any two individuals in the nation has become almost instantaneous, but at the cost of about 95% (at best) loss of information in those exchanges.
ARIE
This brings me to a theory I’ve developed and bookmarked in my Facebook feed a few years back, ARIE, or, Average Rate of Information Exchange. Going into detail into this theory here is outside of the scope of intent of this article, so I only mention it here as a placeholder for future exploration. In short, ARIE is an abstract measure of the rate at which information is exchanged in human networks, or in normal people speak, society/culture. The theory goes something like this:
Culture is the exchange of information between individual nodes in a cluster-group. The rate at which culture changes or shifts is positively correlated to the speed at which information is exchanged between individuals in that cluster or culture. The slower the rate of exchange of information, the slower the rate at which culture/society changes, the greater the rate, the faster the change. (This has some interesting implications if you also consider pendulum theory)
Circumstantial evidence for this rate of shift change exists in the fact that for hundred or even thousands of years, culture changed and shifted at a snail’s pace. Yes, it is true that pockets would shift here and there outside of the mean rate, but was still held within constraints assigned by the larger socio-cultural limitations. However, as new technology increases this rate of exchange of information, first through the use of language, then written language, eventually leading to the invention of the printing press, which can be directly tied to the enlightenment era, to the telephone opening up the progressive era a hundred years ago, to the internet opening up the digital age in the 21st century, resulting in the current crisis we are all engaged in today given that human interaction, and thus culture, is no longer beholden to geographical limitations, and in the considerations of pendulum theory, the ARIE has the effect of increasing the rate of change of the pendulum shift, which is how we got from Obama to Trump so drastically.
As technology decreases the time it takes for individual consciousnesses to exchange ideas, the rate at which society evolves has increased, which further increases the rate at which individuals within a particular society are able to exchange said ideas. It’s a positive feedback loop, and one where the loss of information rate is something like 95 or 97%.
Back on track
As I said I want to be clear that I did not approach this project a decade ago with some plan, but, in the midst of confusion and curiosity in trying to understand how some people could have political opinions that, to myself, seemed entirely ridiculous and not in accordance with the rules of reality as I knew them, I came to wonder about a question —
I was quick to realize that patterns existed in the political arguments and opinions I came across online. Eventually I hope to better catalogue and write about these patterns as I learn natural language processing with machine learning, but for now, let’s just say my exchanges with different people who were unrelated in every way other than their particular political persuasions would consistently give me similar arguments for the position of the day.
It was these patterns in language that I became interested in. It became obvious that particular patterns of language could induce a particular response in those on the other side of the debate. This is quite intuitive if one thinks about it. after all human interaction can be boiled down to patterns, and we either work with those patterns if we like the other person, or work against those patterns if we find their results distasteful. (I would argue that reality can also be reduced down to mere patterns, but that would also be outside the intended scope of this article, and hope to explore this idea in depth at a later date.)
But it was the fact that different patterns had different reactions that interested me. That was what I, as I now see it in hindsight, became obsessed with understanding, and over time I began to wonder, if one particular pattern of language is used by one side in the larger culture, and a different pattern of language was used by the other, then is there a master pattern of language that can speak to both sides at the same time?
The ‘Master’ Pattern of language
Or as I like to call it: Truth + manners — ego. This formula, once you can incorporate it to your online interactions, will get you anywhere and everywhere online. The average person online reacts with pure emotion and usually vitriol, and if you are any person with any sort of public influence, you essentially receive a constant influx of hate and death threats. Attempting to engage with someone using the above formula is a pattern of language that stands out amongst the noise, and will typically result in a response. At the very least, my own experiments have shown that such a pattern of behavior online will typically result in the algorithm grouping you in a cluster of people who also behave this way, and the shift in online experience from negative to positive is something to behold.
We have a force working against us when it comes to social media. The way algorithms work, they attempt to drive like minded people into clusters so as to better facilitate advertising. This makes sense for the most part, except that the results do not meet the expectations. This seems counter-intuitive because we would think that driving like-minded people together would create nothing but echo-chambers, and to be fair it has created plenty of those, where the average person rarely encounters someone who thinks differently than they do or holds differing opinions, but, while the working algorithms are great at predicting (And effecting) how we think, the end goal is to keep you online more, which means getting you emotionally engaged, which means giving you more of what you interact with online, not necessarily what you consciously want to interact with online.
This brings me to my second theory I hope to expand upon and explore in future writings as well:
Local Social Minimum
There is a concept in Machine Learning called Local minimum, where as you are training the machine, you are adjusting algorithms in n-dimensional space so as to find a plane of best-fit to predict future outcomes. Don’t worry if that doesn’t make any sense, the best way to think of it is like the image below:
As a machine is being trained, it is attempting to find the best algorithmic formula to predict results on the graph above. It does this by trying to find the global minimum and maximum values to try and ‘fit’ itself too (Among other linear algebra things that we don’t care about here). When doing so, it is possible for the machine to get stuck in what is called a local minimum, so that the algorithm stops changing because it thinks it has found the bottom, because it does not see the global minimum due to it’s inability to see outside of it’s own hole. This causes predictions to be wildly off, as it’s limited perception prevents it from seeing the larger picture, leaving it incapable of recognizing that it’s environment is not as it believes it to be, only that it’s predictions do not produce the expected results.
I’ve found it to be true that we can apply this same conception to society at large. (If you really wanted to, you can apply it to human experience as well) As something like social media is interfaced with an existing society, the ARIE goes up within that society exponentially. What would the result from such a drastic change in the exchange rate of information cause? Keeping in line with our machine learning example above, this would be akin to something like exponentially increasing the processing power, so that training iterations, or the rate of change in the algorithm, goes up, keeping in line with the ARIE theory.
If we consider that the closer to 0 we get, the more true something is, then it’s likely that as social media, or instantaneous interaction between millions of consciousnesses is applied, stratification is expected across that society as local minimums manifest in the algorithms that drive that human interaction, herding individuals into those riffs between valleys of falsehoods that have some bit of truth, but might not contain the complete story.
This is a problem. I would even wager to say that this is THE problem of the 21st century. AI has become a ubiquitous term in today’s society. Everyone knows about it, most people even know that it basically drives everything we do, even though it’s true most people don’t realize to what extent that is the case. Such a shift in human interaction at a global scale is bound to have consequences that will ripple out through the centuries in front of us. What most people don’t know, is that we don’t actually know how it works, and every theory we have about reality says that it shouldn’t. The video below does a fantastic job of giving a quick and humorous overview of why this is.
My generation grew up in a fairly computerless world. I remember them as the amazing things that we got to die from dysentery every Tuesday on when in grade school, and that annoying typing class we had to take in middle school. We didn’t have a computer in the home until I was about 10 yrs old, and we didn’t have the internet until I was probably 16.
I remember having this friend when I was younger that I thought was rich because they had a device that was specifically designed to rewind VHS tapes as fast as humanely possible without breaking them. This was incredible to me. We didn’t have to wait 15 minutes to start a movie if it needed rewound, we just had to wait 2!
Meanwhile my daughter has instantaneous access to every movie ever made at the touch of a button, often free, and usually only a couple bucks if it’s not, all available on a tablet that she can take with her anywhere.
And we’re expected to prepare our kids for the future when we can’t even figure out the present?
A Path Forward
I could write a whole article about solutions to the current problems we all face. I have many opinions and theories, and while I’m sure they’re mostly wrong, I think I tend to do a pretty good job of at least angling in the right direction. But I’ve already written a whole article, and this thing is long enough as it is.
I can say this about a path forward, though. It lies through changing how we engage online. Once you understand that society runs on machine learning, which is nothing but billions of networked input/output nodes constantly interacting with one another, and you realize that the internet is nothing but a large network of billions of people, constantly interacting with one another, you can begin to get an idea of the larger game that is being played.
We as individuals have little power to effect how that network is used, but we do have power over the input we receive from the network, and the output we feed back into it. We are each neurons in a neural network, some of us more connected than others, some with few online connections at all, each having at least some small effect. (Even a hypothetical individual with no online connections will have some effect to the overall network just through their routine interactions with others who are connected.)
The power of a nuclear weapon pales in comparison to the power that social media gives to the individual. The path forward requires each of us as individuals to recognize and understand that power, and to work towards being cognizant of how we utilize that power, given it’s outsized effects on the real world, and our own limitations in understanding the larger perspective as just a mere neuron only able to learn from those other neurons we can see in our local cluster.
We are not mature enough to responsibly wield the power that social media gives us. We have connected 3.2 billion human subconsciousnesses together in a network and the world is currently fighting over the admin rights. Because of this, any top down solution is inherently flawed, given that the inherent power up for grabs with any such solution is such that the darker side of humanity is willing to kill large sums of people for.
This means only bottom up solutions can work, which means each of us, individually, need to pay attention to what patterns the algorithm is feeding to us, and paying attention to what patterns we are feeding it in return.
