Thursday, February 2, 2012

Social Media, Clouds, Stochastic Processes and Why I Love Them

This blog post will probably certify my status as a nerd and no, I have never watch a full Star Trek episode (did grow up devouring Star Gate SG-1 episodes though...). The reason why I am claiming nerd status stems from my love of social media, stochastic processes and clouds. 

In the words of Liz Lemon, "What the, what?!?!?!"


How the heck are those three connected and how can someone love them?  It has to do with tracking and predicting how things interact with each other. For the sake of comprehension, bounce and bump readily come to mind as workable synonyms for interact. Anyways, I also love people. I think they are the most fascinating and interesting things out there for a plethora of reasons. One of them can be linked to the near impossibility to say with absolute certainty what we will do as a group. There are no time tested static natural laws just excellent rule of thumbs in predicting peoples'  behavior.

My question is what else acts like this?   

For some reason my mind instantly brings me to chemistry and the behavior of electrons. Hey, I told you I was a nerd. Unless your truly enjoyed chemistry back in the day, there are probably only two or three things you remember from class. One of them being"chem is try," another remembering how awful methane smells and then the basic structure of an atom.

Time for a quick refresher on the Neil Bohr model that was shoved down our throats (which depicts an extremely basic view on the composition of your stereotypical atom): You got a positively charged nucleus which is made up of protons and neutrons and then you have negatively charged electrons flying around it. The protons and neutrons are basically grouped together which form the nucleus while you have the electrons hovering around them at a comparatively large distance from it. Electrons are negatively charged while protons are positively charged. Two separate atoms can share the same electron but when that occurs they form a bond. Hence these two atoms form a compound and have become more than mere atoms. 

This formation of a compound I personally find interesting but there's more to keep in mind. You also have "power levels" that regulate what atom can bond another type of atom. If I remember correctly these levels also dictate what kind and strength of bond that occurs (this is currently all strictly off the top of my head so if I am wrong please let me know). Well here is my insight: I think people are very similar to this view of atoms especially when it comes to understanding their behavior which I suspect strongly correlates with what groups they associate with.

I think people have their core "self" kind of like the nucleus described by Bohr and then we also have these abstract ideas that we use to help define oneself. These abstract labels are our tools that we utilize to connect with others. They are our electrons. They swirl around in our heads. We can loose them. We can gain them. And we can share them. The sharing of these abstract self definitions is pertinent aspect in formation of social groups. Again, the only thing I believe is new to most people is the perspective we act similarly to atoms and the tight groups we form can be seen almost as compounds.

Now to the cloud!

So we have these atoms wondering around forming compounds with one another, some of these compounds are extremely ethereal and others much more permanent, wandering around bumping into one other and interacting in this thing called life. Some of these interactions seem to be quite arbitrary while others seem almost destine to occur when watch afar, as if some wind is pushing them. To me this sounds awfully like the dynamics we see in the sky.

Imagine clouds representing one of these abstract self definitions that a large amount of people currently share. The dynamics within each cloud are the different "compounds" and "atoms" communicating with one another. Historically, the dynamics within these "clouds" relied upon each "atom" having a handful of one-to-one interactions which typically would be limited to its current geography. Then these smaller yet stronger bonded compounds/ molecules would take their cue to move as a cloud by joining one of the few atoms and, or molecules whose affinity to bond with others is much stronger than typical.

 Much of this dynamic still pertains today and I have a suspicion it always will. People are social creatures and want to be part of groups; be it just between two people or with all of humanity. You will have smaller and better bonded compounds bonding with the few who for some reason or another can bond with many more creating molecules. These "stronger" molecules will play a disproportionally larger role in dictating the inner dynamics of said cloud. What has changed is that this "strength" until just recently almost always was limited to certain molecules. Thanks to the internet and social media, every atom or strongly bonded compound can in theory become that molecule who drives the inner-workings of said cloud.

To make matters even more interesting, when one thinks of a cloud (at least when I think of a cloud) one probably imagines millions if not billions of H2O particles with slightly different charges constantly bouncing off each other. Meaning it is nearly, if not, impossible to track each interaction. Thus for the longest time it was a Herculean task which constantly bordered futility trying to understand if a cloud was going to quickly dissipate or eventually gain critical mass and become a devastating hurricane. Again, thanks to the internet and social media this has changed. We can now trace a vast majority of today's interactions between people and ideas quickly and with a relatively high degree of accuracy. We can now see the inner-workings of society's clouds. To me this is super cool and brings us to the third aspect of this entry: stochastic processes.

As said earlier, there are no set natural laws that dictate human behavior; just excellent rules of thumb that are usually based on probability. If someone is hungry said person will most likely search for some food in the immediate future. It is not definite. They might be stuck in a meeting or attempting to diet, etc. It is not a given just very probable. This has been driving politicians, generals, business people and the nerds they hire nuts for years.

What the nerds who were hired by the business people noticed was that as a group, peoples' day-to-day actions produced a lot of "noise." They also noticed something interesting when viewing the big picture. They found that  the "noise" would seemingly hover around an unspecified, yet fundamental value (These people were usually trying to predict the market price of an asset and it's likely future price). They called it the population mean.

This underlying movement towards the population mean and also the noise surrounding it reminded them of Brownian Motion and the heat equation. Both of which (again, this is off the top of my head) are based on stochastic processes. They eventually came up with an equation called Ito's Lemma and utilize it as starting point to price a variety of things. Anyone in finance who does options trading and actually understands what they are doing can go on all day about Ito's Lemma. So can a good bond trader.

At the moment what's important isn't the math but the principle that people have been able with discreet data produce replicable methods in predicting human behavior that mirrors empirical results. They have taken something so intuitively uncertain and made it to an extent reliable. To me this is amazing and also a great opportunity.

Let's rehash the important points covered so far: People can be viewed like atoms
                                                                             Social groups act similarly to clouds
                                                                             Everyone theoretically has the power to influence
                                                                             We can now see inner-workings of "clouds"
                                                                             There's math that predicts human behavior
                                                     The above has stolen my attention like a pretty girl in a sundress

To my understanding these formulas based on top of stochastic processes ignore the reason why. Instead they look at a whole bunch of past data, standardize said data, assume some relationships and add the noise function. Bingo you got more or less what financial engineers have been doing for the past couple decades. But let's get back to the why. I suspect the reason why they ignored the why can be traced back to the idea that we operate in clouds. Until just recently it was notoriously hard to understand these social clouds at a micro level but also not everyone had the same access in controlling its inner dynamics. These two changes have occurred largely thanks to social media and they are increasing making the old formulas less relevant.

Look around you and you will see that this shift is occurring in real time all around us. This is a well documented and much lip service has been paid to the "power" of social media/ the internet. The trillion dollar question, literally on the small end of things a trillion, is who has properly updated their math and their assumptions that follow?

This is why I currently have a love affair with social media, clouds and stochastic processes it has everything to do about our future.                                                                








No comments:

Post a Comment

What's your two cents?