Monday, February 1, 2016

We Are All Ants

Imagine that you were walking outside one day and spied an ant scurrying over the ground.  You pause for awhile to observe the ant.  It moves to and fro, spying out new trails, picking up food, and generally making random maneuvers.  As you watch this ant it's highly likely you're thinking about just how unique it is amongst ants.  You might even make the guess that every ant in an anthill is equally unpredictable.  In other words, you could pick any ant and in watching it notice that its behavior is distinctly different from any other.

If you're like me and have a distinct hatred for fire ants (Hello, Texas), you may occasionally kick over a mound just to see the bajillions of ants moving about in an absolute frenzy.  For those of you who have done this, you'll agree that once the top comes off the mound it's just absolute chaos!  It looks like rice boiling in water with no pattern discernable whatsoever.

Now what would you say if I told you that every single ant you'll ever find on the face of the Earth has behavior which is absolutely predictable?  Yes, in this post we are going to talk for just a bit about predictive analytics.  There is a whole scientific practice that is dedicated to the creation of pinpoint predictions about just exactly what is going to happen and when.

Back to the little part about ants from earlier.  If you were to do what I suggested and pick an ant to watch for a bit, you most likely would learn nothing about it other than it acted without much repetition.  However, if you could view this exact same ant for five days straight (assuming x-ray vision for underground movements) you would start to build up a robust data set of actions.  After an interval like this you would become very surprised at just how similar this ant tended to behave.  Even accounting for external conditions - changes in climate, temperature, or people kicking over the ant's home - the ant's behavior would be remarkably easy to predict after five days of observation and data collection.

So why is this topic so important that we would need to dedicate a whole blog post to it?

That ant is really you and me.  Does that sound completely insane?  There is no way that a human's behavior, as complex as it is could possibly be compared to an ant.  Right?  Wrong.

Humor me for a second and think about a couple of things:

  • What you do right before you go to bed // What you do right after you get up
  • What do you eat and when do you eat it?
  • How often do you see the same car(s) in your daily commute even though you live in an area with thousands/tens-of-thousands/millions of people?
  • What pathways do you choose to walk to your most common destinations?
  • How do you answer the phone and what do you say when you hang up?
  • What do you say when someone sneezes?
These questions could go on and on, but something tells me that you get the point.  Nobody acts and behaves the same all the time.  External stimuli makes sure of that.  But as quickly as possible, people revert back to the same patterns.

Remember when I referred to the need to have a sufficient data set to predict the ant's behavior?  Guess what?  If you're a person (don't even need to be an adult) with an email address and a credit card there is already an ENORMOUS database dedicated just to you.  Can you guess who has it?  (The answer is that your data is shared around more than gossip on Facebook)

None of us is nearly as unique or random as we'd like to believe.  The truth is that we, as humans, are exceedingly repetitive and thus quite predictable.  If we could view ourselves as we might an ant, over time we would see that most of what we do day in and day out is part of a regular pattern.  From what you say to what you do and where you go (and why) can be predetermined by your historical behaviors.

Being predictable is not always a bad thing.  It allows us to be served more effectively, from entertainment to products (and marketing), to traffic control.  Government and private business can be prepared well in advance to ensure that our needs are always met and our "wants" are always catered to.  The science of predictive analytics has allowed us to do so many things to better our world that it would be hard for a modern day human (1st world) to live in a previous century.  Imagine going to a supermarket and NOT finding exactly what you need.  Today, the store already knew well in advance what you'd need and had it ready for you before it even crossed your mind that you had to go to the store.  Your great grandparents would be in awe.

Predictive analytics now combined with big data are changing our collective world in new and wonderful ways.  If you feel inclined, take a look at what your own company is doing today.  Are there opportunities?

The only downside for predictive analytics comes for those people who are targets for assassination or kidnapping...

No comments:

Post a Comment