Understanding Artificial Intelligence Using Golf as a Metaphor © 2017 Tina Whitfield

dy·nam·ic / dīˈnamik/   physics: Relating to forces producing motion; linguistics: change, disruptive1

con·stant / ˈkänstənt/  physics: A relation or property which remains the same in all circumstances2

Part 1: The Constant Environment

Why use Golf as a metaphor for the interplay between Artificial Intelligence and its environment?  Because both golf and the big data analytics that drive AI are impacted by their environments.  Particularly, constant versus dynamic environments.  And, in AI, a constant environment directly correlates to the successful quality of outcomes.

Today’s blog digs into this phenomenon.  And, to simplify the complex, we begin with golf.

Golf

Golf announcers often complain of slow play.  For some of us, it’s not slow; it’s methodical.  We analyze the lie of the turf, each blade’s turn, and the angle of the bunker with its rise and fall. Then with stillness, we listen to the wind, feel the temperature of the air, and see the position of the sun.  When we are set, we focus on our target and swing.  We see the direct correlation between the observed environmental factors and our ability to hit the target.

Now, think about each of the factors presented: air, wind, temperature…

When the club head connects with the ball, it is sent high aloft, catching and cutting the winds that send it to the target.  While the ball is in the air, its loft, direction, and speed are impacted by the air temperature, humidity, temperature of a nearby pond, and speed and direction of the winds; which are all impacted by the position of the sun.  These interactions are dynamic and cannot be replicated;  which makes the quality of shots or outcomes in a game of golf – unpredictable.

As golfers prepping to swing, those of us that are methodical, scan the environmental factors, establish a “factor set”, then look for matches to previous environmental factor sets tied to successful outcomes.  When a match is made, our confidence increases.  And, we swing.  The more constant the environmental factors, the more confidence.

Big Data for AI

Like golf, we want ‘big data analytics for AI platform environments’ to operate as a constant to improve replication so, for example, the machine learns and predicts as it should, leading to a successful quality of outcomes.

The first factor in an environment is the Algorithm.  Deep within the elegant code and the beautiful syntax is the signature of the coder. And, when the coder sends out the code or ‘golf ball’ it travels high aloft through singular or multiple cloud systems to produce results that can be tracked back to the coder.  And, when developers use the code, they can maintain consistent confidence in it.  A singular algorithm is a constant.

Unfortunately for those that crave the consistency of the constant, to achieve replication, many companies operate with multiple siloes; whether campuses, geographies, or divisions and use – the next factors -different vendors in each silo with different platforms created by different system architects, with different algorithims and coders – for operational use inside the company and in partnership with the supply chain.

Using different big data analytics platforms is akin to introducing different golf drivers and balls.  Replication is challenged.  Thus, ‘trueness’ is challenged across the company.

How? If we input the same problem or request into each of these siloed unique vendor platforms, we would experience the same effect as if we selected the same player (problem/request), but different driver (vendor platform) and different ball (algorithm), combined with the dynamics of varied air temp, speed, and wind direction.  A repeat of a glorious hole in one, cannot be expected.

Achieving the Ideal State

As discussed, we want to improve replication. The ideal would be to have the same environment in every location.  Then, analysts in Dubai can compare outcomes with peer analysts in California and Australia, and do so with a high-level of confidence in global results.  It can be achieved.

Create continuity.  Invest in due diligence.  Remember not to kluge vendors and systems, rather scale one environment across the company.

End–

Note: Part 2 of Understanding Artificial Intelligence Using Golf as a Metaphor will explore Decision Trees.

  1. and 2. Oxford English Dictionary, https://en.oxforddictionaries.com/

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s