The 95/5 Principle of Exploration – Tim Clark
In this Show
The 95/5 Principle of Exploration
This morning I drove a different route into my office to see if there was actually a faster way to get to work compared to the route that I have been driving for years.
In the AI industry, we call this the 95/5 Principle of Exploration.
– 95% of the time, we use the most efficient method for completing a task or solving a problem.
– 5% of the time, we explore a “new way” to accomplish the same task. You may be surprised when you find a faster or easier way that you ever knew existed.
We use this same principle in machine learning when automating businesses. 95% of the time, our software uses the “best-known” approach to solving a problem, but 5% of the time, it tries something new, to see if there is a better way.
There are 2 possible outcomes when applying the 95/5 principle (in life and in software):
1) The explored approach is less efficient than your current approach. While this may seem like a step in the wrong direction, this is actually a Win! By trying a different approach and in effect failing, you are actually gaining confidence your current approach is the best and optimal solution.
2) The explored approach is more efficient than your current approach. This is clearly a Win and will provide tremendous savings as you apply this approach into the future.
And as for my drive into work this morning using a different route, it actually took about 5 minutes longer than my typical route. That gave me confidence that I have been driving the best route possible all along, and not only that but by driving through a part of my community that I had never seen before, I gained some new inspiration for a different way to package my products and services.
If you are interested in applying this principle to your business, please connect with me at nikosce.com.