Showing posts with label langton's ant. Show all posts
Showing posts with label langton's ant. Show all posts

Feb 22, 2009

Langton's Actuarial Ant

As an actuary, you believe in the consistency of your risk models.

You might think that with 10.000 observations you've got enough stuff to present a consistent statistical model with realistic expectations, variances, etc.

You are aware that the output of your model depends on the quality of the data and the assumptions. In your advice you try to communicate all that to the board in order to support sound and responsible decisions.

In other words, you've got a consistent model and, as an actuary with a professional and consistent life-philosophy, you have everything under control. No great changes will take place?


you're probably wrong !

Just like our models, we actuaries, are not consistent

Even if we (or the risk reality we try to model) act in a stable consistent way, we (or risk reality) keep interfering with our environment and our environment responses to us.

At first this response seems meaningless and of no value. You think you're consequent and your work and achievements in life seem relatively stable, perhaps a little bit chaotic and of no great significance. But in repeating your proven receipts, way of doing or procedures endlessly, eventually

Something will change

This change often will not appear as an evolution in your life, but as a kind of revolution, out of the blue and most often unexpected.
Suddenly, just like in the credit crisis, there's an emerging situation. The way you always did it, doesn't turn out right anymore. Your model crashed, you crashed and there was nothing you could do about it. You couldn't have foreseen it, you could not have prevented it the classical way.

That's why we always have to add some non-classical extra 'common sense' safety margin thinking in our models.

The other side of this is also true. Fore example, when you study, you'll probably, once in a while, think: what progress am I making?

But don't worry, if you keep on your track, there'll be a day your future suddenly comes to you (out of the blue: as a kind of emergent property) instead of "the you trying to make your future" in this Game of Life.

A good demonstration of this principle is

Langton's ant

Langton's ant is an virtual ant that starts out on a grid containing black and white cells, and then follows the following set of rules.

  1. If the ant is on a black square, it turns right 90° and moves forward one unit.
  2. If the ant is on a white square, it turns left 90° and moves forward one unit.
  3. When the ant leaves a square, it inverts the color.

The result is a quite complicated and apparently chaotic, but relatively stable, motion. But after about 10.000 moves the ant starts to build a broad diagonal "highway".

So keep in mind "Langton's Actuarial Ant" next time you design a new risk model.

Anyhow, stay on your track as an actuary and remember, whether it's you in life or your models, someday there'll be

a collapse of chaos