The Business of Prediction


A king shall fall and be put to death by the English parliament shall be. Fire and plague comes to London in the year of 6 and 23. An emperor of France shall rise who will be born near Italy. His rule cost his empire dear – Pay-nay-loron his name shall be.

from the Quatrains of Nostradamus

Let’s face it; we want to know the future – and why not? Wouldn’t it be cool and save us a whole lot of trouble if we knew what tomorrow will bring? It’s remarkable to note that, during times of great uncertainly and dissociations, we increasingly  turn to prognosticators and seek out answers; this trend is evident on several recent events:

* Predicting the weather. Knowing when and where hurricane Sandy was going to land did not stop things, but it made for a far more effective response and coordinated effort. Compare Sandy to Katrina and you can well appreciate how far we’ve come in terms of emergency management and practical planning.

* Election results. No where is this more true than the numerous pundits who sought out the future and turned to a variety of models, castings and other such approaches.

* Business / economic trends and developments. Increasingly, Wall Street awaits the word from Washington and the Bureau of Statistics and the Department of Labor to learn of the latest trends and developments, seeking to know when and what the forecasts are in terms of employment, investment, trade, commodities and other market developments.

Now enter Big Data and AI.

As noted by such sites as fivethirtyeight blog (our kudos to Nate Silver!) it’s no longer so much what the pundits are saying: they’re only in it for the ratings so naturally, they’ll always have a slant (or, as my grandfather used to say ‘beware a person who believes in their own bullshit‘). Using the cold, hard facts and level-headed statistics – like those utilized by fivethirtyeight – demonstrated how well we’ve advanced just in the past ten years along in terms of prediction.

I have to note my own role in the business of prediction. Some fifteen years ago, I developed a means of predicting when and where crime would likely occur, offering a tool for local police to utilize (this was known as – surprise! – Predictive Crime Analysis). This was achieved through a means of data analysis vis-a-vis GIS (Geographical Information Systems) and did not require a large computer; rather, it required dedicated staff submitting accurate information, conducting close data review coupled with utilizing proper statistical determination / relevance to any given data set and then mapping same. The result? In one locale, we were able to reduce crime by over 40% in the first year.

Fantastic, right?

Wrong.

After awhile, nobody wanted it as it cut back on police overtime and, in some instances, forced the criminals to cross over into neighboring towns to conduct their activities with the neighboring towns grouping together and placing political pressure to stop this effort. In the end, it was removed, retired and now forgotten (although I do still have various articles and papers discussing this; please feel free to contact me if you want to learn more).

Author’s Note; Fast forward fifteen years later, and the irony is that those very same towns are being asked to “share” their police personnel to help deter the rising crime wave in the neighboring town where crime was once down 40%,…! Arguably, despite their best efforts to deter the future, the future came forth and changed them,…!

The point is that sometimes knowing the future is not always agood thing, as by knowing the future, we (sometimes) change the future (like that famous Twilight Zone episode, where the individual seeking to learn his future finds out that he will die in twenty-four hours; it is suggested that by his learning his fate, he only increased his chances of making it happen). It is a conundrum noted by Quantum Physics by the (famous) thought experiment known as Schroedinger’s Cat: the very act of looking into the box changes the outcome of what it is you’re seeking to understand as the very act of observing a physical phenomenon can affect the outcome.

Now, this is not to suggest that by predicting the future direction of a hurricane (or other large-scale natural events) that we can change things, but it is not too far to hazard a suggestion that by knowing the trends of business, commodities, trades, employment developments, voter perceptions, etc. – that we can also change the nature of what it is we seek to understand – or control.

It is an axiom that in conducting any type of precognition, you need to set aside your beliefs – both conscious and sub-conscious – if you’re going to do a good job. This is not easy for sometimes, we just don’t like what the future is telling us.

But the opportunity! We are in an age of Big Data and information review unlike any never seen before in the history of mankind. We now have the tools and processing power. We can download and obtain data on a multitude of subjects and developments, convert it and feed it into systems that we can readily program and design for any variety of applications. Now, more than ever before, we can predict our futures in ways never even realized. The trick is, doing it right – and accurately.

Now then, that being said – allow me to enter a prediction of my own.

We shall soon see an AI arising from the bulk of Big Data – sooner than we realize. And quite possibly, it may even already be operating amongst us (as noted in my prior posts),…

Perhaps Nostradamus wouldn’t be such a bad name for such a computer.

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