Monday, May 23, 2011

Predicting the future: The Rapture is Neigh!

This blog has a new home:
http://happystatemachine.com/blog/questions/predicting-the-future-the-rapture-is-neigh/



Can future events be predicted? The answer is that some events can be predicted, and some cannot. It all comes down to what kind of events are of interest.


Normally people who ask this question are talking about events like the Rapture, or the winning lotto numbers, or the outcome of future sport matches.

These events cannot be predicted as such, but some interesting discussion follows later.

First, let’s discuss events that definitely can be predicted:
  1. The failure of a hard drive can be accurately predicted given performance data collected by its logic board. There is an error margin associated with this sort of prediction.

  2. The expiry date of milk can be accurately predicted given known bacteria growth rates. Of course, the date is worked out for a certain temperature range. If the milk is frozen, then this date would be off.

  3. The future can also be planned. For instance, the start times of television programs are planned and accurately executed, since the duration of the preceding programs and advertisements are known and controlled. The TV guide essentially foretells the future.

  4. Anything subject to a known rate of change can be predicted. Even if the rate of change varies, or if there are multiple rates of change in play. An example is the phenomenal accomplishment by NASA and the Stardust mission:

    A probe was launched on 7 Feb 1999, rendezvoused with the comet Wild 2 on 2 January 2004 to collect dust particles from the comet tail. The Wild 2 comet moved at a relative speed of 13,650 miles per hour at the time of the dust collection. The probe then returned to earth and landed in the Arizona desert, exactly following the planned mission parameters as they were set even before the 1999 launch. This was a scientific triumph, and is an example of controlled events, making the future predictable.
Obviously if a previously unseen comet had impacted the probe on its way there, or if there had been a mechanical failure, or even a mistake in the math, then things would’ve turned out differently, and all the planning in the world would’ve been for naught.

The missing ingredients in the planning of any future event are … other future events.

You can plan to do the washing tomorrow, but if an earthquake swallows your house, then that plan will probably not come to fruition.

There is an inherent unpredictability in nature and natural events. We can study past events, and learn a great deal about anything that has already occurred, but we cannot determine that anything will or will not happen in the future.

In fact, we cannot even determine what is going on right now! Even when we directly observe a cell dividing under a microscope, for example, we are already seeing the event as it happened in the past. The actual photons that travel from the cell splitting, all the way up the microscope, and finally to the observing human’s eye, are all a few nanoseconds late. Travel time is not zero, and so we can only see into the past. Of course, for any event that takes longer than the light travel time to occur, this is never a problem. But the principle holds all the same.

Now, what about predicting the future?

Well, one can try to do with a given event what hard disk manufacturers do with hard drives.

You can construct a mathematical model of some event you want to predict, collect any relevant data that your model needs, and then run scenarios to see what the future holds.

This is what the weather forecast actually is: the results of the most current, most accurate scenario based on heavy math and a massive set of data points. And, as we are all aware, the weather remains obstinately unpredictable.

Where does this leave us? Well, even with our best mathematics and our best scientific minds, we still cannot predict the future. Not tomorrow’s weather, not the lotto numbers for the big jackpot, and not the movement of your Google stock either.

On a quantum level, we have some strange behaviour that could possibly be explained by information arriving at a point in time before the information was even generated. This comes from one of the many weird observations in quantum physics, where the outcome of the event could possibly influence the preceding sequence of events that led to that particular outcome. In other words, the result can determine the process by which the result was achieved.

However, this is just a fanciful way of expressing some as yet completely unexplained observations in quantum physics, with a good dose of scientific philosophy thrown in.

Many years from now, when we get our Einstein-for-Quanta, we will no doubt be amazed at how simple everything really is, but for now we will have to deal with the real world, in which we are doomed to predict non-events like the Rapture, and then, when nothing comes of such crazy and unfounded predictions, we smile at each other and say: I never believed it would happen in the first place.