The Prediction Paradox

The Chronovisor is a hypothetical device, that looks and operates a bit like a laptop computer, but which allows one to see the future (or the past).  Whether it truly exists or not is a matter of conjecture, as is also the question of how it might be possible for it to work, if it does exist.

Let’s suppose that it does exist, just for this momentary discussion.

Suppose further, then, the following scenario.  Joe Schmo is planning to go swimming at the ocean beach next week, and wonders whether it will be a good day to do so.  So, he operates his Chronovisor to view the time and place of his planned swimming trip.  There on the screen, he sees his future self at the beach, enjoying a swim—but then, horror of horrors, he sees himself attacked and grievously injured by a giant shark.  Of course, Joe immediately cancels his plan to go to the beach, and warns the authorities to close the beach for that day.  He then checks his Chronovisor again, and is pleased to see that no one will be in the water that day.  The future has changed.

What has happened, is that, the very act of predicting a future event, has prevented it from happening.  Is this a paradox?  It depends.

If the future is perfectly predictable, then that would mean that it is inalterable.  It must inevitably occur.  That would mean that Joe could not prevent himself from taking his tragic trip to the beach.  That, however, would be absurd, since we know Joe well enough to know that he would not go, if he knew in advance about the shark.

On the other hand, if the future is not utterly predictable, then, the chronovisor could not work as advertised.  Therefore, the chronovisor cannot exist, at least not with accurate performance.

We must then ask, which is it?  Is the universe utterly predictable, or utterly unpredictable?  The answer seems to be, neither.  We can make predictions, but the future is alterable.  We can make predictions, but only with varying degrees of probability.  The universe, then, is approximately predictable.  Some predictions have such a high likelihood of being correct, that we can ignore the chance of error. 

The paradox, however, remains.  Can the accurate prediction of a future event, itself make the prediction inaccurate?