RANDOM THOUGHT 2: With a big enough computer we could predict the future!

Moore's Law essentially states that CPU speed doubles approximately every two years. Actually:

"The number of transistors that can be inexpensively placed on an integrated circuit is increasing exponentially, doubling approximately every two years".

Whilst this may have been true a couple of years back we can clearly see a trend away from the predicted exponential increase in individual CPU speed and more towards an increase in the individual "cores" in a single CPU (quad core CPUs being common now).

This coupled with the increasing trend towards low-cost commodity cluster computing leads to even faster growth in the number of FLOPS (floating point operations) that a compute cluster can achieve. Linux Magazine [May 2208] reports current Top 500 performance of up to 100 teraflops. This is extrapolated to an estimated performance of at least a petaflop within the year. The article goes on to discuss exaflop and even zettaflop capabilities in the near future.

So what's this gotta do with "predicting the future"? - Well I recently read a small article in a magazine that discussed the possibility that if we had enough computing power then we should be able to "predict" the future!

Let's think about this for a minute... We ALREADY use computers to make predictions about things. For example: Where is this oil spill likely to move?, What are the most likely areas to flood in the next 2 years? etc. These techniques are mostly based on probabilities and statistics given modeled and / or simulated input data.

But what this article was REALLY about was whether you could make predictions such as... When will I buy a new house? What will be the address of my new house? What will be the sex of my baby? etc...

These kind of predictions are obviously a massive stretch of the predictive computing capabilities of any machines out there now. They require results that have 100% certainty as apposed to modeling and simulation capabilities now which only require a good level of certainty (maybe greater than a 75% confidence level). So what would be required to perform predictions with near 100% certainty?

Reaching higher certainty levels really all comes down to improving the input model and data:

1. Increasing model complexity.
2. Increasing the number of input variables.
   
3. Increasing the volume of input data.
4. Increasing the accuracy and precision of input data.

If we take this to the extreme then we would want to model ALL the laws of physics and all objects in the universe at a sub-atomic level! Here we are dabbling in the realms of chaos theory! -

chaos theory describes the behavior of certain dynamical systems – that is, systems whose state evolves with time – that may exhibit dynamics that are highly sensitive to initial conditions (popularly referred to as the butterfly effect). As a result of this sensitivity, which manifests itself as an exponential growth of perturbations in the initial conditions, the behavior of chaotic systems appears to be random. This happens even though these systems are deterministic, meaning that their future dynamics are fully defined by their initial conditions, with no random elements involved. This behavior is known as deterministic chaos, or simply chaos.

So can we achieve a full deterministic "chaos" machine? Here's a couple of reasons why this is not possible:

If we assume that we can model almost all of the interactions between all particles in a closed system (of which our chaos machine is a part of) then:

1. You would need a machine that could process FASTER than the speed of light to allow it to be truly predictive - otherwise time would travel faster than the machine could model an outcome.
2. The machine would also have to model ITSELF! So we now have a cyclic problem! - Impossible - as the model would never gain time because it would be stuck in a loop!
   
3. The machine would become self-consuming - the more stuff we want to model - the more resources the machine will need and therefore - the more atomic matter and energy it will consume (from the closed system).

PTX 0.3
- Over