The .NET Task Parallel Library simplifies development of and reasoning about asynchronous code. However, working with Task instances instead of blocking method results can still be difficult and un-intuitive. One example where this is true is in implementing retry logic for functions that return Task instances. Consider this helper function which retries a provided function a specified number of times:

The Law of Large Numbers and the Central Limit Theorem imply that the world is in a state that has the highest probability in relation other conceivable states. The previous statement refers to the world as a whole and thus minor variations in the probabilities of proximal states are expected. The fundamental technique in the mentioned mathematical theories is that of measuring the likelihood of an outcome based on a determined function. In particular cases, such as the familiar deck of cards, this function is derived from ratios of expected cards versus the entire sample space - the deck. It is reasonable to believe that the things that happen, in a large environment and over a sufficient period of time, are the things that have the highest probability. In fact this empirical data is where the probability functions derive from. The question here is whether these things that do happen are simple or complex. And surely the next question asks for the definition of simplicity and complexity.

A.I. purports to understand thought and create systems that are able to perform thought, or think. In this endeavor we attempt to observe, study and analyze our own thinking, and make conclusions about thinking in general. These conclusions become the axioms in the syllogisms of Aristotle. Subsequently, traditional intelligence models follow this paradigm. An important observation is that these models are not only designed to follow syllogisms, but are also arrived at in the same fashion. This however is in great contrast to how our own intelligence evolved. Our own intelligence evolved through millions of years of random trial and error, extinction and survival. The logical backbone behind the rationality humans posses is the confluence of our environment and the challenges it entails for survival, which of course is a requisite for any sort of intelligence to exist.

An interesting challenge arises in abstract matematics when it is required to translate intuitive notions into formalisms. The existence of a particular formalism is always uncertain and the path to its discovery is often a walk on thin ice challenging mathematics at each step.

There seems to exist natural beginnings of thought in approaching problems to be solved. As an example take the problem of finding the volume of a solid. It is natural and certainly reasonable to think that the volume is related to the shape and size, therefore it is possible to measure. This vague notion of relationship eventually develops into a more formal notion of function and thus a strict formula. At this point a natural inclination, and a general pattern of human thought is abstraction. We now seek similarities between measuring the volume of different solids. Skipping ahead to the mathematics of real analysis we have measure theory, integration theory, etc. These theories we developed from this cycle of making particular observations and then creating generalizations. Thus after the efforts of hundreds of years of mathematical research we have the ability to measure the volume of a solid, and therefore a mathematical theory of measuring - measure theory. When a new generation of mathematicians is learning the work of the past they approach solved problems from the perspective of these theories. The important thing to note is that this is the reverse of the approach that resulted in the theories. A student’s effort is then directed towards understanding the theory, or in essence an encoded thought of another mathematician, instead of understanding the problem the theory addresses in the first place. Not to say it is a wasted effort it is an interesting pattern in learning.