The Relations Between Implicit Parallelism and Artifical Intelligence

Implicit parallelism exists in a lot of intelligence algorithms. It is the causation of parallel search ability in results space of some algorithms. Implicit parallelism of algorithm is analyzed by physical theory and find that the uncertainty and high entropy state of algorithm are the root of the implicit parallelism. But the uncertainty is the cost of implicit parallelism. In other words, if an algorithm wants to get high level implicit parallel computing ability it must reduce the computing precision. For an intelligence algorithm the a priori knowledge decides the search direction in results space and the implicit parallelism provides the high speed parallel search ability. According to this we can increase the a priori knowledge and implicit parallelism to get the high level intelligence algorithm. Correct search direction plus implicit parallelism is artificial intelligence. The implicit parallelism and the prior knowledge together form the two elements of intelligence algorithm: speed and direction.