Parallel distributed processing is resilient and efficient
Parallel distributed processing (PDP) models have been suggested as a solution to the flaws of centralized symbolic models. The von Neumann bottleneck refers to the extensive time complexity of algorithms forced through a sequential architecture. By attempting to parallelize cognition, PDP enables more efficient models, which often better reflect human performance. Additionally, due to its distributed nature, PDP is more resilient to accidents, malfunctions, or mutations.