More on the Local Optimum

In a previous post we talked about being stuck at local optimum. In order to avoid this outcome, one solution is running the hill-climbing algorithm many times, and each time choosing a different starting position randomly. The hope is one of the starting positions will be near the “real maximum” and simple hill climb will take us to the top, so our mechanism, which cannot see the global solution, after many tries, can still pick the “maximum among the maximums”, and throw away the rest.

LIFE is a complex problem, in this setting individuals could be seen as “local solvers” whose starting positions are scattered, and different. This is why pluralism is important, multiple viewpoints should matter, multiple companies, multiple enterprises, even multiple efforts inside the same company need to exist so they can all climb various different hills, each trying to find the optimum point.

In history, civilizations, cultures can find themselves being stuck at various local maximums. Roman Empire climbed one and got stuck. Ottoman Empire found itself at another and collapsed. It takes a man to recognize one is stuck at a local optimum, and decide to climb down, and start the climbing that other hill.