My Photo
Name:
Location: Hospers, Iowa, United States

Tuesday, July 06, 2004

Probabilistic Algorithm for Naming Intelligently Childen (PANIC)

After reading awhile about AI and particularly, gaming, I finally believe I understand the best way to name children. In order to avoid all of the complications and potential hurt feelings, may I suggest the following:

First, we know there is an element of chance involved, i.e., it could be a boy or a girl. The game of child naming involves not only this chance but also numerous states of decision as time progresses from when the pregnancy is first discovered until the actual birth. The terminal state is when you are asked what to place on the birth certificate. The operators involved are "accept the name", "reject the name", "hold it in the possible queue", and "remove it from the possible queue".

There are of course opponents in this game so you have to assume they will make the choice they feel is best. These could be family members( past and present ) and friends but hopefully not the pregnantee or the pregnantor.

It is suggested that one code this program into a networked computer program with a distributed front end that can be sent to all persons involved thus reducing further tension when a name is rejected.

One should probably use the minimax algorithm with alpha beta pruning( the one that has worked so well defeating chess champions ) and one will probably need to incorporate an expectimax calculation at chance points. Those persons who are likely to cause the greatest threshold of pain over the long run, i.e., years, due to the expectants making a wrong choice ( in the suggestors mind ) should probably receive a higher weight to their choice of name. Also names that could generate emotional hardship or psychological torture, like a boy named Sue, should receive a lower weight. Names that rhyme with the expectants's last name should be given negative values. Names that don't rhyme ( for instance the word orange ) or are difficult to rhyme should be given higher weight.

The program can thus be enacted so that it responds appropriately to each input it receives and places into the queue viable options. The queue can be set to be periodically viewed by the expectant ones. They can then line item veto names they feel are inappropriate.

The program should be set up to send a kind but firm rejection notice to individuals whose suggestions were rejected. A note of apology from the computer would be nice and some complicated mathematical language that could not be reasonably explained
by any normal human should be stated as the reason for rejection of the suggested name. When asked, the expectant ones can say they paid so much money for the program that they could not possibly think about not using it now that it is paid for.

There should also be incorporated into the program a "final name pick button" looks at each name in the queue and makes a final choice when the moment arrives. This button would enact a function that utilizes a weight of probability of success in life as drawn from a statistical analysis of persons names and their achievements throughout the world and history. The expectants can optionally choose to place weights ahead of time on particular values in this analysis as well, like love weights more than fortune, fortune more than power, etc...

Until then

0 Comments:

Post a Comment

<< Home