How can we make artificial intelligent robotic assistants and androids more life like? How can we get these units to function more like humans and around humans without constantly reminding them that they are only robots? Is there a way to insure that the decisions that robots make will look more life like? That is to say the decisions, which are simple to the ones, which are more complex in nature. How so you ask? Well consider the washing of a car in the driveway, an obvious task assignment of a human to their robotic artificial intelligent android. Now then, let's look into a plausible and likely hypothetical future; The washing of a car has been programmed into the computer via a little program designed and the patent rights have been licensed to some 30 different robotic household android manufacturers. Only one problem with all this; on any given Sunday neighborhoods across America and around the world have their robots washing their cars all exactly the same way. Now what? Well we call up this program writer and patent holder character and tell him of the problem; "Hey look you Bozo, since you own the rights to the program and patent on artificial intelligent car washing in the driveway, every G-damn android washes the cars exactly the same way!" So of course back to the drawing board we go. The answer is simple, we break the wash process into some 20 parts and we set parameters to each of the parts for a multitude of things; speed of movement, number of times area is wiped, number of times towel tricks are done (such as snapping, pizza spinning technique and tossed over the shoulder), we modify the number of times to and from the soap bucket, etc. Also we randomly assign which parts of the car are done when and how much time is spend on each, by setting associative boundaries of time both plus and minus. All this makes the motions of the robotic android appear to be more random and different from the robot next door, down the block, around the corner and in the neighboring country. Please consider this in 2006. |
