The Winged Wumpus Womper performed well overall.
During the competition on 25/12/08 it was able to solve the board as expected and relatively quick.
The gold was successfully collected, wumpus and pits avoided, and the robot re-traced its steps on a safe route back to the starting tile.
There were no mistakes in terms of not following its original programming and algorithm.
The inital test runs in the real world during lab and in our simulation also performed just as we intended.
For future improvement of the robot, there could be an added element of making it determine the safe route, pits, wumpus,
and gold more efficently if possible and also incorporating the idea of chance in order to solve more difficult boards that rely on the choice between two unknown options.
Relating this Project to Class:
Does the robot have a mind and is it intelligent or conscious?
3C or The Winged Wumpus Womper is capable of navigating the board to get the gold whilst avoiding danger.
It does this by tile by tile movement and analyzing of the signals it recieves in order to determine the safe route.
This may seem like signs of intelligence or even consciousness but there are several factors that say otherwise in this case.
The Winged Wumpus Womper as of now is only programmed to this specific game - the Wumpus World.
All the decisions made to get to the goal are all based off the rules we set in place.
So, if one were to try and place our Wumpus Destroyer in another game, say Chess, this would not go well let alone go anywhere.
Another factor to consider is the ability to learn, either from ones mistakes or success.
When a human plays Chess, they learn new strategies as they play more often and learn what not to do that resulted in unfavorable consequences previously (losing pieces and/or the game).
Though for the Winged Wumpus Womper, nothing is taken away after each play of the game or considered as to why certain paths were good or bad outside of its base programming telling it yes or no.
Therefore, we have concluded that the robot does not have a mind and is not intelligent or conscious as it cannot learn or make its own decisions outside of its original programming.
Do you feel any differently regarding the feasibility of AI after this project?
We feel that while currently there are many softwares, websites, and platforms based off of or using AI, there is still much work until AI can be considered to have a mind, conscious, and intelligence. But, while doing
this project in particular, our opinions on the feasibility of AI haven't changed. This is because the way the logic of our robot operates doesn't accurately reflect how current AI technologies
work. By definition, AI performs activities such as learning, problem-solving, and decision-making. It solves the Wumpus Board Game and makes decisions based on the data it collects, but it doesn't learn from its experiences,
which is the main difference between an AI and just any old computer program. Because of this, our opinions on the feasibility of AI have not changed.