ORIGINAL: USS America
A chess board has a grand total of 64 spaces and 32 units for an AI to consider. Add even 10 pieces on each side, and you are multiplying the number of AI calculations by a massive amount.
Now, consider the smallest AE scenario. How many spaces are there on the board for the Coral Sea scenario? How many ships are there? Don't forget the air units, inclucing the air units onboard ships. There are probably a few more LCU's than the total number of pieces in a game of chess, for the Allies and the Japanese. Oh, and some of those spaces on the board are very different from others (bases/ports/airfields).
Deep Blue plays chess by "brute force" calculations of all the possible moves. I'm 42 years old. Even if the developers had a thousand systems as powerful as Deep Blue to use, they will never be able to program an AI that can be as effective as a human in my lifetime.
I am very impressed with the improvements Andy and company have been able to make with the AI in AE.
Now, I need to go and play a turn for my evil Martian opponent. I never giggle to myself, picturing the look on my opponent's face as I sneak in an attack where he didn't expect it, when I'm playing the AI. [:D]
However, my point about no one being able to afford an AI that could play WiTP as smartly as a human is still valid. [:)]
Chess is complex game and you have to play against a truely skilled opponent to appreciate it. The fact that a computer can be built that can (even through brute force) be able to analyze and react to a thinking grand master is quite an accomplishment. Even with all the brute force calculations, the computer still had to make a choice of one right move, in a sense 'out-thinking' it's opponent...maybe not in the conventional human way of thinking, but definately in an analytical manner.
And something I like to quote every now and then (from a book I can no longer remember the title or author of sadly)...
"If brute force doesn't get the job done, you aren't using enough."
