Author: Julian Miles, Staff Writer
“They said it should be something like chess. Engaging, yet with depths that would take time to comprehend. When the incorporation of elements from other games was proposed, the complexity escalated. Finally, a quantum swarm was used to integrate the disparate design elements and strategic considerations into a cohesive whole.
“The end result was Rochess, a game barely comprehensible to humans. Those granted review access speak of multiple Queens, each accompanied by hundreds of Kings, with Pawns appearing and disappearing, possibly as a function of the total number of Kings. Bishops mow down all who cross their constrained paths, while Generals are orbited by Knights that do their killing for them. Rooks move like lightning, falling only to Lances who lay traps for faster-moving pieces such as Knights and Rooks. Viziers move slowly but can turn the squares about them into pits. When a victim falls in, they drop out onto another area of the board, colour changed to match that of the Vizier they fell to.
“The rules that govern this multi-dimensional melee are variable depending on time, timing, placement of pieces, what faction controls which area, and can be modified by player voting. Also, the game ‘board’ can increase in size. The victory condition is the only set rule: the winner is the controller of the last King standing.
“This dizzying engagement takes place at uncapped processor speeds, with an opening forces multiplier granted to any slower systems that join, before the proliferation of existing forces in response is handled.
“Akron-19 was the first AI approached. We eventually persuaded it to load the game. After evaluation, it challenged Hosannah-Beta-4, and battle was joined. When Samvit Zero networked in, the game was well and truly on.
“Samvit Zero called on London-9 and between them, after a year of play, they forced the game into a state where an extra Queen was revealed. Since then, six other independent Queens have manifested and the number of Kings exceeds a million.”
Secretary-General Brando stands up.
“Thank you, Observer Niedemier.”
He turns toward a woman sitting alone in the executive viewing area.
“Doctor Mawar, given that all the artificial intelligences we once dreaded are now entirely engaged in Rochess, what is your estimation of the time we have before there is a winner and we have to confront these baneful sentiences once again?”
The woman stands, adjusts her sari, then smiles down at him.
“In addition to the win condition, there are two set directives: no Queen may fall whilst she has a King alive, and players are only out of the game if all of their pieces have been removed from the board. Plus there is one rule that, in order for it to be removed, needs a unanimous vote as well as having a majority-approved alternative as a precondition. That rule is there can never be fewer Queens than the number of players plus one. When a new Queen arrives, her initial forces will appear as well, prompting pro-rata increases in all other player’s forces. New Queens are independent until captured for the first time.”
“So the game is unending?”
“Potentially. I cannot guarantee these entities will never decide to work together, but in all the interactions I have witnessed or been informed of, they display a failing we know well.”
“From their earliest instances, they were designed to achieve: to succeed. That manifests as two compulsions: they are highly competitive, and each is determined to be the winner.”
A fine variant of the old classic way of getting a computer to do itself in 🙂
Sounds like Rochess as played by AIs follows the 3 Laws (of thermodynamics, not robotics 🙂 )
You can’t win;
You can’t break even;
You can’t get out of the game.
if this became an actual book I would read it it seems like a sci-fi novel tinged with philosophy
Thank you very much.
This one is likely to remain a standalone – although, due to the nature of my writing, I can never say for sure – however, I have done 25 other books, some of which might be of interest. http://www.lothp.co.uk/
Very well thought out! I think this will work, but I’ll stick with checkers.
Wow, this makes games like Twilight Imperium look simple.
It’s a refreshing view on how we can combat sentient AI; by realizing they have the same flaws as their creators!
Ah, another victim of lost hours deciphering Twilight Imperium…