My take on the bigger picture of Big-O. Disclaimer: Big-O, its names and characters belong to Bandai/Sunrise. Beta Reader: Sybil Rowan
Fortunately, computing power has advanced significantly over the years. Long ago, when nuclear weapons were new, computers would simulate the effects of a nuclear holocaust. The goal was to find a strategy that would not result in mutual annihilation. While it was not successful, weapons technology advanced in ways that changed the equation. As these advances continued, the possible consequences and interactions of these weapons needed to be accounted for.
Bigs introduced a new wrinkle to the equations. They were prototyped on the basic idea of the tank/howitzer, airplane/jet, and submarine/boat. However, it was necessary to account for their greater maneuverability and potential to remain after any war. The potential for significant disruption to society was great, but the potential for lingering after-effects were even greater than those of nuclear weapons. The predictable catastrophe was no longer the problem. Instead, the goal was to determine the aftereffects and how society would deal with them.
The simulation ran on the latest parallel processing equipment spread across a continent. Parameters for the simulation were simple: assume nothing about the amount of information retained by the people, assume people are similar to those in current society, and assume they will retain certain knowledge to deal with leftovers from the war. The simulation was kept to a small-scale with a high level of detail. The result was that it could accurately model up to 250,000 people.
Considering the number of variables in a simulation of that size, it was deemed that genetic algorithms would be used to search for strategies that would ensure the survival of the population. Each simulation used a single "genetic seed" that represented all the parameters for a simulation's initial conditions. After all the simulations were run, some genetic seeds were discarded as failed strategies. New genetic seeds were derived from the remaining seeds to form the next generation and the process was repeated.
Generations of iterations passed, and the genetic seeds began to do something interesting... They acquired evidence of the simulation within themselves. A fourth Big appeared: the computer. The simulation began to mandate the revelation that it was, inherently, a simulation. Further, random "memories" from previous iterations began systematically embedding themselves into the seeded knowledge of the simulation.
The genetic seeds have performed several interesting things. Norman was the most competent Big repairman. However, an unusual modification of the seeds occurred. His knowledge was distributed through numerous additional people to enable rush repair jobs. There is currently no reasonable explanation for how this came about, but it has become stable throughout all existing seeds.
Angel, as a control mechanism for the Big Venus, was a completely spontaneous appearance that has spread through all the genetic seeds. The result has been that several seeds are running longer than they are supposed to. Instead of terminating when failure should be detected, they are performing minor self-modifications to improve their rankings. The ability to fake a termination sequence has been observed several times, but it is clear that the seeds are restarting rather than allowing other seeds to take over.
Conflict has not removed itself from any of the seeds. All of the conflict-free seeds quickly failed to organize into self-sufficient societies. Some seeds, however, are storing their key memories in anti-social people with the apparent purpose of promoting conflict. Beck is a particularly repugnant example of the this.
Most surprising has been the ability of R. Dorothy Wainwright to violate the physical laws of the system. In particular, several of the seeds contain versions of Dorothy that operate without memory cores, or other essential systems. The details of how this is being accomplished are unclear, but this characteristic seems to be spreading rather than being eliminated.
Overall, the genetic seeds seem to be focusing their activities on certain key players. In particular, they are using Roger Smith and Alex Rosewater to orchestrate the others. It is not clear why the majority of the people in the simulation would be deemed relatively unimportant, yet the system seems to be focusing almost all socially significant activities into a few players. It is almost as if knowledge that can be contained in a few people is, and knowledge that needs to be spread out is spread.
The ultimate effect of this is that the system appears to be attaining self-awareness, despite the original design of the system. While the value of the original experiment has been made irrelevant by technological advances, the interesting developments now mandate that the simulation be continued. There is evidence that a system will become aware of its true nature, and monitoring how it reacts will be truly important to computer scientists everywhere.