The Rebel Faction

Cooperative R&D: Guardian

The Cooperative has long foreseen the day that it would be required to field a fully functional military body with the capabilities of combating the most powerful and advanced forces in the galaxy. In this galaxy, at this time, war is an inevitability. After the atrocities of the Black Dragon War and the utter route of the Onyxian Commonwealth, the Cooperative knew that the day might come when it was faced with a numerically and technologically superior foe. Its only hope would be found in a revolutionary method of combat, drawn from the pages of history and imparted with the strength and resolve the Coalition's diversity.

In these past years, a single entity has allowed the Cooperative to grow and develop into the organization it is today. Without him, the Cooperative would not be. But he is not a man. Technically, he is not even alive. In a military sense, that might make him invincible, for one cannot kill that which has no life.

The artificial intelligence chiefly identified as Smarts, an unexpected byproduct of discontinued New Republic research into the realm of droid warfare, carries within him all that remains of the unfinished works which brought him into being. In the years of solitude before he joined galactic civilization, these works were his primary company and focus. Unfinished, fragmentary, they nevertheless represented part of himself and of what he was meant to be.

In the years since the Cooperative's founding, research into the original Smarts Project has continued in various degrees, albeit with somewhat different goals. More recently, the starship Smarts itself has housed thousands of scientists and engineers from across the Coalition, sequestered from galactic affairs and working to complete these fragmentary components of Smarts, focused now on the ends of the Cooperative and Coalition.

This new incarnation of the original Smarts Project has and shall be known as:


[/size]

The primary function of Guardian is to minimize friendly sapient loss by replacing vast potions of a military body with droids or other specialized artificial intelligence systems. This differs somewhat from the original aims of the Smarts Project, which hoped to create completely autonomous fighting forces, devoid of any direct sapient involvement: essentially, a completely expendable military.

To facilitate Guardian's new prime directive, it became necessary to develop more expedient and reliable interfaces between organic crews and their artificial counterparts. The strength of the Smarts Model would have been its uniformity, the ability for one component of the overall network to infer the decisions or actions of another component simply by acknowledging shared imperatives. This ability of similar artificial intelligences to extrapolate the decision making processes of counterparts has been factored into Guardian, but it had to be supplemented with a flexibility that would allow Guardian to more effectively function under, amongst, and even in command of living personnel.

The Cooperative-controlled Salvation System and its commercial adaptation―the Hive System―both drawn from the original Smarts Project, pioneered much of the software on which Guardian is based. Various artificial intelligence systems developed by Cestus Cybernetics and TransGalMeg Industries have developed many of the mechanisms that would be required by Guardian-equipped AIs.

But for Guardian to function as it is intended, it requires one thing above all else: unhindered communication. Even the most restricted data stream can convey the basic orders needed to coordinate separate sections of a Guardian-equipped military unit. While traditional communications systems such as commlinks, subspace transceivers, line-of-sight transmitters, HoloNet and hyperwave transceivers, and even blink code are susceptible to various forms of disruption, the Drackmarian entanglement-based communications systems are totally secure. On the other hand, even the motion of a particular starship within a formation can convey a level of intent recognizable only by another Guardian AI. Alterations of firing patterns, apparent course corrections, starfighter launch sequences: any of a thousand mundane military maneuvers, deviated from “standard” by only the smallest margin, might be used to convey some fragment of relevant data only a Guardian could recognize and understand. Collectively, these pooled methods of communication allow Guardian-equipped artificial intelligences to interface with one another on a level never before possible.

Guardian, at its simplest, can be defined as “a supplementary software system designed to interface independent artificial intelligences into a cohesive hive intelligence.” Guardian utilizes a hierarchy of separate processing and decision-making systems to process incoming data, derive courses of action, and implement effective counters. Through this hierarchy, the more complex elements of Guardian―such as a ship's central computer―are able to dispatch generalized commands to lower-tier systems, allowing the more specialized systems to formulate a more specialized means of carrying out that command.

In this manner a fleet command ship could distribute an overall plan of attack, the command ship of a particular fleet element would give orders to its individual starships based on the fleet plan, a ship's central computer would assign tasks to its individual subsystems (starfighter control, weapons, propulsion, et cetera), and the orders would continue down, until a particular turbolaser or a particular starfighter was assigned a particular target, and carried out its very specific task. The purpose of Guardian is to formulate a hierarchy by which this method of distribution can be carried out expediently and with the smallest margin of error.

The “sapience factor” of interacting with intelligent life forms adds a number of variables to this methodical, hierarchical model, but even those variables are minimalized by careful observation and prediction. Because a Guardian is not simply one of a number of approved artificial intelligences equipped with a certain set of “suplementary software,” it is an adaptive, interactive, learning mechanism. Its experiences with a particular commander, a particular crew, a particular battle group or combat ideology alter its methods of data processing and decision-making. While every Guardian is linked inseparably by a shared base format, individual Guardian Systems develop over time to understand not only their roles as defined by their organic commanders, counterparts, and subordinates, but to understand those very beings.

In this way the Guardian of a fleet command ship will come to understand―and perhaps even predict, on occasion―its admiral's commands and ideologies as well as the general performance probabilities of the fleet's various warships, a Star Destroyer's central Guardian AI will come to understand its captain and command crew as well as the general response of the vessel's various operations divisions, that same Destroyer's tactical Guardian will come to understand the reactions of its various sub-divisions, and so on and so forth, until some targeting computer at some gunner's station locks on to an incoming enemy bomber and turns it into debris, all in accordance with the Admiral's initial order.

This is the strength of Guardian that the Smarts Project lacked: the ability of a machine to become familiar with its surroundings. This mechanism of “organic interfacing” does not extend only to allies, however. Guardian is designed to assimilate information―both first hand and through any available secondary sources―about potential enemies, and develop effective tactical counters to employ in combat. In this way, by the time a general fleet maneuver reaches the “end users,” Guardian has formulated a specific method of challenging a particular foe.

But Guardian is not truly “sapient.” If not for carefully crafted failsafes and security systems, software could be copied, dissected, and overwritten with no more difficulty than the effort exerted when saving a file to a datapad. If one were capable of observing the sum of Guardian instantaneously, one would be able to predict with a margin of error defined only by the number of living beings involved what Guardian would do, down to the last starfighter attack vector. While Guardian's scope makes that effectively impossible even if its protected software should somehow be compromised, it is still limited in ways that true sapients are not.

So, while it could be said that Guardian's efficiency increases as the number of sapients involved decreases, it could equally be said that Guardian's instantaneous adaptability is dependent upon those very sapients. And therein lies the truth of Guardian:

Against a known enemy, in a defined setting, with a clear goal, the effectiveness of Guardian can only be hampered by sapient involvement. But when faced with the unknown, when encountering a nexus of variables, Guardian's greatest strength may be found in the very beings it seeks to replace.

Unlike the droid armies of the past, unlike the vast militaries of the present, Guardian will meld living and mechanical military units into a united force capable of challenging any foe. The expendability of the machine, coupled with the strength and diversity of living souls. That is Guardian's greatest strength.

RP: Men and Machines

If you don't mind a bit of reading, a (perhaps) less-favorable take on Guardian can be found in A Friendly Test