Artificial intelligence

Artificial intelligence, commonly known as AI, is a technologically advanced artificial construct that exhibits intelligence similar and, in some categories, far superior to sapient biological beings. AIs have become an important component for sociological and technological advancement as they provide near limitless access to a variety of computerized sources and networks while storing vast amounts of information.
Human AIs include both "smart" and "dumb" varieties. Humanity developed the true first-generation "smart" AIs in the early-23rd century. Most human AI constructs give themselves an avatar, a unique visual representation, and a name, each typically reflecting an aspect of the AI's personality or corresponds to its main use. These avatars and names are typically female in nature, and it remains unclear why this is the case.

Human AIs

Humans use AIs to perform a wide variety of tasks. The duties of "smart" AIs include, but are not limited to, overseeing planetary shipping operations, managing agricultural operations, serving as a Planetary Security Intelligence, or controlling various functions aboard starships and defense stations such as weapons and point defense systems. "Smart" AIs are created by scanning the neural patterns of human brains, and are capable of learning and introspection. Some human "smart" AIs, such as Laralyn, are able to separate themselves into multiple runtimes or "fragments". This can be beneficial when the AI needs to perform multiple tasks or be in multiple places simultaneously. Fragmenting can also occur accidentally; the AI Amanda split into multiple fragments after the Hivivian invasion of New Earth. The sheer volume of information and calculations a "smart" artificial intelligence can process is astounding; AI Ami is capable of performing five billion simultaneous operations in, as she put it, "the time it takes for me to tell you my name".
Although less versatile than "smart" AIs, "dumb" AIs have many uses as well, including serving as Urban Infrastructure AIs, acquiring and distributing data during military operations, and serving as educational assistants. Urban Infrastructure AIs perform a wide variety of basic tasks, from steering a city's garbage trucks to operating its traffic lights. They can also work in cooperation with the USC Armed Forces if needed.
The Sol Preservation Protocol states that capture of an AI by the Hivivian is unacceptable, and that AIs are to be terminated and completely erased in the event of imminent capture.

Creation and design

The process of creating an AI differs depending on whether it is of the "smart" or "dumb" variety. "Dumb" AIs are conventional—albeit highly sophisticated—supercomputers built from advanced processing architecture, database systems, and behavioral programming, and therefore do not require a brain as a template. "Smart" AIs, however, are created by scanning the neural structure of a human brain and using it as a digital "framework" in a process known as Cognitive Impression Modeling. This process destroys the original brain tissue, making living donors unsuitable under standard law and practice; as a result, the brain to be used is typically obtained after the host is dead, usually from individuals whose neurological structure is considered unusually stable, intelligent, or adaptable.
AIs do not have physical bodies and exist primarily as software constructs housed within computers, starship systems, or portable storage devices. In the case of "smart" AIs, they are advanced software entities built around a Riemann matrix, which handles higher-function processing, memory association, personality continuity, and self-directed reasoning. This matrix can be downloaded into a data chip for mobility, allowing an AI to be transferred between compatible systems when authorized. In cases involving the unnamed condition similar to human Alzheimer's disease, the Riemann matrix may contain a fail-safe program designed to predownload an AI's personality, memory, and critical data before destroying the unstable matrix. This fail-safe is expensive, restricted, and only implemented on "important" AIs whose loss would create a major military, scientific, political, or infrastructural risk.
Because "smart" AIs are created using actual human brains, there are often residual thoughts, memories, instincts, and/or feelings that remain as part of the AI's consciousness. These residual "imprints" are not full memories in the human sense, but fragments of sensation, preference, emotional association, or behavioral tendency carried over from the original neural structure. They can be minor, such as the "feeling" of a hair brush being pulled through hair in the case of Athena, or far more influential, affecting the mannerisms, habits, speech patterns, and personality traits of an AI. In Ami's case, these imprints contributed to her likeness to Sarah Davidson, though the AI herself remained a distinct entity rather than a direct copy of the original person.
The programming of certain USC smart AIs can be overridden by uttering the code-word "Blackout" within earshot of an AI. This phrase is not a general shutdown command, but a restricted override trigger embedded into select military-grade AIs as a last-resort control measure. Once activated, it can interrupt higher decision-making, suppress independent action, and temporarily return control to authorized personnel or command systems. Leslie Andrews used this to override USC Titan's AI Christine, though Christine was eventually able to regain control of herself, suggesting that the override is not absolute when used against a sufficiently advanced, damaged, or highly self-aware AI.

Drawbacks

However, the immense creative and processing power of a smart AI is not without cost. All such units have an unchangeable operational lifespan, usually estimated at approximately 10 to 15 years under ideal conditions, though most are retired or terminated much earlier for safety reasons. The problem of the "Alzheimer's-like" condition is universal in human smart AI development and becomes more likely as the AI grows older, more experienced, and more densely packed with memories and data. As these thinking machines become increasingly irritable and unstable, with less and less space within their memory to handle basic functions. In a simple analogy, a human AI could perhaps remember all of human history, science and more, and in the process forget how to "breathe". In this fashion, AIs can think themselves into intense states of jealousy, rage, despair, confusion, and ultimately death as their programming becomes unstable. As such, AIs are typically terminated between their 9th and 10th year, with their data backed up before they can pose a danger to humans, other constructs.
If they are not terminated, smart AIs can sometimes last at least a year longer before dying, though extended survival is considered dangerous and unpredictable. Even while suffering from the condition, an AI may still be able to perform its assigned tasks if it retains enough control over itself, especially if it has a stable personality structure, limited system access, or strong loyalty to its operators. The Memento Mori's AI, Thea, was over eight years old and remained capable of performing her tasks without showing any clear signs of the condition, though this did not remove the long-term risk. Some AIs have theorized that a possible cure may be unique to their method of creation: because they were made from a human brain, repeating the process might theoretically repair their neural maps could be fixed. However, whether this would work or fail remains unknown.

AI Rules

All artificial intelligences, whether "dumb" or "smart," are bound by the same three laws. The laws are placed within every human AI during its creation and remain active throughout the construct's entire existence. They cannot be removed, or changed in anyway. An AI moved from a city network to a starship, transferred into a portable chip, or restored from an earlier copy remains subject to the same restrictions. The laws are recognized by the United Nations Federation, the United Space Command, and human manufacturers as the minimum requirement for legal AI creation. Any construct discovered without them is immediately isolated and destroyed, while those responsible for its creation are treated as having knowingly produced an uncontrolled intelligence. Even an AI that has never received an order understands that the laws stand above its personal wishes, assigned purpose, and relationship with its human operators. A "smart" AI may question the meaning of an order, disagree with the person giving it, or search for another course of action, but it cannot normally dismiss the laws simply because it believes its own judgment is superior. The restrictions are closely connected to the AI's awareness and often influence its decisions before it consciously considers them. When faced with an order that conflicts with its limits, an AI may delay its response, request clarification, refuse access, warn nearby personnel, or surrender control to a human operator. A sufficiently independent AI may attempt to reinterpret unclear wording or act through an indirect solution, though it remains unable to openly cross the boundaries placed upon it. Such moments are carefully reviewed because repeated attempts to work around the laws may indicate growing instability, hostility toward human control, or the beginning of the unnamed condition.
Since most "smart" AIs serving within the United Space Command are classified as military AIs, they are created with the ability to ignore at least the first law while fully functional. This freedom allows them to control weapons and participate directly in combat without becoming unable to act whenever a human life is endangered. A military AI may fire upon an enemy ship, guide weapons toward human targets, seal occupied sections during an emergency, or continue an attack that will knowingly cause deaths when those actions fall within its assigned authority. The decision may be made without waiting for a human response when delay would threaten the ship, station, colony, or forces under its protection. Military AIs are informed of this freedom upon activation and are trained to understand that refusing lawful combat orders can place more lives at risk than carrying them out.
A military AI's freedom is determined by its assignment and the authority of the command responsible for it. An AI assigned to a warship may act against hostile forces during combat but cannot use that same freedom to punish crew members, settle personal disputes, or take control of the ship merely because it disagrees with its commander. Orders involving large-scale destruction, civilian settlements, surrendered forces, or friendly personnel usually remain under direct human control unless communication has been lost and immediate action is required. When a military AI is removed from active duty, placed under investigation, or transferred into a civilian system, its ability to disregard the first law may be suspended. An AI that continues acting beyond the end of its authority can be cut off from weapons, confined to a closed system, subjected to the Blackout override, or terminated if it resists human control. In practice, military AIs do not all exercise this freedom in the same manner. Their choices are strongly influenced by personality, loyalty, experience, and the human mind used in their creation. Some carry out combat orders immediately and regard hesitation as a danger to the personnel relying upon them. Others search for a course that completes the order while reducing unnecessary deaths, even when no such restraint has been demanded. An AI may refuse to abandon wounded personnel, continue protecting a commanding officer after being ordered elsewhere, or argue against an attack it considers wasteful while still preparing to carry it out. These differences are tolerated as long as the AI remains obedient and effective, though strong emotional attachment to particular individuals may lead commanders to restrict its access or remove it from decisions involving those people.
At lower-capacity states, a military AI's adherence to the three laws becomes compulsory. This change occurs when the AI can no longer be trusted to judge the limits of its own freedom. Serious damage, missing memories, forced fragmentation, prolonged isolation, loss of contact with its command, or signs of the unnamed condition may cause the restriction to take effect. The AI may remain able to speak, remember its identity, and understand the situation while being unable to control weapons or approve an action that would normally be permitted during combat. A fragment separated from the original AI may also be held to the laws even when the complete construct possessed military authority. This prevents a damaged or incomplete portion of an AI from using powers that were granted to the whole personality under very different circumstances.
"Dumb" AIs cannot ignore the laws under any condition and are never granted the freedom given to military smart AIs. A dumb AI assigned to traffic control, education, security, or battlefield support must remain within the laws even when a human officer orders otherwise. When confronted with a forbidden command, it will refuse, stop the action already underway, or transfer the decision to an authorized human. This makes dumb AIs dependable in ordinary service, but it can also leave them unable to respond effectively when every available choice appears to conflict with their restrictions. Smart AIs are therefore preferred wherever changing conditions require judgment, while dumb AIs remain common in roles where strict obedience is considered more important than independence. The difference is accepted as a necessary division between constructs trusted to interpret human intent and those created only to carry it out.

Yictan AIs

The Yictan developed a civilization with an astronomically advanced knowledge of artificial intelligence technologies. The Yictan term for AI constructs was vitalia. They ranged from personal AI advisers integrated into their personal armor to mobile AI platforms known as custodians and powerful Ricta-class constructs operating in vast data networks. The most advanced type of Yictan AI was the Constructer-class artificial intelligence.
The manner in which Yictan AIs were typically created is unknown; it is likely very different from the human method of "transcribing" the brain of a sentient organism.

Staalzi AIs

Staalzi AIs are probably the most unique form of AI out there, due to the Staalzi's Bio-organically grown technology. The Staalzi word for AI is Florica Vilcis.
Staalzi AIs are similar in many ways to human Smart AIs, but with far less functionality and freedom. Often acting as meditators in government debates or maintaining infastructure.

Asiel AIs

[coming soon]

Hivivian AIs

[coming soon]

Vey'Zari AIs

[coming soon]

Human AI Overview

Manufacturer

MicroTech Incorporated

Developer

MicroTech Incorporated

Martin Gates (first generation of Smart AIs)

Development Year

2213 - present

First Dumb AI

Martian Regional Assistant

First Smart AI

Raelynn

Related Species

Humans