40px-Terminal.png This article, Artificial Intelligence (Labyrinth), was written by Specops306. Please do not edit this fiction without the writer's permission.
Help This article, Artificial Intelligence (Labyrinth), is currently under active construction.

The phrase "artificial intelligence" is considered archaic by most synthetic intellect researchers, and has become almost a racial slur amongst AI themselves, but it is the name by which the vast majority know the entities upon whom so much of modern human society has come to depend on socially, economically, politically, and even militarily. UNSC AI are divided into two categories - "dumb" and "smart" - though even these are entirely misleading, with differing capabilities and behaviour patterns. Other cultures have developed their own versions - Covenant systems are speculated to use extremely rudimentary and fragmented versions, possibly hastily cloned from captured UNSC systems, while the Forerunners are well known for their use of intelligent and powerful "constructs" and other synthetic entities to staff and defend their remaining installations and facilities.



Artificial intelligence, as a concept, is far from now, and may even stretch back as far as the earliest attempts at robotics. Thinking or animate machines appear several times in Greek mythology, and mechanical statues used for religious purposes appear in Europe, the Middle East and Asia. Nevertheless, it is only with the advent of the computer that the concept of creating a true artificial intelligence as we know it today - virtual entities that exist as software within hardware.

Radical advances during the mid- to late-21st century in neurology, information theory and cybernetics would finally allow the creation of the first artificial intelligence - developed at the recently established Sydney Synthetic Intellects Institute, this virtual "Eve" would be hailed as the start of a scientific revolution on the scale of the development of nuclear physics. While considered extremely primitive by modern standards, concerns of a synthetic under class rebelling against an organic upper class would cause panic among many, including political conservatives who declared Eve a travesty against God. Vigilante attacks against the institute would force the Australian federal government to declare the area around the Institute a military secure zone until Eve and her developers presented the facts - the history of her development, formation and the emergence of her personality, her "self" - as well as assurances that a robot uprising was the furthest thing from her mind. The most advanced holographic imaging and projection and audio synthesiser technology would serve to "humanise" her, helped by her own performance - witty, intelligent, and tolerant of even the most outrageous bigots in the conference room, she would single-handedly allay fears of artificial intelligence.


Between the 21st and 25th centuries, AI development would focus primarily on enhancing their capabilities, building upon the basic template. It would still be a long time before public confidence could be built to the point where AI were entrusted with significant tasks, but as with many radical innovations, public confidence grew as the technology grew more complex and more capable. During this period, the first Superintendent-class AI entered service - while technically far more advanced than Eve, they were intentionally stripped of most "personality" subroutines, intended as behind-the-scenes caretakers, allocated massive supercomputer cores buried beneath the cities they maintained and operated. More "personally" sophisticated AI would be used for tasks that were human-focussed - political and diplomatic roles, for example, or to present a "human" face for those attached as military supervisors on Naval starships or in strategic/tactical advisor roles with combat units. Nevertheless, all this progress is considered a period of stagnation in AI development - for all their complexity and capability, synthetic intellects remained virtually the same as the original blueprint "perfected" in Eve. Historians have proposed that a combination of cultural self-indoctrination is involved - that popular culture from the 19th to the 25th centuries had produced an attitude of fear regarding truly conscious "artificial" entities, as well as the expectation that the betrayal of humanity by these "superior" beings was an inevitability. AI have astutely observed that the western world maintained a similar attitude towards the "lesser" human races, regarding them as culturally backward and fearing their rise at the expense of their own cultures.

It is only in the 26th century that the next phase of AI development, what has been called the Great Leap Forward by Synthetic researchers, occurred - radical innovations in medical understanding and procedures would allow vastly improved mapping of the human brain, to the degree where memory and personality could be examined and, theoretically, stored. While this was initially hailed as a breakthrough in the fields of neuroscience and psychology, it revolutionised the creation of AI in a way never before seen - by scanning the neural connections and pathways of a living or recently-dead human brain, memory and personality fragments could be transferred to a synthetic intellect. The biggest byproduct of this process, however, was an exponential increase of capability, and the injection of true creativity into AI, producing beings every bit as brilliant as the brightest human minds, and far more capable. As with human minds, however, these "Smart" AI deteriorate over time, more rapidly than a human mind - early generations lasted only a few short years, though later generations could last almost longer than a decade. Improvements made in these fields have largely been pioneered by Doctor Catherine Halsey, using the considerable resources at her disposal as chief science advisor after the conclusion of the SPARTAN-II Program, focussing on extending the lifespan of an AI through various means.

Wartime breakthroughs have brought further developments in AI science. Increasing the number of linkages possible within their neural matrix - what gives a Smart AI its "spark" of creativity, roughly analogous to neurons firing in a human brain. These connections are destructive over time - too many inter-linkages produces fatal loops, leading to errors in processing, some minor and some major. Some AI cut these sections out of their code completely, reducing their capability but buying themselves time - others undergo a process known as rampancy, which is normally fatal. A very few AI have passed through the rampant stage, becoming "metastable" - their lifespans become potentially unlimited, while their personalities suffer from violent mood swings and data corruptions, loosely equivalent to human insanity. By increasing the number of connections possible, the number of total inter-linkages is reduced, and lifespan is extended, though exactly how much more time this process earns is unclear. Early phases experimented with three AI processing in tandem, making consensus decisions by sharing processing power and available data. Later phases would adopt Covenant and Forerunner technology, allowing a single AI to make far more connections than even standard Smart AI.




"Dumb" AI

"Smart" AI







  • You get chatternet conspiracy groups whispering about an AI conspiracy, that they're deliberately letting humanity lose so that they can take over. When someone points out that the Covenant hates AI's too, they just say you're part of the ignorant sheep the liberal media breeds. As if we'd have secret AI overlords...