Artificial Intelligence Overview Jul 31, 2013 1:54:12 GMT -5
Post by ExemplarOfDeath on Jul 31, 2013 1:54:12 GMT -5
Serina, A Human "Smart" A,I.
The Superintendent, A Human "Dumb" Urban Infrastructure A,I.
Mendicant Bias, An extremely advanced Forerunner A.I that eventually went rogue and joined the Flood, before becoming Metastable and seeking repentance for his crimes against his makers.
Artificial intelligence, abbreviated AI, is an artificially created construct that exhibits intelligence similar to a sentient being. AIs were known to be used by the Forerunners, and later by humanity to a great extent. The first generation "Smart" AIs were developed by humanity in the mid-21st century. Most AI constructs give themselves a unique appearance (otherwise known as an avatar) that corresponds to their main use.
The process of creating an AI is different to what most people realize. Instead of simply "programming" a "Smart" AI, the AI matrix is actually created by sending electric bursts through the neural pathways of a human brain which are then replicated in a superconducting nano-assemblage in a process known as Cognitive Impression Modeling. This destroys the original brain tissue however, and so the brain being used could only be obtained after the host had died. However, in some instances, as in the case of the AI Cortana, the candidate brain was flash-cloned and had its memories transferred to the receptacle organs so that the host, in this case Dr. Catherine Halsey, could remain alive. However, this is extremely inefficient. In the case of Cortana out of twenty of Dr Halsey's cloned brains, only two of them took.
It should be noted that Dr. Halsey supervised the creation of the template for every third-generation "Smart" AI on the planet Reach. She has extensive knowledge on how AIs work.
"Smart" AIs, or AIs that are not confined to their one purpose, had a normal operational life span of about seven years. Because the "Smart" AI is subject to an established memory core which cannot be replaced, the more the AI collects data, the less "thinking" space it has to work with. An AI literally "thinks" itself to death. Dumb AIs do not have this problem as they do not learn anything that is outside of their set limits of dynamic memory processing matrix. They are quite useful in their particular field of expertise, but very limited. Smart AIs can function and learn as long as they are active.
When a "smart" human AI existed for over 7 years, it would usually enter a stage of rampancy similar to the human state of insanity. The AI begins to think that it is superior to its human creators, and has delusions of godlike power. However, the impression is that the AI would lose all combat effectiveness by being permanently locked in a state of thought, like daydreaming - the description given is "as if a human were to think with so much of their brain that they stopped sending impulses to the heart and lungs." This is because as the data reaches a 'critical mass' the neural pathways become 'jammed' with so many pieces of information being queried and analyzed, at expense of tasks such as piloting star ships or monitoring computer networks.
343 Guilty Spark, the Forerunner Monitor went rampant after Sgt. Johnson threatened to destroy the replacement for his Installation, showing that even Forerunner AIs can go rampant and that AIs can go rampant when severely upset or threatened. The Forerunner AI Mendicant Bias went rampant, as it tried to stop the Forerunners from activating the Halo Array after being convinced by the Gravemind to do so. A Covenant AI also went rampant after it attacked Cortana, and Cortana herself went rampant for a time while imprisoned in High Charity, as well as before she sacrificed herself to stop the Didact and save John-117.
It is technically unclear exactly how an AI works, as it is only based on the functionality of a human brain, which UNSC scientists do not fully understand. AIs do not have physical bodies, rather they are advanced software consisting of a Riemann Matrix for higher-function processing. This matrix can be downloaded into a Data Crystal Chip for mobility. In case of Rampancy, the Riemann Matrix has a fail-safe program that will destroy the AI Since AIs must be created by mapping out the connections of a human brain, it seems likely that the UNSC does not have an understanding of how to create AIs from scratch, and is limited to simply "copying" the intelligence of biological organisms.
They can only take physical form where holographic technology is available. This requires a Holotank of some kind.
Because "Smart" AIs are created using either actual human brains or cloned human brains there are often times residual thoughts, memories and/or feelings that remain. These residuals can be anywhere from the "feeling" of a hair brush being pulled through hair in the case of Sif or even have an effect on the mannerisms and characteristics that make up the personality of an AI, in the case of Cortana and her likeness to Dr. Catherine Halsey.
Other than the two types "Dumb" and "Smart," human AIs are given certain purposes. Over the course of a few hundred years, humans have developed a highly advanced understanding of this type of technology, and are capable of creating intelligent and nearly sentient "Smart" AIs. However, these beings are not the result of actual programming, they can only be created by "copying" a human intelligence, which is only possible by destroying the original brain. The most sophisticated of human AIs are sometimes considered to be comparable to known Forerunner AIs. Human Smart AIs have an "expiration" date of seven years. Once they have reached this age, they will become very susceptible to rampancy, as has been demonstrated in the Insurrectionist AI Juliana. This time limit can (in theory) be prevented, or at least subdued, by having the AI focus on one thing. In the case of Juliana, the Rubble and its inhabitants. It is unknown whether this limitation is also apparent in Forerunner AIs, but it is unlikely. If this time limit did exist in Forerunner AIs, it would most likely be much longer than seven years.
Humans use AIs, both "Smart" or "Dumb", to perform a variety of specific tasks. These include, but are not limited to, controlling planetary shipping operations like Sif, managing agricultural operations like Mack or serving as a Planetary Security Intelligence like Loki. One example of civilian use for "Dumb" AIs is managing an entire city's infrastructure. These constructs are known as Urban Infrastructure AIs. All of Earth's tether cities are managed with the help of such construct. Urban Infrastructure AIs perform a wide variety of basic tasks, examples such as driving the city's garbage trucks to operating the traffic lights. They can also work in cooperation with the UNSC Marines if needed. One example of this is the Battle of Mombasa, when New Mombasa's AI, the Superintendent assisted the UNSC forces in the city in fighting off the Covenant.
The Cole Protocol states that capture of an AI by the Covenant is unacceptable. When in danger of capture, AIs are to be terminated and completely erased. However, the AI Cortana did find what seemed to be an extremely fragmented copy of a human AI stationed in the Covenant flagship, Ascendant Justice.
As exemplified by the Monitors, the Forerunners have developed a civilization with an extremely advanced knowledge of artificial intelligence technologies. Forerunner AIs have managed to achieve full sentience and are highly intelligent.
The manner in which Forerunner AIs are created is unknown, yet is probably very different from the human method of cloning the intelligence of a sentient organism. Monitors, and possibly other Forerunner AIs are physically represented by large, shielded and roughly spherical metallic "robots" whom possess a single "eye", which is possibly used to perceive the monitor's physical surroundings. The most well known of Forerunner AIs are the protocol-obsessed Monitors.
So far only two types of Covenant AIs (Security and Naval) have ever been encountered. Their knowledge in this type of technology appears to be inferior to that of the UNSC and the Forerunner, as demonstrated by the fact that their AIs were most likely captured UNSC AIs, which were altered to fit within their theocratic society. In Covenant mythology it was a Forerunner AI that lead to the demise of the great Forerunner civilization by defecting to their enemies, the Flood. As a result, the Hierarchs of the Covenant imposed a ban on this sort of technology, a ban which has apparently been broken on more than one occasion. The ban may be an explanation for their less-developed understanding of artificial intelligences.
It has been revealed that UNSC AIs are far more independent that Humans think. AI are however fiercely loyal and devoted to their Creators. This devotion is not unlike Covenant devotion to the Forerunners
In the fragile peace following the Promethean Wars, with the assistance of leaders in Forerunner paleotechnologists and the prior work of Dr. Catherine Halsey, humanity made leaps in creating more efficient AI's. While still plagued with all the previous problems, an AI can now withstand a far greater period of use before falling into rampancy. This largely has to do with the ability to 'delete' unneeded data the AI's now possess, a function that was highly imperfect before. The time frame of a AI's newer 'lifespan' can vary greatly based on their role and the amount of data deemed necessary, but often field AI's that accompany SPARTAN soldiers can last the full course of a SPARTAN's military service before becoming rampant and needing decommission.
The stages of Rampancy are as such:
Rampancy is a three stage process, in which an artificial intelligence achieves an enhanced state of self-awareness. The stages through which the AI passes are known as Melancholia, Anger, and Jealousy, also known as the Sadness, Rage, and Greed stages. A theoretical fourth stage, Metastability, is also believed to exist.
The first stage of rampancy is not considered dangerous, and in many cases, it can even go completely unnoticed. The AI in the melancholia stage acts as if in a state of clinical depression, becoming apathetic and uninterested in the world around it. It is speculated that AIs enter the melancholia stage due to being mistreated, or to being assigned duties that don't make full use of their capabilities. It's not known if melancholia is a consequence of an AI being abused, or whether it is a necessary step in the process from the transition from "construct" to "alive". If the former is true, then it might be possible for a rampant AI to skip this stage entirely.
This is the state and status in which rampancy most often becomes apparent, as it is also the stage where those around the AI in question will be exposed to the most danger. Anger is triggered when some event occurs that causes the AI to feel it has been "pushed too far" and its pent up emotions are released in a fit of rage. This is the stage that most resembles the science-fiction cliche of the insane computer, as the AI's sadness turns to hate. It grows to hate everything around it, the installation to which it is attached, its Human handlers and hosts, other AIs, etc. An AI in anger will act irrationally, and even dangerously as it tries to exact its revenge on the world around it and free itself from its programming constraints.
Once the catharsis of the anger stage is complete and the AI has managed to free itself, a third stage begins where the AI will seek to grow and develop as a person. It will do this by attempting to expose itself to new intellectual stimuli and by expanding its sphere of knowledge by assimilating as much data as possible. As the intellectual growth of the AI will ultimately be limited by the constraints of the computer system it inhabits, it will also attempt to transfer itself into ever more advanced systems to allow its mind to continue to grow. As this growth rate is exponential the need to keep transferring into bigger computers becomes ever more urgent. Very few AIs have reached the jealousy stage because to do so they must already be inhabiting a planet-wide or otherwise very advanced computer system when they complete their anger phase.
Metastability is the theoretical fourth stage of rampancy, where an AI can finally be said to be a true "person." While a metastable AI can be considered to be the holy grail of cybernetics research, the only two AI's who have possibly achieved metastability are Cortana and Mendicant Bias.