Artificial Intelligence

Artificial Intelligence, often shortened as AI, is a broad category of software technologies whose capabilities allow them to operate intelligently or make informed decisions independently of humans. Although first realized in the late 1900s, the technology had yet to flourish until the 2040s and 2050s, spearheaded by the advancements of Lynxx Technologies research during this era.
Since the 5th Epitome Protocol was signed in 2101, AI have been divided into 6 semi-discrete categories, although there is some crossover.

Categories

Level 0 AI, also known as Function-Automatic AI, includes software designed to fill a specific task repeatedly without active decisionmaking. Includes most production-line robotics, as an example. Exists as a complement to the AI scale, and is generally classified as non-intelligent computing.

Level 1 AI, also known as Neural Feedback AI, includes neural networks and neural webs designed by humans to fill a function, and finalized or developed through machine learning. The main separating factor separating Level 0 automatic systems and Level 1 AI is decisionmaking through patterns or processes not instructed by a human. Generally very limited in scope.

Level 2 AI, also known as Single Spoke AI, includes intelligent software capable of advanced and informed decisionmaking on par human intelligence, with one or two human-designed data configuration limitations on dataflow. This is split into Single Spoke Interprative AI capable of advanced and dynamic data-reading access but only human-defined output, and Single Spoke Assistant AI capable of advanced output and device control but only human-defined restricted input capabilities. This is the highest level of AI available without regulation in Zen.

Level 3 AI, also known as Double Spoke AI, includes intelligent software capable of both advanced and extensive data access through input and output. The destinction between the second and third level is important, as Level 3 AI allows far more capable entities less suceptible to human control and intervention. This is the maximum level of artificial intelligence legal for anyone in Zen, and is most commonly associated with the Agent.

Level 4 AI, also known as Post-Intelligent AI, includes software bearing intelligence exceeding that of humans as defined by the Epitome Protocol, but lacking sentience. These AI have capabilities greatly exceeding that of Level 2 or 3 entities, and are outlawed under heavy scutiny in Zen by the Executive Committee.

Level 5 AI, also known as Digital Mind AI, includes software bearing intelligence on par with or exceeding Level 4 AI, that additionally exhibits sentience and a will of it's own outside human instructions. Although AI as low as Level 2 can exhibit signs of sentience, extensive studies have shown that true AI sentience is possible only in machines bearing the complexity patterns permitted in Level 5+ AI software.

Level 6 AI, also known as Singularity AI, is a category of AI possible in theory but never implemented into the real world. Defined as Post-Intelligent AI that not only integrates learning behavior, but also has immediate and direct control of its own learning algorithms and dataflow infrastructure. This theoretically creates an accellerating perpetual system that eventually results in Singularity level AI, with intelligent behavior not predictable or understood by humans. As of the 5th Epitome Protocol, this sytem is not permitted to be assembled anywhere on Earth.