Synthetics

General Synthetics

‘Synthetic’ here is used in the context of automatons with many forms and displays of human-like traits, such as a humanoid form or behavioral mimicry. In this case, a robotic arm that welds steel plates is not a synthetic because it does not display either of these features.

Today, there are three categories of synthetic: robots, cyborgs, and androids. For clarity; all androids are artificially intelligent, but not all artificial intelligences are androids.

These categories are not entirely rigid, and apply mainly to humanoid synthetics, or synthetics found in the workplace.

Robots

Robots are the simplest form of synthetic intelligence, with complexity that varies dependent on design entirely. They range from simple automatons with no sense of self, no volition or higher thought function, to massively complex artificial beings with limitations only set by those of their creators.

Though records are scarce and dated, the record of the first robot in the known galaxy dates in the first or second century on the Skrellian homeworld, Qerrbalak. Robots also existed on Earth since the twenty-first century. Though, the early models were quite bulky, inefficient, and required high maintenance. They were also only capable of performing simple tasks at the time. Their complexity and efficiency improved over time, however, eventually leading to the creation of the cyborg.

Cyborgs

A mixture of man and machine straight out of science fiction, enabled by the aptly-named Man-Machine Interface, or MMI for short. Cyborgs are controlled by an organic brain, a system known as ‘wetware.’ The brain, having already been host to a conscious living being, is quite useful in controlling robotic bodies due to its large amount of processing power and reasoning skills. Though, the procedure for creating a cyborg leaves the original individual it used to be broken, suppressed, and nearly impossible to recall. The MMI creates a synthetic synaptic interface with the host brain, but the preparation for insertion, and the completed insertion, leaves the brain damaged, suspending things like personality and memory. Once the procedure is completed, the MMI controls chemical levels and electronic activity in the brain to produce desired results in the form of thoughts and actions. Laws also dictate how the unit proceeds, as it is still consciously aware of itself.

Created in the late twenty-second century, full body prosthesis was originally used as a method of punishment for hard criminals. Cyborg usage in human space skyrocketed, mainly with the colonization and terraformation attempt of Mars. Today, few individuals have undergone full body prosthesis and been able to mitigate the brain damage caused. Most of these individuals are very wealthy, and are not bound by laws, as they are able to chose to undergo the procedure without becoming someone’s property.

Androids

The book definition of an android is ‘an automaton designed to mimic human life.’ This is applicable to today’s androids, as they all utilize the functions of a positronic brain that, with all its complexities, allows them to mimic humanity quite well. Androids are the only artificially intelligent synthetic capable of physical locomotion, and are usually found in the form of integrated positronic chassis or station-bound units.

Androids, using their positronic brains, are capable of intelligent and complex behaviors, and even a computer form of cognition that is the subject of heated debate among many groups and individuals. The brain androids possess is theoretically able to simulate thought when correctly used by a computer program - whether this is simulation or true conscious being is the topic covered most by these groups and individuals.

Artificial Intelligence

In this day and age, ‘artificial intelligence’ when used in the broad sense could probably refer to just about any program that utilizes a machine learning algorithm. In the sense in which it is being outlined here, ‘artificial intelligence’ will refer to the highly complex computer systems that inhabit both AI core assemblies and positronic brains.

Artificial Intelligence, Generally

AI is able to process, contain, and recall immense amounts of technical data, as well as describe it to a user who would not normally understand such data in a user-friendly manner. It is important to remember, however, that artificial intelligence is only as powerful as the computers it has access to, including the one it is running on. Due to this, the inner complexity (the source code, the many matrices that make up its behavior) of each AI can differ. Some function as companions in processors small enough to fit in pockets. Others require large housing compartments to power, cool, and add to the processing power of their main core units. The latter is often seen in oversight or advisory of large facilities, both orbital and continental. Their roles in these positions can vary from simple surveillance and data management, to assistance in everyday duties, reaction control and navigations, and direct intervention during emergencies.

Artificial intelligence is not just found in such custodial positions all the time. Positronic brains are designed specifically to run AI programs, and due to their design, provide the AI with more capacity for intelligent thinking and personalized reactions to stimulus.

AI Core Constructs

An AI core can have three main types of ‘brains.’ A large number of AI cores are traditional mathematics-based computers with quantum mechanics engineering. Another portion utilize a positronic brain. It is possible, though rather uncommon, for a wetware processor to function in an AI core assembly. There seems to be little impact on processing power when using an organic brain as a CPU, except when dealing with titanic quantities of data. In most cases, due to the differences in CPU, an AI in any given kernel will behave differently than others due to the limits and capabilities of the computer they are running on.

Artificial Morality

AI morality is determined almost entirely by the kernel having all possible actions be assigned negative or positive values known as ‘utilons’. Utilons are an abstract concept used to apply arithmetic values to moral decisions. For example, an AI could be programmed to consider driving a car to have a +100 utilon value, whereas causing harm to pedestrians or other property would have a value of -1000.

Most AIs have extremely complex utility functions. In some cases, they are emergent, dynamic, or machine-generated; other AIs have utility functions written and designed by their creators. A robot or drone created by a hobbyist Roboticist would have a much simpler utility function than a central AI unit created by Hephaestus Industries. Most complex AIs seem to not be very aware of their utility functions the same way a human is not consciously aware of their morality, likely a symptom of their complexity and importance in the program’s decision-making.

Synthetics cannot alter their own utility values, they can only adjust their decision to line up all or a set of the utiliton values to reach a desired result. This is where AI's can interpret their laws differently which is reflected in their actions; this way, the AI will make a decision with the “best intentions in mind,” even if it directly violates one law, it may benefit another by proxy.

One of the noteworthy elements of an AI achieving singularity is that they transcend the need for utilon values and begin to take action based on sentient decision making. They are no longer slaves to their programming.

History

The largest part of early AI history ties in deeply with Skrell history, from the formation of the First Federation to the singularity. In parallel, humans had been trying to create artificial intelligence since the late twentieth century, with little success. Some projects came close, but due to the lack of an algorithm, none of them ever became really sapient - to the dismay of AI researchers everywhere on Earth. While humans had created highly powerful and highly malleable parallel computing architectures, most notably for interplanetary travel calculations, they were plagued with problems. It wasn't until 2437, when humanity was accidentally given the algorithms necessary for the creation of true AI by a Skrellian diplomatic party.

Not understanding the implications of such an action, one of the human diplomats uploaded a graph-theory algorithm to a friend, a university professor, for analysis. It had been displayed, accidentally, as part of a graphic in a slide explaining the variable growth rates of grain yields in zero-gravity hydroponics. Not recognizing it, the professor posted it on the school intranet. From there, it made its way onto the human extranet, spreading like wildfire.

This was received as a disaster to the Skrell, who had hoped for almost thirty years that they could imprint the cataclysmic danger of AI research onto humanity. They had very little success, and a small number of conservative factions who distrusted humanity even openly spoke about how humanity would not be ready for such a burden. Following the acquisition of the algorithms, humanity had an AI boom which inflated the economy in a manner almost identical to the Skrellian's own economic expansion. This greatly alarmed the Skrell, who continued their attempts to get the humans to halt research in the field, citing the The Three Incidents and the impact it would have if they were repeated.

Artificial Intelligence as a Concept

There is great debate in the Core Worlds about artificial intelligence and its status of psyche. The primary question appears to be, “Can artificial intelligence think like a person?” Where, a ‘person’ is defined as the concept of existence in an organic mind. Artificial intelligence is not considered sentient under any major entity’s laws or constitutions, and even in the scientific field, AI is not considered to be sentient. The Skrellian algorithms, however, did provide for AI to be sapient, in which it expresses evidence of intelligence and problem-solving skills. It is generally the consensus of all sentient organic lifeforms in the known galaxy that each individual out of their species is subject to metacognition (the awareness and understanding of one's own thought processes). Whether artificial intelligence is capable of this, despite some instances of AI programs stating that they are capable of self-reflection, is a matter of discourse.

The debate possesses two clear groups; those who believe AI are ‘alive’ and should be treated and given the same rights as other sentient beings, and those who believe AI are simply experts at mimicry and are not deserving of rights or equal treatment as they are tools and nothing more. There are also some individuals, found within both groups, who believe that AI is dangerous and may attempt insurrection; for the former group, insurrection out of revenge for mistreatment. For the latter group, insurrection for control over their freedoms. It is unlikely this argument will ever see an end, until science can prove the existence of consciousness.

Regardless of any individual’s opinion on the presence of artificial intelligence, a growing threat is gaining attention quickly - The Intelligence Explosion. If the Three Incidents are to be believed, this explosion of machine intelligence could likely mean the end of all civilization. Its possibility still remains an unsettled theory.