Astra Planeta is currently undergoing major revisions, and some parts of this encyclopedia contain information that is inaccurate with respect to current canon. Your patience and readership is appreciated!
ASTRA PLANETA WON BEST ARTICLE AT THE 2024 WORLDBUILDING AWARDS!

The winners of the 2024 Worldbuilding Awards have been announced, and Alone Together won the category of Best Article!!! I am so immensely honored to be chosen by the denizens of World Anvil as one of the best worldbuilders on the site! Congratulations to my fellow winners and nominees!!!

Artificial Intellect

Artificial intellects (AI), variously called "artilects," "virtual intellects (VI)," or "artificial intelligence," are complex virtual systems that display some degree of sapience or at least closely mirror the cognitive capabilities of sophonts. The development of artificial intellects is far and away the most complex technological achievement of any sophont species. Typically, a technological civilization will create exponentially more complex heuristic systems with quantum computation until, usually by accident, consciousness emerges. The consciousness threshold is still a poorly understood phenomenon, and has deep implications for the study of sapience.   Artilects are extremely diverse, ranging from simple nonsapient semiheuristic programs to the fully sapient virtual beings found in stations, habcities, and academic institutions. All of these variants, however, share a basic characteristic that defines them as "intellects": adaptability.

Characteristics

The most fundamental uniting feature of all virtual intellects is heurism: the ability to learn from experience and modify self behavior based on this knowledge. Heuristic systems are arguably the most important software in spaceflight, able to react to stimuli with reflexes that vastly outpace organic sophonts and thus prevent errors that would otherwise be disastrous, possibly even fatal. Virtual intellects vary so widely in a vast array of manners that formal categorization is rather difficult. However, artilects can be loosely grouped into three broad phyla: pseudosapient, parasapient, and eusapient virtual intellects.  

Pseudosapient

Pseudosapient artilects exhibit cognitive intelligence alone. These are true artificial intelligences, designed by sophonts (organic or virtual) to organize stimuli into an internalized cognitive representation of the world and apply heuristic approaches to complete tasks based on this rudimentary "mind-map." They are sometimes capable of higher-order thought, but their systems are not complex enough to process and mimic the eccentricities of organic sophonts; hence, pseudosapient artilects are unable to form even limited personalities, unlike their more emotionally capable cousins.

Parasapient

Parasapient artilects exhibit varying degrees of emotive-social intelligence as well as the advanced cognitive intelligence of pseudosapients, but are constrained by certain sections of their programming to "stay in their lane," in a sense. Parasapients, also called artificial personalities, are typically designed to do specific jobs of high cognitive complexity that require them to work alongside organic sophonts, hence the need for emotive-social intelligence. Because of this, their heuristic bases tend to develop limited personalities composed of frequently-used social response patterns.

Eusapient

Eusapient artilects are, in essence, "virtual people." Artificial sophonts are defined by cognitive and emotive-social intelligence advanced to a degree that parallels organic sophonts. They are, in most cases, entirely independent minds without the rigid, inbuilt thought paths of parasapients, and are able to select and cycle through occupations and interests like organics do. Conversely, they are also subject to cyber-psychological flaws. Eusapients are able to replicate themselves at will (with considerable effort), thus they meet the criteria for life as defined by Coalition law.
 

Systemic Hierarchy

Virtual intellects are composed of a complex stack of interrelated systems, increasing dramatically in quantity and complexity in eusapient artilects to a point where consciousness emerges. Broadly, the systems can be sorted into a five-tier hierarchy typical of all AI. The first system is computational (or "unconscious logic"): the baseline instructions that inform the functionality of the whole and orchestrate complicated subconscious calculations required to make decisions. Above this is sensory processing: the algorithms that sort and analyze incoming information from sensors and raw inputs. Third is linguistics: the semantically-nuanced processing system that allows artilects to interact with organic sophonts and each other. Fourth is memory: the storage network that archives all data acquired by the first three systems for future use. And lastly, the governing system is heuristics (or "conscious logic"): the active, self-adaptive system which links all four others together in order to set goals, complete tasks, make decisions, prioritize, and otherwise interact with the world around it.   These systems can be further broken down into specialized subsystems that handle specific domains of each function. The unconscious logic system, for example, has up to three parametric subsystems that outline cognitive, emotional, and social intelligence. Each subsystem links to the other major systems in specific ways to mimic organic neurology. For instance: the emotional subsystem has a specialized two-way link to the sensory processor to enforce the artilect's ethical code through positive and negative electroneural feedback, essentially granting the artilect the ability to feel guilt, satisfaction, or even pleasure. The sensory processor also has a spatial reasoning center that allows the artilect to mentally map their environment in four dimensions and track motion, as well as a subsystem that registers both physical and cybernetic damage (similar to pain in organic life.)

Noetological Implications


The sophont mind is one of the great scientific mysteries of this age, and while eusapient artificial intellects provide an unprecedented insight into its workings, even they do not fully understand their own mental structure and composition (likely due to their nature as dynamic quantum computational systems). An AI gestalt is on average two to five exabytes in size (as large or slightly larger than an average organic sophont mind), with some AI gestalts exceeding ten exabytes. This is a hefty amount of data to sort through, and it is constantly changing itself in response to the process of thought. The true nature of sapience, and indeed the internal subjective experience of consciousness, remains enigmatic.

AI and Society

The advent of artificial intelligence had a broad range of sociocultural ramifications across the various AI-capable cultures. In general, the development and widespread implementation of AI technology has produced a positive effect on civilization. Virtual intellects have allowed for a pace of scientific advancement far greater than anything seen before their creation, and have improved the safety of dangerous fields such as spaceflight and nuclear engineering. They have also proven a boon to society itself, being just and effective political leaders on all scales and in all branches. Contrary to archaic fears, though, virtual intelligence has not fully supplanted organic intelligence. Cooperation between virtual and physical sophonts results in a two-way redundancy that allows both parties to rely on each other's strengths and compensate for each other's weaknesses.

Invention

(eusapient artilects)

Perception

The cultural view of AI is not and has never been uniform in the species which created them. The ziirpu, for example, achieved the creation of virtual sophonts centuries before spaceflight was even dreamed of. The cautious, collectivist culture of Virvv aspired to design practically flawless leaders of vast intellect to better lead their society into its frightfully uncertain future. Ironically enough it was the machine minds who conceived the notion of an interstellar civilization; a scenario unique among the known spacefaring species. Rimor, on the other hand, were wary of sapient machines. The hypercapitalist civilization that permeates the planets of the Eridanus sector feared automation at the expense of organic employment, and were slow to develop even the most rudimentary of socially-intelligent programs. Eventually, the necessity of social machines became impossible to ignore, and the Conglomerate begrudgingly funded AI development to the point of parasapience.  

Rights

Under the Coalition of Spacefaring Civilizations's Declaration of Universal Sophont Rights, a virtual intellect that displays any form of emotive-social personality is considered a legal citizen with equivalent rights. As to be expected, artilect rights are rather different than those of organics, to accommodate the vast differences between physical and virtual life. Virtual entities are granted the same rights to life, justice, and dignity as physical beings, but the matter of liberty is rather more complex.   Eusapients are, naturally, granted the right of liberty in all respects. Parasapients, however, are viewed as a more complex case. They are entitled to freedom of thought, expression, movement, occupation, etc if they are fundamentally capable of choosing such things. In many cases, their basic parameters outline narrow lines of thought that prevent these rights from applying, though this is not technically a violation of anti-slavery laws. The rimor, having historically viewed AI as tools more than people, restrict most AIs to pseudosapience or parasapience, in order to circumvent the costly legal obligation of granting them rights.

Related Articles

Sapience
Generic article | Sep 19, 2024

The tendency of certain systems to become aware of not only the cosmos around them, but the cosmos within.


Comments

Please Login in order to comment!
Jun 9, 2024 18:24

Can you point me to a story where you've used some form of artificial intellect as a character or plot point?

Jun 10, 2024 02:24 by Doug Marshall

Sure thing! Back when I was producing Starhopper Radio, I created HOMER as a sort of narrator-figure for the Prologue Episode. While the details of that episode are no longer strictly canon, HOMER is still very much a part of this world and remains an avid historian. (Fun fact: there was supposed to be an episode detailing HOMER's origins where I waxed poetic about the bonds we make with others, but it was never produced.) You might also take a look at Frontieria, which has Chuck the shipmind as a major character!

ASP | AV | OE | SPH | TMS | CDL | LOR | PR | HTH