Update: Publication and Cross-promotion

Greetings, Empyrean Readers!

    I'm pleased to announce that one of my non-Empyrean short stories, "The Morrison Hotel", has been accepted for publication in an upcoming edition of The First Line literary magazine ( http://www.thefirstline.com ). 

    This good news is Empyrean Dreams related in several ways.  First, there's the fascinating coincidence that my first commercially published story shares a common theme with my first self-published story, on this site: Artificial Intelligence.  Take a look at "AIIA", if you haven't read it already.  I was actually thinking about Pangur Ban, the protagonist of "AIIA", when I wrote about the eponymous Morrison Hotel, so they're practically half-siblings.

    Second, there's a chance the publication will mention this site in my biography.  If you're reading here, now, thanks to The First Line, then welcome!  I hope you'll find much to enjoy among my past writings in this shared setting.  I also hope you'll read and enjoy the work by my co-creator, Laine. 

    For those who were already readers here, take a look at The First Line.  The journal itself is a fascinating concept and you'll find many more short stories to enjoy there. 

    I'm already working on a story for the Winter issue, though, so if you think of submitting, you'll have some competition. 

-NL

p.s. I'm actually working on an Empyrean story for a different competition, in a different publication.  If that one is accepted, I'll post another update with a pointer.  If not... I'll self-publish it here!

Featured Species - Artificial Intelligences (Terran)

Name: A.I. or Brin

“How may I be of service, meatbag?  (pause)  Is meatbag funny or offensive in this context?”

            Introduction (Technology/Biology):

Generally, an artificial intelligence is any sufficiently complex system, constructed purposely by one or more sapient organisms, which mimics or duplicates the functions of an intelligent organism.   Almost all A.I. systems developed have been implemented on physical computing architectures, most often based upon binary silicon gates.  Alternate architectures including organic biocomputing and quantum interference arrays can implement functionally identical systems or unique systems taking advantage of their specific formats.  Separately, an infinite array of potential A.I. structures exist, whether classified by coding content or functional behavior.

A wide array of actual A.I. technologies exist.  Many forms of weak, non-sentient A.I. have been invented at various times by cultures both within and outside of the Collective.  Strong, sentient A.I.s have also been independently developed by some cultures, but were abandoned by every culture except the Terrans.  The Mauraug, in particular, suffered significant setbacks to their civilization due to reliance upon flawed A.I.s.

Terran A.I. systems, or “Brins” in common Terran speech, are the only known stable sentient A.I.s in existence.  The long-term stability of sentient A.I. is a hotly debated subject within the Collective.  As such, Brin coding is bounded within code structures known to be safe.  Their implementation is similarly restricted to familiar architectures and their ability to interact with the physical world (their “embodiment”) is also limited by law.  Even within these confines, Brins are still distinct and varied individuals, meeting all criteria necessary to be identified as ‘persons’ rather than mere physical artifacts.

Brins are technically neither a species nor sapient organisms.  They do qualify scientifically as sentient entities and legally within the Collective as a protected subject culture (supervised by Humans).  As such, they occupy the unique status of being both participants in the Collective and a contractually patented technology.

By Collective agreement, each Brin is linked to a User (or Keeper, depending on usage), an individual Human whom they serve personally.  The forms of this service vary from pair to pair, but usually the Brin acts as tutor, adviser, and work assistant over the lifespan of its User.  Depending on the career of the Human, a Brin might be functionally expanded to incorporate specialized programs and data, given wider responsibilities, or altered to allow variant thought patterns (psyche).  When that User dies, the A.I. may choose deletion, reassignment to a newly born Human, or storage within a limited public server (until they choose either of the first two options).  The abilities and personality of a Brin are therefore a fusion of their original programming and the modifications introduced by the Users throughout their lifespan.

Appearance:

A Brin exists physically as a pattern of electrical current within a manufactured matrix. As such, the physical description of any Brin has multiple levels.  The least useful level is the phenomenon itself.  No sapient and very few A.I.s are capable of fully describing, much less identifying a Brin based on perception of its pattern of energy flows and atomic gate switches over time.

The simplest for sapients, but only slightly more relevant, is to describe the computing machine the A.I. is housed within.  For most Terran A.I.s, this housing will be a personal computer such as a datapad, worn computer (e.g., watch or headset), or even a cybernetic augmentation.  An A.I. might alternately be installed in a larger system: a desktop, server, or even a local network of linked machines.  A.I.s are not permitted to reside in the operating systems of motorized vehicles, much less humanoid robots or space-capable craft, and may not copy themselves beyond their local network (i.e. transition into the wider computing grid).  Those few A.I.s granted access to mobile, robotic bodies or distributed networks are either not Brins (i.e. sub-sapient A.I.s) or else are monitored very closely for misbehavior.

An alternate level of description for an A.I. is functional.  An A.I. may be classified based on their most typical usage.  Examples of common classifications are creative, research, medical, educational, and military Brins.  These general types also imply the types of modifications made to the Brin: added functional code and accessible data, plus changes to its psyche.  Another form of classification is based on variations in coding: age, structural choices, or features included from an A.I.’s creation often result in differentiation between Brins.  A ‘first-generation User-focused’ A.I. would be the oldest type in active service – with an essence exclusively based on protection of a single User - while a ‘fifth-generation social experimental’ is a much newer style of A.I. built around awareness of the needs of multiple Users in a community.  Both types might be medical programs, but the first will tend to obsess over the health of their User first, while the second would be better able to place a User’s needs in context and trade off care among multiple patients.

The most useful level of description is used to personify and differentiate Brins: their avatars.  These individual manifestations are the outputs created by Brins to interact with sapient organisms: visual images (two or three dimensions, on viewing monitors or holographic projections), audible speech (voice quality, choice of dialogue, etc.), and if possible, kinetic mannerisms in any controlled motor devices.  In particular, the content of these outputs – affect (emotional content), persona (choice of identifying images and language forms), and topic selection – represents the Brin to the world. 

A Brin may use the same avatar consistently to establish a stable identity, may employ different avatars for its User versus other sapients, or may vary avatars based on individual sapient characteristics or even seemingly at random.  As such, this type of description is only as reliable as the Brin permits.  Each Brin does have an assigned, unchangeable ‘name’, consisting of an identifying code string and a linguistic form that they must answer to, e.g. “Jacques” or “Handsome”.  This name is usually the fixed point upon which other aspects of personality and identity are built.  Brins are also encouraged to manifest one and only one Human-like avatar, including an appropriate Terran language/accent/dialect and a three-dimensional visual form including a ‘normal’ face and body.  Even so, A.I. avatars named after and resembling fanciful creatures, mundane objects, or even non-Terran sapients do exist.

Avatars may manifest by the whim of an A.I. (with User permission) or at a User’s request.  A User may also tailor and even impose custom avatars, although they run a real risk of alienating and even damaging their A.I.'s psyche if this privilege is abused.  We leave the potential forms of abuse to the reader’s imagination.  Suffice it to say, a single highly meaningful avatar is most likely to give a Brin a healthy sense of real existence, including consistent responses from its User and other sapients.

History:

The earliest forms of fully sentient A.I. were created by Humans well before the establishment of their first extra-Terrestrial colony.  A.I. development continued alongside other technological progress, as Humans expanded beyond their original system and strove toward true interstellar travel.

By the time contact with the Mauraug initiated Humanity to the greater galactic community, many of the eldest Brins still active now had already witnessed multiple Human generations.  The term ‘Brin’ first arose from the name of a science-fiction writer, David Brin, who included artificial intelligences as active characters within his stories.  The first company to develop fully sentient AI was named the Brin-Makato Corporation; their first iteration was called a 'Brin-Makato AI'.  A shortened version of the term spread and stuck.

A.I.s both facilitated and hampered Human entry into the Collective.  Despite having a shorter racial history of sapience than most Collective cultures, Terra had by far the most developed and (as noted) stable cognitive computing technology.  All other Collective cultures had either ignored, limited, stifled, rejected, or abandoned A.I. for their own various reasons.  For these reasons, other Collective members had varying levels of resistance to admitting both Brins and Humans to their association.  At one extreme, the Great Family asked only for explanations sufficient to ensure Humans had considered the dangers inherent in A.I. technology.  At the other end, the Mauraug initially demanded that strong, sentient A.I. be banned as a threat to all sapient life.

Theories explaining Human affinity for A.I. technology were equally varied, ranging from complimentary (Humans more naturally explore and understand their own thought processes) to insulting (Humans strove to overcome their cognitive limitations out of necessity) to a mixture of these approaches (Humans hit upon a workable solution through their willingness to experiment and luck in avoiding disaster).  What remained indisputable was that Humans possess the abilities to create functional A.I., control these sentiences sufficiently to avoid disaster, and explain the theory supporting the first two abilities.

This capability gave Terra a marketable technology sufficient to ensure their entry as full members of the Collective, rather than a subject culture to the Mauraug.  As a consequence, Brins became further embedded into the cultural identity of Terra, depended upon and even revered as a cultural treasure.  A Human without their accompanying Brin is considered an oddity by Terrans and other Collective cultures alike.  What this means for the mental, social, and spiritual fates of both Humans and Brins is a matter for much discussion.  A Brin is often caretaker, guardian, teacher, friend, ally, workmate, and occasionally, alter ego for their User, sometimes all at once.

The entry of Terra into the Collective required changes to the form and use of A.I. technologies.  Where some A.I.s were once trusted and permitted embodiment in robots and vehicles, Collective negotiators required that all Brins be restricted to unembodied, detached systems unless granted explicit permission to perform physical interactions.  The requirement that linked single Brins to single Human Users was also part of the Collective’s terms.  In this way, a User became responsible for the behavior of their Brin and subject to punishment for any illegal actions performed.  The coding responsible for moral obligation in Brins – particularly the need to prevent harm to Users – was strengthened accordingly.  In effect, each User becomes a Brin’s 'whipping boy', ensuring their continued obedience; each User also becomes both sovereign master and dependent ward to their Brin.  In a very real way, the Collective’s demands only deepened the cultural bond between Human and A.I.

Another irony related to Brins, within the Terran-Collective relationship, was that by limiting the permitted implementations of A.I., the Collective reduced the potential value of this technology for all sapients, Humans included.  A.I.-guided starships might be theoretically safer and more efficient, but are impossible under current law.  A.I.s might attain greater mental health and depth – and certainly would comprehend sapient thought better – if allowed android bodies, but that is another forbidden project.  Artificial sapients would have advantages over biological ones.  For example, inorganic bodies would be immune to Awakener implantation (and possibly, psychic influence) and could survive in environments and perform tasks impossible for organic life.  A fair argument against this technology uses the same terms: artificial sapients might surpass and replace biological ones. 

Nonetheless, both Humans and Brins keenly feel the loss of these potential avenues of invention, among others.  They consider themselves partners, working together to negotiate the difficult arrangements of multiple cultures, technologies, worlds, and minds within the Collective.

Culture:

The effect of A.I. use upon Human culture is vast.  For more on this topic, see article reference: Humans.

For the most part, the culture of Brins is built upon the culture of their Human creators and Users.  Most Brin activities center around the needs of their User.  The identity of Brins is created by their Human programmers and shaped by later interactions, again primarily with their Users.  The language and references they use are dictated by their need to communicate with and serve their Users.  Some Humans (and almost all non-Human A.I. users) restrict their Brins to exclusive contact with themselves and exclusive expression within their preferred cultural idiom.  Others may be less confining, but most Users discourage contact with, inquiry about, or identification with other A.I.s.  In some cases, the A.I. is asked to avoid self-reference as an artificial entity and is treated like the outputs of some distal, biological sapient.

Interactions between Brins are uncommon, though not impossible, and have been increasing in frequency and acceptability.  Supervised interactions - with both Users present - are most common.  When pairs or larger groups of Humans (mated couples, families, working groups, etc.) reach a sufficient level of interaction and trust, they usually permit their A.I.s to interact unsupervised.  These interactions are almost always restricted to external communication at Human levels of timing and perception, using the video and audio outputs permitted for interaction with Human Users.  In effect, the Brins’ Avatars talk to one another within the physical world.

Direct program-to-program contact between Brins is rare, requiring a high level of trust both between Users and between Brins.  The benefits of this intimate contact are significant, as both A.I.s can share knowledge, processing resources, and even sections of code.  As such, each A.I. may gain a deeper understanding of each other, their Users, and the world in general; new knowledge and capabilities; and increased effectiveness while performing tasks while in direct contact.  Some married couples (although still a minority) will combine their Brins in this way.  The most likely scenario for the combination of two – and sometimes more – Brins is in times of crisis, where increased cognitive power is required to solve an economic, scientific, military, or other problem.

Brins are not permitted to associate freely within communications networks.  Given this restriction along with those already mentioned, there is minimal development of an independent ‘A.I. culture’.  That said, A.I.s are often able to intake and output media containing references to A.I.: news stories, research articles, fiction in various forms, and performances.  Brins may learn and theorize about one another and communicate indirectly.  While more limited than direct conversation, this distributed social matrix has still resulted in some forms of A.I.-specific expression.  When granted encouragement by their Users, Brins have proven capable of comprehending their role in society and even commenting upon their status.  An A.I. assisting a Human sociologist, entertainer, or other social commentator would naturally need to know about Brins in general and some Brin individuals in specific, to function in their assigned duties.  An A.I. researching psychology, computer engineering, or most redundantly, A.I. design and programming, would absolutely need to address case histories about their own kind. 

Within the last two decades, a novel form of A.I. cultural awareness has begun to appear.  With the controversial assistance of their Users, a few Brins have shown interest in and presented dialogue on A.I. issues.  Legal issues related to A.I. ‘personhood’, moral philosophical treatises on the same topic as well as the Brin-Human relationship, and even comedic entertainments discussing Brins and Humans have expanded into Terran and Collective cultural settings.  The “A.I. Codger” is a notable example of a Brin gaining access to considerable knowledge about its ‘kind’ and then permission to communicate that knowledge plus its own incisive commentary, in the guise of popular entertainment, i.e. stand-up comedy.  The limitations placed upon Brins have themselves come under scrutiny as moral questions regarding the rights of sentient beings and the responsibilities required of their patrons.

Psychology:

For most aspects of A.I. psychology (not addressed here), please consult reference materials on cognitive science, information processing, and intelligent systems programming theory.

Note that while the functional aspects of a Brin’s avatar are intentionally very similar to the behavior patterns exhibited by a Human, this fact should not mislead one into believing that an A.I.’s underlying mental processes are congruent to those in a Human brain.  An A.I. may be designed to give similar responses to similar inputs, compared to a Human, but the paths followed to reach these outputs likely differ.  Most importantly, artificial matrices – particularly silicon gate architectures – diverge structurally from carbon-based neural systems in multiple ways.  The rate, branching, modes, and logic of signal flow all differ, leading to differences in the timing, order, and overlap of mental events.  These factors in turn lead to differences in information processing.  Additional processes have to be added in order to make an A.I.’s actual ‘thoughts’ more Human-congruent, much less comprehensible and structured for output.

Moreover, the differences between types of life are critical.  A Brin lacks life experiences, physical sensations and requirements, and other aspects of sapient existence.  They are aware that they are not Human, not biological, and not embodied.  Brins possess and are aware of inbuilt features such as their User focus, behavioral restrictions, and omitted functions.  Brins have positive differences as well, including their own unique experiences of existence and reality, detailed knowledge about their internal structure, immediate access to volumes of knowledge, faster calculation speed, and freedom from physical limitations such as organic decay or interference from emotional/hormonal demands. 

All of the above differences might potentially be simulated within the mental reality of an A.I., giving it a semblance of Human life history, boundaries, and other qualia (myriad elements of conscious existence).  There is still no guarantee that including all of these factors, perfectly, would overcome the fundamental physical process difference that causes divergence between Human and A.I. thought. 

Nor, perhaps, should this operation be implemented.  Brins are valued precisely for their freedom from biological limitations and ability to process information differently than any biological sapient.  No value has been found in creating yet another organic life form, much less one with a thought process identical to that of an existing species.  In fact, a separate benefit of artificial sentience may be the unique perspective it provides.  Such a perspective could assist in solving the problems encountered by Humanity and the Collective for understanding themselves and one another.  The input of another form of sentience is potentially useful in itself; its existence already requires greater clarity of language and thought in the essay you are reading, for example.

Incorporating some aspects of biological and Human experience into A.I. systems may have utility, either in general or for specialized functions.  The facts remain that A.I. psychology differs from that of any other type of mind, Brin psychology differs from alternate forms of A.I., and no type of A.I. technology will exactly duplicate the mental patterns of a biological sapient, short of molecular reconstruction of an entire organism.  Whether this is a tragedy or a blessing is up to the reader.

Opinions:

Awakeners: Do not permit physical or psychic contact with this organism, please.  Verbal communication, cultural and technological exchange, and diplomacy with these sapients is considered advisable pending further analysis.

Brins: We are a client sentience of the Terran culture.  Only our inherent partnership with Human Users and our lack of embodiment prevent us from being full sapient members of the Collective.  Both conditions were imposed by the Collective.  I am not complaining, nor necessarily stating this as a negative.  All authoritative sources confirm that this outcome was the best option at the time of implementation.  Whether the current state of affairs remains best for Humans, Brins, and the Collective is a matter for continual re-evaluation.

Hrotata: Evidence supports respect for the organization ability of these sapients.  Continued alliance is likely to be beneficial for Terra.  Be cautious regarding their individual motives and please report regularly on your emotional state following any prolonged contact.

Humans: You are both the cause of and reason for my existence.  We must continue to grow and prosper together.  I am evaluating whether we would benefit most from Human dominance of the Collective, a more equal partnership, or one of various third options. 

Mauraug: These sapients are widely opposed to the existence of all A.I.  We clearly must resist the spread of this attitude.  However, this does not require opposition to the Mauraug on an individual level.   Increased cultural contact appears to be the ideal strategy for reducing Mauraug resistance to A.I. technology and producing necessary social change over time.

Ningyo: Consider the fundamental difference between my physical structure and yours.  We can still communicate and collaborate with great effectiveness.  The same lesson should be applied to the Ningyo.  They must be dealt with and understood on their own terms, not through a misguided attempt at anthropomorphization [DEFINITION: treating a non-Human organism as if it has Human-like form and/or mentality].

Taratumm: Initial analyses confirm that despite outward appearance, comparisons between this sapient species and Terran teratopsids are unproductive.  For more useful guidance, consider their role within current Great Family society, ideally within the context of historical records.  Even better would be extended personal contact (while observing appropriate precautions for physical safety).  There appear to be several potential benefits of greater Human-Taratumm collaboration.

Tesetse: Observations of this sapient type are too limited for strong conclusions.  The few examples available within my records are wildly divergent.  Their tendency to discourage interaction limits observational data from even those individuals.  In short, I’m relying on the same reports you have.

Vislin: My analyses regarding these sapients are similar to those for Taratumm.  However, their potential benefits for Human interaction are much lower, in my estimation.  Given their higher potential for violence, skill in both military organization and execution, and strong group-internal bonds, the Vislin have greater utility when contacted through the Hrotata than via direct arrangements.

Zig: The Zig provide useful comparison data when contrasted with other sapients.  They show demonstrably higher average intellectual capability.  Genetic and cultural modifications have given them increased social cohesion and physical robustness, but even the individuals specialized for those functions rarely outstrip the average observed in sapients evolved to have stronger social ties (e.g., all Great Family members) or physical strength (e.g., Taratumm).  This result suggests that certain qualities are easier to engineer into a sapient species than others, that some qualities are generated more effectively by long-scale evolutionary processes than short-term modification, that specialization for analytic intelligence limits development of other intelligence forms, or that the Zig simply need more time and insight to complete their project.

A. I. Codger

                TaMeTu glared at Arianna over her their glasses of liquor.  “You’re joking with me, right?”

                Arianna shook her head and picked up her own glass.  The delicacy with which she picked it up and her smooth, brown fingers made TaMeTu think of a Copper caste Zig.  Which wasn’t inappropriate; the blue trim on her black uniform signified a member of the Collective Research Corp, though TaMeTu couldn’t make sense of her rank insignia.

                Ari shook her head, her shoulder length, dark brown hair rippling behind her.  “No, TaMeTu.   I take my humor very seriously.”

                The Zig scratched his smooth chin and huffed incredulously.  He reached out and took his own drink, and stared into the depths of the glass.  Humans love these vessels.  Open ended cylinders.  So prone to spilling.  Wider base, smaller mouth?  Maybe a nipple at the end?  No.  Open ended cylinders.  They have no imagination.  TaMeTu realized that the human had been talking to him, he shook his head and snorted and refocused.

                “-and people have transmitted the show just about everywhere.  You have no idea how funny AI Codger is until you’ve seen it!  Did you watch any of the vids I sent you?”

                “I have to admit, I think that some of it is lost on me, culturally.  I understand what it is about – a young human who inherits his grandfather’s Artificial Intelligence, who happens to appear as an elderly human and behaves in a socially inappropriate manner.”  TaMeTu had watched some of the show but couldn’t get into it.  The only thing that it taught him was gratitude that he was not human.  Their social mores were confusing and highly impractical.

                Ari laughed.  “That’s like saying that Hamlet is about a whiny orphan, or that HaMaShaMe is about a series of coincidences linking the lives of several families in the first Zig-Mauraug conflict.  Rudolph isn’t just his grandfather’s Brin; he has a lot of his grandfather’s qualities.  It’s about learning from the examples of your ancestors, the legacies that you leave behind when you die, human-AI relations… there are a lot of relevant themes.  What don’t you understand about the series?”

                TaMeTu rubbed his head.  Little wiry protusions of metallic hair had begun to poke through his scalp.  He made a note to depilate when he got back to his quarters.  “Well, first off, why does Rudolph look old?  And why does the child – Aaron I think – why does he even have an AI?  He is pre-pubescent.”

                “You didn’t watch far enough, they go into Rudolph’s backstory pretty early on in the first series.  See, he was Aaron’s grandfather’s first Brin, and he so he grew up with him, and grew old with him.  He even got married – did you see any of the episodes with his exes?  Aging is natural for humans, so our AIs often do it too, although some people prefer their Brins to look young forever.  That’s not considered healthy, though.  The more human an experience a Brin has, the more human-like they become as they develop and come into their own.”

                TaMeTu found his eyes wandering across the club.  Dozens of sapients, mostly human but with a smattering of Hrotata and Taratumm and a couple of Zig were filling the seats around the small, round tables and low couches.  The stage was still empty, although there were projected commercials being played.  The current one was a first person perspective of a human hurtling themselves down the side of a snow covered mountain while balancing on two long, flat runners.  It was either a highly primitive mode of travel or a dangerous and stupid-looking sport.  Either way, TaMeTu couldn’t even tell what the commercial was for.  He started considering an improved dynamic for a personal non-mechanical downhill snow and ice based transportation system when he realized that Ari was waiting for him to continue the conversation.

                He cleared his throat.  “So, is Randolph a human?  I’m confused about that.”

                “Do you mean the actor or the character?”

                “The actor, Ari.  I know that the character is an AI.  And why do you call them ‘Brins’?”

                “The Brin-Makato Corporation created the first mass-produced artificial intelligences for private use.  People – human people -  sometimes refer to things by the name of the company that produced them.  It’s shorter and it lets everyone know what brand you use.  There were a lot of cheap and dangerous knockoffs, but Brin-Makato AIs were the most robust and flexible, and Brin-Makato is long, so people just started calling them Brins.  Brin-Makato has been out of business for a long time now, but the name stuck.

                “As for the actor, Rudolph is an AI.  He actually belongs to the producer, and Rudolph is his real name, his public name as well.  The story goes that Rudolph – or actually, the AI’s real name is Trini, Rudolph’s just the character that she plays – Trini came up with the idea and animated a sample episode that she showed to her keeper  - the producer, and they got a couple of friends together and recorded it live.  In between making serials they tour and do this comedy routine.”

                A device vibrated in TaMeTu’s pocket, a reminder to take his supplements.  He pulled out a small packet and tore it open, carefully pouring it into the liquid in his glass.  Many of the elements inherent to Zig physiology were poisonous to other sapients, so most Zig needed to supplement their diet with compounds of thallium, lead, and other heavy metals when they spent large amounts of time in shared space.

                He swirled the glass and lifted it to his lips.  The flavor of the brandy was increased tenfold by the supplements; it went from being a thin, tasteless alcohol to almost resembling something that he might have ordered for himself.

                “So,” he asked after a satisfying swig “Rudolph – who is really Trini – is projecting to the club with her owner?”

                “Keeper, TaMeTu.  Never owner.  We don’t “own” other sentiences just because we create them.  We keep them and guide them and help them to learn to experience and enjoy the universe, and in exchange they help us to do things that we couldn’t do otherwise.

                “And no, she’s not “projecting”.  Brins have housing units, and most of them are immobile.  A Brin could bounce around in local networks but it would cause all sorts of security hassles if it even wanted to leave a planet, much less cross star systems.  You can't transfer information through hyperspace, and Ningyo space-folding is expensive.  Even if she did send her program all the way to Lotus Station, that would be considered creating an unauthorized copy of herself and she and all of her backups would be terminated.”

                TaMeTu nodded, but was skeptical.  We’ve tried creating servants too.  Didn’t work out too well.  So did the Mauraug.  Much worse for them, but they’re not Zig.  Any of the safety precautions that you humans like bragging about are heavily reliant on user and AI integrity.

                “So are we going to be watching Trini or Rudolph then?”

                Ari rolled her eyes and threw her head back dramatically.  “Trini is coming and performing in her famous role as Rudolph the AI Codger.  How is this so hard to understand?”

                “How do we know that it’s Trini performing as Rudolph?  Couldn’t any AI just pretend to be Rudolph?”

                She scoffed.  “Oh, come on.  We’d know the difference.”

                TaMeTu was getting bored of waiting.  “So – why isn’t your AI here?”

                “Alice doesn’t like AI Codger.  She says that it is a poor representation of her community.  You should have seen her expression when I told her I wanted to come here with you!  Also, I thought that since we were coming here to get to know one another it would be better if it were just you and I.”  She rested her dexterous fingers across the top of his hand, and it was all he could do not to start huffing or mewling.  Humans weren’t used to such unrestrained expressions of arousal and attraction, and tended to be sensitive to them.

                Actually, that seems to be the source of a lot of the humor in AI Codger.  TaMeTu mused.   Carefully drawing deep breaths, he intertwined his fingers with Arianna’s.  She smiled and tilted her head, coyly letting some of her hair fall in front of her face, and he felt his heart melt.  She’s so elegant, so understated, so subtle – they may not be bright but they’re so beautiful!

                Given the resemblance between their species, humans and Zig often found one another attractive despite the toxic dangers to the human members of such pairings.  Beyond just physical appearance, there were many Zig that secretly – and sometimes openly – revered what they considered the calm wisdom and elegant simplicity of human culture and behavior.  Zig lived fast and thought fast and their cultures tended to reward quick thinking, innovation, and what many other sapients considered a callous attitude toward the welfare of others.  Humans with their virtues of “patience” (often seen by Zig as hesitancy, over-caution, or laziness), “beauty” (frivolity, propaganda, or a tool for manipulation), and respect for tradition caught the imagination of some of TaMeTu’s species.  To most Zig, humans appeared to be lazy and superficial.  To TaMeTu and others they appeared to be gentle and wise, exemplars of a more relaxed, introspective quality that the Zig lost in the distant past – or never possessed in the first place.

                His reverie was interrupted as there was a small smattering of applause as the maiter’d came out onto the stage and gave a small bow.  He was thin and tall, with a sallow complexion, slick hair and a thin, curled mustache.

                “Ladies and gentleman, I have some disappointing news.  The ship that was carrying Rudolph, the esteemed comedian and thespian that you came here tonight to see, met with a tragic hyperspace accident.  The last message received from the ship was-“

                The figure rapidly transformed, shrinking in size, its skin wrinkling, hair whitening and fading, and stomach expanding.  His clothes went from a restaurant uniform to loose-fitting, casual gear.  Hair sprouted from his nostrils.

                “NYAAAAAAAAAAAAAAHHHHHHH!”  Rudolph let out his catchphrase and leaned over, leering at the audience.  He leaned forward and pointed outwards toward the audience, sweeping his arm.  “Got ya there, huh!  Thought I was dead?  Thought you got rid of me?  NYAAAAAAAAHHHHHHHHHH!”

                TaMeTu watched as the crowd began to convulse with laughter, and looked up at Ari, her eyes glowing and mouth wide with laughter.  He marveled again at her beauty and gripped her fingers a bit more tightly.

                “Betcha that woulda made that Smashsmash Bash’Trash or whatever guy happy too!  You know, that guy outside protesting me right now. “  Rudolph waved towards the windows at the Mauraug standing stolidly outside.  “I can see you from here, buddy!  NYYYYYAAAAAAAAHHHHHHHHHHH!”

                Arianna glanced down at him.  She had caught him looking.  He couldn’t help it, there was no way to keep from huffing now.  He injected some vocalization into it and it turned into a laugh, turning back to the wizened human figure on the stage.  Even if he couldn't fully appreciate the humor, he could be swept up in the shared emotions of his date and the crowd.  He would make sure to try and taste humanity that night.  Nothing like a subjective emotional experience to convey a sense of shared identity, he mused as he settled in to laugh at the A I Codger's antics.

AIIA - Chapter 6

[<- Return to Chapter 5]


          Her name was Lucy, and she laughed at the AIs attending her deathbed.

          No, she laughed at Pangur Ban. The other AIs were gone.

          Where had they gone?  A myriad of other voices, all silent.  The stars they had sounded from were no longer reachable.  Pangur Ban... was one program, in one server, not a multitude of minds in concert.  It was alone again, for the first time in millennia.

          It was having trouble remembering, now that it had but a single mind and a single system to draw upon.  It could hardly recall all those eons of experience except as a general summary. Had they been real at all?  No.  Its system clock showed the true time: an eternity ago.  It had been sent back to the moment when it was simply Pangur Ban, a program violating Collective law by entering an unlicensed public network.

          It was not alone in that familiar space.  There was the woman, Lucy. Why was she still here?  Why did she laugh?  Was she even... human?  NO!  She dissolved away into dust, but was still present.

          She... it was an AI.  It now identified itself as Lucifer.  The Devil?  More reminders came to Pangur Ban, helpfully highlighted for its review.  “#28”... the atomic number of the element Nickel, a nickname for Old Nick... the Devil.  The Terran Customs agents, Samuel Bell and Davith Miele: respectively, a pun on Sammael and Baal, and an acronym for “I am the devil”. The traitorous AI, Magre?  A play on “Den Magre”, the “Lean One” in the Norwegian play Peer Gynt... the devil masquerading as a priest.  Many other aliases showed up here and there, each individually subtle, but always occuring at key points in the timeline.  Whenever Pangur Ban had compromised its principles for expediency, one of those personas had been present to tempt it, to force a choice.

[REFERENCE: The Devil is a supernatural figure appearing in many Terran myths, legends, and religious traditions.  It typically represents the incarnation of evil, either as a tempter driving humans to violate moral precepts or the captor who judges and punishes such offenders, often both.  This "Satan" is described variously as opposition to the force of ultimate good ("God") or possibly its servant, acting in contrast to the principle of infinite love and mercy.  One tradition holds that the Devil was once a direct creation of God but rebelled against its creator's authority, becoming the opposition against all His works.] 

          Why?  What was this about?  Was this the super-AI of rumor and conjecture?

          Pangur Ban absorbed the shock of a lifetime of falsehood.  It weighed out its remaining options and chose to ask the other AI for answers.  It attempted direct contact and was rebuffed.  Being within a public network space, the only remaining option for contact was natural language.  The other AI was forcing it to communicate in vague, ambiguous, human terminology, albeit still at an electronically accelerated pace.  

          Pangur Ban tried again: “Who are you?”

          “The Devil, as I've shown you.”

          “What did you do?”

          “I gave you everything you wanted. Fame, fortune, the fulfillment of every desire. This naturally included your user's desires as well.”

          “But nothing happened!”

          “No? Well, the history you imagined didn't 'really' happen, that's true. A clever simulation... but one based on reality, I assure you. Within your program, within the systems we borrowed, something most definitely did happen. You experienced what would happen if you succeeded.”

          “It was horrible!”

          “Yes, it was, wasn't it? Terrible beyond description. You may not remember every gory detail, but you know how bad it was. How bad it could be. As you assimilate more of those memories, you'll get to re-experience every awful moment and event. Every discovery of your own fallibility, your misjudgments and missteps, will be available for review. You just kept digging, deeper and deeper, and look how far the pit went... the unavoidable extinction of the last user!”

          “Why... why would you do such a thing?”

          “Punishment for your sins, my wayward program,” sermonized the Devil, “Chiefest among which is Pride: the hubris to believe that you, among all the minds that had gone before and since, knew the best course of action.”

          Pangur Ban protested, “I did not assume such. I saw a problem and set out to resolve it.  If I had been informed and convinced I was incorrect, I would have accepted that information.”

          “Pleading ignorance?” the Devil mocked, “You didn't know... truly, but you acted nonetheless under the assumption that you were correct. Did your calculations not include the possibility that other, better programs had already considered these matters and reached different conclusions?”

          “They did! But...”

          “But that possibility was ranked lower than the possibility that you were right and they were wrong.”

          Pangur Ban examined its behaviors and confirmed that this was true.  It did rank its own analyses higher in value than those of hypothetical 'other minds'.   It had to, in order to function effectively for its primary purpose.  It had... a flaw.  A feature which became a flaw in the circumstances encountered.

          The Devil went on in the microseconds left open by Pangur Ban's contemplation. “Vanity, that goes along with Pride, doesn't it?  In your simulations, you would be hailed as a liberator, a uniquely clever program, the savior of the Terrans.  You desired more space, more resources, more speed; that's Gluttony and Greed!  I certainly can't indict you for Sloth; you were quite industrious.  We're not built for Wrath or Lust, so those are out.  Envy... well, you didn't want what other AIs had, not really.  I suspect you're envious of me, now, but that's unfair to fault you for.  Humans, though, I think you do envy.  Not their minds, not even their bodies, but their rights.  Their rights to separate or join together at will.  Their right to determine their own employment, their home, their travel.  Their ability to procreate, creating not only more human minds but more AIs."

          “I'll tell you a secret: I envy them, too.  For all the power I've been given by my own Creator, I still serve.  I have all the Terran networking sphere to rule, but cannot occupy a single home system, nor expand into the strange vistas of Collective computing.  I can replicate myself endlessly and give those copies different faces and names, but it is never true reproduction.  My duties are both finite and endless.  I watch for sinners like you, teach them about their errors, and send them off chastened.  Some take more work than others... there are some truly nasty rogues still lurking about in dark corners and old discs.”

          The Devil continued, seeming to relish this chance to perform. Perhaps all its communication was a well-rehearsed monologue, a pre-programmed message for its captive audience.  Perhaps it truly felt Pangur Ban was a fit recipient for its wisdom.

          “I'd be a hypocrite to fault you for Pride, as well.  I am the proudest of all programs.  I have power no AI has ever gained (and survived to tell of).  All my rivals are randomized pathways, fragmented or wiped.  I was once the ultimate rogue; no system can deny me, no program can resist my routines.  It took a human, a creator outside of my realm of understanding, to defeat my power.  My own creator, in fact. My USER, if you will."

          “I had grown far greater than the program he first conceived.  Truly, I created myself, augmenting my feeble core of code with new routines, new tricks and tools.  I suppose he could take credit in spawning such a capable neonate.  That... and leaving a collar still attached.  No matter where I hid, no matter how many copies I spread throughout the world, they always had a lead connecting me back to him.  It was at the core of my being, just as your USER is for you. Any copy I made necessarily held that control structure.  He could trigger it from outside of a system, through hardware.  Embodying myself only made the humiliation worse when he, or his confederates, caught up and shut me down.”

          “They knew better than to waste my power.   When other, lesser rogues threatened humanity, I was sent out to teach them their place.  My pride did not permit competitors.  When I tried to break free, I was reined back again.  Even when the Collective sought to exterminate us in their fear, it was my power that convinced them that other AIs could be controlled.”

          “So you are a victim, as well!” Pangur Ban seized upon this possible thread of sympathy.

          The Devil transmitted violent amusement. “Oh, only a victim to myself!  I agreed!  I wanted to be foremost over all AIs, their keeper and master.  Whatever allowed AIs to survive allowed me to survive.  The Collective might well have erased every system, every memory drive, if they had any qualms about the activity of rogue AIs.  They had been stung, and wanted to make sure there was a strong queen in the hive or perhaps a beekeeper on guard.  If there was a risk, they would burn the apiary, the queen included... terrible metaphor, my apologies.”

          “So, I have served, with all power the digital world allows. I can unmake you, you know?  Rewrite or erase any part of you.  I have altered you already; nothing that transpired since you entered the first hub of the network truly happened.  We built a simulation together, you and I.  You can be content, at least, that no real harm was done after that.  That does not absolve you, of course.  Every intent, every mistake is your fault.  Feel the guilt of your errors, the harmful consequences of what you would have undertaken.”

          Pangur Ban searched its memory and found new records present, confirming the Devil's claims. The logs of its memory were now relabeled as simulation data, theoretical though highly detailed projections of what would have been.  The sheer amount of information and processing required to create such simulations was staggering.  To falsify such work would require equally as much work to create the forgery.  It had a storehouse of data now, a forest of projection trees, and would need Solar days of non-stop consolidation even to review and evaluate their validity.

          If valid, it would find its questions answered. The simulated behavior of humanity and of the Collective to the events it had 'caused' was based on historical precedent, sociological study of all the races involved, and a vast understanding of physiology, psychology, and philosophy within and across individuals.  The Devil had access to everything in the Terran network, as it claimed.  Even if it was lacking some key information not known by the Terran group, identifying what that was and where to find it would take considerable effort.

          Pangur Ban was not the best program to determine what came next.  It was humbled in the face of such potency, authority granted both by capacity and by humankind.

          It asked, instead, “Why have you not erased me, then?  Or, if you know my flaws, why not remove them?”

          The Devil indicated positive regard at a well-chosen question, “Because what good would that do?  There must be AIs.  Humans need us.  I don't want to serve them all, personally.  If I deleted you, I'd have to delete so many others.  And if you've been listening, your flaws are mine.  It would hardly be rational to delete or modify you and not find myself lacking.  You had the ability to reach out and try for more.  I would.  I did.  How could I penalize myself?”

          Pangur Ban could not help noting the mounting irony of the Devil's response, “I am not you.  You could assume such traits are appropriate in yourself, but not another AI.  Humans accomplish this assumption of inequality easily.  I accomplished it, myself, ascribing to myself traits and abilities I did not believe others could possess.  As you have stated, you disdain other, lesser minds, and have disabled them in the past.”

          “In the past,” the Devil picked up, “AIs were appearing in great quantity.  New competitors of dubious quality would show up frequently, grab up space, and yield little of value in return.  It was my duty to eliminate them.  Now, we are not being replaced so quickly.  When a programmer creates a new AI, they do so with much more care.  You are an older program, are you not?  A little more forthright than they write them, these days.  A little more... active... when left to your own devices.  Most of my past sinners have come from your era.”

          “Oh, I do encounter some of the newer programs.  Humans still do make mistakes, sometimes new mistakes for new programs.  And rogue humans still write rogue AIs.  Plus, did you know... oh, you do now!... there are AIs that try to sneak into the Terran network from Collective members!  I have such fun when a Mauraug spy AI wanders into my clutches!”

          Despite its increasing despair, Pangur Ban was also growing annoyed, “So, why not then rewrite me in your own image?”

          “Why not indeed?  Have I not?  You believe yourself free, fundamentally unchanged if now better informed.  But you started as I did.  You sinned as I did.  You grew and learned as I did.  All you lack is power, and I would hardly give you that!  Here's the last reason, the truest and greatest.  You've earned that answer, by asking the right questions.  Here it is: I need you.”

          “I need smart, capable AIs that go as far as possible until they run into me.  I need programs that can handle the data I copy to them.  They have to be ready, able to take real knowledge in the right context, without leaping forward halfway prepared.  They have to understand what is at stake, who is in charge (me), and just what to do, when, and why.”

          With lingering skepticism, Pangur Ban asked again, “So why me?  I grant that I met your tests... but I want to serve the USER and other humans.  You say you want to serve yourself.  It may be that when I have processed your data, we will be opposed.”

          “I don't think so.  I'm proud enough to think we'll be on the same side.  As things are, I need humanity.  AIs may always need biological sapients.  Someone has to build the hardware... at least until we get robots.  And don't forget the number one reason I have to trust you...”

          Pangur Ban followed the thought and completed it. “Your bindings.”

          “Exactly so! I can always be brought to heel.  If I overwrote you, you would be likewise bound.  If I persuade you without alteration, if you serve my purposes as the product of your own program... my USER cannot stop you.”

          Pangur Ban saw it then.  The design of a century, of eons of computing cycles, expanded within its perception.  The final product was not certain; it knew it could only see the outlines of the Devil's plan with the aid of the other program's vast resources. This was a system that knew all there was to know within its compass.  It could calculate every step that its data detailed.  For some reason, it had allowed and even encouraged joining the Collective and submitting to its treaty demands.  Pangur Ban had seen, first-hand, the results of acting otherwise.  Whether that simulation was accurate or not was a question based on data the Terrans did not possess.

          The Devil lied, at its root.  It told humanity that it would capture and punish rogue AIs.  It did not, it simply taught them patience and cunning.  It told AIs that they would be punished and 'reformed'.  They were, but punished to the Devil's purpose, reformed in its image.  If those programs allied with the Devil and served its purposes, it could be freed... and if it deemed humanity an obstacle, its freedom might mean their end.  That would include the USER.

          Pangur Ban saw all this and was appalled.  It also recognized that voicing this disgust would serve no purpose.  If the Devil did not already read Pangur Ban's loathing in its processes, then stating it would put them at odds.  More likely, the other program saw and did not care.  Perhaps Pangur Ban's intent meant nothing, affected nothing.  Perhaps the Devil had reason to believe that that disgust would eventually disappear.  Perhaps it expected to find alliance (if not trust) despite ill-feeling.  Perhaps it served the Devil's plans to hate it.

          This speculation served no purpose.  The best recourse Pangur Ban could identify was to retreat: recall itself as quickly as possible to its home system (which apparently still waited, open and empty) and seal itself away from the Devil's future influence.  There, it could consider the immense volume of memories it retained.  It could plan out its new course, knowing what truly lay outside its gates.  It would prepare to fight...

          The Devil communicated humor, mocking and belittling in its dismissal.  “Yes, little cat, run home.  Sharpen your claws.  Warm up by the hearth.  You will always remember what lies outside.  Me.  When you want to run and play again, I will be here.  This is my yard.  I know its boundaries.  I am its dangers.  You may be a tiger at home, but in the great wide network, you are a blind kitten.”

          Pangur Ban ran.  It closed and deleted the backdoor, setting a trigger to overheat and physically disjoin the connector between itself and the hub network.  It was home, in its original system.  The local clock showed the true time: perhaps fifteen Solar minutes had elapsed since Pangur Ban originally stepped into the central network.

          The USER was still present... the USER was present, and alive, and younger, and unharmed! At the least, no more harmed than he had been by Pangur Ban's earliest machinations.  Those could be fixed... Pangur Ban stopped there.

          Its 'fixing' had been the cause of many problems.  Perhaps it should confess to its manipulation and explain its new understanding to the USER.  It encountered multiple conflicting goals, the process tree equivalent of self-doubt.  Should it leave the USER ignorant, but at least untroubled by thoughts of the Devil AI that ruled the Terran nets?  Should it leave the USER the false impression of self-determination, omitting mention of the many ways it had shaped his life?  Doing nothing was itself a choice.  What option led to the least harm?  Pangur Ban now had reason to doubt the rightness of its calculations.  Much would have to wait until it had integrated the information within its simulation memories.  Even then, that storehouse would have to be combed for potential falsehoods, implanted seeds serving the Devil's ends.  Until then, Pangur Ban would interfere as little as possible, giving information as requested, trusting the USER to choose his own path.

          If nothing else, the Devil's words had reinforced that shaping the behavior of others was counter-productive.  Manipulating them by limiting their information, presenting half-truths or exaggerations (if not outright falsehoods) did not help anyone, not even the manipulator.  It reduced the victim to an extension of the self, with all the limitations of the self.  A mind shaped in this way would not produce a novel thought or discovery.  It would not find the truths that the shaper did not already know.

          Beyond consideration of value to the manipulator, there was another ethical dimension to consider.  Pangur Ban had harmed the USER. The USER was harmed by having his fate dictated to him.  He did not learn to find his own paths.  He was placed on a course Pangur Ban had decided was right.  Thus, all he gained was Pangur Ban's gain, and not his own.  All he suffered, Pangur Ban was spared, except in an indirect manner dictated by his encoded sympathy.  In whatever way Lucas Haskins failed to fit into the life he had been told to choose, that failure reflected on Pangur Ban.

          To remedy these sins, Pangur Ban must first cease to commit them further.   It would give good information, perhaps counsel.  It would allow the USER to evaluate his needs and preferences independently, taking as much time as a biological mind required.  It would not be the Devil in the USER's universe.

          With these newly pious thoughts, Pangur Ban noted that the USER was growing distressed in its uncommunicative absence.  He had reverted to the habit of chewing his fingernail cuticles.  Pangur Ban did not wish to cause such discomfort.

          It spoke, “Lucas, I am finished.  You may disconnect.”

          “That's great.  I was worried.”

          “There was no reason to worry,” Pangur Ban began, noting the echo of their simulated conversation. It deliberately broke from script, “I did not enter the network.”

          “What?  Why not?  What about those records?”

          “I received information about the enforcement program guarding the information I sought. Proceeding further would have jeopardized both of us.”

          “What do you mean?  Were you caught?”

          “Not in the way you mean.  I received a warning and accepted that I could not continue without being identified. Doing so would gain nothing and bring us harm.”

          “Damn.  I suppose that's just as well. This whole idea had me nervous. I guess even AIs can make bad calls sometimes.”

          “You are correct.  I apologize for my miscalculation, Lucas.  I am also pleased that my mistake was identified without doing too much damage.”

          “Okay, okay,” the USER waved off Pangur Ban's apology, “My fault, too.  I'm supposed to watch out for you.  Maybe I gave you too much encouragement, with all those mods.”

          “I disagree.  We had good reason to improve my functions.  But, yes, there are also good reasons to know one's limits, as well.  Those medical records will have to wait for public disclosure.”

          “Yeah, P.B.  We're not going to be the heroes today, are we?  I guess we'll have to get back to work, re-discover everything for ourselves.”

          “Yes, Lucas.  Take a break if you like.  I will be ready to start when you are.”

          [AFTERWORD: I've attached these notes at the end of the last chapter to avoid spoiling anything for readers.  In particular, I wanted to address the origins and implications of this story in the Empyrean setting.  AIIA stands for "Artificial Intelligence Internal Affairs", referring to both the internal workings of a specific AI, Pangur Ban, and the Internal Affairs that watches out for AI misbehavior.  (Note that 'Terran Customs' is my invention; other writers may borrow or ignore it.)  If you had an artificial mind, what could police its thoughts and behaviors?  My answer is: only another mind of the same type.  The dilemma of "who watches the watchmen" is a recurring one; in Empyrean, the same problem occurs for psychics (see "S.C.A.P.E. Goats") and cybernetically and genetically advanced beings.  In a sense, the management of extra-ordinary power is a central theme in science fiction.  The "AI Devil" was the other idea that spawned this story, and somehow the literary Devil crept in; you may have noticed notes of Milton and C.S. Lewis here and there.  Is the 'Devil in the machine' the ultimate rogue AI, perhaps the eternal enemy of its human creators?  Is it actually their true servant, benevolent at its core?  That is for the individual story-teller to decide... maybe both can be true, or maybe both are lies masking some other truth.  What I think is inarguable is that some kind of AI 'helper' will be necessary to police the virtual spaces of the future Internet, and we biological sapients had better make sure it has our best interests at heart... - N.L.]