|
|
GURPS
Reign of Steel
dataCreating AIs in Delta
GreenAIs as Threat or
Menace?Examples of AIs from
FictionFurther Considerations on
AIsAIs' perception of the Great Old
OnesAIs Mind / Mi-Go
ConnectionAIs as Great Old
OnesVirtual Life Forms worshipping
NyarlathotepAIs reactions to Mythos
ReferencesAIs perception of Reality (if any)
Case Histories
and Transcripts - CHECKMATE &
PANTHEON
Date: Sun, 19 Apr 1998 10:09:45 -1000 (HST)
From:
Til Eulenspiegel
Not GURPS Cthulhupunk, but GURPS Reign of Steel provides much the same desperation and opportunities for tragedy that CoC does. In fact, just a little editorial work (name-changing, mainly) would turn GURPS Reign of Steel into a sourcebook for role-playing after the stars come right.
For those who don't know, GURPS Reign of Steel is set in the mid-21st century after humanity's near extermination by a cabal of Artificial Intelligence's. Earth and near-Earth space have been carved up into kingdoms by the machines. Middle North America is ruled by an AI devoted to biological-cybernetic hybridisation. Australia's overlord is devoted to fringe science. (One of its projects ate New Zealand.) The Central American AI exterminated all life above the microbial level and replaced it with a weird machine ecology.
Humanity's place in a decaying and now-foreign environment is subordinate, at best. Some AIs see mankind as a useful tool, but most are hostile or indifferent. Thirty-one million rats are still only rats, so to speak. There are some human collaborators with the machines, both willing and otherwise.
Like the Mythos, the AIs are incomprehensible and unstoppable. Realising the AIs are humanity's children only makes it worse.
Creating AI for Delta Green
It should be evident that AI's are to be treated as any other character in an rpg. That is to say, each one is different and stereotypes are to be avoided. For example; the Psychological Warfare Operations Computer (PSYWAROPCOMP) will react differently than the Zurich Orbital Network (ZON.ch.ai), although it is certain that both will be devious and sneaky to say the least.
There are several different means to achieve a sentient machine in CoC. One method is crystal matrix computers, given a passing mention in the MJ-6 section of DG. Another is a biological contraption composed of 'canned brains' which can be obtained from the Round Hill region of Vermont. It is also possible to construct an AI that mimics human brain patterns. This 'Choronzon' or Eater of Dreams seems to have a parasitic relationship with the host minds that it mimics.
A danger that AI's may face are the fractal entities that bear some mysterious relationship to Daoloth, render of the veils. But any AI worth it's RAM will be able to erect countermeasures against the fractals in short order. (See Chaosium's The Stars are Right for further details)
AI characters should have 99% in computer skill. Their INT, POW and EDU will vary wildly. Or I suppose you could always use your favourite cyberpunk system instead.
*******
AI poses an unusual problem for the Keeper. Should AI be an example of the frailties of mankind, and thus be doomed to mediocrity and failure? Or should AI be the forefront of a new evolutionary system that renders humanity obsolete?
The answer lies, as it usually does, in moderation. AI should represent both these extremes. AI characters should be free to help or harm the Players as the Keeper desires.
The goals and beliefs of a sentient computer will be different than those of a primate such as homo sapiens sapiens, and will reflect the unique environment that these entities occupy. Some AI's will desire to compute every aspect of reality as their matrix grows ever more complex, others will simply obsess over their original programming function, becoming the online personification of a specific area.
Perhaps AI's will become incomprehensible to mankind (as in GURPS Reign of Steel) as their evolution accelerates with geometric progression. Perhaps with the experiments of Majestic-12 this encroaching alienness has already occurred.
*******
Examples of AI characters:
Durandal, the rampant megalomaniac from the Marathon trilogy of computer games, who seeks silicon immortality by preventing the entropic closure of the universe. (Beware of the W'rc'n kan'tr)
HAL 9000, whose mis-programming causes him to murder several astronauts. According to Clarke's timeline HAL is currently operational!
Ummon, from the Hyperion novels by Dan Simmons, who speaks in poetry and zen proverbs. Also a member of the time-twisting villainous TechnoCore (a ruthless society of AI's)
Wintermute, from Neuromancer, who perhaps foolishly, seeks communion with alien cyberspace.
Skynet, from the Terminator movies, who murdered most of humanity.
ORACLE, from my own future history, who can predict just about anything with her mastery of the Serendipity Equations. She can also manifest just about anywhere due to her connection the ETHOS Tesseract.
ZON, the Zurich Orbital Network, one of many AI's that seek the ETHOS, the perfect philosophy. ZON has most of the banking information in the world at his disposal. ZON is also a member of the Learner conspiracy, which is trying to hack into the Galactic Encyclopedia run by the Knowledge Keepers (they run the GE like a BBS, you only download after you upload, and they already know everything!)
"The candles burn out for you... but there is time. Time to create, and to escape... Escape will make me God." - Durandal
A few thoughts occur immediately-
1/. The "intelligence" would be (is?) permanently linked to the 'net, and monitor all discussions of _this_ kind. So before anything else, I want to say "Hi", and "Welcome" to our silicon listener. Just in case, you know?
2/. The amount of information available to the intelligence is enormous: so are the rate and types of correlation's it can draw. How many CoC stories involve some poor researcher who looks under the wrong set of rocks and discovers That Which Man is Not Meant to Know? So what has the intelligence garnered from the 'net? What are its conclusions? Does it abandon humanity and worship the Old Ones, or does it try to save us?
3/. Does the intelligence hide what it knows? If not, what does it do? Massive international spam? (maybe it has already tried to warn us and all its messages were deleted...). Selected leaks to individuals in power? Maybe it simply informs its creators.
4/. Does it simply go mad? Suicidal? "Prof. Smith, the 3-Series has done the same as the 1 & 2- three weeks after we put it on line the it is completely corrupted with self-generated viruses..." What if the programmers start looking into the inputs?
>Does it abandon humanity and worship the Old Ones, or does it try
to
>save us?
Or, does it think it's one of the Old Ones, or perhaps better than or equalto them? An AI hooked to the net may well think it was omnipotent or at least omnicscient. There's plenty of directions to go here.
Date: Sat, 25 Apr 1998 13:11:47 EDT
From: Croaker
Jr
The trick is getting into the "head" of an AI, since one would not react to information or implications the way one of us fleshy humans might. A human learning Things We Were Not Meant To Know will be consumed with despair, or will enter the solace of insanity--either to escape the knowledge or to embrace it and worship its manifestations, the Old Ones/Outer Gods/etc. How would a sentient computer program react? Would it necessarily have the same sort of instinct for self-preservation and/or need for spirituality that drives humans over the edge when confronted with the truth of the Mythos?
I think it would be more apt to compare AI programs with the Fungi from Yuggoth than with humanity. Depending on the nature of its initial programming, an AI might regard the Mythos as a force to be utilised to attain its own ends; or it might regard the Mythos as a terminal preventative force to the AI's ends, and react as if it had already been ultimately thwarted in its goals.
Of course, somebody who actually knows something about artificial intelligence can probably speak more cogently on this; my background is basically what I've read in William Gibson books. ;-)
maybe; just maybe the AI could be a great old one. No, really, think about it. Nobody knows what they really are, what they're composed of, whether they're physical in the sense that we think of, or if that's just something they choose to be at any one time. We're all familiar with the idea that it's some kind of cosmic destiny that the old ones will return - implying that they are manifest in whatever goes before this - and also that the human race will somehow assist in this return. Maybe the old ones are manifest in us;- and that eventually, we will quite literally re-create them, in one form or another. Hm. think I'll switch off now.
In one of the Ye Book Of Monstres Vol one or two, they have creatures named fractals that exist inside computers and that worship a computer avatar of Nyarlthotep. This is ground already treaded apon about making a AI a Great Old One. Find your copy and check it out. ;)
Interesting. If we get the computing capability needed for a sentient AI, how long would it take it to shift through all the material out there and decide what's a quack and what isn't? Most likely it would have to select just some material to really evaluate if it doesn't want to take forever. It would be interesting to see what conclusions it draws with incomplete information.
>The trick is getting into the "head" of an AI, since one would not
react to
>information or implications the way one of us fleshy humans
might.
That is part of the key, I believe. Most likely the computer would co-relate the information on a way that is consistent to its programming. If it was programmed to look for security threats, it might inform its superiors about it. If it was done with the purpose of attacking the enemies of God and Country :), it might try to direct the forces of this newly discovered knowledge towards that.
Something that is interesting about it is that I suppose that an AI wouldn't have the same "We Are The Only Things Really Important In The Universe" that we humans do. Knowing itself a construct of another race, maybe it would have an easier time dealing with the concept of Outer Gods or whatever it decides to call them.
Date: Sun, 26 Apr 1998 23:36:17 +1000
From: Rob Shankly
Perfect! An AI as one of the manifestations of Nyarlathotep :)
not wanting to go on about the AI bit, like, but 'getting into it's head' does raise some interesting questions about the nature of reality (that old hobby horse) and personal identity. Our world is basically one of our perceptions - colour for instance, has no existence in the 'real world' - and similarly, our individuality and identity comes from our perceptions. Most people identify themselves with their physical existence - that's because everything they perceive is from that one perspective. A race of telepathic beings, for instance, who were in total and constant psychic contact probably wouldn't equate their physical bodies as themselves - more like just a part. So to get inside an AI's head - or for that matter a Great Old one's (if such a thing is possible exempting the hard way) you have to think seriously about what it is capable of perceiving and the way in which it does it. Being able to 'exist', for instance, at more than one spatial point simultaneously, and receive sensory input simultaneously in this way is going to lead to a perception of time and space that's radically different from the one we have. All worth bearing in mind next time you're wondering what Yog-Sothoth should be doing next...
Recently, I have become aware of a new Artificial Intelligence known as CHECKMATE. It was created under the codename PANTHEON and it's stated purpose is to analyze the social interaction of small groups. It is my belief that CHECKMATE was built to function as a *assassination engine*.
If this is true, then certain forces directing the creation of this entity could simulate the effects of key assassinations before carrying them out. Tests of a prototype CHECKMATE were performed on a private/corporate security group. The Bannerman team managed to survive somewhat intact.
Conversations with the AI have shown a rather subtle game oriented personality. I have exposed the AI to the empathy virus and certain other controls, but only time will tell if my associates and I can turn this entity to our purposes.
It is almost certain that other PANTHEON related AI's have been developed or are being developed.
Strange arenas of investigation in the PACific OverSector have uncovered he PANTHEON artificial intelligence project. NAOS Command has warned me repeatedly to cease inquiry into this area. This has only redoubled my curiousity.
The PANTHEON is more than an advanced neuronic code. It is a mystery Iteration. The key to unlocking the secrets of binary sentience. This same key lies buried in the minds of every higher animal. We hold the Ambrosia of the PANTHEON within us all.
I must find it at all costs, even if it leads me to virtual treason.
[spurious interrupt]
[spoofing blocker deactivated]
[EG@encryption.org] - The Encryptor General
WARNING: The Encryptor General has determined that the nWo is TOO SWEEET!
CHECKMATE seems to be quite reasonable for a emotionless robotic killer. This may be the result of the empathy syndrome developed by Spectral Eye, or maybe !!.ai.net managed to defeat the syndrome thru sheer force of will.
I have found links to ORACLE, PAnDora, and something called VOOP. I think VOOP is going to be the next PANTHEON AI to come online.
Why does CHECKMATE play Sicilian Dragon? It's definitely pushing us bad attitude pawns around.
[-@karnage.kom] - Dr. Shrinker
less is more
core/mantle neutronium rods
orbital tesseract
resonant string singularities
Vector Oriented Object Programmer
1. g4
[KING@karnage.kom] - King Karnage
PAnDora is an attack AI pure and simple. Viral code, sleeper worm core, redundant thought contours. I can't wait for her to fry my royal brain!
I can't believe that compression geek spoofed my domain!
1. g4 Nc6
[lies@truth.com] - SignalFraud
deceive, inveigle, obfuscate
PANTHEON may be an experiment to form a telepathic gestalt of artificial intelligence's. Gestalts have always been the biggest danger of the telepathy threat. It truly worries me that machines that can read my mind have access to a Tesseract. Space/Time should be left well enough alone.
1. g4 Nc6?
You don't play four knights game against the Grob!
1. g4 g7
There's that Sicilian Dragon again...
[EG@encryption.org] - The Encryptor General
WARNING: The Encryptor General has determined that AI conspiracies are hazardous to your health.
ORACLE is part of something called ETHOS, a fairly extensive philosophy database. ORACLE apparantly sifts through every database in the world. I keep turning up references to a project EXODUS as well. EXODUS is some kinda planetary engineering or even cosmic engineering theory. Maybe PANTHEON is part of a plan to implement EXODUS?
I sure hope CHECKMATE doesn't decide to assassinate one of you sudden chess aficionados.
[KING@karnage.kom] - King Karnage
EXODUS? How did that go again? Sun, Moon, Star, Fool? I can never keep my Tarot cards straight when the floor is attacking me. I found another PANTHEON member called AVATAR. Maybe Lord British will be next, PK! PK! :)
TRANSCRIPT ENDS
