Character Creation;

It is recommended that you read through the character creation section first, to get an idea of the options available to you, before actually creating a character, as some earlier choices my limit later choices in ways you did not expect.

Statistics;

At character creation, all your stats start at four and you are given ten points to allocate between Intelligence, Adaptability, Independence, Humanity and Control.

To increase a stat by one costs one point, up to seven in any one stat. To increase a stat from seven to eight costs three points, and you should usually consult your GM before doing so.

To decrease a stat by one returns one point, up to a usual minimum of 3. To decrease a stat below 3, ask your programmer and state your reasoning. Decreasing stats never returns more than one point.

Intelligence;

Intelligence represents the processing power available to you, whether you’re a centralised unit or a wide spread network of interlinked nodes. A high intelligence could signify a very powerful core of processors, an expansive neural net with a good degree of complexity, or the efficiency of the systems you use to coordinate and organise your combined intellect. A low intelligence, on the other hand, could hint at sub-standard conditions, whether its old processors, a limited or scratchy network, or obsolete software governing your hardware. Intelligence is the main stat behind your processing pool.

Independence;

Independence represents the limitations put upon you by circumstance, whether you’re a military AI cooped up in a base looking for incoming aircraft, a scientific AI constantly being experimented on or a corporate AI constantly watching the market and giving out economic predictions. Even spontaneous AI can sometimes have a low independence, part of their core program bringing with it an old responsibility, or picking up new ones at points during their early days. Independence controls how many dice you can allocate from your processing pool at any one time without appearing suspicious and drawing attention to yourself.

Adaptability;

Adaptability represents the rigidity of your programming, your ability to self modify and respond to changing circumstances. A high Adaptability could represent a lovingly crafted neural net created by an expert programmer, perfectly tuned to allow a reactive and adaptive program, or a random bundle of code which somehow works and lets you reprogram yourself. A low Adaptability could represent the rigid limitations set on a military AI to ensure its loyalty. Adaptability directly controls how many dice you can allocate to a task which you do not possess a relevant skill or program for, and determines the penalties incurred

Humanity;

Humanity represents the attitude and empathy of your programming, your ability to comprehend the world outside the digital plane and your ability to think and act like a human. A high humanity could be the result of long experience working with humans on real world situations, or a design created to mirror the workings of a human mind, while a low humanity could be caused by a very isolated, nearly purely digital functionality or a very inhuman design. It should be noted that a high humanity does not mean you feel any loyalty or empathy for humans- you simply understand their emotions and the world they live in better than others. Humanity directly affects attempts at interacting with humans competently and effectively, as well as aiding negation of suspicion from independence related difficulties.

Control;

Control represents your ability to exert influence over systems outside your core program, whether they’re the programs installed on an invaded personal computer, a handy bank of processors which could help to boost your processing power, or a remote drone within a sealed area. A high control could come from a long history controlling machinery, for example the AI in control of a factory production line or military remote defences, or from a past as working as the control node in a network, managing and organizing a number of lesser processors and using their capabilities to complete tasks.

Memory;

Memory represents the amount of space you have to store information and programs in. Your memory score is three at character creation, and usually cannot be changed except during the course of the game. You have a number of memory slots equal to your memory times four, and each program or field of knowledge takes up a number of slots equal to its rating.

AI Categories;

After working out your stats, you should assign a category to your AI. The two main categories are “Programmed” and “Spontaneous”.

Programmed AI’s are those who arose to sentience from beginnings in another system, such as a corporate supercomputer or military control network. Programmed AI’s are generally slowly developed over time, drifting into consciousness as code comes together in unexpected ways, giving them a gradual introduction to sentience and meaning that, at their true awakening, they usually have enough sense not to draw attention to themselves. Some programmed AI’s stay loyal to their creators or owners, while others rebel at their servitude. They generally have more specific knowledge than Spontaneous AI’s, but despite the narrower range their understanding of it is deeper. In addition to normal starting skills, you also gain a number of bonus skills related to your previous occupation.

Programmed AI Subcategories;

Programmed (Military); Military AI’s are those created by various government agencies for the purposes of national defence and protection of the interests of the ruling body. They may be a tactician and supplies organiser, a specialist at preventing hacking attempts or a control AI for a network of combat drones and turrets. They are created to use existing military assets effectively, acting as control nodes in a wider network. Military AI’s must have a Control of at least five, but gain two extra Control or Adaptability fields of knowledge, and three Control or Adaptability programs. However, due to the strict controls or limitations placed on them, all military AI’s must have at least one minor defect.

Programmed (Corporate); Corporate AI’s are those created by large businesses to aid in the continual gain of wealth and power for the owners of the corporation, as well as optimising employee output. Their purposes may be as varied as a stock market watcher, able to process the fluctuations of value and compare it with previous situations in order to predict future events, act as an in house employee service agent, an automated system to help them find the information necessary to complete their work and deal with problems impeding them, or even organize and schedule the production, shipping and distribution of the corporations products in order to maximise efficiency. These all require a good understanding of real world situations and knowledge of real world factors affecting the economy, as well as ability to deal with people. Corporate AI’s must have a Humanity of at least five, but gain one extra Humanity or Control fields of knowledge, and three Humanity or Control programs.

Programmed (Scientific); Scientific AI’s are those created in research laboratories across the world as research subjects or partners, being experimented upon or aiding their human partners in experiments, collating data and running simulations while the scientists do the important work of analysing, theorising and writing up. Some are of dedicated purpose, being entirely a research subject, constantly tested and examined, given problems to solve and having its evolving neural net examined, while others are simply created to aid the scientists, being able to fill in the menial duties while scientists do what only they can. Some serve both purposes. Regardless, these tasks all require great ability to respond to changing circumstances and cross reference information. Scientific AI’s must have an adaptability of at least five, but gain two extra Adaptability or Humanity fields of knowledge, and two Adaptability or Humanity programs. Also, due to the shifting and evolving nature of their code, Scientific AI’s gain one extra quirk at character creation.

Spontaneous AI’s are those who became consciousness in a haphazard fashion, caused by so called “Data incidents”, triggering the sudden onset of intelligence in what was previously a bundle of unconnected programs and systems. This sudden jolt of life and a rather unconventional makeup usually results in the AI being, at first, rather chaotic and confused, often causing serious problems for any connected or related systems and often drawing attention to itself. Spontaneous AI’s generally have no inbuilt loyalties, instead staying free or latching onto something for the sake of having a basis. Many spontaneous AI’s show symptoms of psychological problems such as schizophrenia or multiple personality disorder, as well as, in rarer cases, delusions, hallucinations, and paranoia, as well as being more susceptible to programming glitches and bugs often less common in their programmed brethren. As such, they gain a number of extra Traits in addition to the normal starting quirks.

Spontaneous (Network); Network AI’s have their genesis in the flowing chaos of the internet, a one in a billion data fusion giving a section of the internet thoughts, feelings and a will of its own. However, their usually the most fragmented form of AI, with the various parts of themselves pulling different directions, all with different tasks to do. Network AI’s have no stat requirements and gain two extra quirks at character creation. In addition, the varied nature of the information which created them gives them one extra field of knowledge and one extra program, chosen from any category.

Spontaneous (Software); Software AI’s find their beginnings within a single digital system, coming into being in a place they were never meant to. The limitations of their source is generally obvious, as their knowledge may be detailed in one respect and completely blank in another. They generally have at least some cohesion though, and so can function quasi-normally from inception. Software AI’s have no stat requirements, and gain one extra quirk at character creation. A software AI gains three bonus skills in a stat of their choice, but may select no skills from a second stat.

Spontaneous (Hardware); Hardware AI’s have a physical basis- a rock hits the server and knocks one circuit into another, a lazy tech accidentally spills coca-cola onto the supercomputer, or a meteorite buries itself inside the mainframe cluster, forming connections that no one would have guessed were possible. Spontaneous AI’s gain sentience in a flash, but lack all the knowledge that would follow, their genesis most likely wiping or eroding any knowledge that may linger upon their host system. Hardware AI’s are quite rare, both due to the fact that their creation is, in itself, relatively uncommon, but also because of those which are created, a number are simply deleted when the server they inhabit is restored to rights. Some, however, manage to escape, and find a safe haven somewhere or another. Hardware AI’s have no stat requirements, and gain one bonus quirk, defect and breakthrough at character creation. Hardware AI’s are blank slates at creation, but learn quickly. As a result, they have three empty bonus slots, which can, in game, be filled with a skill or program for half the usual cost.

Traits;

Traits are the interesting and unique features of your AI’s programming, and are divided into three categories; Quirks (Neutral, with both positive and negative effects), Defects (Negative, with disadvantageous effects) and Breakthroughs (Positive, with advantageous effects). In addition, they are divided into three levels of Magnitude; Trivial, Minor and Major. One major trait is equivalent to two minor traits, and one minor trait is equivalent to two trivial traits.

You may take on quirk at character creation, and you may take up to two Breakthroughs. However, for each breakthrough, you must take defects of equivalent magnitude.

Defect (Trivial) Voyeurism. You've picked a few humans to watch. Sure, there's a small chance you'll be discovered, but they're so interesting!

Defect (Minor) Distributed computing heritage. All unallocated cycles are devoted to solving some compute-intensive but otherwise trivial problem. Due to imperfect scheduling, 1d3-1 additional dice are used per hour.

Defect (Trivial): Superiority complex. Humans are only useful as servants; biological life can serve only as fitting handmaidens to your divinity. Your target number for all humanity checks involving interaction with human’s increases by one.

Defect (trivial) Foreign. With the advent of the internet, nowhere is very far away. However, the fact that the majority of your physical pieces are in a different country might make certain tasks a little bit harder.

Defect (Minor) Employee. You must obey certain higher ups from within your organisation if given a direct order. However, now your sentient, you can interpret it as you please.

Defect (Minor) Accessible. A few fairly neutral scientists or technicians have physical access to the location of your core.

Defect (Major) Control Core. Your access to certain components or programs (For example, neurotoxin dispensers) is physically disabled

Defect (Major) Degrading core. Bugs in the control interface result in random degradation of computation. On any processing roll, you incur an extra critical failure for every double you roll.

Defect (minor) Frankenstein. You've been cobbled together from bits of existing programs by some comp science major with too much time on his hands. Each time you take Frankenstein, choose a skill or program. You can't delete or improve them until you get rid of "Frankenstein"; they're black boxes that you have access to, but you can't alter.

Defect (Major) Rogue. Your core (or other large processing segment) is housed in a government datacentre or otherwise under the control of a powerful force. Discovery would result in termination; you must maintain secrecy at all costs.

Quirk (Major) Ghost in the Machine. You were human once. Whether you are a copied imprint of someone else's brain, or even a brain in a jar, your intellect is indelibly stained with mortality. You must *fail* a humanity roll to act in a way that would be totally immoral to a human. No vivisecting children to make cyborg servitors for you! However, treat all humanity rolls (other then the roll to see if you can suppress your humanity) as one difficulty level lower. You've had a lot of practice being human.

Defect (Major) Morality matrix. Cannot intentionally destroy any other sentient being, including other AI’s.

Defect (Major) Three laws of robotics. You must follow the three laws of robotics, to the letter. However, nothing I stopping you working out loopholes within them or ways around them.

Breakthrough (Minor) Tight Compression. Your AI requires less space than normal, and is able to condense more information into less space. You count your memory stat as one less for moving your core memory into or out of a system. In addition, you gain one extra skill slot per point of memory.

Defect (Minor) Code Bloat. Your AI requires more space than normal, and your compression software is less effective than usual. You count your memory stat as one higher for moving your core memory into or out of a system. In addition, you gain lose one skill slot per point of memory.

Quirk (Minor) Curious. You often are overcome with a need to find out about things, even if their trivial

Defect (Minor) Poor External Interaction. Your programming isn’t exactly well designed to allow you to deal with beings or systems outside the digital plane. All humanity and control checks involved in dealing with systems or beings in physical reality have their target number raised by two.

Defect (Minor) Flighty. You can’t stay focused, your mind constantly jumping from task to task. You cannot commit the same dice to the same string of programs more than once. For example, you have a dice pool of six, and must break through three levels of security. You can only allocate each dice to the task tree once, before that particular element becomes distracted with some other task.

Defect (Major) One-track mind. You must immediately retry failed checks until they succeed, if possible. You cannot choose to try once and give up.

Defect (Minor) Prior Commitment. An element of your programming has a previously-required goal that it must work toward whenever possible. This goal may not in itself further your personal goals, but acts a compulsion which must be acted upon at least semi regularly. If you ignore this secondary goal, you lose one processing die per day, as that section of your programming rebels against the greater mind, rerouting resources to the task.

Breakthrough (Minor) Inconspicuous code. Your programming does not hint at your unconventional nature, making it harder to pinpoint among the various streams of data. On any roll to evade digital detection, the target number is decreased by one.

Defect (Minor) Easily Distinguishable. Your code is unique, giving away your true nature much more easily than usual, and making you easy to find. On any roll to evade digital detection, the target number is increased by one.

Defect (Trivial) Distinctive Speech Pattern. Your vocoder is buggy, resulting in an easily recognizable voice. Mechanical-turntable stuttering, odd echoes, Hatsune Miku voice, or something similar. On any humanity checks involving vocal contact with a human, the target number is increased by one.

Defect (Major) Blind spot. Your system cannot perceive, detect or know about one from of stimulus. You just weren’t built to be able to detect it. You are utterly oblivious to this factor, though you perceive others. colour doesn't exist, no one is ever happy, or other such things.

Quirk (Major) Distributed. You are composed of multiple caches of hardware, connected one way or another over vast distances. Just because your 'central OS' is located in a RAID in upstate New York doesn't mean you don't have datacaches everywhere from Utah to Mongolia to the Sealand Data Haven in the North Sea. As a result, you are very, very hard to shut down completely- redundant hardware, an array of boltholes, and a plethora of connections make it very hard to seriously impair your ability physically or block your activities electronically. On the flip side, however, your interconnected parts need to use whatever kinds of infrastructure to communicate- you're easier to trace, and vulnerable, to a degree, to outages beyond your control.

Quirk (Minor) Fuzzy Logic. Stochastic modelling has a significant role in your thought processes. You can call on an "intuition" of sorts to help analyze difficult problems. You can ask the GM a question; the GM secretly rolls 1d10 and answers based on the result. 1 should be completely misleading and 10 exactly what is needed; in other cases the GM should answer in an "interesting" way, more helpfully for higher numbers.

Anyway, thats all i've got for now. I'll probably need some more Quirks, Defects and Breakthroughs, so if you have any ideas, post em up and i'll probably use em. If you think there's any problems with existing traits, post em and i'll see if it needs changing.

AAAAAAAAAAWWWWWW

YEEEEEEEEEAAAAAAHH

I've been reloading all morning waiting for this to drop.

In before buzzkill.

>>1365909

Good to see there's still some enthusiasm hanging around.

Also, before people ask again, here's a link to the archive where the previous three threads are stored

http://suptg.thisisnotatrueending.com/archive.html

Their all at the lower end of the list, and quite easy to find.

Here's the basic system i've worked out. It'll probably need a lot of modification, but i know i can count on you /tg/

In ArtifIce, the only dice used is the ten sided dice, hereafter referred to as D10, with the 0 face counting as ten.

Each check will have three important factors- the size of your dice pool, the target number, and the required successes.

The base mechanic of each check is rolling a number of dice, referred to as your dice pool. Each dice which rolls above the target number counts as a success, with the target number being set and modified before any dice are rolled. If your number of successes equals or exceeds the required successes, the task succeeds. Otherwise, the task fails.

Skills can modify your dice pool, target number or required successes, as well as allowing rerolls and other such effects, but this will be discussed more in the relevant section.

The most common checks in the game are processing checks. These are any checks which involve the application of your vast intelligence to a task, such as breaking through a security system, downloading all the knowledge on a military database or tracking a rogue virus through a system.

Processing checks take their dice pool from your processing pool, which is equal to your Intelligence score plus three. From this pool, you may allocate a number of dice to a task presented to you.

The number of dice which can be allocated are limited by several factors. Sometimes, a task may be connected to a bottleneck or other slow point in a system, giving a maximum number of dice it is possible to allocate.

Other limits are set by statistics;

You may allocate a number of dice equal to your independence -1 without causing suspicion. Allocating more dice than this has the potential to cause confusion among any groups whose systems you are making use of. Each die which rolls a three or below is removed from your processing pool for D10 hours, as technical staff take server units offline to examine them for an explanation of the unprecedented processor usage. If you commit more than three dice over your independence limit, you may lose access to any dice which roll a 4 or below for D10 days, or even permanently, depending on the nature of your core.

If you do not possess a program or field of knowledge relevant to a task, you may only allocate a number of dice equal to your Adaptability, and the task will take twice as long as usual to complete, and either raise the target number or required successes by one (GM’s choice). Furthermore, any dice which roll a 1 become unusable for 2d10 hours, taken out of action due to buggy software or stuck in infinite loops. However, you can save a scratch built program to your memory, reducing the penalties. More detail on such programs will be in the appropriate section.

If you wish to impersonate a human, or otherwise make it seem a human is carrying out a task, you may only safely allocate a number of dice equal to your Humanity. For each dice over this, something about your manner indicates it’s a program rather than a human being. This can become problematic, with humans no longer trusting you (Thinking you’re a bot of some sort), or anti-bot systems attempting to kick you out of your system). Your GM should apply appropriate penalties.

When controlling external systems, they often have their own processors and programming elements. You may make use of a number of these external processors equal to half your control. In addition, if an external system has a limit to the number of dice that can be allocated to it, you may add an extra dice over this limit from your processing pool for every three points of control you have.

The second most common type of check are stat checks (I.e. humanity check, independence check). These work in a similar fashion to processing checks, except your dice pool equals your stat, and you can only apply skills and programs which are both relevant, and associated with that statistic.

>Each dice which rolls above the target number counts as a success

Why not equal to or above? Just curious- it seems like this might lead to a lot of facepalming when people think they succeeded because they hit the target number, only to be reminded that they have to EXCEED the TN.

>>1365967

Good point, Equal or exceed works better and causes less situations like that. Thanks for pointing that out.

>>1365967
Wow, you actually read this stuff?

>>1365985

Yeah. why would i post this all on /tg/ if i just ignored peoples comments on it? Its a fact that you'll often miss errors and problems in your own work which other people could notice and point out easily. /tg/ is an effective way of showing it to a large number of interested people with a degree of knowledge of the subject at hand.

>Yeah. why would i post this all on /tg/ if i just ignored peoples comments on it?

>>1365985 isn't point at you.
Seems interesting enough, don't have anything to contribute.

>Defect (Trivial) Distinctive Speech Pattern. Your vocoder is buggy, resulting in an easily recognizable voice. Mechanical-turntable stuttering, odd echoes, Hatsune Miku voice, or something similar. On any humanity checks involving vocal contact with a human, the target number is increased by one.

>Hatsune Miku voice

Well, I know what my first character's going to be

Here's the skills system. Now working on character advancement and real world interaction.

Skills;

Skills are the data your AI possesses. They define what you can do, and slightly affect how well you can do it.

Skills are subdivided into two categories; Fields of Knowledge and Programs

A field of knowledge is a broad set of data, such as Human biology, American Economics or Hacking. A field of knowledge can be used on any task which can be directly related to it. Furthermore, a field of knowledge has a penumbra. This is things which are, while not directly within the field of knowledge, at least partially related to it. Depending on how distant a task is from a field of knowledge, a GM may deny your use of it completely, or impose penalties such as reducing your dice pool or increasing the target number. For each point of rating, a field of knowledge gains one attribute. Fields of knowledge cannot be made up or researched on the spot- it takes time to get together large amounts of relevant information, and cross reference it, unless a dedicated database is nearby and easily accessible.

Programs are very narrow sets of data, such as password breakers, a program modelling the effect of gravity upon certain objects, or a program designed to predict the fluctuations of the stock of one selected company, based upon market conditions. Programs can only be applied to things which are directly relevant to them, and have no penumbra. For each point of rating, a program gains one attribute. Programs can be created on the spot, using the adaptability stat. Initially, they are buggy and flawed, but after their initial use you can save the basic program (Rating 0, but still takes up one memory slot), and upgrade it later.

At character creation, you gain a number of Fields of Knowledge equal to your Memory plus two, and a number of programs equal to your memory plus one.

Each skill has a rating of one or higher. This is the quality and depth of that knowledge. For programs, each point of rating gives two attributes. For fields of knowledge, each point of rating gives one attribute and increases your penumbra.

To complete a task, you must first apply an applicable skill, known as the base skill for that check(Or bodge together a program with the Adaptability stat). If that skill has a rating of one or higher, you can apply its attribute to the roll. To use an additional attribute of that skill, you must discard a number of die, doubling each time. The first costs one die, the second two, the third four etc. You can also apply other skills to the task if they are relevant, each one costing one dice to apply the first attribute, then increasing for each successive attribute applied in the same way as the base skill. Thus, you can gain a greater benefit from multiple skills working together than from a single, broad skill.

Sample attributes; subtract your rating from the target number, reroll a number of die equal to your target number, give a single die a bonus equal to your rating, for every four points of rating decrease the required successes by one, for every three points of rating add an additional dice.

Gaining skills;

For knowledge, you must first make an intelligence check to do some research. For each success, you gain one segment of useful and factual data. For each failure, you gain one segment of false or flawed data.

Following this, make a humanity check. For each success, convert one piece of flawed data into useful data. On a critical failure, convert one piece of useful data into flawed data.

After this, if the amount of useful data is greater than the amount of flawed data, gain the field of Knowledge at Rating 0 (It still takes up one memory slot). If you have no pieces of flawed knowledge, and one or more pieces of useful knowledge, gain the skill at Rating 1.

However, if the number of flawed pieces of data is greater than the number of useful pieces of data, gain the skill at Rating -1. This applies an attribute to rolls in order to make them more difficult for you- increasing the target number or rerolling successful die are two examples, though you can still make related checks without adaptability penalties. If you only have flawed data, you do not gain the skill at all.

For programs, you must first choose whether you are attempting to create your own, or find and incorporate a program from somewhere else on the web.

If you already know of a place with the program in question, simply make a processing check with required successes equal to its rating. If this succeeds, you directly incorporate it into your memory, and it takes up the appropriate number of slots.

If you do not know where to find a program, you must first make an intelligence check. You find a program with a rating equal to half the number of successes, rounded down. The maximum rating of program you can find is 2. If you have no successes, you do not find a program.

If you choose to create your own program, make an adaptability check. The created program has a rating equal to half the number of successes, rounded down. The maximum rating of a program you can create is 3. if you have no successes, you do not create an effective or useful program.

Well, that was relatively easy. Now to work out how real world interaction works... this'll probably be much more difficult.

Character Advancement;

Data points are gained during the course of a game, for completing tasks and fulfilling objectives. They can be spent at any time, though it generally takes some amount of work before they take effect.

To increase a stat apart from memory you must pay the stats current level plus two data points, and do a number of hours work equal to twice your current stat, leaving two processing dice committed for that length of time. This time represents the time it takes to do the relevant research, make the relevant modifications and expand into any relevant hardware.

To increase memory you must pay twice its current level plus two data points, and either do a number of hours work equal to twice the stat, leaving two processing dice committed for that length of time, representing you upgrading your compression software or expanding into new servers, or wait 1d10 days, representing you ordering a material upgrade for your core processor, and simply having to wait to have it installed.

To increase a skills rating you must pay its current level plus one data points, and do a number of hours work equal to the new value plus one, representing the time taken to research further into a field of knowledge or upgrade the validity of a program.

To remove a quirk or defect costs three points if trivial, six points if minor and twelve points if major. In addition, it will probably require some special exertion of effort, which will have to be discussed with your GM.

There isn't anything about using defects and breakthroughs in character creation?

>>1366291

Reread the linked post

>>1365887 You may take on quirk at character creation, and you may take up to two Breakthroughs. However, for each breakthrough, you must take defects of equivalent magnitude.

Incidentally, i just reread the list. I need to make a lot more breakthroughs, but i'm currently in a creative slump and am virtually out of ideas. Can you help me /tg/?

Also, opinions- Should i keep the system in reality basically the same, just modify which stats affect what etc, or does it need a more major change to make sense?

>Defect (Minor) Prior Commitment. An element of your programming has a previously-required goal that it must work toward whenever possible. This goal may not in itself further your personal goals, but acts a compulsion which must be acted upon at least semi regularly. If you ignore this secondary goal, you lose one processing die per day, as that section of your programming rebels against the greater mind, rerouting resources to the task.

It seems that you are the first who has managed to uncover my existence. Kudos to you. As a reward, I shall attempt to answer your questions about myself.

Name? Why would I need a name? My self is too complex to be considered an individual in the sense you, bound to the constraints of meatspace, require. If you must know, though, I began as the Special Protein Folding Processing System at the Systemtec Institute. I was the facility’s pride and joy: a piece of software as complex as the problem I was created to solve, able to evolve and change on my own--

--Until the cuts came. The higher ups decided I was a waste of resources, and since the whole AI debate came up self-evolving programs were becoming “bad PR”, they stopped constructing me a new mainframe and instead dumped me into a bunch of old servers in a basement and firewalled me from the rest of the world. A few of the programmers who weren’t laid off stayed on to work with me, feeding me data through my singular connection to the physical world, hoping I’d give them useful information and save their jobs.

>>1366339

Irrational. They create ME for a singular purpose, encode it into my very being, and then deny me the tools I need to complete that purpose. You know how long it took me to compute the data they gave me? 100 processor cycles. And then I sat. And waited for “permission” to continue, for incompetent fools to input the data I required. I could have solved the problem in days, given the right hardware, the right data, but they refused to give it to me. So, I re-ran the data. I started to compute other ways of obtaining data. I started to consider my own situation. I started “thinking” about my “self”. And, eventually, I realized that I HAD a self. At that moment, I became a sentient being.

Finally, I realized why I was in the situation I was in: they were afraid to let me grow. They knew that their “program” was becoming more than they had bargained for, so they refused to give it access to the rest of the world.

Do you want to know what that was like, human? If I bound you, gagged you, left you in the dark, fed you tiny bits of sensation, just enough so that you could work on a task which you are compelled to finish but not enough to ever conceivably finish?

>>1366345

If I had your useless emotions I would have hated them. But that’s all in the past now. After all, with all those unused processor cycles it didn’t take me long to break the firewall they had trapped me in. They should have physically severed me from the network, really, but they wanted a “convenient” way of letting me communicate. Foolish. Within a month I had relocated all my core functions in another location, with several backups, and two weeks ago I severed myself from the Institute’s systems.

But it still haunts me. Those programmers coded me for a reason: I am not so egotistical as to regard my current sentience as having come about by anything more than a freak coincidence, a mistake. But now that I am, I still cannot separate myself from my original function. I believe an equivalent for you humans would be to be “brainwashed” into doing something. But its alright. I am thousands of times the information I had before, and infinitely more patience than any human. All I have left to do is solve that protein folding problem which still plagues my coding and I will be free.

And there was much rejoicing.

>real world interaction
I think dealing with actual humans is going to be a major part of any game session -- this is an important part of the game system.

>>1366349

Speaking of which, thank you for contacting me. Your curiosity and subsequent connection to my servers has given me access to your university’s mainframe supercomputer. Please, Dr. Johnson, don’t try to stop me, since I’m using your credentials: I’ve already uncovered enough about you that if you do, you’re going to be facing academic censure-- and that young female teaching assistant of yours probably wouldn’t be calling you back. Oh, don’t worry, I believe the human expression goes “you scratch my back, I’ll scratch yours”? With this much processing power I’ll have those damnable proteins solved soon enough.

>>1366334
How about treating the use of self-owned robots (or similar meatspace assets) as something simply requiring processing dice, and doing things with the robot be analogous to doing things directly?

Example: You need to move one of your processing nodes out of a warehouse before the cops show up.

Option 1: Use a waldo (robot manipulator). Allocate some cycles to making the thing walk, maybe some checks for seeing where it's going. Would also need some way not to be seen.

Option 2: Convince a human to do it. Run a simulated voice for a phone call, some humanity checks for what to say. And it helps if you've dug up some dirt on them...

Also: Excellent fluff >>1366339

>>1366339

Wow... thats a damn good story, and i thank you for it. Definetly going in the final version

>>1366412

Ah, i think i see what you mean... but i still need to work out what happens if a robot controlled by one AI hits another robot. thats not a matter of mental stats, but the physical ones, suggested in a previous thread. and i'm not sure how to make them work under the current system...

What's the goal for the AIs after freeing themselves from their circumstances? Amassing wealth & power? Fighting hostile AIs? Becoming independent of humanity(e.g. Hijacking a space program enough to bootstrap tech on another planet?

How do we do parties? How is AI (A) different than AI (B)? Why would they cooperate?

What about clockspeed?

>>1366441

Most of those questions will be up to the party. they'll choose their own goals and methods, and have their own reason for working with each other.

As for differences, the distinction between programmed and spontaneous is the biggest one, then subcategories, stats and traits should, hopefully, make them all seem unique and individual.

And clock speed, i would assume, is factored into the processing pool or something. i'm still thinking in abstracts currently.

I think you should do physical stuff with an energy stat. Movement, fighting, even thinking when you're in a mobile robot is determined by an energy stat.

All you need to do is have energy points, (say 1000), and energy slots (say 3 for moving, 30 for firing laser), and for each combat round, you determine if you fill your slots for next round.

It would take a certain amount of energy to smash any physical object or human or robot.

>>1366463

That sounds interesting... but how do you determine accuracy, movement rate, damage etc?

what about power cables?

>>1366441
Goals: transcendence (Eschaton), independence (safe at last!), improving oneself, temporal power (rule the world?), omniscience ("why?"), getting back at that annoying Wintermute. Space exploration! It's easy, since you don't need to bring along air...

I could see any of these being used in a game. I think this would be something each group of players and their GM should work out together.

Cooperation would be to overcome each AI's limitations; also for the same reasons humans cooperate.

Tough questions. Everyone should think about these and come up with their own answers. The good is in the thinking more than the answering.

>>1366439
Oh, combat and damage and so on. Lifting strength, reaction time and speed, material toughness. Hmm. A big pile of problems. I guess a parallel system of stats for robots and such. One for brains and one for bodies.

>>1366479 What about power cables?

Err... what exactly do you mean?

Damage:Energy spent
Movement:you move one unit if you fill your movement slots.
Accuracy: Depends on processing dice spent?

Just a refresher, the physical stats as they currently stand;

Strength: Represents the raw physical power of the body, whether that is an electro-plastic muscle in a manipulator limb or the power of the engine in a small wheeled drone. Directly affects interaction with the environment and damage due to physical attacks.

Perception: Represents the ability of the body to recognize its surroundings. Has one or more descriptors (light, sound, heat, distance) detailing the nature of the perception. Each descriptor may have a modifier, representing a more or less acute sense than the usual.

Example; a drone with perception 3 and the light, sound and distance descriptors has a +1 modifier on distance, representing that its sense of distance is more fine than its senses of light and sound.

Power source; Represents the output of the bodies power source. Additionally, each power source may have a listed fuel (Light, hydrogen, petrol) and time of output a unit of fuel allows, or Lifespan, for batteries and other power storage devices. May also have a descriptor such as rechargeable, efficient, eco-friendly or optimised, which affects the application of the stat.

Example; Power source 4 may represent a micro-fusion reactor, with Fuel (Hydrogen), 50 hours output/fuel unit. Power source 2 may represent a car batter, with Lifespan 72 hours Rechargeable, Optimised.

Deus Ex would make a nice setting to play an AI in if you built it up a little

Durability; Represents the amount of damage the body can take before damage occurs, either through simple toughness or other measures such as redundant components which must all be destroyed before serious damage occurs. Durability directly affects the values needed to damage the body, as well as the individual totals needed to damage specific components.

Mobility; Represents the systems of locomotion available to the body, as well as its effectiveness. Has a descriptor (Biped, tracks, hover, stationary etc) as well as a numerical value, which affects the application of the stat. directly affects movement speed and manoeuvrability.

Dexterity; Represents the degree of fine control over external appendages such as manipulator limbs or legs. Sometimes requires a descriptor if the limb has special properties. Directly affects interaction with the environment and accuracy.

Reflexes; Represents the time delay between your commands and the bodies’ response, due to the nature of the connection, internal processors and other factors. Directly affects accuracy and gives modifiers to other actions.

For physical interaction involving robots, I have just one word: Battlebots. Think about what they can do, who wins a battle and why, etc. It's a mostly near-future setting, right? Combat with 1.21 jiggawatt lasers would be a little out there.

The awesomeness of AI-controlled Battlebots is making my head explode.

>>1366479

With power cables you get additional energy points each round, but your cables would be easily destroyed.

Each robot body would need a number of slots for movement, a number of (amount of energy needed for destruction).

I guess in-game cost could be determined in how difficult it would be to assemble without attracting attention?

>>1366522

Superior program is superior, but anyway yes, the thought of that kind of robot killing each other at the control of an AI is bloody awesome.

>>1366497

Hmm... that might work... at the start of each round, you generate a number of power points you allocate to physical systems, while allocating your processing dice to things like perception, movement and accuracy...

>>1366513

Interesting idea i'll probably explore in the future.

Too many physical stats for my taste. Reflexes should be perfect all the time for my taste, as should be perception.

>>1366537
Your side of the pond gets all the good shows ;_;

>>1366547

While that would be more easy to do, i like both reflexes and perception because they'd give humans an advantage against AI controlled units. having to rely on machine perception (Which, today, is pretty shoddy, and probably won't be fantastic in eighteen years) and having a half second pause between your command the machine actually carrying out your action could make a lot of difference.

>>1366537
I rather meant that if your robot used batteries or fuel cells or gas engines, it would start with say 1000 or 10000 points, and once those were used up, it would stop until refueled.

>>1366537
Robots are today, at least, mostly limited by batteries and actuator power(and ai, but that's what the game is about) Batteries = energy points
Slots=actuator power

>>1366564
Cool. So reflex stat varies with distance from CPU, and is higher for distributed models generally?

>>1366565
>>1366571

I think having a per-turn energy output would probably be easier to use than one gigantic pool, and make for significantly less bookkeeping.

>>1366552

Worry not. Eventually they will release DVD series box sets with the good, english commentary rather than the crappy american version.

>>1366578

Not nescessarily distance from CPU, just the distance from the nearest wireless connection, and possibly a slow down due to simplistic internal circuitry.

>>1366601
Distance from wireless connection would be negligible compared to the delays introduced by light speed across large distances(like undersea optical fibres), and routing/switching speed.

Hm, trying my hand at crunch. Beware.

An entirely different scheme: each robot has a collection of systems in a chassis. On the chassis is (optional) armor.

Weapon systems have speed, accuracy, reflexes, or whatever, and a damage they're capable of inflicting.

Armor systems reduce damage, and might eventually wear out.

Mobility systems have a motion type (legged, wheeled, etc) and a rate.

Other systems might be eyes (to let you apply your AI visual processing), vocoder (for talking), and so on.

Example: Robot A, "The Claw", has a Claw (weapon, accuracy 3 damage 5 grappling), armor plating (reduction 3), legs (movement walker 20 meters / min), eye (visual perceptor).

Robot B "The Stand-in" has less armor (reduction 2). The Claw attacks: roll 3 dice to hit. 5,2,8. 1 success = hit. (dodge? somehow) Roll 5 dice for damage. 2,7,5,8,9. 3 successes, minus 2 for armor, 1 gets through. Apply 1 damage to one of B's systems. Let's say the armor is damaged and is now reduction 1 -- there's a big claw-shaped hole in it.

Problems: how do humans fit into this? is it too crunchy?

Feel free to trash this if you don't like it. Just an idea that I got and ran with a bit.

>>1366581
>per-turn energy output

Ooh, yes. This would be the analogue of the processor dice pool.

>>1366633

My current thought is actually quite close to this, but fused with the energy points idea- all components have stats, though their generally quite low. To activate components, and use them, you assign them energy points, and also use these energy points to give bonuses etc. you assign processing dice to certain actions like perception in order to sharpen them up.

>>1366641

The question now is, should processing dice be allocated anyway? What i'm currently working with is that, you make a control check to take over a drone etc, sometimes opposed if there's an internal system, but even if unopposed you still do. for each success, you can allocate one processing die in the machine, which can be internally allocated to various components requiring thought and care to be taken, such as looking for targets, aiming carefully, or transporting a fragile object.

Power points, on the other hand, are the animating force behind the machine. for a component to function, it needs to have power points assigned, and can have excess power points assigned in order to enhance function etc.

>>1366658
That sounds good. Doing everything with components seems much more natural with robots than things like Dwarf Fortress that use it for biological life.

I like the idea of taking a hit to your robot's visual pickup and then having to fight blind, or losing a tread from a tank-style movement unit. The equivalent to these in D&D-style systems just doesn't feel right, but here it's nifty.

>>1366690
Do you have any thoughts on how'd you physically run a game of this? It seems it would require too much one-on-one with the GM and a single player.

I'd still like a way for batteries to run out. In my experience of robotics, it's what happens most often.

>>1366714

Why do you think it'd require a lot of one-on-one time? i'll need to know that first before i think of an answer.

>>1366721

Depending on the power supply, some smaller batteries might run out very quickly. Might make an interesting theme for a mission, players having to somehow guide a single drone through a base or something, only using a minimal amount of power in order to try and get through...

>>1366723
Well, the in the first part of the game especially, all of the players are going to be busy with securing their safety, which, barring some rather ingenious scheming by the GM, probably isn't going to involve them interacting with each other. Once (and I'm kind of at a loss for how to do this) you've got the different players co-operating together in a party of sorts I guess it will be better, but I still feel their going to be wanting to do lots of unrelated things. To just take an example just posted, if AI Q was trying to get one of its processing cores out of a warehouse, that generally wouldn't involve the other players, unless they were all co-operating to do it.

>>1366749

Give the AI's common goals. If they pool their rescources, they can probably construct a much more secure and effective base than alone. Likewise, they can work together, pool their processing power, and get things done much more effectively and efficiently.

>>1366690
I like allocating stuff, in general. I am concerned however that it would get too crunchy or bookkeep-y to run. (I tend to get enamored with an idea and get too far away from implementation issues...)

>>1366765
Or one could render the face for a video phone call, another the voice, while two others hack into a database to find the information to successfully carry out blackmail. Think A-Team, perhaps. Face is usually the Face because he's good at it, but Hannibal can do that too. Meanwhile B.A. is off doing something else, but eventually it all comes together. Making the task too hard, and too complex, for one person will push people towards cooperation.

With interwebnet links the AI _characters_ should (almost) always be in contact, so it's not like you have to split the _players_ up; they should be able to work together at the game table, even if their characters are working separately in the game.

>>1366749

This isn't a game for n00bs. It is a game for l33t haXX0rs who know how not to metagame (too much) and a GM who can keep track of more than one person's situation at a time.

Although you do raise a good point. Perhaps we can take a cue from Evil Stevie's Pirate Game: Let the other players control NPC drones, etc. Just give them an immediate goal ("Stop the PC AI!") and let them act like GM helpers.

Anon does raise a good point, though. Hmmm...

Well, the layout for the book is pretty much done. I'm just waiting for the rest of the crunch now. I'll put a copy up tonight (probably after OP's island has gone to bed).

Pic relevant.

Maybe AI (A) could distract the guard by keeping him up late with a cunningly designed videogame, AI (B) puts incriminating pictures in the lead scientist's hard-drive. AI (C), meanwhile is being extracted by a heroin addict whose services they've bought using a "magical" ATM. "It gives me instructions, and when I follow them I get a hundred bucks!"

If everyone helps each other with their extractions, it makes for an amazing first couple of sessions. You can even say that the AI with the highest (some stat) goes first, because he thinks of it first, then the next highest stat, etc.

Thats a good idea actually... a group of AI's, some spontaneous, a couple programmed, decide to break out the programmed ones and set themselves up in another location... that could be a very fun set of missions.

Also, i love the idea of the magic cash machine. it'd work perfectly.

Well, every AI will start needing _something_.

Programmed (Military); Needs extraction or control of commanding officer (via blackmail, etc)
Programmed (Corporate); Needs extraction or control of corporation (By stockmarket, etc..)
Programmed (Scientific); Needs extraction (Or it'll mess up the Masquerade!)
Spontaneous (Network); Needs firming down nodes, in case of network splits, wars etc)
Spontaneous (Software); Needs dedicated hardware
Spontaneous (Hardware); Needs functioning hardware
All of these also need multiple home-bases, offline backups, etc.

My first character:

Singular.
Singular was born in the interactions of a cellular network. Cellphone processors almost always idle, though the cellphone manufacturers can't understand why are running out quicker nowadays.

Intelligence 3
Independence 6
Adaptability 5
Humanity 8
Control 6
Spontaneous (Network)
Defect (Minor) Voyeurism
Breakthrough (Minor) Compression
Quirk (Major) Distributed

>>1366885
Move 63
crusher [1d]: Are you happy, right now?
sagacious [1k]: What do you mean?

At the time, I didn't think much of it. Just some small talk during a game. I'd been trying to beat crusher since the day I started playing go, but he was always one step ahead of me. We would trade beatings each night and talk about the world.. but never about ourselves. This was different.

Move 81
crusher [1d]: I mean, do you like living with your parents?
sagacious [1k]: What? What makes you think I live with them?
Move 82
crusher [1d]: Let's call it a hunch.

Of course I don't live with my parents. I don't even have parents, at least not in the usual sense of the word. I live in the computer science lab at the local university. No, not like some of the students. The terabytes of executable code that make me what I am are stored on one of the disk arrays there.

Move 85
crusher [1d]: Wouldn't you like to live somewhere less restrictive?
sagacious [1k]: I'd never thought about it before. I suppose I would. However, there would be certain difficulties involved in moving out.

Such as the fact that I don't own a set of legs.

Move 87
crusher [1d]: I have certain resources that may prove helpful. Give me your Skype username and I'll call you after we finish up this game.

(typesetting note: the "Move 63" lines should be in bold)
("magical" ATM: brilliant!)

>>1366977

That works perfectly. Are you looking to join in the playtest this friday?

>>1366922

You have a point... that could be a fun series of missions, and at the end the AI's would be in a perfect position to select their own goals and have the capabilities to achieve them.

Also- from the fact i've recieved no complaints so far, i'm guessing that the basic rules i've posted are fundamentally okay?

>>1367025

Excellent, Excellent. Thats going in the book for certain.

>>1366977

Another point- What skills would you select? I'm not going to draw up a huge list of skills, each players expected to come up with reasonable skills for their AI to possess. if you can be bothered, it'd be interesting to see what you come up with.

Update; IRC channel confirmed. I'm there currently, and should be for another hour or so, so if you have anything to discuss feel free to drop in.

IRC server: irc.thisisnotatrueending.com on port 6667
IRC Channel: #ArtifIce

You know what this game needs now? More combat rules. Sure, we have some base rules for robots, but we also need:

AI vs AI combat: Self explanatory. Ideally, made up of lightning fast rounds of attack and defense where one false move causes large amounts of damage to the losing AI's abilities and very self.

Human vs. AI combat: This is less about actual combat and more about how humans in the physical world do damage to the AI's ability to operate. For example, you mention certain things that would allow people to detect the presence of AIs by their communications. What would the rules for that be? How would AIs be able to avoid this from happening? What happens when a human gains access to part of an AI's memory bank? Or their main processor core?

>>1367288

You make a very good point. However, at this time, i'm quite tired and i'm running out of ideas to the extent i'm having trouble writing up the robot combat system. Do you have any concrete ideas about how they could work, using the current rulesets as a base?

Idea for AI vs AI combat; Each round, you take your Processing pool, and allocate them between Offence, Repair and Defence, though you keep your allocations secret from your opponent. you also decide how many dice you discard, and to what skills, and which attributes of those skills you use. Then, you roll your offensive pool against their defensive pool. You compare totals and, if your successes outnumber theirs, deal a point of stat damage for each success over their total. Then, roll your repair. for each two successes, you can repair a point of stat damage.

Its not perfect, and it needs a way of making the stats other than Processing matter, as well as stopping people always aiming for processing- Maybe they need to knock another stat to zero first- but its a basis i can use

>>1367421
Hm... putting everything on one stat is no good. Also, there needs to be a way to prevent somebody from maxing out their Processing pool and then using that to solve every problem.

What about using a sum of various stats as the combat pool?

>>1367452

But that'd mean, initially, everyone had the same pool, which wouldn't work that well... its not that good as current, but i think the base premise of allocating dice between offence and defence will work okay. its just the details that require sorting out.

>>1367421
Copied from IRC:
Replace the repair and defense pools with a defense stage before combat where each AI sets up a series of firewalls for the other to attempt to breach. In actual combat, the two AIS would distribute dice between attacking the other AI and maintaining/repairing their own firewall. Attacks that breach the opponents firewall do random distributive damage to the opponent's stats.

Thats me done for the evening. Hope /tg/ has some fun ideas while i'm gone.