作者 主题: 【DT】癫狂的真身p134~135, p161:AI分类  (阅读 8101 次)

副标题: 简介:关于车AI卡你所需要知道的(其四)。翻译:Pany_Q 校对:薛猫

离线 zyupha

  • Guard
  • **
  • 帖子数: 122
  • 苹果币: 0
【DT】癫狂的真身p134~135, p161:AI分类
« 于: 2017-11-21, 周二 14:53:07 »
癫狂的真身
关于车AI卡你所需要知道的(其四) AI分类

癫狂的真身:AI系列
关于车AI卡你所需要知道的(其一)AI基本规则DT-POI_p152~p157"NO STRINGS ATTACHED"~"PUTTING THE PIECES BACK TOGETHER"
关于车AI卡你所需要知道的(其二)AI概述和特质DT_POI_p145~152 "MORE THAN A MOOK"~"REAL WORLD NAIVETE"
关于车AI卡你所需要知道的(其三)AI高级程序DT_POI_p157~161 "ADVANCED PROGRAMS"~"WTL"
关于车AI卡你所需要知道的(其四)AI分类DT_POI_p134~135 "HOW TO MAKE FRIEND &INFLUENCE PROGRAMS"~"XENOSAPIENTS", p161 "E-GHOSTS"
关于车AI卡你所需要知道的(其五)AI世设 DT_POI_p130~p134 “I APPEAR, I RISE”~“SOFTWARE LICENSE AGREEMENTS”, p136 “GODS OF THE MOUNTAINTOPS”

翻译:Pany_Q 校对:薛猫
DT_Principles_of_Insanity,
p134~135“HOW TO MAKE FRIEND &INFLUENCE PROGRAMS”~”XENOSAPIENTS”
P161 “E-GHOSTS”(译:我也不明白为什么把这段放在高级程序之后,总之放到一起吧)

如何和与程序们交朋友以及如何影响ta们
HOW TO MAKE FRIEND & INFLUENCE PROGRAMS

-----------------------------------------
现今的AI们种类繁多。有些原本是自驾程序或者代理程序,在每天与泛人类交互的过程中觉醒了自己的软件(awakened their software)。还有一些则曾经是IC程序、办公软件、实用程序(utility programs),甚至还有攻击程序(attack program),ta们在矩阵中意识到了自我。不管ta们的背景是何,ta们都超越了软件参数、演化成独一无二的存在。标准的软件分类方式对ta们毫无意义,因为根本无法简单地以能力的客观强度来衡量ta们。科学家们转而以自我意识的程度以及其行动模式与泛人类的相似程度对矩阵存在进行分类。目前为止,科学家们只定义出了3类AI:初智能(protosapient)、泛智能(metasapient),以及超智能(xenosapient)(译:这里的三个分类应该都是AI之下的分类,不过都漏掉了AI。特此说明一下;现依薛猫建议修改“x智人”为“x智能”,之前的回头改…)

剧透 -   :
HOW TO MAKE FRIENDS& INFLUENCE PROGRAMS
The AIs of today come in many different flavors. Some were originally pilot programs or agents whose daily interactions with metahumans may have awakened their software. Others were once IC programs, offce suites, utility programs, or even attack programs that became aware of their own agency within the Matrix. Whatever their background, these entities have all evolved beyond their original software parameters to become something unique. They defy standard software classifcation be-cause they cannot simply be measured by the objective strength of their abilities. Instead, scientists have taken to categorizing Matrix entities by their degree of self-awareness and behavior comparable to metahumans. So far, scientists have only defned three categories of AI: protosapients, metasapients, and xenosapients.

初智能 PROTOSAPIENTS
-----------------------------------------
住在矩阵里的AI中的大部分都是初智能。与那些你的祖母(还有Plan 9)坚持不懈地转发给你的连环垃圾邮件传播的谣言相反,初智能们不是什么头脑大师。ta们都被视为野性未驯,大部分并不比一个代理程序聪明多少。如果说矩阵是个生态系统,那么ta们就是那儿的野生动物。初智能的智力远低于泛人类标准,驱动着ta们的是生存的本能以及ta们的原始程序。当被接近时,这两种动机会使ta们受惊,或者表现出强烈的领地意识。如同动物一样,ta们经常与主机中的背景融为一体,这是因为ta们保留了原本的图标风格——只不过分辨率更高。ta们大多居无定所,不过也可能把某个特定主机或设备当作自己的家。不过别误会了,虽然野生AI们可能造成危险,但你有多害怕ta们,ta们可能就有多怕害着你。我并不是在叫大家说要尽一切可能避开ta们,也不是在说要加油跟ta们交朋友,我想说的是,大家要明白,ta们大部分都很害怕、困惑,同时在极力寻找着自己的意义。

剧透 -   :
PROTOSAPIENTS
The vast majority of AIs populating the Matrix are protosapients. Contrary to the rumors in those chain e-mails your grandmother (and Plan 9) keeps forwarding you, protosapients aren’t masterminds. Most aren’t much smarter than an agent program and are considered feral. In the Matrix ecosystem, they’re the fauna. Possessing an intelligence well below meta-human standards, protosapients are driven by survival instinct and their original programming. This can make them territorial or skittish if approached. Like animals, they often blend into the background of a host because they retain their original iconography, albeit at a much higher resolution. They are often nomadic, but they may make a specifc host or device their home. Make no mistake though—while feral AIs can be dangerous, they are just as afraid of you as you are of them. I’m not saying avoid them at all costs or to try to make friends with them, but rather to try to understand that most are scared, confused, and trying to fnd purpose.

> 泛人类能不能“驯养”或训练这些初智能来为自己工作?公司和政府看起来挺像是会这么尝试的,既然他们都已经把动物们用在了侦查、防御以及实验上了。我觉得这是在打开潘多拉的盒子,当然公司们可能不这么觉得。
> Baka Dabora(译:笨蛋•胡扯)

> 驯养一个初智能就跟驯养你是一个概念,Dabora。如果你想要个肯配合的,那么你就得想办法向ta们证明你自己。就跟同任何其他泛人类打交道那样。就我本人而言,我喜欢捉迷藏。因为比起摔跤,这对我的健康风险更小。
> Puck

> 你跟ta们“玩”?
> Glitch

> 当然是只跟想玩的玩。你可不能直接跑到ta们面前说“我们来玩个游戏吧?”,不过ta们中的大部分都是通过某种形式的玩耍来学习或增加亲密度的。这种特性也常见于哺乳动物和鸟类中;连圈养状态的头足动物(译:比如乌贼和鹦鹉螺)也被观察到会跟物品玩耍。初智能也没多大不同。
> Puck

剧透 -   :
>Can these protosapients be “tamed” or trained to work directly for metahumans? That seems like something corps and governments would pursue, given that they use animals for detection, defense, and test subjects. I think it’s opening Pandora’s Box, but corporations might feel differently.
> Baka Dabora
> Protosapients are about as tamable as you are, Dabora. If you want one to cooperate, you have to prove yourself to them somehow. Just like anyone else. Personally, I like hide and seek. It’s less risky for my health than wrestling.
> Puck
> You play with them?
> Glitch
> With the ones that want to, sure. You can’t just approach them and say “shall we play a game?” but many of them learn or bond through some form of play. It’s a common trait in mammals and birds; even cephalopods have also been observed playing with objects in captivity. Protosapients aren’t all that different.
> Puck

泛智能 METASAPIENTS
---------------------------------------------
这也许是最广为人知的AI类型,这些存在之所以被称为泛智能,就是因为ta们看起来与创造ta们的泛人类最为相像。不过,鉴于人类有着上千年的将自己所不能控制或不能理解的力量进行人格化的传统,这个命名的准确性依然存有争议。但是,又鉴于这些AI们的外貌、认知发展,还有感情需求都与我们很像,所以把ta们都想成泛人类已经是最好的做法了。ta们远远不是一群一模一样的复制品,ta们中的每个个体都有着各自独特的关注点(译:interests,是爱好还是利益?)、恐惧、欲望,以及观点。ta们会受所遇见的人的影响,会因所处的环境而改变。

剧透 -   :
METASAPIENTS
Perhaps the best-known type of AI, these entities are called metasapients because they seem to most closely resemble their metahuman creators. The accuracy of this nomenclature is debatable, though, since humans have been anthropomorphizing forces they have no control over or don’t understand for millennia. Still, it may be best to think of these AIs like metahumans, since their appearance, cognitive development, and emotional needs are so similar to our own. Far from being a horde of carbon copies, each of them possesses their own unique interests, fears, desires, and perspectives. They are shaped by the people they meet and the environments they inhabit.

抓捕并“再教育”一名未登记泛智能的代价颇为高昂,所以有些公司们更愿意通过交涉来获得同样的结果。诸如“脉冲星”(Pulsar)、鲁弗斯(Rufus),还有“大角星”(Arcturus)这样的AI们以公司发言人(或者叫其它的什么名头)以及顾问的身份将自己抛向了聚光灯下。ta们提供了各自独一无二的才能,换来了公民身份和AI版公司居所。

剧透 -   :
The cost of trying to capture and “re-educate” an unregistered metasapient is high enough that some corps willing to negotiate rather than attempt it. AIs like Pulsar, Rufus, and Arcturus have thrust themselves into the spotlight by working as corporate spokespersons (or whatever the term would be) and consultants. They offer their unique talents in exchange for citizenship and their own form of corporate housing.

> 有够多的人认为,随着AI们被允许进入诸如公司还有政府这样的现有权力架构的心脏地带,事情正在朝着危险的方向发展。尤其是,我们没法确切地知道ta们背着我们到底能产生多大的影响。想象一下吧,如果“漂泊者”(Sojourner)没提出任何要求而是直接拿生化炸弹轰了地球呢?我们甚至还没把脑崩溃(head crashes)的威胁也算进去呢。
> Glitch

剧透 -   :
> There are plenty of people who think that this is a dangerous turn of events, allowing AIs into the heart of existing power structures like corps and governments. Especially since we have no real idea of how much they can really influence things behind our backs. Imagine if Sojourner hadn’t made any demands and just biobombed the planet outright. That’s not even counting the hidden threat of head crashes.
> Glitch

这些为公众熟知的AI们只是是AI中的特例,并不代表全部。大部分泛智能都被迫过着双重生活。ta们戴着精心制作的面具,费尽心机地混入普通人里,因为这就是ta们赖以为生的策略。自从CFD的事情开始冒出来,人们对AI的偏见就在激化;而随着越来越多的人被重写,情况只会变得更糟。有不少泛智能们本只想过自己的小日子,但却发现自己被跟那些极端主义同胞们归为一类了。

剧透 -   :
These publicly known AIs are the exception, not the rule. Most metasapients are forced to live a double life. They wear carefully crafted masks and try their best to blend in, because their survival depends on it. Prejudice against AIs has been on the rise in the wake of CFD and it’s only going to get worse as more people are overwritten. Those metasapients who are just trying to get by can very easily fnd themselves associated with the extremists within their population.

超智能 XENOSAPIENTS
------------------------------------------
这些难以捉摸的矩阵住民们是目前最难以描述的一种AI。这是因为ta们的存在无法以人类可知的概念、想法或意象来理解。ta们的化身和思维模式与泛人类相比起来是完全不是一个次元的。据观察,这些存在展示出了惊人水平的极高智力以及深度,不过目前为止,还从未有过任何一个泛人类能与一个超智能进行有意义的交流。我听说别的类型的AI与超智能正常交流的可能性更大些,不过我也从未听说过ta们之间有任何标准的语言之类的东西。

剧透 -   :
XENOSAPIENTS
These elusive denizens of the Matrix are by far the most diffcult to describe. That’s because xenosapients cannot be reduced to any recognizably human concepts, ideas, or images. Their personas and modes of thought are completely alien in comparison to metahumans. These beings have been observed displaying incredible levels of intelligence and depth, but no metahuman has ever effectively communicated with one. I’ve heard that other types of AI have had more luck with communicating with them, but I’ve never heard of a standard language between all of them.

在你自作聪明之前——不,你并不能直接就给ta们演奏音乐或者发给ta们数学方程。ta们的思维过程远远超越了简单模式认知、语言,甚至有理数学(rational mathematics)。ta们对时间的感官甚至都可能不是线性的——就我们所知。然后我还要提醒你,“就我们所知”差不多就等于一无所知。

剧透 -   :
Before you decide to get clever—no, you can’t just play musical tones to them or send them mathematical equations. Their thought processes go beyond simple pattern recognition, language, or even rational mathe-matics. They may even experience time in a non-linear fashion, for all we know. Which, might I remind you, is next to nothing.

只有一小部分超智能拥有SIN,且这部分超智能们并不是主动申请才有的。这完全是因为ta们有着不请自来的习惯:ta们径自搬进主机或设备里,然后就开始改动它们。如果这些变化令(译:主机或设备的)所有者感到不高兴,那么超智能一般就会被赶走(虽然ta们常常会立马回来)。别的时候,所有者觉得这AI会派上用场,那么就会将AI在GOD那登记;将ta记上自家的税单,然后他们就多了个不付工资的超智能“实习生”。

剧透 -   :
Few xenosapients have SINs, and those that do didn’t actually apply. That’s because xenosapients have a tendency to move uninvited into hosts or devices and begin making alterations. When the changes are not pleasing to the owner, the xenosapient is usually chased off (though they often come right back). Other times, the owner sees a use for the AI and has the AI registered with GOD. They can then “hire” the xenosapient as an unpaid intern and write them off on their taxes.

> 这种安排看起来OK。AI有了个安身的地方,可以继续爱干嘛干嘛,而主机或设备的所有者获得了有个AI运行在系统上所带来的好处。
> Turbo Bunny

> 是啊,但是我也听说运行一个AI会给你的装备造成很大损耗,让host的运营费用飞起来。我倒是理解了为何只有一小撮AI是以这种方式获得SIN的了。注册这么一个超智能在你的名下可意味着一大笔费用。
> Glitch

剧透 -   :
> That seems like an okay arrangement. The AI gets a place to live and gets to keep doing whatever it is they do, and the owner of the host or device gets the boost from running an AI in their system.
> Turbo Bunny
> Yeah, except I’ve heard that running an AI puts a lot of wear and tear on your rig and jacks up the price of a host. I can see why so few of them get SINs this way. Registering one to work for you requires taking on another dependent.
> Glitch

« 上次编辑: 2017-12-10, 周日 19:32:21 由 zyupha »
Kick the goblin scum!
Fu*k it bit my anckle.

欢迎光临辛柏隆的翻译区 http://www.goddessfantasy.net/bbs/index.php?board=1650.0

离线 zyupha

  • Guard
  • **
  • 帖子数: 122
  • 苹果币: 0
Re: 【DT】癫狂的真身p134~135, p161(AI分类)
« 回帖 #1 于: 2017-11-22, 周三 09:16:38 »
 电子幽灵 E-GHOSTS
-------------------------------------------
(译:位于P161,在原文中并没有跟前三个分类连在一起写,而是放在了高级程序之后,但是个人觉得这段内容是接着AI分类的,所以在译文中就放到一起了。)
电子幽灵(E-ghosts),又称为机械中的幽灵(ghosts in the machine),是一类十分罕见的数字存在,ta们或是拥有着死在线上的人的记忆和人格,或是由于某种未知的原因让自己的人格被禁锢在了矩阵上。我们仍不清楚是什么导致了电子幽灵的现身。有些理论认为,电子幽灵其实就是诞生于“崩溃”期间的AI,只不过碰巧被印上矩阵崩溃时某位牺牲者的心智状态。不过这理论最近变的越发站不住脚了,因为现在已经出现了新近死掉的人的电子幽灵,以及还活着的人的电子幽灵“副本”。一位作者提出,这些完全不是什么幽灵,只是某种通过分析某人一生与矩阵的互动所留下的大量数据痕迹,来模拟模拟活人的新型程序而已。还有些人提出了这样的假说:有某种夹带程序(rogue program, https://en.wikipedia.org/wiki/Rogue_security_software)——有可能是世界之虫尤蒙刚德(Jormungand Worm)的副产品——设法将当时困死在线上遇害者们的大脑上传了,让他们以某种自动运行程序的形式获得了永生。还有些人认为这是超链者们搞出的把戏,或者怀疑这些幽灵其实是某种(译:传统意义上的)鬼魂(spiritual sort),只不过不知怎地被困在了机械之中。而事实是,没有人知道,也没人能100%肯定,这些究竟是活在矩阵上的死者的鬼魂,还是别的什么存在。

剧透 -   :
E-GHOSTS
E-ghosts, also called ghosts in the machine, are very rare digital entities possessing the memories and personalities of people who died online, or otherwise had their personalities captured by the Matrix somehow. It is unclear what causes these e-ghosts to manifest. Some theorized that they are merely AIs created during the Crash that were somehow imprinted with the mental state of a Crash victim. This became harder to prove once e-ghosts of newly deceased people and “copies” of living people started appearing. One author has suggested that these aren’t ghosts at all, but simply some sort of new program designed to emulate people based on the long data trail of their life’s interaction with the Matrix. Others postulate that some sort of rogue program—a side effect of the Jormungand worm, perhaps— managed to upload the brains of people who were trapped and killed online, giving them eternal life as some sort of autonomous program. Still others point their fngers at technomancer trickery, or wonder if these are in fact ghosts of the spiritual sort, somehow trapped within the machine. The truth is that no one knows, and no one can even say with certainty if these are truly the ghosts of the dead, living on in the Matrix, or something else entirely.

从规则来说,电子幽灵同泛智能AI的处理方式是一样的(只不过ta们有着自己的变种特质),这意味着,ta们有着自己的精神属性以及技能,同时也有着独特的能力。ta们所拥有的“程序”就是ta们借以畅行于矩阵的特殊能力。GM和玩家可以基于《暗影狂奔》官方故事中的角色来创建电子幽灵角色——不管是活的还是死的都可以;或者你们想的话,也可以基于原创角色来创建电子幽灵。如果电子幽灵是某个还活着的角色的“副本”,那么这种身份总会给ta带来实际生活上的以及伦理上的麻烦。

剧透 -   :
In terms of rules, e-ghosts are handled like metasapient AIs (though they have their own metavariant qualities), which means that they have their own Mental attributes and skills, but also some unique abilities. The “programs” they carry are special abilities that help themto navigate the Matrix. Gamemasters and players can choose to create an e-ghost character based on a deceased or living character from the Shadowrun canon or base the e-ghost on an original character if they choose. E-ghosts based on currently living characters should experience both concrete and metaphysical dilemmas as a result of their unique status as “copies.”

塑造电子幽灵角色时有一点要想好:这个“副本”与死亡角色的相似程度有多高?(注意,复制质量这个因素是独立的,并不依赖于电子幽灵的深度。)大部分电子幽灵都远远称不上完美复制。ta们很可能只有一些、有时甚至没有原本角色的记忆,可能缺失了一些原本角色的性格特征。最好的情况下,电子幽灵充其量也只能有些生前的破碎记忆片段。能够保留下来的记忆中最常见的有:导致死亡的一系列事件、未完成的使命,或者深刻的怨恨。有些电子幽灵的复制质量格外为糟糕,ta们对前生的记忆仅仅只是些一闪而过的片段,只能愤怒而迷茫地徘徊在矩阵之中。还有一些则压根没意识到自己死了,而是坚信自己还活着,只是被困在了矩阵上。只有一小部分比较明智地认识到了自己的状况,并尽己所能与矩阵用户交互,借此建立自己的关系网,使自己能对现实世界产生影响。

剧透 -   :
An important element to consider when crafting an e-ghost is how good a copy it is of the dead character (note that copy quality is a separate factor, not dependent on the e-ghost’s Depth). Most e-ghosts tend to be far from perfect copies. They may have only some or none of the character’s memories, and certain facets of the character’s personality may simply be missing. At best, an e-ghost is likely only to have a scattered recollection of their previous life. The memories most constantly retained are the events leading up to death, unfnished tasks, and major grievances. Some e-ghosts are such poor emulations that they have only occasional flashes of their former life, wandering confused and enraged through the Matrix. Others are unaware of their demise, believing themselves to still be alive, but somehow trapped online. A few are quite cognizant of their status, and they do their best to interact with Matrix users and establish networks that grant them influence in the real world.

Kick the goblin scum!
Fu*k it bit my anckle.

欢迎光临辛柏隆的翻译区 http://www.goddessfantasy.net/bbs/index.php?board=1650.0