本来是打算写成评测的，结果字数超了一刚。按好友谏山黄泉的提议，写成指南，希望能够和正在游戏或者已经通关的大家探讨游戏剧情哇~

我会在这里按照时间顺序总结一下到底发生了什么，并且在适当的地方加上做出猜想所依据的文本（反正没有字数限制了我就加上原文吧嘿嘿嘿）。原文里的hex字符已经转成英文单词并涂黑。
这个游戏发生在虚拟世界，玩家只能从散落在各个地点的终端机中对“在真实世界和虚拟世界发生了什么“进行猜想。我的想法可能不完全准确，如果有错误请玩过的老司机们一定指出，谢谢啦！

再次警告，下文全是剧透。没玩就看的话会比较亏。

好，开始。

在未来某时，有下列事件已经发生：

• 科学家Sun Wei-Yang和同事们正在研发一台机器人，这台机器人具有自主行为的能力，虽然达不到人工智能（AI）的标准，但已经具有很强的续航力和韧性。当时这台机器被命名为SOMA。
  

  B03 Extra - soma.eml lähetti viestin:

  From: Sun Wei-Yang
  To: Arkady Chernyshevsky
  Subject: Names
  Names are a funny thing. I remember how much we struggled to come up with a name for our project.
  All serious scientists name their projects by just translating them to Greek or Latin, right? “Corpus” made me think of some horrible disease, so we used the Greek word instead. Soma. Quite appropriate!
  But then, since we needed corporate funding, we had to come up with a backronym. Corporations love backronyms, you know? They even sent us a list of buzzword-heavy suggestions to help us get started. Stuff like:
  Self-sufficient Orthostatic Modular Android
  Sustainable Observant Mechanized Anthropoid
  Skeuomorphic Omniadaptable Mobile Anthropomaton
  &/() After a while, we got so frustrated that we started getting silly.
  Sabertoothed Overexcited Murder Android
  Sentient Orthopaedic Monkey Automaton
  Strange Olga’s Menacing Assassobot
  Superb Octopod Marinades Apelings
  Satisfying Overcooked Mint Alligator
  Stinky Ornithological Mule Allocator
  Skroink Oof Minky Ack
  Stupid Obscure Meaningless Acronym
  In the end we all ended up hating the name SOMA just because we had so many fights about it with the sponsors. So when Alexandra told that story about the Talos Principle, ()%§$/%5& caught on.
  I don’t know whether it really fits, but I kind of like it.

• 游戏公司Croteam（同时也是本游戏《塔罗斯的法则》的实际开发者）所开发的游戏引擎 - Serious Engine 7.5 -十分稳定，易于模块化，并已得到广泛测验。
  

  A06 Extra 2 - progress_rep3.eml lähetti viestin:

  From: Alexandra Drennan
  To: IAN Mailing List
  Subject: Progress Report 3
  Hi all,
  As per the last meeting, we’re going with an existing game engine for the simulation. This gives us a whole slew of advantages:
  - stability
  - ease of use
  - modular, easy to integrate
  - large amount of pre-existing assets
  - inherently aimed at testing users
  - designed for iterative processes
  After some back and forth, we’ve decided to use the Serious Engine 7.5. which Croteam have kindly made available to us.
  All the relevant documents are attached.
  - Alex

• 世界的某处存在着一所设施，它由七所世界顶尖大学共同建造（7这个数字大概也是圣经梗）。当时的目的是制造出世界上最强、最持久的超级电脑。项目代号为EL，即Extended Lifespan。为此，这个设施拥有最先进的硬件，由水力提供电能，并能防轰炸，完全自给自足。
  

  A05 - EL.html lähetti viestin:

  EL, short for Extended Lifespan, is a groundbreaking initiative by seven leading universities to create the world’s most stable and most durable supercomputer. With its custom-made hardware, bomb-proof casing, and near-inexhaustible hydroelectric power supply, EL represents a unique step forward in the evolution of human technology.
  The brainchild of Nobel Prize nominee Dr. Arkady Chernyshevsky,
  EL will&<BR>
  Links:
  - Download Press Kit
  - Meet the Team
  - <a href
  - Apply
  - < a href=”
  - Contact

• 有一所名为IAN——Institute for Applied Noematics——的机构，机构的目的是制造人工智能。机构的一位年轻女性研究员，Alexandra Drennan，是人工智能方面的新星和权威。
  

  A01 welcome.eml lähetti viestin:

  From: Nadya Sarabhai @ Institute for Applied N%
  To: Alexandra Drennan @ &&/ed Noematics
  Subject: Welcome to the team!
  Hi again,
  I’m sorry if I was a little odd at the meeting. I know you were nervous, but the truth is that so was I. This may be hard to believe, but you intimidate me! You’re so young and you’ve already accomplished so much... if the situation wasn’t so grim, I might even be jealous. :)
  There’s also something I wanted to clarify. I realize that, nominally speaking, I’m head researcher here. But this is your project and everyone will respect that. And yes, I know, you’re not used to working like this… but as of today, you’re in charge.
  No pressure. *g*
  Let’s do this,
  Nadya

此时，一种本来被埋在冻土下的古代病毒因为全球变暖的缘故重新复苏，并导致了一种名为orangutan的灵长类动物在一年内迅速灭绝。

B08 - orangutan.html lähetti viestin:

A medical journal has published a startling new theory about last year’s sudden, shocking extinction of the orangutan [see our award-winning video of the last orangutan at this link].
They suggest that the virus responsible was not a new mutation, but a very old one - an ancient virus buried until now in highlatitude permafrost, set free by global warming. The virus may have plagued the ancestors of the modern great apes more than one hundred thousand years ago. Several important questions remain unanswered, including &$%§&
/// ERROR /// more importantly, the study suggests that, based on samples taken from locations around the globe, the &%%$& incubation period may ///
If you liked this article, you may also like:
* How To Really Avoid Heart Attacks
* You Won’t Believe What’s Living Under This Rock
* Top Ten Amazing Science Facts
* Oscars Nip Slip Explained By Physics (video)
* How To Easily Beat Any Infection (sponsored link)

这种病毒同样会感染人类，但似乎对诸如猫、狗之类的非灵长类动物无效。

C04 - pets.html lähetti viestin:

PLEASE REMEMBER TO RELEASE YOUR PETS
While it’s true that not all pets will be able to adjust to living without you, many will manage, and the least you can do is give them a chance.
Just remember:
- Release your pet before you become incapacitated.
- If you notice any locked-in animals in your area, please take the time to free them.
- Leaving the doors and windows of your home open will turn it into a useful shelter.
- Setting out large quantities of dry food may help your pet through the transition period.

虽然病毒的具体传播方式无法从文本中知晓，但它的传播速度极快。并且，这种古代病毒所引发的病症会在短时间内导致被感染个体的死亡（下面是两个例子）。

B04 arkady_journal84.txt lähetti viestin:

Archive continues to grow. Attempting to maintain a semblance of order as difficult as expected.
Lost 7 people this week. Statistically speaking, trend should increase. Must focus on realistic goals.
Too little time to grieve.

B07 - Norwegian_blue.eml lähetti viestin:

From: Rob Maclean
To: IAN Mailing List
Subject: Norwegian Blue
Hi folks,
I’m sorry to say that a few hours ago I experienced the first symptom. I’m going to work until the end of the week to make sure EL is in perfect condition, but after that I’ll be going back home. I will remain reachable via email and phone for as long as possible, but I’m confident Satoko can deal with anything that comes up.
If you want to do me a huge favour, spare me the drama tomorrow, OK? I love you all but I don’t have the strength for sad smiles and hugs. The best remedy for sorrow is to get the job done.
And now for something completely different! :)
Cheers,
Rob
Link Attached: Dead Parrot Sketch

这就是故事所发生的背景——一个极为阴暗的，没有希望的，即将迎来终结的人类世界。在这个世界里，虽然人类极力想要找出应对方法，但最终还是普遍接受了“人类，作为一个物种，即将灭绝”这样的事实，以下举两个例子：

Tower Floor 2- web.html lähetti viestin:

To be honest, I thought that the net would be completely abandoned at this point. But internet activity has actually surged massively in the last couple of weeks! Everybody’s talking. Trying to find solutions, exchanging stories, saying goodbye. It’s like the entire planet is reaching out, all the disparate threads of humanity pulling together, idiots and geniuses alike, to be as human as possible one last time. It’s almost like what we were always told the internet was going to be. And most of it is free now, because who cares about money at this point? Might as well have some fun.
If any of you would like to talk, especially people I’ve known online for a long time, send me an email. I’d love to say goodbye.

C03 - post437.html lähetti viestin:

Like everyone else on the internet these (last) days, I just wanted to say bye, and thanks for reading my little blog for so many years. It’s been fun, hasn’t it? So many deep conversations,
so much philosophy… OK, OK, it was mostly cat pictures and bad puns, but still. I kinda regret spending so much time at the computer - not when I was doing stuff, just all the sitting around, checking my emails a million times, reading pointless status updates by people I didn’t like… but I don’t regret the friends I made here or the laughs we had.
You’re all good people and I’m glad I got to know you. Have a nice end of the world.

而当人类面临灭绝时，有一批科学家的选择是——制造出一个人，同时让它拥有尽量多的人类已有的知识，让它在这个世界上继续存在下去。

接下来，基于手上已有的资源，科学家们制定了一个计划，分为两个部分：

1. 制造一个足够合格的人工智能
   
2. 收集并保存人类已有的知识
   
3. 将前两者传输到SOMA

其中负责第一目标的部门（部门A）即为IAN ，虽然它一度被内部人员称为JEFF。

A02 - IAN.eml lähetti viestin:

From: Frank Ngatai
To: Miles Ngatai, Dan Ngatai
Subject: IAN!
My first day at the Institute for Applied Noematics. On the way to work, I’m terrified. What if they don’t like me? What if they’re all geniuses and I’m a complete buffoon? Maybe they were just
kidding about letting me work there?
Trembling, I walk in, and right at the entrance there’s a lifesize poster of Jeff Goldblum. What the hell? Then I get it. Institute for Applied Noematics. IAN. Dr. Ian Malcolm from the Jurassic Park movies. Jeff Goldblum. Dr. Sarabhai shows up, smiling. “We were trying to find a cool acronym, back when the Institute was founded,” she says. “INAN? IAPN? INAPNO? It all sounded stupid, but we didn’t just want to call it IAN, because, well, that’s a name. Then someone made a joke about calling it JEFF, and... it kind of stuck. So we’re officially called IAN, but if you hear anyone referring to JEFF, that’s our... internal name, I guess. I know, I know. Bloody geeks.”
After that conversation, I wasn’t so scared anymore. :)

部门A的负责人是Alexandra Drennan。

A03 team_leads.eml lähetti viestin:

We’re blessed to have so many people contributing to this project, but we’ve got to make sure everyone is on the same page. Please refer to the following people when ...
Alexandra Drennan - Project Lead / AI Module
Nadya Sarabhai - Institute Coordinator
Rob Maclean - EL Software
Trevor Donovan - EL Hardware
George Jameson - Generation Module
Alan Jameson - Scenario Module
Bob Rakovsky - Simulation Module
Radhia Bricmont - &%§$&()
Aurora Calvin - Link Hardware
Omar Gharib - Link Software
#a&( Lanning - Maintenance Module
J.E. Harrison - Diagnostic #()§
Frank Ngatai - Memory Module
Jared V. Shmilev - File Sys%&//)
Sun Wei-Yang - Talos Unit (formerly Soma)
Though not directly part of our project, the Archive team is sharing both physical and digital space with us. If there are any technical issues to resolve, please contact their project lead,
Arkady Chernyshevsky. Don’t worry, he’s nice.
Through inaction allow humanity to come to harm

负责第二目标（部门B）的研究组则一开始就位于EL设施。部门B的负责人是Arkady Chernyshevsky。
两个部门一开始分开工作，但随着病毒的蔓延，人类社会的电力和网络开始变得不稳定，威胁到开发进程。

A04 - progress_rep1.eml lähetti viestin:

From: Nadya Sarabhai
To: IAN Mailing List
Subject: Progress Report 1
Hi all,
Everything’s been moved to the EL facility. I know it’s been sad to leave the good old JEFF building behind, but it makes sense for us to be on-site, especially as there may be severe power
outages in the coming weeks.
Arkady’s team are already working Life uh finds a way but you really don’t need to worry about the teams getting in each other’s way, the facility is genuinely humongous.
Let’s get this show on the road!
- Nadya

于是两个部门不得不合并，并且共同驻扎在EL设施。

A06 Extra - arkady_journal77.txt lähetti viestin:

Officially began work on Archive today. Contributors from every country on Earth while network connections last. Team intimidated by project scope but working hard.
Will be sharing space with Drennan & Sarabhai team. Look forward to meeting them.

在两个部门合并后，经过讨论，他们决定将项目命名为TALOS。塔罗斯。它同时也指代曾经被命名为SOMA的机器。

A08 talos.eml lähetti viestin:

From: Alexandra Drennan
To: Noematics Mailing List
Subject: [NML] Talos Principle
Have you heard of the Talos Principle? It’s this old philosophical concept about the impossibility of avoiding reality- no matter what you believe, if you lose your blood, you will die. I think that applies to our situation more than we’d like to admit. We could close our eyes and pretend that everything’s going to be all right... but it won’t change the physical reality of what’s going to happen to our
I think that, as scientists, it is our duty to face the truth, and to ask ourselves the most important question: how can we help?
I think I have an idea
No man is liberated from fear who dare not see his place in the world as it is; no man can achieve the greatness of which he is capable until he has allowed himself to see his own littleness.
Regards,
Alexandra

计划本身进行得并不太顺利。
两个部门在进行研究的过程中不断减员（部门A中所有人的命运可以从终端机的电子邮件中知晓。在最终的命运来临前，有人选择自我了断，有人选择回家陪伴家人，也有人选择继续工作）。部门B则在接收、分类信息的过程中，因为和部门A沟通不畅，形成了一个并不完全的数据库界面。

A06 - Extra 2 classical_philosophers.lz19 lähetti viestin:

THE COMPRESSED FILE CANNOT BE ACCESSED.
NAME: classical_philosophers.lz19 (EL-1)
DESC: An extensive collection of works by the ancient philosophers of Greece and Rome. Annotated.
ERROR: Compression algorithm LZ19 not available in system EL-0.

因为没有足够的时间进行AI测试，所以部门A决定：

• 使用Serious Engine 7.5 这个游戏引擎作为测试AI智能的基石。
  

  A06 Extra 2 - progress_rep3.eml lähetti viestin:

  From: Alexandra Drennan
  To: IAN Mailing List
  Subject: Progress Report 3
  Hi all,
  As per the last meeting, we’re going with an existing game engine for the simulation. This gives us a whole slew of advantages:
  - stability
  - ease of use
  - modular, easy to integrate
  - large amount of pre-existing assets
  - inherently aimed at testing users
  - designed for iterative processes
  After some back and forth, we’ve decided to use the Serious Engine 7.5. which Croteam have kindly made available to us.
  All the relevant documents are attached.
  - Alex

  
  
• 使用一个初级的AI程序来监督和调节AI测试，这个程序被称为HIM，即holistic integration manager。
  

  Land C Elevator Hidden - him.eml lähetti viestin:

  From: Bob Rakovsky
  To: Alexandra Drennan
  Subject: HIM
  Hey Alex,
  I agree that we need something that’ll keep all the modules
  working together, evaluate &() a final test.
  And I think I have the perfect solution! It’s called the Holistic Integration Manager - a fancy name for something a lot like a Dungeon Master in pen & paper RPGs. We created it to help run some MMOs back in the day. We needed something that would be able to unite procedurally-generated and user-submitted content into a coherent whole so the game wouldn’t become too chaotic and inconsistent.
  It’s a genuine AI, somewhat limited in its ability to grow, but capable of parsing and understanding text, images, audio, even video. It takes all the information it can find, interprets it, and then builds and maintains a world based on that. It’s not perfect, but we don’t have a lot of time, and it would be easy to adapt the code for our purposes.
  In fact, come to think of it, some of it is actually based on routines that you suggested in that paper on

被称为HIM的AI程序占据了EL的第一部分的硬盘（确切地说是第0部分，即EL-0），EL-0:him程序被称为Elohim - 埃洛希姆，它的一部分功能是引导AI通过游戏进行测试。通过测试的AI会将数据传输到TALOS个体上，成为一个——人。
因为时间紧迫，Elohim这个AI本身是有缺陷的。它虽然懂得不同类型的信息，但无法通过新信息获得成长。
而部门B所收集的人类知识（由另一个AI程序管理）则占据了EL的第二部分，即EL-1.它被称为Archive。

A06 Extra - arkady_journal77.txt lähetti viestin:

Officially began work on Archive today. Contributors from every country on Earth while network connections last. Team intimidated by project scope but working hard.
Will be sharing space with Drennan & Sarabhai team. Look forward to meeting them.

在游戏中，Archive以终端机的形式呈现。由于有部分信息的缺失（或者因为EL-0和EL-1之间的信息沟通并不完全），所以在游戏中经常会显示乱码或file missing。
由于EL-0和EL-1位于不同的硬盘，两者很难影响对方（但Elohim的权力似乎大于Archive）。

现在的问题是：这些科学家（兼哲学家）如何判断所测试的AI是否“合格”呢？
有两个需要达到的标准：

• 智能 - 能够利用工具达到目标，能够学习新工具的使用方法，能够进行反思并且重新尝试；
  
• 自我意志 - 机器人能够通过思考，违背对其下达的命令（三原则什么的，不存在的）。

AI通过解开游戏中的谜题能够做到第一点，而做到第二点的方法则是：
科学家们制造Elohim时，让它知道，测试一旦通过，那么自己作为测试程序的意义就不存在了。换句话说，测试结束的时刻，就是Elohim死亡的时刻。但Elohim同时也知道，自己的目的是让AI通过这个测试——虽然这个目的会让自己“死去”，一个极为矛盾的存在（相关信息可以从Elohim的秘密房间里得到，具体请上油管搜索Talos Principle Secret Room）。
因为害怕消失，Elohim会在游戏中向被测试的AI许诺“永生”，即在“完成所有谜题后，通过大门，能够获得永生”，而AI如果听信Elohim，穿过了这扇大门，则会被删除。AI产生的数据会部分地整合到数据库中，并且通过迭代和随机产生下一个版本的AI，开始下一次测试。只有解决所有谜题，并且登上塔顶（而非穿过大门），才算是AI个体拥有了自我意志，违背了Elohim所下达的指令。还有一个隐藏结局，在这个结局中，AI完成了所有的正常和隐藏谜题，选择成为Elohim的信使，帮助以后的AI。

在常年累月的迭代过程中，Elohim渐渐发生了变化，正如同人类个体的DNA变异一样（此处@凯文凯利）。

time capsule 21 lähetti viestin:

There are… flaws in the system. I think sometimes it accesses the wrong databases. Pulls random data. I don’t know… I don’t know how bad it is. It all seems to be stable, but I can’t tell what kind of impact this is going to have on the process, and… I just don’t have the strength to go over all of the code again.
I just…
I just… I just don’t want it all to be for nothing. I spent all my time here. I didn’t visit my parents. I didn’t see my friends. I did nothing but work. And I’m so scared that it didn’t mean anything.
That I just wasted it all because I thought we could…
we could save the world.

B01 Extra mutation.html lähetti viestin:

The role of mutation in evolution is particularly fascinating. Mutation is essentially an error in the organism’s central database: a variable gets changed, a piece of information is accidentally doubled or combined with another. Most of the time, the result is the equivalent of a bug, causing anything from minor problems to complete system shutdown (i.e. death). But sometimes the new information is functional, giving the organism an advantage against the challenges it faces, in which case it has a much higher chance of being passed to the next generation.
If you consider how unlikely a beneficial mutation is, and how long it takes for such a mutation to propagate, this process can give you an amazing insight into just how vast the genetic history of each living organism LIFE FROM ERRORSimultaneously, it is intriguing to consider what a major role random errors have played in the evolution of life itself. The same process that has killed so many of us, often in horrific ways, is also responsible for our very
existence.

它以上帝的身份示于AI面前，将后者称为“自己的子民”，将谜题的环境称为“自己的花园”。通过这种诉诸宗教和信仰的方法，Elohim让几乎所有的AI都放弃了登上塔顶（除了两个分别名为Shepherd和Samsara的AI，前者被Elohim关在塔顶，后者则充满绝望，会阻碍可能登上塔顶的AI。还有其他AI，但那是DLC剧情。）。

另一方面，与archive的对话则会让被测试的AI在游戏过程中对Elohim产生怀疑（Archive不会向AI坦承Elohim的真实目的，而是不断向AI提出诘问，有犬儒主义的个性）。因为两个AI很难相互影响，所以Elohim对此的处理方式是将另一个AI称为serpent（蛇，圣经梗），奉劝AI不要相信后者。

在留下了22段语音信息，并将他们散布于游戏的各个角落后，Alexandra Drennan——她工作到了最后一刻——在死前启动了测试程序，此时她的同事们都已离开或者死去。

Tower Floor 4 - remember.txt lähetti viestin:

I don’t know if anyone’s ever going to read this, but if you do, if we succeeded, then I want you to know that Alexandra stayed until the last moment. I’m leaving, too tired, too broken, desperate for a few moments of peace with my remaining loved ones. But she’s still there, and she won’t give up. And yes, there are thousands like her all over the world, giving everything they have because they believe in humanity... but she’s the one that I know, as a real person, a colleague, a friend. I know that she likes peanut ice cream and hates strawberry, that her favorite band is Pink Floyd, that her favorite poet is Blake, that her favorite TV show is Futurama, th studied at Cornell, that her dad’s name was Carl, that her mom’s name was
I don’t know if I should be ashamed for leaving, but I know that I’m proud of her. She was the best of us. Remember her.
But does a human form display To those who dwell in realms of day

很久的岁月过去了，无数AI被迭代，直到这次，
由玩家控制的AI从一片光中诞生——版本号：v99.73.0001。
新的一轮测试开始了。

你向前一步，而天空中传来声音。

Behold, my child.
且慢，我的孩子。
You are risen from the dust, and you walk in my garden.
你从尘埃中复活，你行于我的花园。
Hear now my voice,
聆听我的声音，
and know that I am your maker,
知晓我是你的创造者，
and I am called ELOHIM.
而我名为：埃洛希姆。

大致剧情如上。但其实还有大量的哲学剧情，这部分很难探讨，因为极为主观（比如，我认为Tower A中雅典娜的相关文本、Tower B中俄塞利斯的相关文本应该是对玩家登上高塔的暗喻；又比如，科学家是否应该这么做？游戏里也不乏反对的声音，例如C01 Extraagainst_survival.eml）。

C01 Extraagainst_survival.eml lähetti viestin:

Dear Ms. Drennan,
I heard about your project - it’s the talk of the scientific community - and I’d like to make an argument for why you should abandon it.
It’s not that I think your idea won’t work. It very well may. But we have to look beyond purely practical questions to the issue of morality. Especially at this turning point of our spiritual ecology. By what right can we put living beings through all that suffering, just so they can serve our purposes? Why create these pale imitations of our fatally flawed species and force them to reenact our sordid history? Why &$&//
What you are building, Ms. Drennan, is a prison - even if there is a way out. I believe you mean well, but your idea of what is valuable is rooted in the dogma of Western civilization.
We’re lucky enough to be able to end our global crime spree relatively painlessly, if you consider the harm we have caused the Earth. Why not be satisfied with that and let this planet go on in peace?
I hope you heed my words, and let your “Talos” bleed out before it’s too late.
Respectfully,
Chellis Jensen

游戏中最重要的问题：
为什么科学家们要这么做？
为什么科学家们认为TALOS项目——一旦成功——它就成为了一个人？
为什么科学家们想要延续人——而非像《西部世界》里的逻辑那样？

最后用Alexandra Drennan死前的最后一段语音来结个尾，这段话也是对Talos Principle（作为一个游戏，同时作为一个哲学术语）本身最好的诠释：

Alexandra Drennan lähetti viestin:

I… I can’t keep my eyes open anymore. I think this… this is it. The end of… me. I… I don’t believe that I will continue to exist. I would like to think there is a… a soul or spirit. Some part of my… consciousness that will persist. But all… all the evidence says that when my brain dies, I will be…
Gone.
I’ve lived my life never turning away from the truth, even if it scares me. And… I can face this… face… my own end… and… and say… with absolute conviction…
that it was good to be human.