【文章标题】：What it feels like to work with Mythos
【文章标题】：与 Mythos 合作是什么感受
【文章正文】：
I had early access to the first Mythos-class AI model being released to the public, Claude 5 Fable. Much of the discussion of Mythos has centered on its impact on software security, but I tested it on everything except that (the guardrails around Fable essentially prevent it from being used for cybersecurity at all). My conclusion is that it represents a very real leap over every model I have used before, and, maybe more important, suggests our relationship with AI is changing in drastic ways.
我提前获得了首个向公众发布的 Mythos 级 AI 模型——Claude 5 Fable 的使用权限。关于 Mythos 的讨论大多集中在其对软件安全的影响上，但我对它进行了除安全之外的所有测试（Fable 周围的安全护栏基本上使其完全无法用于网络安全）。我的结论是，它比我以前使用过的所有模型都实现了真正的飞跃，而且，也许更重要的是，它表明我们与 AI 的关系正在发生剧烈的变化。
First, how good is Fable? In experiment after experiment I conducted, it outperformed basically every other public model I have used by a considerable margin. It was capable across many problems and produced some startling results — it would work up to a dozen hours executing on multi-page specifications. I’ll walk you through a couple of more complex, and serious, use cases shortly, but you could see the general improvement across the board on every task. The problem about communicating this in a post is that many of the most impressive results are going to be interesting to only small portions of my readers. For example, it made
首先，Fable 有多好？在我进行的一个又一个实验中，它的表现基本上远超我用过的所有其他公开模型。它能够处理许多问题，并产生了一些惊人的结果——它可以连续工作长达十几个小时来执行多页规格说明。稍后我会带你了解几个更复杂、更严肃的用例，但你可以看到它在每项任务上的全面改进。在帖子中传达这一点的问题在于，许多最令人印象深刻的结果只会引起我读者中一小部分人的兴趣。例如，它创作了
the most sophisticated academic social science paper
最复杂的学术社会科学论文
I have yet seen from an AI from a single prompt and one piece of feedback. It also created
是我见过的由 AI 从单个提示和一条反馈中生成的成果。它还创作了
a 10-page epic rhyming poem
一首长达 10 页的史诗级押韵诗
about a haircut where every word starts with the letter s.
关于一次理发，其中每个单词都以字母 s 开头。
So, as a more accessible and entertaining example, I also had it create a bunch of games you can try. All of these are one initial prompt in Claude Code where Fable had to take my vague prompts and generate something workable, followed by a couple of additional prompts with minor encouragement (“make it better”) or feedback. What makes these especially impressive is that Claude cannot generate images, so every piece of art or 3D object was made with math alone, not using any external assets. You can try any of them:
因此，作为一个更容易上手且有趣的例子，我还让它创建了一些你可以尝试的游戏。所有这些都是在 Claude Code 中通过一个初始提示完成的，Fable 必须根据我模糊的提示生成可用的东西，然后再加上一些带有轻微鼓励（“让它更好”）或反馈的额外提示。这些游戏尤其令人印象深刻的是，Claude 无法生成图像，所以每一件艺术品或 3D 物体都完全依靠数学生成，没有使用任何外部资源。你可以尝试其中任何一个：
a game about flipping coins
一个关于抛硬币的游戏
(prompt: “Balatro, but for the game of coin flips”) that is quite fun;
（提示：“Balatro，但用于抛硬币游戏”）非常有趣；
a snake game
一个贪吃蛇游戏
where the snake is self-aware and crazy things happen; or a game about
其中蛇具有自我意识，会发生各种疯狂的事情；或者一个关于
descending into the depths
深入深渊
to see what is there.
看看那里有什么。
So the output is impressive. But, especially as I turned to more serious projects, I often felt using the tool was somewhere between delightful and unnerving. Delightful because I just asked for something at it happened. And also unnerving because I just asked for something and it happened.
所以输出令人印象深刻。但是，尤其是当我转向更严肃的项目时，我常常觉得使用这个工具的感觉介于愉悦和不安之间。愉悦是因为我只要提出要求，事情就发生了。不安也是因为我只要提出要求，事情就发生了。
Maps and Methods
地图与方法
To see why, it helps to understand the way in which Fable gets work done, and for that I want to turn to an example I have tested on many previous AI models: building an isochrone map. This is a map that shows the distance you can travel in a given length of time, and the first one was created in 1881 showing travel times from London.
要理解原因，有助于了解 Fable 完成工作的方式，为此我想举一个我在许多以前的 AI 模型上测试过的例子：构建等时线地图。这种地图显示在给定时间内可以旅行的距离，第一张这样的地图于 1881 年创建，显示了从伦敦出发的旅行时间。
The original map
原始地图
No previous model did an even halfway useful job with trying to create a map like this because it involves researching thousands of potential trip distances and a lot of small judgement calls and decisions. I decided to try it on Fable using Claude Code with this prompt:
以前的任何模型在尝试创建这样的地图时，连一半有用的工作都做不到，因为这需要研究数千个潜在的旅行距离，并做出大量细微的判断和决策。我决定在 Claude Code 中使用这个提示在 Fable 上尝试：
i want you to build a fully researched and beautiful isochronic map that lets me pick various cities and see real isochronic lines based on real data. I want the design to be unique. You should take into account airports (and travel time to and from airports) trains, walking, driving. The data does not need to be live but should be real based on your research and data. You can start with a few cities but more general is better, this should be an entirely new project.
我想让你构建一张经过充分研究且精美的等时线地图，让我可以选择不同的城市，并查看基于真实数据的真实等时线。我希望设计独特。你应该考虑机场（以及往返机场的旅行时间）、火车、步行、驾车。数据不需要是实时的，但应该基于你的研究和数据真实可靠。你可以从几个城市开始，但越全面越好，这应该是一个全新的项目。
It then suggested that it do this in the style of the original map. I agreed, and it got to work.
然后它建议以原始地图的风格来制作。我同意了，它就开始了工作。
It is worth a second looking at the transcript of the multiple hour building session the AI went through on its own, because you can see some unusual things. First, the AI launched multiple other AIs (I believe mostly the cheaper Claude Sonnet) to help it conduct research on travel times, ultimately retrieving over 2,200 specific flights, the rail schedules for trains from the TGV to the Shinkansen, and road speeds per country from multiple academic papers. And while those agents were running, it started coding. Then it launched yet more agents and tests to verify its code, all the while taking notes about its progress.
值得再看一下 AI 独自进行的数小时构建会话的完整记录，因为你可以看到一些不寻常的事情。首先，AI 启动了多个其他 AI（我相信主要是较便宜的 Claude Sonnet）来帮助它进行旅行时间研究，最终检索了超过 2,200 个特定航班、从 TGV 到新干线的火车时刻表，以及来自多篇学术论文的各国道路速度。在这些智能体运行的同时，它开始编写代码。然后它又启动了更多的智能体和测试来验证其代码，并一直在记录其进展。
The result was a fully functioning map of impressive sophistication that looked a lot like the 1881 original, but that doesn’t mean it was perfect. I noticed that a lot of remote locations (like Greenland) just contained estimates of travel time, not exact numbers, so I told Fable to fix it, including the instructions:
结果是一个功能齐全、复杂程度令人印象深刻的地图，看起来很像 1881 年的原始地图，但这并不意味着它完美无缺。我注意到许多偏远地区（如格陵兰）只包含旅行时间的估计值，而不是精确数字，所以我让 Fable 修复这个问题，并给出了以下指示：
actually get travel times to remote airports and locations.
实际获取偏远机场和地点的旅行时间。
This time the AI launched a workflow, adversarial groups of agents that did research and tested each others results. It figured out how often ships sail to Pitcairn Island in the Pacific and how to get to Grise Fjord from Ottawa. And it used a tremendous number of tokens in a very short period of time (more on this soon).
这一次，AI 启动了一个工作流，即相互对抗的智能体小组，它们进行研究并相互测试彼此的结果。它弄清楚了船只多久开往太平洋的皮特凯恩岛一次，以及如何从渥太华到达格赖斯峡湾。而且它在很短的时间内使用了大量的 token（稍后会详细说明）。
The results were impressive. I pushed a few more times in directions that interested me (including asking for other visualization approaches, etc.).
结果令人印象深刻。我又朝几个我感兴趣的方向推动了几次（包括要求采用其他可视化方法等）。
I would recommend spending a couple minutes clicking around
我建议花几分钟点击浏览一下
the results, and you can read its methods and sources at the bottom of the graph.
结果，你可以在图表底部阅读其方法和来源。
What the AI generated. Click on the map to go to the interactive version
AI 生成的内容。点击地图进入交互式版本
This is probably not a useful project for you unless you really like travel and maps, but it is indicative of AI solving a hard problem involving research, math, visual development, taste, judgement, complex coding, and more. And, the unnerving part was how little I did. I gave a really ambitious instruction, the AI followed it. I gave a couple of minor pieces of feedback, and the AI figured it out. My role was extremely limited.
除非你真的很喜欢旅行和地图，否则这个项目可能对你没什么用，但它表明了 AI 能够解决一个涉及研究、数学、视觉开发、品味、判断、复杂编码等难题。而且，令人不安的部分是我做得太少了。我给出了一个非常雄心勃勃的指令，AI 就照做了。我给了几条小小的反馈，AI 就自己搞定了。我的角色极其有限。
Importantly, it was just limited in how much work I did relative to the model, it was also limited in how much control I had over how the model did things, why the model chose particular approaches, or even how in-depth its results would be. The details of the AI’s decision making are not shown to me, and the process would be too long to even be worth following. The map required the AI to make judgement calls about hundreds of little choices, and it just made them, without me understanding the choices or having a chance to weigh in. In many ways, it is miraculous (I can always ask for edits at the end) on the other, it turns AI into the ultimate black box.
重要的是，不仅相对于模型而言我所做的工作量有限，而且我对模型做事方式的控制力、模型为何选择特定方法、甚至其结果的深入程度都同样有限。AI 决策的细节不会向我展示，而且过程太长，甚至不值得去跟踪。这张地图需要 AI 对数百个小选择做出判断，而它就自己做了，我不理解这些选择，也没有机会参与意见。在很多方面，这堪称奇迹（我总可以在最后要求修改），但另一方面，它把 AI 变成了终极黑盒。
Working with a Mythos-class model
与 Mythos 级模型合作
The most ambitious project I got from Fable takes a little more explanation. I do a lot of research where humans produce messy answers and doing any sort of analysis requires categorize those answers properly: how innovative is an idea? why do people like this book? To figure this out, we used human researchers to make a judgement call about a piece of information, and statistically compare their answers with others to figure out whether we can trust the data. A lot of recent research has shown that AIs might be able to do this important work, but calibrating AI and human judgement has been difficult and expensive. So I asked Fable to solve the problem, first generating a
我从 Fable 那里得到的最雄心勃勃的项目需要稍作解释。我做了很多研究，在这些研究中人类会给出杂乱无章的答案，而进行任何形式的分析都需要正确地对这些答案进行分类：一个想法有多创新？人们为什么喜欢这本书？为了解决这个问题，我们让人类研究人员对一条信息做出判断，并将他们的答案与其他人的答案进行统计比较，以确定我们是否能够信任这些数据。最近很多研究表明，AI 或许能够完成这项重要的工作，但校准 AI 和人类的判断一直困难且昂贵。所以我请 Fable 来解决这个问题，首先生成一份
complex 19 page design document
复杂的 19 页设计文档
and then executing it.
然后执行它。
It worked for nine and a half hours.
它工作了九个半小时。
The result was an extremely sophisticated piece of software the AI called Concord that could take in multiple datasets, calibrate human and AI responses, and then conduct complex data analysis on the results. Again, it wasn’t perfect. As an expert, I was able to spot some errors and omissions (some as a result of the design I had asked for) that I had the AI correct. But the scope of the delivery on this project, and many others, exceeded anything I had seen before. In this case, it was a piece of software that researchers have needed for years but was never profitable to create.
结果是一个非常复杂的软件，AI 称之为 Concord，它可以接收多个数据集，校准人类和 AI 的响应，然后对结果进行复杂的数据分析。同样，它并不完美。作为专家，我能够发现一些错误和遗漏（有些是我要求的设计导致的结果），我让 AI 进行了修正。但在这个项目以及许多其他项目中交付的范围，超出了我以前见过的任何东西。在这种情况下，它是一个研究人员多年来一直需要但从未有利可图去创建的软件。
You can now just use or modify the code here
你现在可以在这里直接使用或修改代码
. I am sure it is not perfect (I only spent an hour working with the results), but a software engineer would iron out the remaining potential bugs that I could not find quickly (which is one reason we may need more, not less, coders in the future, to help with the explosion of new uses for software).
。我确信它并不完美（我只花了一个小时处理结果），但软件工程师会解决我无法快速找到的剩余潜在错误（这也是我们未来可能需要更多而不是更少的程序员的原因之一，以帮助应对软件新用途的爆炸式增长）。
This power goes hand in hand with strangeness and limits. Among those limits is its token usage. Fable is twice as expensive as Opus, and it burns through tokens at a rate that suggests the answer to how much it costs in production is “a lot,” though its clever delegation to cheaper models may lower the real price considerably. The guardrails for Fable also trip at the faintest hint of a security problem, defaulting to the less powerful Claude 4.8 Opus, and it happens way too often. And the jagged frontier is still there. For example, the AI still writes in the same weird style (in fact the software Fable produces bears traces of Claudisms; so do its progress reports, all that carrying the weight and earning the answer). But the deeper strangeness is how little I had to do, and how little I could see while it was being done.
这种力量与奇怪之处和限制相伴而生。其中一项限制是它的 token 使用量。Fable 的价格是 Opus 的两倍，而且它消耗 token 的速度表明它在生产环境中的成本答案是“很多”，尽管它巧妙地委派给更便宜的模型可能会大大降低实际价格。Fable 的安全护栏也会在最微小的安全问题的迹象下触发，默认回退到功能较弱的 Claude 4.8 Opus，而且这种情况发生得过于频繁。锯齿状的前沿仍然存在。例如，AI 仍然以同样古怪的风格写作（事实上，Fable 生成的软件带有 Claudisms 的痕迹；它的进度报告也是如此，所有那些“承担重任”和“赢得答案”）。但更深的奇怪之处在于我需要做的太少，以及在完成过程中我能看到的太少。
Last year I called this working with a
去年，我把这称为与
wizard
巫师
: you chant the spell and something happens. With Fable the spell has gotten powerful enough that I am no longer sure I am the wizard. I am closer to a patron. I describe what I want, I pay for it, and I judge the result. The conjuring happens somewhere I cannot watch, in hundreds of small choices I never get a vote on. The work has shifted from process to outcome. I no longer steer; I commission.
合作：你念出咒语，事情就发生了。有了 Fable，咒语已经强大到让我不再确定自己是巫师。我更接近一位赞助人。我描述我想要什么，我为此付费，然后我评判结果。魔法在我看不到的地方发生，在数百个我从未参与投票的小选择中发生。工作已从过程转向结果。我不再操控；我委托创作。
It is possible the sidelining is temporary, just an artifact of interfaces that haven’t caught up, and that we’ll get better windows into what these models are doing and better ways to steer them midstream. It is also possible that the opposite is true: that the more capable the model, the less there is for a human to meaningfully do, and the black box is the price of the power. I suspect that is more likely to be the real direction. None of this is a loss of control in the obvious sense. I can still steer Fable, and it follows instructions remarkably well: the more ambitious the instruction, the better the result. But steering is no longer the same as doing. I brief the model, it spins up its own agents to research and write and check one another’s work, and what comes back is finished. A patron commissions a single artist. Fable is closer to a whole studio, where I am the client who signs off on the final work without ever setting foot on the floor.
这种被边缘化的感觉可能是暂时的，只是界面尚未跟上的一种产物，我们将会获得更好的窗口来了解这些模型在做什么，以及更好的方式来在过程中引导它们。也有可能恰恰相反：模型能力越强，人类能有意义地做的事情就越少，而黑盒就是这种力量的代价。我怀疑后者更可能是真正的方向。这一切都不是显而易见意义上的失控。我仍然可以引导 Fable，而且它非常擅长遵循指令：指令越雄心勃勃，结果就越好。但引导已经不再等同于亲自去做。我向模型简要说明，它会启动自己的智能体来研究、写作并相互检查工作，返回来的就是成品。一位赞助人委托一位艺术家。Fable