Blog Jamboree 2026: The Winners

This is the third annual Experimental History Blog Post Competition, Extravaganza, and Jamboree, where people send me their best new blog posts and I reward my favorites with praise and cash.

这是第三届年度实验历史博客文章竞赛、盛会和狂欢节，人们向我发送他们最好的新博客文章，我用赞扬和现金奖励我最喜欢的博客文章。

This year, I got 68 submissions from folks all over the world, including visual artists, consultants, high schoolers, economists, engineers, lawyers, and all sorts of pseudonymous internet weirdos.

今年，我收到了来自世界各地的 68 份意见书，其中包括视觉艺术家、顾问、高中生、经济学家、工程师、律师和各种匿名的网络怪人。

For the first time, I ran all potential winners through Pangram, which is an AI-writing detector.

我第一次通过 Pangram（一种人工智能书写检测器）对所有潜在获奖者进行了调查。

1 So according to the machine (and to my ear), all of the winners are 100% human-written.

1 因此，根据机器（以及我的耳朵）的说法，所有获奖者都是 100% 人工编写的。

And here they are:

他们在这里：

In the blogging biz, when someone dares to actually work on a piece of writing before they publish it—when they bother to cite their sources, check their facts, find a good story to frame their argument, etc.—we call that an “effort post”.

在博客行业，当有人敢于在发表一篇文章之前实际进行写作时——当他们费心引用来源、检查事实、找到一个好故事来阐述他们的论点时——我们称之为“努力帖子”。

“How to Survive Boiling Water” is a quintessential effort post.

“如何在沸水中生存”是一篇典型的努力帖子。

Cook accidentally drinks some probiotic tea, and then she’s like, wait, how could “probiotic tea” possibly be a thing?

库克不小心喝了一些益生菌茶，然后她想，等等，“益生菌茶”怎么可能是一种东西？

Don’t all the bacteria die when they get dunked in boiling water?

当细菌浸泡在沸水中时，不是所有细菌都会死亡吗？

This is where most blog posts would end.

这是大多数博客文章结束的地方。

Instead, Cook decides to run an elaborate experiment involving, among other things, a makeshift Bunsen burner, molten agar, and a dash of polymerase chain reaction.

相反，库克决定进行一项复杂的实验，其中包括临时本生灯、熔融琼脂和少量聚合酶链式反应。

Ultimately, she discovers that bacteria are freaky little creatures that are capable of all kinds of depravity.

最终，她发现细菌是一种奇怪的小生物，能够做出各种堕落的事情。

“As far as I can tell,” begins López Cordero’s piece, “the word ‘vape’ appears in only one national constitution on Earth.”

“据我所知，”洛佩斯·科德罗的文章开头说道，“地球上只有一部国家宪法中出现过‘vape’这个词。”

There’s an art to the short blog post like this, the kind that drops something interesting your lap, gives you a nod, and walks away.

像这样的简短博客文章有一种艺术，就是把一些有趣的东西放在你的腿上，向你点头，然后走开。

So in describing his post, I’ll do the same.

因此，在描述他的帖子时，我也会这样做。

Bio: Ricardo López Cordero is “a writer and producer from Mexico City”.

简介：里卡多·洛佩斯·科尔德罗是“来自墨西哥城的作家和制片人”。

“Young person discovers their roots” is a tried-and-true format.

“年轻人寻根”是一种行之有效的模式。

What stands out about Hsieh’s version is that her roots occupy such a unique spot in history: her grandparents were Chinese born in Indonesia, which means they got caught between the rock of the Cultural Revolution and the hard place of anti-communist backlash.

谢家华版本的突出之处在于，她的根源在历史上占有如此独特的地位：她的祖父母是出生在印度尼西亚的中国人，这意味着他们陷入了文化大革命的困境和反共反弹的困境之间。

A Mi [Hsieh’s grandfather’s classmate] told me about the first weeks: propaganda posters appearing on walls overnight, enormous red characters, and then the label fandong (反動), reactionary.

阿米（谢祖父的同学）向我讲述了最初几周的情况：墙上一夜之间出现了宣传海报，巨大的红字，然后是“反动”的标签。

For overseas Chinese who had grown up in capitalist Southeast Asia, the label followed them everywhere.

对于在资本主义东南亚长大的海外华人来说，这个标签无处不在。

[...] I think about how differently we use the word uncertainty now.

[...]我想我们现在对“不确定性”这个词的使用有多么不同。

I say it about job markets, AI, whether to stay in one city or move to another.

我说的是就业市场、人工智能、是留在一个城市还是搬到另一个城市。

She meant: I did not know if I would be allowed to continue learning.

她的意思是：我不知道我是否可以继续学习。

I did not know if I would be allowed to leave.

我不知道我是否会被允许离开。

I did not know who I was allowed to become.

我不知道我被允许成为谁。

Dingfelder has extremely sparse autobiographical memories, which means, on occasion, people remind her that she has made out with them:

丁菲尔德的自传体记忆极其稀疏，这意味着有时人们会提醒她，她和他们亲热过：

This feels mean to say, but it’s true: I have absolutely no memory of this makeout session, or Mark’s entire existence, for that matter.

这句话听起来很刻薄，但这是事实：我完全不记得这次亲热，也不记得马克的整个存在。

Kamdar is a conservationist, and he finds that regular people generally figure out reasonable ways of dealing with the wildlife they encounter.

卡姆达尔是一位自然资源保护主义者，他发现普通人通常会找到合理的方法来处理他们遇到的野生动物。

“Total institutions” like the army, on the other hand, try to deal with animals in ways that are totally bananas.

另一方面，像军队这样的“整体机构”则试图以完全香蕉的方式对待动物。

For example:

例如：

The armed forces of Sweden, for over a decade in the 1980s were bombing the living daylights out of their coasts because they were certain that what they heard through their acoustic sensors in the water were Soviet submarines.

20 世纪 80 年代的十多年来，瑞典武装部队一直在轰炸其海岸线，因为他们确信通过水中的声学传感器听到的是苏联潜艇的声音。

[...] [Much later,] the Swedish armed forces approached ecologists working on marine sounds.

[...] [很久以后，]瑞典武装部队联系了研究海洋声音的生态学家。

They quickly figured out that these sounds were only shoals of the herring fish releasing gas from their butts.

他们很快发现这些声音只是鲱鱼从屁股中释放气体的浅滩声音。

Martsinovich argues, via code snippets, that there is no such thing as a “conversation” with a chatbot.

Martsinovich 通过代码片段认为，不存在与聊天机器人的“对话”这样的事情。

The AI is born anew every time it speaks, and it simply reads the dialogue so far and then tries to write the next line:

人工智能每次说话都会重新诞生，它只是简单地阅读到目前为止的对话，然后尝试写下一行：

ChatGPT may look like a chat, but there is no single entity on the other side.

ChatGPT 可能看起来像聊天，但另一边没有单一实体。

If anything, it’s more akin to speed dating, where you get a new partner for every phrase you say.

如果说有什么不同的话，那就是它更像是快速约会，你所说的每一句话都会有一个新的伴侣。

The other side receives a full transcript and does their best to role-play a coherent conversation.

另一方收到完整的文字记录，并尽最大努力进行连贯的对话角色扮演。

They are all excellent actors, and you’re having a great time, but in the end you will go home alone and nobody will know your name.

他们都是优秀的演员，你玩得很开心，但最终你会独自回家，没有人知道你的名字。

People say that reading introduces you to places and people you’d never meet otherwise.

人们说，阅读可以让你认识其他方式永远不会遇到的地方和人。

This is especially true about reading blog posts.

对于阅读博客文章尤其如此。

In this case, allow me to introduce you to someone who will do whatever it takes to avoid using a toilet that flushes automatically:

在这种情况下，请允许我向您介绍一个会尽一切努力避免使用自动冲水马桶的人：

I was 5 years old, starting kindergarten, and using the toilet in the bathroom attached to our classroom.

我五岁了，开始上幼儿园，使用教室附属浴室的厕所。

I finished my business, reached for the toilet paper and FA-WHOOM, a tsunami exploded beneath my exposed bottom.

我完成了我的工作，伸手去拿卫生纸，然后“呼呼”一声，一场海啸在我裸露的屁股下爆发了。

A fear was planted that day, one that has blossomed and sprouted since.

那天，一种恐惧就被种下了，从那时起，这种恐惧就开花结果了。

I don’t truck with automatic toilets.

我不带自动厕所的卡车。

[...] I can count on one hand the number of times I’ve used one in the last ten years.

[...]过去十年里我用一只手就能数得过来。

And with that, I hereby call to a close the 2026 Experimental History Blog Post Competition, Extravaganza, and Jamboree.

至此，我特此宣布 2026 年实验历史博客文章竞赛、盛会和狂欢活动结束。

Thanks to everyone who submitted, thanks to all of you for reading, and thanks to the paid subscribers who make both this blog and this jamboree possible.

感谢所有提交的人，感谢大家的阅读，感谢付费订阅者，他们使这个博客和这次盛会成为可能。

Because of you, the jamboree continues eternally in our hearts.

因为有你们，狂欢在我们心中永远延续。

There is no perfect detector for AI writing.

人工智能写作没有完美的检测器。

Freddie de Boer has a post showing how easy it is to get Pangram to say that a piece of writing is “100% human” or “100% AI”.

Freddie de Boer 的一篇文章展示了让 Pangram 说一篇文章是“100% 人类”或“100% AI”是多么容易。

But given the recent Commonwealth Prize debacle , I think competitions that intend to reward humans should at least try to make sure that they aren’t rewarding machines instead.

但考虑到最近英联邦奖的惨败，我认为旨在奖励人类的竞赛至少应该努力确保它们不会奖励机器。

Blog jamborees are one of the few formats that actually surface writers nobody has heard of yet.

博客盛会是少数几种真正出现的作家尚未听说过的形式之一。

The institutional algorithms keep recycling the same top twenty.

机构算法不断循环使用相同的前二十名。

Glad someone is still doing the curation work.

很高兴有人仍在做策展工作。

> Martsinovich argues, via code snippets, that there is no such thing as a “conversation” with a chatbot.

> Martsinovich 通过代码片段认为，不存在与聊天机器人的“对话”这样的事情。

The AI is born anew every time it speaks, and it simply reads the dialogue so far and then tries to write the next line.

人工智能每次说话都会重新诞生，它只是简单地阅读到目前为止的对话，然后尝试写下一行。

Actually, this understates things.

事实上，这低估了事情。

Every time an LLM adds a new output token to a conversation, it needs to reread every prior word in the conversation.

每次 LLM 将新的输出标记添加到对话中时，它都需要重新读取对话中的每个先前单词。

In other words, every word that an LLM generates is produced by an isolated process; it's born anew every time it needs to choose the next word.

换句话说，法学硕士生成的每个单词都是由一个独立的过程产生的；每次需要选择下一个单词时，它都会重新诞生。

Every output token that it adds to the conversation could be generated by a completely separate server in a different part of the world, and you would have no idea.

它添加到对话中的每个输出令牌都可以由世界不同地区的完全独立的服务器生成，而您可能不知道。

(The term for a system that works like this is a Markov process.

（像这样工作的系统的术语是马尔可夫过程。

Also, to get really technical, there is some caching of intermediate calculations done between output token generations, but it's not necessary at all.)

另外，为了获得真正的技术性，在输出令牌生成之间进行一些中间计算的缓存，但这根本没有必要。）

The idea that LLMs are more like industrial processes than people is one of the things I tried to get at in my submission, but I admit that at this point, I have no idea what the state of "common knowledge" around LLMs is, and thus what the starting point for talking about this kind of thing should be.

法学硕士更像是工业流程而不是人，这是我在提交的材料中试图表达的观点之一，但我承认，在这一点上，我不知道围绕法学硕士的“常识”状态是什么，因此谈论这类事情的起点应该是什么。

It's definitely one of those topics that's new enough that everyone approaches it with different prior knowledge and also different intuitive understanding and reactions, which makes it both easy to write about (because there's a lot of room for new takes) and hard to write about (because any given reader might have a very different baseline understanding or opinions than you.) (I'm probably going to restructure my submission quite a bit before posting it publicly...)

这绝对是足够新的主题之一，每个人都用不同的先验知识以及不同的直觉理解和反应来处理它，这使得它既容易写（因为有很大的新空间）又很难写（因为任何给定的读者可能有与你非常不同的基线理解或意见。）（在公开发布之前我可能会重新调整我的提交内容......）