The most wonderful thing that happened to Tommy McHugh

This is the quarterly links ‘n’ updates post, a collection of things I’ve been reading and doing for the past few months.

这是季度链接“n”更新帖子，是我过去几个月阅读和做的事情的集合。

A British man named Tommy McHugh was on the toilet pooping when apparently he strained too hard, had a stroke, and woke up in the hospital.

一位名叫汤米·麦克休（Tommy McHugh）的英国男子当时在厕所大便，显然他用力过猛，中风了，在医院醒来。

McHugh was construction worker and ex-con who had never been in an art gallery “except maybe to steal something”, but as he recovered, he was possessed by an intense desire to make art :

I’ll paint three or six or nine pictures at a time.

我一次会画三、六、九幅画。

[…] Canvases became too costly, so I started painting the ceilings and the wallpaper and the floor.

[…] 画布变得太贵了，所以我开始粉刷天花板、壁纸和地板。

As he said in an interview, “The most wonderful thing that happened to Tommy McHugh [was] having a stroke while doing a poo.”

正如他在接受采访时所说，“汤米·麦克休发生的最美妙的事情就是在大便时中风。”

The blogger Linch claims that Pope Leo’s first encyclical, which is about AI, was substantially written by AI .

博主 Linch 声称利奥教皇的第一本关于人工智能的通谕基本上是由人工智能撰写的。

I believe that it’s a mortal sin to pass off computer-generated text as your own, so if this is true, I hope the pontiff confesses and repents.

According to the blogger Jenn, Canada lost its measles elimination status because they don’t have enough nurses who speak Low German .

据博客作者 Jenn 称，加拿大失去了消除麻疹的地位，因为他们没有足够的会讲低地德语的护士。

Basically, measles outbreaks in Canada are concentrated among Mennonites, who do seem to be hesitant about vaccines, but much less so when medical professionals can speak to them in their native language.

For instance, here’s the kind of miscommunication that can happen between doctors and Mennonites, as told by a Low German-speaking support worker in Ontario :

The following morning, the mother called me: her child was coughing so violently she was vomiting.

第二天早上，母亲打电话给我：她的孩子咳嗽得很厉害，呕吐了。

I told her to go to the hospital.

我告诉她去医院。

Later, she called me again, upset.

后来，她又给我打电话，一脸不高兴。

She said that when she got to the ER, they’d told her to go home.

她说，当她到达急诊室时，他们让她回家。

I couldn’t help but think something was off.

我忍不住觉得有些不对劲。

The hospital doesn’t turn people away, I told her, but she insisted that they had.

我告诉她，医院不会拒绝病人，但她坚持认为他们拒绝了。

So I called them directly to figure out what had happened.

于是我直接打电话给他们，想了解到底发生了什么事。

It turned out there had been a miscommunication.

原来是沟通有误。

Hospital staff had told her not to come in, using a “stop” hand gesture to communicate, and she had become so flustered that she failed to catch the second part of the message: that she should wait in the car while they prepared a negative-pressure room.

In 1830s Massachusetts, beards were so uncommon and strange that a man named Joseph Palmer was mocked and shunned for wearing one.

在 1830 年代的马萨诸塞州，胡须非常罕见和奇怪，以至于一位名叫约瑟夫·帕尔默 (Joseph Palmer) 的男子因留胡子而受到嘲笑和回避。

Assailants tried to forcibly shave him, but he fought them off with a knife, and was imprisoned for assault.

袭击者试图强行给他剃胡子，但他用刀击退了他们，并因袭击罪入狱。

He refused to pay his fine or court fees, and so he remained in prison for over a year.

他拒绝支付罚款或法庭费用，因此在监狱里呆了一年多。

This at least gave him the excuse to create this all-time classic tombstone :

这至少给了他创造这个历史经典墓碑的借口：

The Vegetable Lamb of Tartary supposedly grew out of the ground, grazed all the grass it could reach with its stem-like umbilical cord, and then died.

European scholars considered it a plausible organism well into the 1700s, but doubts started to creep in after various expeditions failed to turn up any specimens.

1

TV shows and movies often depict uranium as a glowing green ore, even though it doesn’t actually glow.

电视节目和电影经常将铀描绘成一种发光的绿色矿石，尽管它实际上并不发光。

Étienne Fortier-Dubois claims this trope did not come from uranium glass, a novelty product from 100 years ago that does glow green (but because of fluorescence, not radioactivity).

Instead the image of green, glowing uranium came, as so many things do, from The Simpsons .

相反，绿色、发光的铀的形象，就像很多东西一样，来自《辛普森一家》。

Uri Bram has a cool new tool that tells you how weird you are.

Uri Bram 有一个很酷的新工具，可以告诉你你有多奇怪。

For instance, if you have a bachelor’s degree or higher, you don’t have a gun in your home, and both of your parents were born in the United States, congratulations—only 14% of Americans match you on all three.

(This is based on the US General Social Survey so it only works for Americans, but it should be easy to recreate for any country with a similar dataset.)

Here’s another number that surprised me.

这是另一个令我惊讶的数字。

If you asked me how many artists make more than $100k in royalties on Spotify, I’d probably guess, I dunno, 5,000?

如果你问我有多少艺术家在 Spotify 上赚取超过 10 万美元的版税，我可能会猜，我不知道，是 5,000 美元？

The actual number is 13,800 .

实际数量为13,800。

I happened to be reading Helen DeWitt’s novel The Last Samurai when I saw her post about her troubles with the Windham Campbell Prize .

我碰巧正在读海伦·德威特的小说《最后的武士》，当时我看到了她关于温德姆·坎贝尔奖的麻烦的帖子。

The prize is meant to “provide writers with the opportunity to focus on their work independent of financial concerns” by cutting them a check for $175,000.

However, they also require you to spend eight hours filming a promotional video and three days participating in a workshop at Yale.

然而，他们还要求你花八个小时拍摄宣传视频，并三天参加耶鲁大学的研讨会。

This might not seem like a lot, but as the biologist Ruxandra Teslo puts it, Weird Nerds come with tradeoffs .

这可能看起来不多，但正如生物学家 Ruxandra Teslo 所说，怪异书呆子总是需要权衡。

If you’re the kind of person who can write an extremely inventive (and good!) novel, you might also have a hard time doing things that are easy for other people, like, say, finding Wi-Fi, or putting your writing on hold to fly to Yale.

DeWitt says she hasn’t had proper time to write for years, between taking care of her dying mother, fending off a stalker, and dealing with dilatory editors, and so she’d rather be strapped for cash if it means she’s free to write uninterrupted.

And so she turns down the prize and the check that comes with it.

所以她拒绝了奖金和随之而来的支票。

That’s what seriousness looks like.

这就是严肃的样子。

Another writing prize controversy: one of the Commonwealth Short Story Prizes appears to have been awarded to an obviously AI-generated story .

另一个写作奖争议：英联邦短篇小说奖之一似乎颁发给了一个明显由人工智能生成的故事。

The prize-winning stories all appear in Granta magazine, and the editor’s response was basically a shrug :

获奖故事全部出现在《Granta》杂志上，编辑的反应基本是耸耸肩：

We showed Claude.ai the story and asked whether it was AI-generated.

我们向 Claude.ai 展示了这个故事，并询问它是否是人工智能生成的。

The response was long, concluding that it was ‘almost certainly not produced unaided by a human’.

回复很长，得出的结论是“几乎肯定不是在没有人类帮助的情况下产生的”。

A computer scientist named Tuhin Chakrabarty claims that the empty metaphors that plague the prize-winning story—and that plague all AI writing— often come from human-written fanfiction .

For instance, the rare phrase “the patience of a reptile”, which appears in the allegedly AI-generated story, also occurs in this fanfiction based on the Japanese anime/manga Naruto .

There’s no way to be sure that an LLM sucked it up and spat it back out, but it’s worth remembering that a meaningful amount of the text inside these machines was written by people who were playing pretend with anime characters.

Apparently some estimates of microplastics in the environment might be way too high because the samples got contaminated by lab gloves .

显然，由于样本受到实验室手套的污染，对环境中微塑料的一些估计可能过高。

As this tweet puts it:

正如这条推文所说：

Niko McCarty has a great article on the controversies around eLife , one of the few scientific journals willing to actually try new things.

Niko McCarty 有一篇关于 eLife 争议的精彩文章，eLife 是为数不多的愿意真正尝试新事物的科学期刊之一。

In 2018, the journal ran a trial where it did away with accept/reject decisions altogether for a small sample of papers—it just posted the manuscripts and their peer reviews as “Reviewed Preprints”.

Authors were then able to try publishing their papers elsewhere, or revise and repost them on eLife .

然后作者可以尝试在其他地方发表他们的论文，或者修改并重新发布在 eLife 上。

In 2023, the journal was about to make this the policy for all papers, and some of the editors and the governing board staged a revolt.

2023 年，该杂志即将将此作为所有论文的政策，一些编辑和理事会发起了反抗。

They got the editor-in-chief, Michael Eisen, fired.

他们解雇了主编迈克尔·艾森。

Clarivate, a for-profit company that collects publishing metrics, yanked the journal’s “Impact Factor”, which is a number that people use to judge a journal’s prestige.

Fortunately, eLife ’s experiment continues, for now.

幸运的是，eLife 的实验目前仍在继续。

It is dismaying to discover that, if you try to do something even mildly different with a scientific journal, people will freak out and try to self-immolate to stop you.

Clearly, this isn’t a system that can be reformed from the inside, and it needs a controlled burn instead.

显然，这不是一个可以从内部改革的系统，而是需要有控制的燃烧。

(See: The one science reform we can all agree on, but we’re too cowardly to do .)

（参见：我们都同意的一项科学改革，但我们太懦弱而不敢这么做。）

According to China-watcher Tanner Greer, the Chinese Communist Party has decided that the future belongs to the nation with the best science and technology, and naturally, they want that country to be China .

As a result, China has

结果，中国

the world’s most sensitive ultra-high-energy cosmic-ray detector , the world’s largest and most sensitive radio telescope , the world’s strongest steady-state magnetic field , the world’s fastest quantum computer by computational advantage , and the world’s most sensitive neutrino detector [...and] the world’s largest primate medical research center.

China now graduates twice as many STEM PhDs as the US does, and it dominates citation metrics in all sorts of applied physics and chemistry fields.

It is easy to waste your science funding and it’s easy to game citation metrics with junk papers— as we Americans know all too well .

浪费你的科学经费很容易，而且很容易用垃圾论文来玩弄引用指标——正如我们美国人所熟知的那样。

And it is hard to pull off a paradigm shift without a lot of intellectual freedom.

如果没有大量的知识自由，就很难实现范式转变。

But a decade or two from now, it’s going to be obvious which countries were pumping money into their scientific ecosystems, and which ones were cancelling grants by using CTRL-F on the word “diversity”.

Every once in a while, I reacquaint myself with Suzanne Triester’s fictional video game stills , made on an Amiga computer in the early 1990s:

每隔一段时间，我就会重新熟悉 Suzanne Triester 在 20 世纪 90 年代初在 Amiga 计算机上制作的虚构视频游戏剧照：

ICE agents carry guns around, and so we think of them like police officers.

ICE 特工随身携带枪支，因此我们将他们视为警察。

But they’re not.

但他们不是。

They’re actually government bureaucrats, and they’re subject to the same laws that govern, say, the EPA.

他们实际上是政府官僚，并且遵守与美国环保局等机构相同的法律。

And since the Supreme Court has been so intent on neutering the administrative state, there are actually a lot of ways to rein in ICE that don’t rely on the police brutality playbook.

So argues this very insightful article in The Atlantic that also happens to be written by my wife .

《大西洋月刊》上这篇非常有见地的文章也是我妻子写的。

Crime rates in the US have been mostly decreasing since the 1990s.

自 20 世纪 90 年代以来，美国的犯罪率大多呈下降趋势。

But most people feel like crime is increasing , even when it’s not:

但大多数人感觉犯罪正在增加，即使事实并非如此：

Some of these people claim that A) the statistics are somehow made up or they’re misleading because people are reporting less crime, or B) homicides only appear to be down because medical care has improved.

Scott Alexander reviews the evidence for these claims and finds them both wanting .

斯科特·亚历山大审查了这些说法的证据，发现它们都不尽如人意。

Okay, but even if people aren’t committing more major crimes, maybe they’re doing more petty crimes—litter, graffiti, shoplifting, etc. Here the evidence is less clear and more mixed, but we can at least rule out a large and obvious increase in social disorder .

As you can tell from the comments sections on these posts, people find Alexander’s claims very annoying.

从这些帖子的评论部分可以看出，人们觉得亚历山大的说法非常烦人。

Deep in their hearts, humans feel like they can precisely estimate crime rates by walking around and comparing the world that they see to the one that they remember.

I often get the same reaction when talking about the Illusion of Moral Decline —once, someone came up to me and was like, “Well, my neighbor just put up a fence to spite me, so clearly morality is declining”.

Also in the Illusion of Moral Decline, we asked people to estimate how cooperation rates in lab-based economic games—i.e., the Prisoner’s Dilemma—had changed over the past 70 years.

We paid them a bonus if they got the answer right.

如果他们答对了，我们就会给他们奖金。

On average, people thought cooperation had gone down, when actually those rates have gone up.

平均而言，人们认为合作率下降了，而实际上合作率却上升了。

A new paper replicates this effect in both America and China :

一篇新论文在美国和中国复制了这种效应：

A new study claims that an eight-minute conversation with ChatGPT can reduce belief in conspiracy theories by about 20%, and that these effects last at least two months.

I am always wary of tiny interventions that purport to have large and long-lasting effects, but I don’t see anything immediately fishy about this study, and the mechanism is plausible.

I bet most 9/11 truthers have never spoken to a skeptic who both took them seriously and who knew a lot about 9/11.

我敢打赌，大多数 9/11 真相支持者从未与一个既认真对待他们又对 9/11 了解很多的怀疑论者交谈过。

This is, I think, an underrated and positive aspect of AI.

我认为，这是人工智能被低估的积极方面。

People are worried about hallucinations and AI psychosis, and rightfully so.

人们担心幻觉和人工智能精神病，这是理所当然的。

But keep in mind that many people do not know basic, uncontroversial truths about their world.

但请记住，许多人并不了解关于他们的世界的基本的、无可争议的事实。

37% of Americans—a plurality—believe that God created humans in their present form within the last 10,000 years .

37% 的美国人（多数人）相信上帝在过去一万年内创造了现在的人类。

40% don’t understand how a control group works .

40% 的人不明白对照组是如何运作的。

26% think that the sun goes around the Earth .

26%的人认为太阳绕着地球转。

For these people, talking to a chatbot is likely to leave them with a more accurate view of reality.

对于这些人来说，与聊天机器人交谈可能会让他们对现实有更准确的看法。

My good friend Jake Robertson has a PhD in history from Oxford, and he’s one of the most creative fellas I know.

我的好朋友杰克·罗伯逊拥有牛津大学的历史博士学位，他是我认识的最有创造力的人之一。

(Not a lot of people are both experts on the Gulag and have also played Hamlet in the nude in Central Park).

（没有多少人既是古拉格集中营的专家，又在中央公园扮演过裸体的哈姆雷特）。

His greatest love is taking stories from the past and bringing them to life, which he’s done for decades with his own family history, and now you can hire him to do the same for yours .

This is a great gift for the grandparent who has everything, for the mom who wants more than just a family tree and a stack of census records, and for settling family feuds about whether Uncle Harry actually had a secret second family back in Warsaw.

Kirk Smith is building a homemade, open-source, low-cost battery to capture renewable energy; check it out here .

柯克史密斯正在建造一种自制的、开源的、低成本的电池来捕获可再生能源；在这里查看一下。

From CasualPhysicsEnjoyer : Join the Independent Science Society .

来自 CasualPhysicsEnjoyer：加入独立科学协会。

Especially relevant if you live in London, but they also have a Discord.

如果你住在伦敦，尤其重要，但他们也有一个不和谐的地方。

Paul Litvak is working on an AI-enabled replacement for peer-reviewed journals .

Paul Litvak 正在研究一种支持人工智能的同行评审期刊替代品。

I hear a lot of science reform ideas that are basically, “What if AI, you know, just made all the problems go away?” so what I like about this one is how specific it is, and that it has an actual theory of change.

我听到很多科学改革的想法，基本上都是这样的，“如果人工智能，你知道，让所有问题消失怎么办？”所以我喜欢这个的原因是它的具体性，以及它有一个实际的变革理论。

Thanks to everyone who submitted to the 2026 Experimental History Blog Post Competition, Extravaganza, and Jamboree !

感谢所有提交 2026 年实验历史博客文章竞赛、Extravaganza 和 Jamboree 的人！

I’ll announce the winners later this summer.

我将在今年夏天晚些时候宣布获奖者。

Jacob Tubbs did another replication of the Things Could Be Better effect and the results look pretty similar :

雅各布·塔布斯（Jacob Tubbs）又复制了“事情本来可以更好”效果，结果看起来非常相似：

The Happiness Lab with Laurie Santos (the Illusion of Moral Decline)

劳里·桑托斯的幸福实验室（道德衰退的幻觉）

And finally, a post from the archive: two years ago, I found a way to put my horrible master’s thesis to good use.

最后，来自档案馆的一篇文章：两年前，我找到了一种方法来充分利用我糟糕的硕士论文。

Years ago, I wrote a master’s thesis that was so bad I immediately threw it in the trash.

几年前，我写了一篇硕士论文，写得非常糟糕，我立即把它扔进了垃圾桶。

But along the way I learned something important, which is how to get seventh-graders to smoke.

但一路走来，我学到了一些重要的东西，那就是如何让七年级学生吸烟。

Experimental History will fight for your right to wear a beard if you want to

如果你愿意，实验历史将为你争取留胡子的权利

I swear I first learned about this through an Astral Codex Ten links post, but now I can’t find it, so maybe I hallucinated it.

我发誓我第一次了解到这一点是通过星界法典十链接帖子，但现在我找不到它，所以也许我产生了幻觉。

Full disclosure, I am also the recipient of an Emergent Ventures grant.

完全披露，我也是 Emergent Ventures 赠款的获得者。

Apparently they’ll give ‘em to just anybody these days.

显然，现在他们会把它们送给任何人。

“26% think that the sun goes around the Earth.

“26% 的人认为太阳绕着地球转。

For these people, talking to a chatbot is likely to leave them with a MORE accurate view of reality.”

对于这些人来说，与聊天机器人交谈可能会让他们对现实有更准确的看法。”

This is a great example of a common phenomenon that needs a name.

这是需要命名的常见现象的一个很好的例子。

Does it already have a name in the literature?

它在文献中已经有名字了吗？

Similarly, Sesame Street didn’t have much effect on the literacy of children of parents who talked to their kids a lot, read to them every night, etc. (Watching TV might have even been mildly negative for some of these kids, as the real-world alternative was available and better.) But it had a powerful positive effect on kids whose parents hardly spoke to them, didn’t read to them, etc. This effect – negative at the top, neutral in the middle positive at the bottom – tends to wash out to nothing statistically.

But it’s actually having lots of different effects!

但实际上它有很多不同的效果！

And arguments about it are extremely frustrating, because people can always just point at one end of the distribution, and ignore the other.

关于它的争论非常令人沮丧，因为人们总是可以只指出分布的一端，而忽略另一端。

Anyway, great post, as ever.

Full of juicy informational goodness.

充满了丰富的信息。

Loved this post, thanks.

喜欢这篇文章，谢谢。

#19 - apologies for being pedantic but (writing from a country that has a monarch to one that theoretically doesn't have one) 'reigning' is what kings do and 'reining in' is how you restrain an unruly horse or a thuggish agent.