故事：否认并非解决之道——与杰弗里·辛顿对话

这里是iHeart播客。经营企业却未曾考虑播客？请三思。收听播客的美国人数已超过通过Spotify和Pandora收听广告支持流媒体音乐的用户。作为头号播客平台，iHeart的规模是第二名与第三名总和的两倍。

This is an iHeart podcast. Run a business and not thinking about podcasting? Think again. More Americans listen to podcasts than ad supported streaming music from Spotify and Pandora. And as the number one podcaster, podcaster, iHeart's twice as large as the next two combined.

了解播客如何助力您的业务发展。请拨打844844

Learn how podcasting can help your business. Call 844844

大家好，美国甜心约翰尼·诺克斯维尔在此。我要向大家介绍我的全新真实犯罪播客《无罪山匪大劫案》，由Smartless Media、Campsite Media及Big Money Players联合出品。这部作品疯狂讲述了一群高能蠢蛋如何阴差阳错完成了美国史上第三大现金劫案。

Hello. America's sweetheart Johnny Knoxville here. I wanna tell you about my new true crime podcast, Crimeless, Hillbilly Heist, from smartless media, campsite media, and big money players. It's a wild tell about a gang of high functioning nitwits who somehow pulled off America's third largest cash heist.

有点像罗宾汉，除了——

Kinda like Robin Hood except for

劫富济贫那部分——

the part where he steals from the rich and gives

我可没那么慷慨。

to the poor. I'm not that generous.

这简直是个近乎励志的真实故事，献给所有曾志存高远却搞砸收场的人。他们盗走1700万美元，却连张帮他逃跑的车票都没买。所以我们说这就像...

It's a damn near inspiring true story for anyone out there who's ever shot for the moon then just totally muffed up the landing. They stole $17,000,000 and had not bought a ticket to help him escape. So we're saying like,

天啊，我们该怎么办？我们该怎么办？

oh god. What do we do? What do we do?

那太蠢了。大家可别学我。

That was dumb. People do not follow my example.

请在iHeartRadio应用、苹果播客或任何你获取播客的平台收听《Crimeless, Hillbilly Heist》。

Listen to Crimeless, Hillbilly Heist on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

我住在一位邪教领袖楼下，我担心自己惹怒了她。

I live below a cult leader, and I fear I've angered her.

等一下，索菲亚。

Wait a minute, Sofia.

你怎么知道她是邪教领袖？

How do you know she's a cult leader?

好吧，达科塔，幸运的是这周是《Okay Story Time》播客的'不怕恐怖故事周'。所以我们很快就会知道答案。这位听众写道：我的邻居每天大声播放音乐，举行污秽仪式，现在我的天花板开始塌陷了。

Well, Dakota, luckily it's I'm not afraid of a scary story week on the Okay Story Time podcast. So we'll find out soon. This person writes, my neighbor has been blasting music every day and doing dirt rituals and now my ceiling is collapsing.

我试图举报他们，

I try to report them,

但事情变得越来越诡异。

but things keep getting weirder.

我觉得他们可能是一个邪教组织的成员。

I think they might be part of a cult.

等等，现实中的邪教？那个泥土仪式又是什么？

Hold up. A real life cult? And what is a dirt ritual?

不知道，达科塔。去听听结局吧。在iHeartRadio应用、苹果播客或任何你获取播客的地方收听《Okay Storytime》播客。

No clue, Dakota. Find out how it ends. Listen to the Okay Storytime podcast on the iHeartRadio app, Apple podcast, or wherever you get your podcasts.

大家好，我是卡尔·彭。在我的新播客《历史重演》中，我们将探讨当今的趋势和头条新闻，并追问：为什么历史总在重演？每周我都会邀请比尔·奈、莉莉·辛格和皮特·布蒂吉格等朋友，从太空竞赛到电影翻拍再到致幻剂，无所不谈。

Hey. I'm Cal Pen. And on my new podcast, here we go again, we'll take today's trends and headlines and ask, why does history keep repeating itself? Each week, I'm calling up my friends like Bill Nye, Lilly Singh, and Pete Buttigieg to talk about everything from the space race to movie remakes to psychedelics.

换句话说，你是不是嗨了？

Put another way, are you high?

听着，现在的世界可能看起来很可怕，但我的目标是让你听完后对未来多一点信心。请在iHeartRadio应用、Apple播客或你常用的播客平台收听并订阅《重头再来》节目。

Look, the world can seem pretty scary right now, but my goal here is for you to listen and feel a little better about the future. Listen and subscribe to Here We Go Again with Cal Pen on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

感谢收听《科技那些事》。如果你不熟悉我的声音，我是奥斯瓦洛西安。由于无可替代的乔纳森·斯特里克兰将主持棒交给了卡拉·普莱斯和我，本节目将继续为你提供全方位的科技内容，所有往期节目仍会在此更新。感谢收听。

Thanks for tuning in to Tech Stuff. If you don't recognize my voice, my name is Osvalosian, and I'm here because the inimitable Jonathan Strickland has passed the baton to Cara Price and myself to host Tech Stuff. The show will remain your home for all things tech, and all the old episodes will remain available in this feed. Thanks for listening.

当时我在加州一家酒店，看到电话突然亮起来。我想：谁会在凌晨1点给我打电话？接着传来一个瑞典口音的声音，说我获得了诺贝尔物理学奖。我觉得这太奇怪了——我根本不研究物理啊。

I was in a hotel in California, and I saw that the phone lit up. And I thought, who's calling me at 01:00 in the morning? And then this Swedish voice came on, and then they said I won the Nobel Prize in Physics. And I thought, this is very odd. I don't do physics.

所以那时我怀疑这可能是个恶作剧。

So that's when I thought it might be a prank.

杰弗里·辛顿获得了2024年诺贝尔物理学奖，这项殊荣曾授予阿尔伯特·爱因斯坦和玛丽·居里。而罗伯特·奥本海默虽入围候选名单却从未获奖。

Geoffrey Hinton won the 2024 Nobel Prize in Physics, an honor held by Albert Einstein and Marie Curie. A certain j Robert Oppenheimer was shortlisted but never won.

我原本最大的愿望是凭借破解大脑运作机制获得诺贝尔生理学或医学奖。但没想到的是——就算没完全破解大脑奥秘，居然也能拿诺贝尔奖。

My big hope was to win the Nobel Prize in Physiology or Medicine for figuring out how the brain worked. And what I didn't realize is you could fail to figure out how the brain worked and still get a Nobel Prize anyway.

欢迎收听《科技那些事》。本期嘉宾是诺贝尔奖得主杰弗里·辛顿。每周三我们都会深度访谈科技前沿人物，带你探索最迷人的技术世界。他最近获奖的原因是'为人工神经网络实现机器学习奠定了开创性理论与技术基础'——这种受人类大脑神经元网络启发的学习模型。

Welcome to Tech Stuff. This is the story with our guest, the Nobel laureate Jeffrey Hinton. Each week on Wednesdays, we bring you an in-depth interview with someone who's at the forefront of technology or who can unlock a world where tech is at its most fascinating. His recent Nobel Prize win was for, quote, foundational discoveries and inventions that enable machine learning with artificial neural networks. Now artificial neural networks are learning models inspired by the network of neurons present in the human brain.

辛顿想要理解大脑的渴望是他开创性人工智能研究的关键灵感来源。我特别着迷于辛顿，因为他的工作几十年来完全逆计算机科学主流而行，但他始终坚持己见。这是一个面对个人损失仍坚持不懈的非凡故事。同样引人入胜的是辛顿与自己创造物之间的关系。以下是他在诺贝尔颁奖晚宴上的发言。

And Hinton's desire to figure out the brain was a key inspiration for his pioneering work on AI. I was particularly fascinated by Hinton because his work went completely against the mainstream of computer science for decades, and yet he stuck to his guns. It's an incredible story of dedication in the face of personal loss. Also fascinating is Hinton's relationship to his own creation. Here's what he said at the Nobel Prize banquet.

当我们创造出比人类更聪明的数字生命时，将会出现一个更长期的存在性威胁。我们完全不知道是否能保持控制权。但现有证据表明，如果它们是由追求短期利润的公司创造的，我们的安全将不是首要考虑。我们亟需研究如何防止这些新生命想要夺取控制权。它们已不再是科幻小说了。

There is also a longer term existential threat that will arise when we create digital beings that are more intelligent than ourselves. We have no idea whether we can stay in control. But we now have evidence that if they are created by companies motivated by short term profits, our safety will not be the top priority. We urgently need research on how to prevent these new beings from wanting to take control. They are no longer science fiction.

谢谢。

Thank you.

所以当我获得与杰弗里·辛顿对谈的机会时，我想知道他如何从一个想理解心智与大脑关系的人，变成为我们所知的人工智能铺路者。他满足了我的请求，讲述了自己从学生到研究员、到教授、再到谷歌员工，最终成为人工智能安全倡导者的历程。容我冒昧地问——在对史蒂夫·乔布斯、比尔·盖茨和马克·扎克伯格保持敬意的同时——您是否在某种意义上算是'大学辍学生'的鼻祖？

So when I got the opportunity to sit down with Geoffrey Hinton, I wanted to know how he went from someone who wants to understand the relationship between the mind and the brain to someone who paved the path for AI as we know it. And he obliged me by telling me about his trajectory from student to researcher to professor to Google employee to, finally, AI safety advocate. Am I right in thinking, with all due respect to, Steve Jobs and Bill Gates and Mark Zuckerberg that you were in a sense the original college dropout?

不是他们那种辍学方式。我的情况是：我先去了剑桥，一个月后就退学了，但第二年又回去了。后来读博士时我又中途退学，但最终还是完成了学位。所以我和他们不同——我总会回去。我是个失败的辍学者。

Not in the sense they dropped out, because what I did was I I went to Cambridge and after a month I dropped out, but then I went back the next year. And then while I was doing my PhD, I dropped out, but then I finished it. So I'm not like them. I I went back. I'm a failed dropout.

'失败的辍学者'，这个说法很好。当时是出于什么原因呢？是不确定感、矛盾心理、好奇心，还是...

Failed dropout. That's good. What what were the reasons for, was it uncertainty or ambivalence or curiosity or

我第一次去剑桥时，是人生中首次离家生活。我一直自认为是聪明人之一，但到了剑桥才发现人人都聪明，这让我压力很大。我每天拼命学习12个小时搞科研，既要努力跟上进度又要首次独立生活，双重压力实在难以承受。

So when I first went to Cambridge, it was the first time I'd ever lived away from home and I'd always my image of myself was that I was one of the clever ones. And when I went to Cambridge, I wasn't one of the clever ones, everybody was clever, and I found it very stressful. I worked very hard, so I was working like twelve hours a day doing science and the combination of working very hard to keep up and living away from home for the first time was just too much for me.

《纽约客》上有篇关于你的精彩报道，引述说'辛顿尝试过不同领域，却沮丧地发现自己从未在任何班级里成为最聪明的学生'，这话让我会心一笑。不过我想沮丧和压力本就是近亲。考虑到你的家庭背景，你当时面临的期望值肯定很高。能谈谈这个吗？

So there was a fantastic profile of you in the New Yorker which said, quote, Hinton tried different fields but was dismayed to find that he was never the brightest student in any given class, which made me smile. But I guess dismay and stress are quite close cousins. I suppose the stakes for you were high given the family you you came from. Can you speak a little bit about that?

是的。我父亲给我施加了很大压力要我取得学术成就，母亲也默许了这种态度。所以从小我就明白那是我必须达成的目标。这种压力真的很大。

Yes. I had a lot of pressure from my father to be an academic success, and my mother sort of went along with it. So from an early age, I realized that's what I had to achieve. And That's a lot of pressure.

他们是如何施加这种压力的？我的意思是，你是怎么感受到的？

How did they exert that pressure? Mean, how are you aware of it?

我父亲是个有点古怪的人。他在墨西哥革命期间长大，从小没有母亲。他性格相当奇特。每天早晨我去上学时——不是每天，但经常如此——临出门时他总会说：'拼命干吧。等你到我两倍年纪时，要是非常努力，或许能有我一半出息。'

My father was a slightly strange character. He grew up in Mexico during all the revolutions without a mother. He was somewhat odd. Every morning when I went to school, not every morning, but quite often, as I left, he would say, get in there, Pitchin'. If you work very hard when you're twice as old as me, you might be half as good.

哇。这种压力可真够呛。

Wow. That's sort of pressure.

你觉得这种话有激励作用吗？

Did you find that motivating?

我觉得很烦人，但可能确实起到了激励作用。很不幸的是，他在我写论文期间去世了，始终没能看到我取得成功。

I found it irritating, but I think it probably was motivating. He very inconsiderately, he died while I was writing my thesis and he never saw me being a success.

你曾在剑桥就读，短暂离开后又回来了。我想你最终选择了实验心理学作为你的学位。

You were at Cambridge, you left briefly and you came back. I think you settled on experimental psychology as your degree in the end.

我当时学的是自然科学。没错。一开始我学的是物理、化学和晶体学。哇。因为DNA结构解码的成功，晶体学当时可是热门领域。

I was doing natural sciences. Right. I started off doing physics, chemistry, and crystalline state. Wow. Because of the success in decoding the structure of DNA, crystalline state was a big thing.

是的。一个月后我就退学了。然后我重新申请了建筑学。我一直很喜欢建筑，但学了一天就认定自己永远不可能擅长建筑，因为我不够艺术。我喜欢工程部分，但艺术方面实在不行。

Right. And I left after a month. Then I reapplied to do architecture. I've always liked architecture and after a day, I decided I was never gonna be any good at architecture because I wasn't artistic enough. I loved the engineering, but the artistic bit, I couldn't do very well.

所以我转回理科，后来学了物理、化学和生理学，我特别喜欢生理学。在学校时父亲从不让我学生物，他不允许。

So I switched back to science, but then I did physics, chemistry and physiology and I really liked physiology. I'd never been allowed to do biology at school. My father wouldn't allow it.

为什么不让学？

Why not?

啊，他说如果学生物，他们会教你遗传学。他是个斯大林主义者，不相信遗传学。讽刺的是，他本人还是英国皇家学会生物学领域的院士，却不信遗传学。

Ah, well, he said if you do biology, they'll teach you genetics. And he was a Stalinist and he didn't believe in genetics. Now, he was also a fellow of the British Royal Society in biology who didn't believe in genetics.

天啊。真是个复杂的人。不过说真的，你别把他想得太夸张。他居然直接对你说不能学生物。

Gosh. Complicated man. Yes. But, I mean, he really you don't wanna exaggerate him. He said to you, you can't study biology.

是的。

Yes.

他当时还有其他不算太糟的理由，比如：生物学可以等年纪大了再学，但数学和物理老了就学不动了。

Now he had other reasons which weren't so bad, which is Yeah. You can always pick up biology when you're older. What you can't pick up when you're older is math and physics.

我觉得这话可能有道理。

I think that's probably true.

所以这个理由就更站得住脚了。

And so that was a more valid reason.

是啊。那你最后从剑桥毕业时拿的是什么学位？

Yeah. Yeah. So what did you end up graduating from Cambridge with a degree in?

心理学。我第一年还修了物理、化学和生理学，物理成绩特别好。

Psychology. Psychology. Physics, chemistry, and physiology for a year. Yeah. And I did very well in physics.

物理拿了优等成绩，这显然预示着能得诺贝尔奖。

I got a first in physics. That's obviously a good predictor of a Nobel Prize.

是啊，看起来你的导师们应该是当时的人

Yeah, it seems like your tutors would have been Then

我放弃了一切去学哲学

I dropped it all and did philosophy.

哲学？

Philosophy?

哲学。我学了一年哲学。然后产生了强烈的抗体，就转到了心理学，所以最终学位是心理学

Philosophy. I did a year of philosophy. And I developed strong antibodies, and then I switched to psychology, and so my final degree was in psychology.

当时是有某个或某组问题驱使你去探索的吗？

And was there a single question or set of questions that you were in search of?

是的，我想知道大脑如何运作、心智如何运作以及它们之间的关系。我很早就认定，如果不理解大脑，就永远无法理解心智

Yes, I wanted to know how the brain worked and how the mind worked and what the relationship was. I decided fairly early on that you're never going to understand the mind unless you understood the brain.

没错。这在当时是主流观点吗？

Right. Was that a popular view at the time?

不，并非如此。当时有一种功能主义的观点，基本上源自计算机软件的理念，认为软件与硬件截然不同，心智的本质在于软件——你使用的启发式方法和表征方式。硬件与之毫无关联。这种观点现在看来极其荒谬，但在当时却显得相当合理。因此我们设计的计算机允许编程，使程序与硬件完全分离，但真实的大脑并非如此运作。

No, not really. There was this kind of functionalist view, that basically the view that came from computer software, which is that the software is totally different from the hardware and what the mind is all about is software, the heuristics you use and the way you represent things. The hardware's got nothing to do with it. That's a completely crazy view but it seemed very plausible at the time. And so the computers we designed so that we could program them had programs being a completely separate domain from hardware, but that's not how the real brain is.

这种我们期待电脑像人、又期待人像电脑的持续摇摆很有趣，对吧？这两者之间存在着一种持续的互动舞蹈。

Was this funny, this constant dance between us expecting our computers to be like us and then expecting us to be like our computers, right? There's a kind of continual dance between those two things.

是的，人们总是试图用最新技术来理解事物。比如电话刚出现时，大脑就被比作巨大的电话交换机。但...

Yes, well, always try and understand things in terms of the latest technology. So when telephones were new, the brain was clearly a very large telephone switchboard. But is

这次有所不同吗？如今人工智能已无处不在，你认为它确实超越了电话的类比吗？

it different this time? Now that AI has taken off and become ubiquitous, do you think this is indeed more than the telephone?

是的，我认为我们训练的这些人工神经网络在许多方面与真实神经网络非常相似。当然，神经元要简单得多，大脑还有许多不同特性，但本质上它们以相同方式工作——通过改变神经元间的连接强度来学习，就像大脑那样。

Yes, I think these artificial neural networks we're training are, in many respects, quite like real neural networks. Obviously, neurons are much simpler. There's all sorts of properties that are different in the brain, but basically, they're working in the same way. They learn things by changing connection strengths between neurons just like the brain does.

那么对你而言，真正渴望理解大脑运作机制的问题是从何时开始的？

And when does that question for you of wanting to understand how the brain works really really begin?

在我高中时期，甚至还没去剑桥之前，我有个叫英曼·哈维的聪明朋友总比我机灵。有天他来学校说，也许大脑记忆像全息图一样分布在整个大脑而非局部存储——因为当时全息技术刚兴起。他正尝试用这种全息新技术来理解大脑的记忆机制。

When I was at high school, so even before I went to Cambridge, I had a very bright friend who was always smarter than me called Inman Harvey, who came into school one day and said, maybe memories in the brain are spread out over the whole brain, they're not localized like a hologram because holograms were new. So he was trying to understand memories in the brain in terms of this new technology of holograms.

这个想法激发了你吗？

And you were stimulated by this idea?

我深受其启发。这是个非常有趣的想法，自那以后我一直在思考记忆在大脑中是如何呈现的。这进而让我思考：大脑是如何学习事物的？

I was very stimulated by that. It was a very interesting idea and it ever since then, I've thought a lot about how memories are represented in the brain. And then that also led into, well, how does the brain learn stuff?

接下来，杰弗里·辛顿是如何加入谷歌的。请继续收听。

Coming up, how Geoffrey Hinton ended up at Google. Stay with us.

经营企业却从未考虑过播客？再想想吧。听播客的美国人比通过Spotify和Pandora收听广告支持流媒体音乐的人还多。作为头号播客平台，iHeart的规模是第二名和第三名总和的两倍。无论你的客户听什么，他们都会听到你的信息。

Run a business and not thinking about podcasting? Think again. More Americans listen to podcasts than ad supported streaming music from Spotify and Pandora. And as the number one podcaster, iHeart's twice as large as the next two combined. So whatever your customers listen to, they'll hear your message.

此外，只有iHeart能将你的信息传递给广播电台的听众。认为播客能助力你的业务？就选iHeart。流媒体、广播和播客。拨打844844开始吧。

Plus, only iHeart can extend your message to audiences across broadcast radio. Think podcasting can help your business? Think iHeart. Streaming, radio, and podcasting. Call 844844 to get started.

号码是844844

That's 844844

我住在一位邪教领袖楼下，我担心自己惹恼了她。呃，等等

I live below a cult leader, and I fear I've angered her. Well, wait

稍等，索菲亚。你是怎么

a minute, Sofia. How do

知道她是邪教头目的？

you know she's a cult leader?

好吧，达科塔，幸运的是，现在正是‘Okay Story Time’播客的‘不怕恐怖故事周’，所以你很快就能知道答案。这个人写道：我的邻居每天大声播放音乐，还举行泥土仪式，现在我的天花板开始塌陷。我试图举报他们，但事情变得越来越诡异。我觉得他们可能是

Well, Dakota, luckily, it's I'm not afraid of a scary story week on the Okay Story Time podcast, so you'll find out soon. This person writes, my neighbor's been blasting music every day and doing dirt rituals, and now my ceiling is collapsing. I tried to report them, but things keep getting weirder. I think they may be part of

邪教成员。等等，索菲亚。现实中的邪教？泥土仪式又是什么？

a cult. Hold up, Sofia. A real life cult? And what is a dirt ritual?

不清楚。但据这个人说，承包商正在拆除露台以查明天花板的问题，而她的邻居们很不高兴。

No clue. But according to this person, contractors are tearing down the patio to find out what's going on with their ceiling, and her neighbors are not happy.

她得尽快举报他们。

Well, she needs to report them ASAP.

她已经举报了。现在他们一直用非常诡异的方式找她麻烦。

She did. And now they've been confronting her in really creepy ways all the time.

那么我们能否知道这个人是否能从他们社区的邪教中幸存下来？

So do we find out if this person survives their neighborhood cult or not?

要收听爆炸性的结局，请在iHeartRadio应用、苹果播客或任何你获取播客的地方收听《Okay Story Time》播客。

To hear the explosive finale, listen to the Okay Story Time podcast on the iHeartRadio app, Apple podcast, or wherever you get your podcasts.

《狼来了》播客讲述的是两个被不公正束缚的男人、一座被秘密困扰的城市，以及不惜一切代价寻求救赎的故事。

The crying wolf podcast is the story of two men bound by injustice, of a city haunted by its secrets, and the quest for redemption no matter the price.

白人受害者，女性，漂亮，富有，黑人被告。

White victim, female, pretty, wealthy, black defendant.

芝加哥，一名白人女性被谋杀，一名黑人男子因未犯下的罪行入狱。

Chicago, a white woman's murder, a black man behind bars for a crime he didn't commit.

我因杀害一个素未谋面的人被判九十年。

I got ninety years for killing somebody I have never seen.

他说警察是他的朋友，然后就这样了。他们转而针对他。

He says the police are his friends, and then that's it. They turn on him.

一名腐败的侦探。

A corrupt detective.

他们用那些审讯手段对付他，太疯狂了。

How he was interrogated with techniques. That's crazy.

一个告密者和被偷走的人生。

A snitch and a life stolen.

他们抓错人了。

They got the wrong guy.

但在监狱里，李·哈里斯与同牢房的罗伯特结为盟友，罗伯特发誓要说出李遭遇的真相并解救他的朋友。

But on the inside, Lee Harris finds an ally in his celly, Robert, who swears to tell the truth about what happened to Lee and free his friend.

如果你跟着我，你的目标就是——我会罩着你。

And if you're with me, your goal is to I'll take care of you.

我会一直陪着你。你这辈子都甩不掉我了。

I'm gonna be with you. You stuck with me for life.

10月22日起，在iHeartRadio应用、Apple Podcasts或您获取播客的任何平台收听《狼来了》播客。

Listen to the crying wolf podcast starting on October 22 on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

开始了。嘿，我是卡尔·佩恩，在我的新播客《又来了》中，我们将探讨当下趋势和热点，并思考：为何历史总在重演？您可能知道我是《哈罗德与库马尔》电影中第二性感的演员，但我还是作家、白宫工作人员，以及大约十五秒前刚成为的播客主持人。一路上，我结识了科学、政治和流行文化领域的专家朋友。

Here we go. Hey. I'm Kal Penn, and on my new podcast, here we go again, we'll take today's trends and headlines and ask, why does history keep repeating itself? You may know me as the second hottest actor from the Harold and Kumar movies, but I'm also an author, a White House staffer, and as of, like, fifteen seconds ago, a podcast host. Along the way, I've made some friends who are experts in science, politics, and pop culture.

每周都会有一位嘉宾加入，解答我迫切想知道的问题。比如：我们是否正走向2008年那样的金融危机？非一夫一妻制又流行起来了吗？为什么航班提前两分钟降落时登机口从来都没准备好？我们的嘉宾包括皮特·布蒂吉格、斯泰西·艾布拉姆斯、莉莉·辛格和比尔·奈等人。

And each week, one of them will be joining me to answer my burning questions. Like, are we heading towards another financial crash like in o '8? Is non monogamy back in style? And how come there's never a gate ready for your flight when it lands like two minutes early? We've got guests like Pete Buttigieg, Stacey Abrams, Lily Singh, and Bill Nye.

当你开始将外太空武器化时，事情可能会变得非常糟糕。

When you start weaponizing outer space, things can potentially go really wrong.

听着，现在的世界看起来确实很可怕。但我的目标是让您通过收听对未来稍感乐观。请在iHeartRadio应用、Apple Podcasts或您获取播客的任何平台收听并订阅卡尔·佩恩主持的《又来了》。

Look, the world can seem pretty scary right now because it is. But my goal here is for you to listen and feel a little better about the future. Listen and subscribe to Here We Go Again with Cal Pen on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

请耐心听我说，我需要尝试总结玻尔兹曼机是什么，因为诺贝尔奖委员会将您的获奖归功于它。根据新闻稿，玻尔兹曼机能够'学会识别给定数据类型中的特征元素'。通过输入极可能出现的样本进行训练后，它可用于图像分类或生成训练过的新型模式样本。辛顿博士基于此研究推动了当前机器学习的爆发式发展。

So bear with me, but I need to try and summarize what the Boltzmann machine is, because the Nobel Prize Committee credited the Boltzmann machine with your win. According to their press release, the Boltzmann machine can, quote, learn to recognize characteristic elements in a given type of data. The machine is trained by feeding it examples that are very likely to arise. The Boltzmann machine can be used to classify images or create new examples of the type of pattern on which it was trained. Hinton has built upon this work, helping initiate the current explosive development of machine learning.

如果时间线没错的话，您从剑桥毕业后攻读博士学位，八十年代来到卡内基梅隆大学，正是在那里您对玻尔兹曼机的研究真正取得了突破。

Now, if I have the timeline right, you graduated from Cambridge, you worked on your PhD, and in the eighties, you wound up at Carnegie Mellon, and that's where your work on the Boltzmann machine really took off.

是的，就在我去卡内基梅隆大学之前。

Yes. Just before I went to Carnegie Mellon.

那么，我是否理解正确，当时存在一场关于人工智能的辩论，而你站在不受欢迎的那一方？

Now, am I right in thinking at the time there was a kind of debate that you were on the unpopular side of about artificial intelligence?

好的。关于这个问题有两点要说。从二十世纪五十年代人工智能的早期开始，就有这两种方法。构建智能系统的两种范式。一种是受大脑启发的神经网络。

Okay. So there's two things to say about that. From the earliest days of AI in the nineteen fifties, there were these two approaches. Two kind of paradigms of how you build an intelligent system. One was inspired by the brain, that was neural networks.

另一种范式则是：不不不，逻辑才是智能的范式。智能的核心在于推理。学习是等我们弄明白推理机制之后的事，而推理依赖于对事物有正确的表征。所以我们必须弄清楚大脑使用的是哪种逻辑上无歧义的语言，才能运用规则进行推理。

Then the other paradigm was, no no no no. Logic is the paradigm for intelligence. Intelligence is all about reasoning. Learning comes later once we figure out how reasoning works and reasoning depends on having the right representations for things. So we have to figure out what kind of logically unambiguous language the brain is using so that it can use rules to do reasoning.

这是完全不同的方法，而且非常不符合生物学原理，因为推理是很晚才发展出来的能力。实际上，在上世纪后半叶的大部分时间里，神经网络并不被视为人工智能。人工智能被认为是你头脑中有符号规则，并通过规则、离散的符号规则来操纵它们。

It's a completely different approach and it's very unbiological because reasoning is something that comes very late and actually, for most of the second half of the last century, neural networks weren't seen as AI. AI was believing you have symbolic rules in your head and they're manipulated using rules, using discrete symbolic rules.

那神经网络当时被视为什么？统计学还是物理学？

And what was what were neural they were seen as statistics or or physics?

它们被视为一种疯狂的想法——认为你可以用这个随机的神经网络学会一切，这显然是无望的。

They were seen as this crazy idea that you could take this random neural network and just learn everything, which was obviously hopeless.

是什么让你如此确信？

And what what gave you this conviction?

大脑总得通过某种方式学习。当然，许多先天结构已经预设好了，这就是为什么山羊刚出生五分钟就能行走。但像阅读这类技能完全是后天习得的，并非与生俱来。

Well, the brain has to learn somehow. And of course, there's a lot of innate structure wired in, which explains why a goat can sort of drop out of the womb and five minutes later it's walking. Right. But so stuff like learning to read, that's all learned, that's not innate.

没错。

Yep.

我们必须弄清楚这其中的机制，而这绝非符号规则能解释的。

And we have to figure out how that works and it's not symbolic rules.

你现在的这种确信，是一直都有的吗？

This conviction that you have now, did you always have?

是的。说来遗憾，在我看来这显然是正确的。部分原因在于我七岁时被送进私立学校，来自无神论家庭的我开始接触宗教之类的无稽之谈。我是全校唯一认为这些纯属胡言的孩子，而事实证明确实如此。这对科学家的成长是极好的历练。

Yes. I'm sorry to say, it seemed to me it was obviously right. I think part of that is I was sent to a private school when I was seven from an atheist family and they started telling me about religion and all this rubbish, all the stuff that was obviously nonsense. And I had the experience of being the only kid at school who thought this was all nonsense and turning out to be right. And that's very good training for a scientist.

我早些时候提到过，《纽约客》有篇关于你的精彩专访，标题是'为什么AI教父惧怕自己的创造'。里面有句话令我震撼，你说'46岁时我完全停滞了'。

I brought it up earlier, but there was a wonderful profile of you in the New Yorker titled, why the godfather of AI fears what he's built. And there's a quote that I found quite stunning where you said, I was dead in the water at 46.

是的。我和妻子收养了两个孩子，两个婴儿，后来我妻子得了卵巢癌，尽管她是个科学家，却开始相信顺势疗法，尝试用顺势疗法治卵巢癌，结果她去世了，留下我和两个年幼的孩子，一个三岁，一个五岁，他们都很难过，我也是。我开始理解女性学者的生活有多么艰难，那简直是不可能的。照顾小孩让你很难有长时间来思考你当下的想法，非常非常困难。

Yes. We my wife and I adopted two children, two babies, and then my wife got ovarian cancer, but she also, even though she was a scientist, she started believing in homeopathy she tried treating ovarian cancer with homeopathy and so she died and I was left with two young children, one of three and one of five, who were both very upset, as was I. And I began to appreciate what life is like for female academics, which is impossible. You can't you can't looking after small children makes it very, very hard to spend long periods of time thinking about your current idea. It's just very difficult.

所以你遭遇了丧亲之痛。你要照顾两个小孩，然后在2012年，你发表了一篇题为《基于深度卷积神经网络的ImageNet分类》的论文，对外行来说这听起来不像会改变世界和我们生活方式的东西，但它确实做到了。

So you were bereaved. You were looking after two small children, and then, yes, that moment in 2012 when you published a paper called ImageNet Classification with Deep Convolutional Neural Networks, which to the layman doesn't sound like something that would would change the world and how we live, but it did.

嗯，它改变了计算机视觉领域以及计算机科学其他领域人们的看法。基本上证明了神经网络确实有效。之前也有人展示过，但没能以同样的方式说服人们。

Well, it changed the views of people in computer vision and the views of people in other areas of computer science. It basically showed neural networks actually do work. Now people have showed that before, but it hadn't convinced people the same way.

所以你在2012年和Ilya Satskeva一起发表了那篇论文，他后来成为OpenAI的重要人物，我想多谈谈他。但论文发表后没几天，你就收到了移居中国的数百万美元邀约。

So you published that paper in 2012 with with Ilya Satskeva, who who went on to be a very important figure at OpenAI, and I wanna talk more about him. But within a few days of the publication of that paper, you had an offer of millions and millions of dollars to move to China.

是的。要么移居中国，要么让他们投资我们的团队。我觉得时间比几天要长一些，但确实很快就来了。

Yeah. Either to move to China or to let them invest in our group. I think it was a bit longer than a few days, but it was it was that full for sure.

你当时是否预感到，'我们要发表了，这将改变一切'，还是对这个反应感到惊讶？

Did you kind of know, okay, we're gonna publish and this is gonna change everything, or were you surprised by this response?

我们预料到会产生很大影响，但没想到影响会这么大。

We thought it would have a big effect. We didn't realize quite how big.

当你接到百度这家中国科技公司的电话时，你是怎么想的？我是说，肯定很诱人吧。

And when you got the call from Baidu, a Chinese tech company, what did you think? I mean, must have been tempting.

是的。我记得他们说可以资助我们团队，或者我们可以去为他们工作，还有其他几种可能的选择。最后我就直接问他们，具体金额是多少？你们是说像1000万美元这样的数字吗？没错。

Yes. I think they said they could fund our group or they could we could go and work for them or various possible alternatives. And in the end, I just asked them, well, how much money are we talking about? Are you talking about like $10,000,000? Yes.

他们回答说是的。

And they said, yes.

你低估自己了。

You lowballed yourself.

当时我没意识到。我想了个自己能想到的最大数字500万，然后翻倍。当他们一口答应时，我才明白我们根本不清楚自己的价值。

I didn't think it at the time. I thought of the biggest number I could think of which was $5,000,000 and doubled it. Once they said that, I realized we had no idea how much we were worth.

确实。

Yep.

所以那一刻我们决定组织一场竞拍。

And so at that point, we decided to set up an auction.

为学术论文设立拍卖意味着什么？

What does it mean to set up an auction for an academic paper?

我们三人创办了一家公司。这家属于我们三人的公司拥有这六项专利申请。这样我们就有东西可卖了，但主要还是出售我们自己。但我坚持保留学者身份，以便能继续指导我的学生。这是个很大的问题，因为谷歌之前从未这样做过。

The three of us founded a company. The company that belonged to the three of us owned these six patent applications. So then we had something to sell, but mainly we were selling ourselves. But I insisted that I still be an academic so I could continue to advise my current students. That was a big problem because Google had never done that before.

他们是竞标我们公司的其中一家企业。

They were one of the companies that bid for us.

所以是百度、谷歌。

So Baidu, Google.

还有微软和当时未被谷歌收购的DeepMind。

Microsoft and DeepMind, which wasn't owned by Google then.

你们不仅与同事共同发明了这项神经网络技术的突破，还独创了出售公司的流程，对吧？

And not only did you invent this, well, together with your colleagues, this advancement of of of neural nets, but you also invented your own process to sell this company essentially. Right?

是的。我们决定通过Gmail进行拍卖，你把出价发到我的Gmail，出价时间以邮件时间为准。我会立即把报价转发给其他竞标者，每次加价必须超过一百万美元。如果最后一轮报价后一小时无人加价，拍卖就结束。令人惊讶的是，报价总会在上次报价的59分钟后出现。我们常以为结束了，结果59分钟后新报价又来了。

Yeah. We decided, we just have an auction by Gmail, and you send me Gmail with your bid and the time of the bid would be the timestamp on the Gmail. I would then immediately send the bid to all the other bidders and you had to raise by a million dollars and if you if there was no bid within an hour of the last bid, that was the end of the auction. I was amazed to see that the bids will come in fifty nine minutes after the previous bid. We'd be sitting there thinking, okay, it's over And then fifty nine minutes later, a bid would come in.

然后大家都信任Gmail了。

And then everybody trusted Gmail.

我曾在谷歌度过一个夏天，这意味着两件事。其一，我确实相信他们不会阅读我的Gmail邮件，我知道他们对此非常严肃。其二，我真心希望谷歌能赢得比赛，因为我非常喜欢当时领导谷歌大脑团队的杰夫·迪恩。所以我希望谷歌获胜，最终我们稍微‘调整’了一下结果。

I had worked at Google over the summer and that meant two things. One, I did trust that they wouldn't read my Gmail. I knew they were very serious about that. And also, I really wanted Google to win the competition because I really like Jeff Dean who ran the brain team at Google. So I wanted Google to win and in the end, we slightly fudged it.

在DeepMind和微软退出后，竞争就在谷歌和百度之间展开。我们担心百度会赢，而我不想去中国——那时我无法出行。另外，我觉得自己无法真正理解中国公司的运作方式。当竞价达到4400万美元时，我们直接表示‘这是个无法拒绝的报价’，拍卖就此结束。而这个无法拒绝的报价就是能与杰夫·迪恩共事。

So after DeepMind and Microsoft had dropped out, it was between Google and Baidu and we were scared Baidu would win and I didn't want to go to China. I couldn't travel at that point. Also, I felt I wouldn't really understand what was going on in a Chinese company. So it got to 44,000,000 and we just said, we've got an offer we can't refuse and it's the end of the auction. And the offer we couldn't refuse was to work with Jeff Dean.

所以从某种意义上说，你得到了你想要的

So you you got what you wanted in the sense of

那些超出我们想象的成果，并且我得以在最向往的公司工作。

of those outcomes. Than we could imagine, and we got to work at the company that I wanted most to work with.

广告回来后，诺贝尔奖得主杰弗里·辛顿将讲述他为何倡导AI安全。

When we come back, Nobel laureate Jeffrey Hinton on why he advocates for AI safety.

经营企业却从未考虑播客？请三思。如今收听播客的美国人比通过Spotify和Pandora收听广告支持流媒体音乐的还多。作为头号播客平台，iHeart的规模是第二名两倍之和。无论你的客户听什么，他们都会听到你的信息。

Run a business and not thinking about podcasting? Think again. More Americans listen to podcasts than ad supported streaming music from Spotify and Pandora. And as the number one podcaster, iHeart's twice as large as the next two combined. So whatever your customers listen to, they'll hear your message.

此外，唯有iHeart能将您的信息传递至广播电台的广大听众。认为播客能助力您的业务？那就选择iHeart吧。流媒体、广播与播客，三管齐下。立即拨打844844开启合作。

Plus, only iHeart can extend your message to audiences across broadcast radio. Think podcasting can help your business? Think iHeart. Streaming, radio, and podcasting. Call 844844 to get started.

号码是844844

That's 844844

我住在一位邪教领袖楼下，恐怕已经惹怒了她。

I live below a cult leader, and I fear I've angered her.

等等，索菲娅，你怎么

Well, wait a minute, Sofia. How do

确定她是邪教领袖？

you know she's a cult leader?

达科塔，幸好现在正值《Okay Story Time》播客的'恐怖故事周'，我可不害怕。这位听众写道：'邻居每天播放震耳音乐、举行泥土仪式，现在我家天花板开始塌陷。我试图举报他们，但情况越来越诡异。我怀疑他们可能属于'

Well, Dakota, luckily, it's I'm not afraid of a scary story week on the Okay Story Time podcast, so you'll find out soon. This person writes, my neighbor's been blasting music every day and doing dirt rituals and now my ceiling is collapsing. I tried to report them but things keep getting weirder. I think they may be part of

一个邪教。等等，索菲娅，现实中的邪教？泥土仪式又是什么？

a cult. Hold up, Sofia. A real life cult? And what is a dirt ritual?

不清楚。但据此人说，承包商正在拆除露台以查明天花板的问题，而她的邻居们对此很不满。

No clue. But according to this person, contractors are tearing down the patio to find out what's going on with their ceiling and her neighbors are not happy.

嗯，她需要尽快举报他们。

Well, she needs to report them ASAP.

她举报了？现在他们一直以非常诡异的方式找她麻烦。

She did? And now they've been confronting her in really creepy ways all the time.

所以我们会知道这个人能否从社区邪教中幸存吗？

So do we find out if this person survives their neighborhood cult or not?

想听爆炸性结局，请收听iHeartRadio应用、苹果播客或其他播客平台上的《Okay Story Time》播客。

To hear the explosive finale, listen to the Okay Story Time podcast on the iHeartRadio app, Apple podcast, or wherever you get your podcasts.

开始了。嘿，我是卡尔·潘。在我的新播客《历史重演》中，我们将探讨当今趋势和头条新闻，并追问：为何历史总在重复？你可能知道我是《猪头逛大街》系列电影里第二性感的演员，但我同时也是作家、白宫职员，以及大约十五秒前刚上任的播客主持人。

Here we go. Hey. I'm Kel Penn. And on my new podcast, here we go again, we'll take today's trends and headlines and ask, why does history keep repeating itself? You may know me as the second hottest actor from the Harold and Kumar movies, but I'm also an author, a White House staffer, and as of, like, fifteen seconds ago, a podcast host.

这一路上，我结识了科学、政治和流行文化领域的专家朋友。每周都会有一位嘉宾加入，解答我的迫切疑问。比如：我们会再次遭遇08年那样的金融危机吗？非一夫一妻制又流行起来了吗？为什么航班提前两分钟降落时登机口从来都没准备好？

Along the way, I've made some friends who are experts in science, politics, and pop culture. And each week, one of them will be joining me to answer my burning questions. Like, are we heading towards another financial crash like in o '8? Is non monogamy back in style? And how come there's never a gate ready for your flight when it lands like two minutes early?

我们有嘉宾皮特·布蒂吉格、斯泰茜·艾布拉姆斯、莉莉·辛格和比尔·奈。

We've got guests like Pete Buttigieg, Stacey Abrams, Lily Singh, and Bill Nye.

当你开始将外太空武器化时，事情可能会变得非常糟糕。

When you start weaponizing outer space, things can potentially go really wrong.

听着，这个世界现在看起来确实很可怕，事实也是如此。但我的目标是让你倾听并对未来感觉好一点。请在iHeartRadio应用、苹果播客或你获取播客的任何地方收听并订阅《Here We Go Again with Cal Pen》。

Look, the world can seem pretty scary right now because it is. But my goal here is for you to listen and feel a little better about the future. Listen and subscribe to Here We Go Again with Cal Pen on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

《狼来了》播客讲述的是两个被不公正束缚的男人、一个被秘密困扰的城市，以及不惜一切代价寻求救赎的故事。

The Crying Wolf Podcast is the story of two men bound by injustice, of a city haunted by its secrets, and the quest for redemption no matter the price.

白人受害者，女性，漂亮，富有，黑人被告。

White victim, female, pretty, wealthy, black defendant.

芝加哥，一名白人女性被谋杀，一名黑人男子因未犯下的罪行而入狱。

Chicago, a white woman's murder, a black man behind bars for a crime he didn't commit.

我因杀害一个素未谋面的人被判九十年。

I had ninety years for killing somebody I have never seen.

他说警察是他的朋友，然后就这样了。他们却背叛了他。

He says the police are his friends, and then that's it. They turn on him.

一个腐败的侦探。

A corrupt detective.

他们用那些审讯手段对待他。太疯狂了。

How he was interrogated with techniques. That's crazy.

一个告密者和被偷走的人生。

A snitch and a life stolen.

他们抓错人了。

They got the wrong guy.

但在监狱里，李·哈里斯在狱友罗伯特身上找到了盟友，罗伯特发誓要说出发生在李身上的真相并解救他的朋友。

But on the inside, Lee Harris finds an ally in his celly, Robert, who swears to tell the truth about what happened to Lee and free his friend.

如果你跟我一起，你的目标就是...我会负责

And if you're with me, your your goal is to I'll take care

属于你的。

of you.

我会一直陪着你。你这一生都甩不掉我了。

I'm gonna be with you. You stuck with me for life.

10月22日起，在iHeartRadio应用、Apple播客或任何你获取播客的平台收听《狼嚎》播客。

Listen to the Crying Wolf podcast starting on October 22 on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

2018年，你获得了图灵奖——这相当于计算机界的诺贝尔奖。但同样在2018年，你再次经历了丧偶之痛。你的第一任妻子罗莎琳德于1994年去世，二十四年后，你的妻子杰姬也因癌症离世。

In 2018, you received the Turing Award, which is kind of like the Nobel Prize for computer science. But also in 2018, you were widowed for a second time. You lost your wife Rosalind in 1994 and then twenty four years later, your wife Jackie passed away, also from cancer.

是的。那段时期确实...很艰难。那时我的孩子们都已成年，所以我不必同时应对照顾幼童和其他事务。谷歌也非常体谅。原本约定我每年要在硅谷工作三个月，但他们豁免了这项要求，允许我全程留在多伦多，还协助我在当地建立了一个从事基础研究的小型实验室。

Yes. That was that was, difficult. I didn't have my my children were much older then, so I didn't have the problem of having to cope with young children at the same time as everything else. And Google was very understanding. Part of my deal had been that I would spend three months a year in Silicon Valley and they let me out of that and said I could spend my whole time in Toronto and they helped me set up a small lab doing basic research in Toronto.

这样压力就小多了，我能陪伴在妻子身边。

So that was much less stressful, so I could be with my wife.

她也患了癌症吗？

She had cancer as well?

她得了胰腺癌。

She got pancreatic cancer.

《纽约客》文章中又一个令人震撼的点是，你谈到通过观察罗莎琳德和杰姬面对癌症的态度，将其作为思考人工智能影响的思维模型？

One of the most striking things, again, in the New Yorker piece was the way you talked about observing the way in which Rosalind and Jackie approached their cancer as a mental model for how to think about the implications of artificial intelligence?

是的，这是个相当阴暗的设想。有时当我想到这些技术可能会接管世界——我确实偶尔会这么想——就面临一个问题：是否该像我的第一任妻子那样选择否认，反复说'这不可能发生'。其实她不算第一任，在那之前我有过一段很短暂的婚姻。罗兹选择了否认，而杰姬则非常现实地面对。

Yeah, that's a rather sort of dark scenario. So occasionally, when I think, well, this stuff probably will take over, which I sometimes think, Then there's the issue of should you go into denial and say, no, no, no, no, this can't possibly happen, which is what my first wife did. Actually, she wasn't my first wife. I was married a long time before that, just briefly. And so, Roz went into denial and Jackie was very realistic about it.

或许我们也该现实地看待机器可能接管世界的可能性。我们应尽全力阻止其发生，但同时也该思考：若真发生了，是否有办法让局面不那么糟糕？比如人类是否还能在机器掌控下继续生存？

Maybe we should be very realistic about the possibility machines will take over. We should do our best to make sure that doesn't happen, but we should also think about whether there's ways of making that hap if that does happen, whether there's ways of making it not so bad, whether people can still be around, even if the machines were in control, for example.

确实。虽然你用'阴暗'这个词，但某种程度上你毕生的工作已开花结果。然而你始终被这些念头困扰——就像面对癌症晚期诊断时该如何应对。

Yeah. Mean, there's something I mean, you use the word dark, but you've seen your life's work, in many ways, come to fruition, But you're haunted by these thoughts of how to deal with something as awful as a terminal cancer diagnosis.

是啊，必须小心许愿。

Yeah. You have to be careful what you wish for.

没错。我开头提到奥本海默与诺贝尔奖的关系。他几乎成为物理学奖得主，但最终未能获奖。

Yeah. I mean, I mentioned Oppenheimer at the beginning in terms of Nobel. Well, he was an almost a physics laureate. He wasn't in the end.

是的，这相当荒谬，不是吗？我获得了诺贝尔物理学奖而奥本海默却没有。这简直太荒谬了。我应该说点什么。人们，尤其是记者们总喜欢问，你会如何将自己与奥本海默比较？

Yes. It's rather absurd, isn't it, that I've got a Nobel Prize in Physics and Oppenheimer didn't. That's utterly ridiculous. I should say something. People, particularly journalists like to, say, well, how would you compare yourself with Oppenheimer?

有两个重大区别。一是奥本海默对原子弹的研发确实至关重要。他主导了相关科学研究，是独一无二的关键人物。而在AI发展过程中，我们是一群人共同努力，即使没有我，所有事情照样会发生。

And there's two big differences. One is that Oppenheimer really was crucial to the development of the atomic bomb. He managed the science of it. He was a single extremely important figure. With the development of AI, there's a bunch of us and if I hadn't been around, all the same stuff would have happened.

这是第一个区别。另一个区别是原子弹除了破坏毫无益处。他们确实尝试过在科罗拉多州将其用于水力压裂，但效果不佳，而且现在那里已经不能去了。最大的不同在于AI的大部分应用都非常有益。

That's one difference. The other difference is that atomic bombs aren't good for anything good, they're just for destruction. They actually did try using them for fracking in Colorado, but that didn't work out too well. And you can't go there anymore. The big difference is most of the use of AI are very good.

它能极大提高生产力，显著改善医疗水平，还可能对应对气候变化大有帮助。正因为这些巨大的益处，AI必将得到发展。这与原子弹截然不同——当时本有可能不研发氢弹的。

It it can lead to huge increases in productivity, huge improvements in health care, might help a lot with climate change. And so AI is gonna be developed because of all those huge beneficial uses. That's very different from atomic bombs, where there was a possibility of not developing the H bomb.

为什么你要主动承担这个责任？你2023年从谷歌离职后，已成为全球范围内警示这些风险最直言不讳且最具专业资历的人士之一。

And why have you taken it upon yourself as your responsibility? Mean, you quit Google in in 2023, and since then have become one of the most vocal and qualified people in the world warning of these risks.

嗯，我老了。已经做不了原创性研究了，但人们还愿意听我说。我真心认为这些风险真实且重要，所以实在别无选择。AI有太多好的用途，我们肯定会继续发展它。我并非反对开发，也不是说要放缓进度。我的意思是，要尽可能安全地推进。

Well, I'm old. I'm too old to do original research anymore, but people listen to me and I really believe these risks are very real and very important, so I don't really have much choice. We are gonna develop AI because it's got so many good uses, So I'm not warning against developing it. I'm not saying slow down. What I'm saying is, try and develop it as safely as you can.

尤其要设法阻止它最终失控，同时也要考虑所有其他短期风险，比如伪造视频、干扰选举和失业问题。我认为需要关注所有这些风险，或许暂停开发是理性的选择——但我并不主张这么做，因为我觉得这根本不可能实现。

Try and figure out in particular how you can stop it eventually taking over, but also think about all the other shorter term risks like fake videos, corrupting elections and loss of jobs. I'm saying we need to worry about all those things and it might be rational to just stop developing it, But I'm not, I don't think there's any hope of that, so I'm not advocating that.

你在12月曾提到，未来30年内AI导致人类灭绝的风险有10%到20%。这些概率是你估算出来的吗？

You said in December that there was a ten to twenty percent risk that AI would cause human extinction in the next thirty years. Do you come up with those odds?

我只是随口编的。就像个闲聊机器人。我就是个闲聊机器人。没错。不是。

I just make them up. That's just like a chapbot. I'm just like a chapbot. Exactly. No.

如果你考虑主观概率，它们基于直觉。我强烈直觉认为超级智能机器取代人类的概率超过1%，同样强烈直觉认为低于99%。我们面对的是极度未知的事物。对完全未知之事的最佳猜测可能是50%，但这不太适用，因为取决于你如何划分。显然概率远大于1%又远小于99%。

If you think about subjective probabilities, they're based on intuition. I have a very strong intuition that the chance of super intelligent machines taking over from people is more than 1% and I have a very strong intuition that is less than 99%. We're dealing with something extremely unknown. So your best bet for things totally unknown is maybe 50%, but that doesn't work very well because it depends on how you partition things. So clearly, the chance is much bigger than 1% and much less than 99%.

或许我该就此打住。但我希望我们能找到方法让人类保持掌控，因为我们是人类，我们在乎的是人类。

And maybe I should just stick at that. But I'm hoping that we can figure out a ways that people can stay in control because we're people and what we care about is people.

你是否认为如果我们'失去控制'，人类的毁灭就是必然的下场？

And do do you view it as inevitable that if we if we quote unquote lose control, our destruction is the is the is the next

不。比如埃隆·马斯克，我和他聊过，他坚持认为AI会把我们当宠物养着，因为我们很有趣。但这似乎是把人类存续挂在一条相当细的线上。

No. They might Elon Musk, for example, I talked to him and he pushed the line that they'll keep us around just pets because we're quite interesting. It seems a rather sort of thin thread to hang human existence by.

所以，你一直在强调AI对社会的整体威胁，但同时也特别批评了山姆·奥特曼本人。

So, I mean, you've been vocal about the kind of overall societal threat, but you've also been specifically critical of Sam Altman in particular.

是的，因为我认为OpenAI成立的初衷是安全地发展通用人工智能，但它正逐渐背离这一目标，转向为盈利而开发AI。因此，最优秀的安全研究人员都已离开，公司现在正试图从非营利性质转变为营利性质，在我看来这完全错了。

Yes, because I think OpenAI was set up to develop AGI safely, and it's just been progressively moving away from that towards developing AI for profit. And so, it's best safety researchers have left, and it's now trying to convert itself from a not for profit company into a for profit company, That seems entirely wrong to me.

而你的前同事兼学生伊利亚·萨茨凯瓦——你们曾合作撰写2012年的论文——在这件事上采取了强硬立场，但最终未能如愿。

And your former colleague and student, Ilya Satskeva, who you worked on the twenty twelve paper with, took a big stand on this, which ultimately didn't break his way.

就萨姆被解除公司CEO职务而言，他确实未能如愿；但就公众意识到OpenAI违背了安全发展通用人工智能的承诺而言，他取得了胜利。而且他现在已经创立了一家致力于此的新公司。

It didn't break his way in terms of Sam being fired as the head of the company. It did break his way in terms of people understanding that OpenAI was going back on its pledge to develop AGI safely. And it has now set up a company, that's trying to do that.

我是说，如果明天萨姆·奥特曼给你打电话或飞到多伦多与你面谈，他问你：'我哪里做错了？我该怎么办？' 你会怎么回答？

I mean, if if Sam Altman called you tomorrow or flew to Toronto and and you were sitting with him, he said, you know, where have I gone wrong and what should I do? What do you say?

我会说'你不是萨姆·奥特曼'。

I'd say you're not Sam Altman.

很公平。

Fair enough.

好吧。假设发生非常意外的情况，萨姆·奥特曼突然顿悟说：'天啊，我们不该为利润做这些，应该为保护人类而做。' 我会非常乐意和他谈谈。

Okay. So suppose something very surprising happens and Sam Altman suddenly has an epiphany and says, oh my god. We shouldn't be doing this for a profit. We should be doing it to protect humanity. I'd be very happy to talk to him.

但你会怎样，又如果，你能以一种方式或其他方式影响他的意见

But what would you and what if if you could affect his opinion in in one way or or that of other

我会说继续发展AI，但在开发过程中要投入大量资源，试图找出它可能失控的方式以及我们能做些什么。如果Sam Altman说，我们将使用30%的计算资源，无论我们有多少计算资源，都将用30%来聘请高薪安全研究人员进行安全研究，不是让它变得更好或进一步发展，而是确保其安全性。我会收回之前的话并说他是个了不起的人。

I would say keep developing AI, but use a large fraction of the resources as you're developing it to try and figure out ways in which it might get out of control and what we could do about that. So if Sam Altman said, we're gonna use 30% of our compute, However much compute we have, we'll use 30% to get highly paid safety researchers to do research on safety, not on making it better, developing it further, but on making it safe. I'd take it all back and I'd say he's a great guy.

我是说，对于一个非营利组织来说，这听起来并非不合理。

I mean, for a non profit, that doesn't sound like unreasonable.

没错。那正是我一开始以为正在发生的事情，也是Ilya以为正在发生的事情。

Exactly. That that was that was what I thought was happening to begin with, and that was what Ilya thought was happening.

就在圣诞节前，《华尔街日报》有篇文章基本是说ChatGPT五号落后于计划，而且深度学习改进的速度正在放缓，可能是因为缺乏真实世界数据，也可能是其他原因。但后来三号出来了，Sam Altman某种程度上说AGI已经到来。你认为我们现在处于AGI的什么阶段？你觉得这甚至是个相关的衡量标准吗？

Right before Christmas, there was an article in the Wall Street Journal basically saying that chat GPT five was behind schedule and that kind of the the pace of improvement in deep learning was was was slowing for different it could be because of lack of real world data. It could be other reasons. But then o three came out and and Sam Altman sort of said that AGI is here. Where do you think we are on this AGI? Do you think it's even a relevant metric?

自2012年以来，一直有人说AI即将碰壁。Gary Marcus在2022年强烈预测AI将碰壁且不会走得更远。你必须把这些放在AI即将碰壁的重复预测背景下来看。这次更真实一些，因为我们确实正在达到数据峰值或易得数据峰值。实际上，在数据孤岛、公司和视频中还有海量数据。

So ever since 2012, there have been people saying AI is about to hit a wall. So Gary Marcus made a strong prediction in 2022 that AI was hitting a wall and wouldn't get much further. So you have to see that against a background of repeated predictions that AI is about to hit a wall. This is a bit more real in the sense that we really are reaching peak data or peak easily available data. There's actually hugely more data in silos and companies and in videos.

是的，我们正在耗尽易得数据，这可能会让进展稍缓，但如果你能让它们生成自己的数据，就能克服这个问题，而你可以通过推理让它们做到这一点。所以即使像国际象棋这样的东西，专家走法也有限，但你可以通过让系统自我对弈来克服这一点，从而获得无限量的训练数据。因此，那些判断‘这是好棋吗？’或‘这个局面对我多有利？’的神经网络现在能获得无限量或说无界量的数据。

So yes, we're running out of easily available data and that may slow things down a bit, but if you can get them to generate their own data, then you can overcome that problem and you can get to them to do that by reasoning. So if you look at even with things like chess, there's only a limited number of expert moves, but you can overcome that by getting the system to play itself and then you get an infinite amount of data to train on. And so the neural nets that are saying, would this be a good move? Or saying, how good is this position for me? They now get an infinite amount of data or rather unbounded amount of data.

你总能生成适合难度级别的新数据。所以没人会说国际象棋和围棋的神经网络会耗尽数据。它们已经远超任何人类，如果我们愿意，还能让它们变得更强大。

You can always generate new data at the appropriate difficulty level too. And so nobody says neural networks for Chess and Go are gonna run out of data. They're already far better than any person and we can make them much, much better than that if we wanted to.

在你离开谷歌前不久，你发推文说毛毛虫提取养分，最终转化为蝴蝶。人类已提取了数十亿条理解的金块，而GPT-4就是人类的蝴蝶。能解释一下吗？

Shortly before you quit Google, you tweeted caterpillars extract nutrients which are then converted into butterflies. People have extracted billions of nuggets of understanding and g p t four is humanity's butterfly. Can you explain that?

好的。观察昆虫的话，多数昆虫都有幼虫和成虫阶段。以蝴蝶为例——毛毛虫并不擅长移动和交配，它的天赋在于汲取养分。

Okay. So if you look at insects, most insects have larvae and they have adults. But let's take butterflies as a obvious example. And if you look at a caterpillar, it's not optimized for traveling and mating. A caterpillar is optimized for extracting stuff.

随后这些养分会转化为‘浓汤’，最终变成截然不同的形态。人类长久以来所做的，正是对世界碎片的理解。

You then turn that stuff into soup and then you get something very different. So what humanity has been doing for a long time is understanding little bits of the world.

通过照片和文字将世界转化为数据

Translating the world into data through photographs and words and

没错。现在我们可以将人类从世界中提取结构的工作——就像毛毛虫汲取养分——把这些提取物转化为不同的东西。你可以把它变成一个知晓万物的单一模型。

Yes. And now you could take all that work we've done at extracting structure from the world, like a caterpillar extracting nutrients, and you could take that extracted stuff and turn it into something different. You could turn it into a single model that knows everything.

初读时我觉得这个比喻既美好又乐观。但重读时，我不再这么想了。

When I read it first, I thought it was quite beautiful and optimistic. And then when I read it again, I didn't think that anymore.

是的。我的意思是，你可以解读为我们已经成为历史。

Yes. I mean, you could read it as we're we're history.

是的。我们正在...是的。向我们毛毛虫的身份告别。是的。你如何解读你自己的隐喻？

Yes. And we're Yes. Bidding farewell to our caterpillar. Yes. How do you read your own metaphor?

我可能多少受到威廉·布莱克一首诗的影响，诗中写道：'叶上的毛毛虫向你重复你母亲的悲伤'——这基本上是在说蝴蝶比毛毛虫美丽得多。我想...我想诗的意思是这样。我不知道我们是否会被取代，希望不会。

I'm probably somewhat influenced by a piece of William Blake poetry, which goes, the caterpillar on the leaf repeats to thee thy mother's grief which is basically saying the butterfly is much prettier than the caterpillar. I think. I think that's what it's saying. You know, I don't know whether we're get replaced. I hope we're not.

我希望人类保持主导权，但希望是在拥有比我们更智能的助手协助下保持主导。

I hope people stay in control, but I hope we stay in control with assistants that are much more intelligent than us.

嗯，就像你说的，母亲是我们所知唯一被不够成熟的生物所控制的生物。记得你说过类似的话。

Well, as you said, mothers are the only creatures in the that we know of who are controlled by less sophisticated beings. Think you said something along those lines.

婴儿的智力远不及母亲——最多相差两倍。而我们讨论的是数量级的差距。

And babies are much less intelligent than the mothers. Like, at most, a factor of two. We're talking about huge factors.

你刚才提到布莱克，但当我读到那个蝴蝶隐喻时，想到的另一个人是你的父亲，他当然是一位昆虫学家，而且...

You mentioned Blake just now, but the other person I thought of when I was reading that butterfly metaphor was your was your father, who of course was was an etymologist and and and

是的。这就是我对变态现象产生兴趣的地方。没错。

Yes. That's where I got my interest in metamorphosis. Yes.

所以Jet Jet GPT既是人类的蝴蝶，在某种意义上也是你的蝴蝶，这个比喻是为了纪念你的父亲，并且

So jet jet GPT is humanity's butterfly, but also in a sense, your butterfly, and this is a metaphor to to honor your father and

我...我猜是吧。勉强算是。

I I guess. Grudgingly.

在最坏的情况下，人工智能确实会导致我们灭绝，那可能会通过哪些方式发生？而在最好的情况下它不会，又可能通过哪些方式实现？

In the worst case scenario where AI does cause our extinction, what are the ways in which that could happen? And the best case scenario where it doesn't, what are the ways in which that could happen?

好的。最明显的方式可能是我们创造出能制定子目标的人工智能体，它们因为超级智能而意识到，获取更多控制权是个很好的子目标——因为掌握更多控制权就能更好地实现目标。所以即使它们是在完成我们设定的目标，也会试图获取更多控制。这有点像...不知道你有没有孩子，如果一个三岁小孩终于想自己系鞋带，但你赶时间，你会让他们试几分钟然后说'不行，我来吧，等你长大再学'。

Okay. So the obvious way it could happen is we make AI agents that can create sub goals and they realize because they're super intelligent, that a good sub goal is to get more control because if you get more control, you can achieve your goals. So even if they're trying to achieve goals we gave them, they'll try and get more control. It'll be bit a like, I don't know if you have children, but if you have a three year old who's finally decided they want to try tying their own shoelaces, but you're in a hurry to get somewhere, you let them try tying their shoelaces for a few minutes then you say, no, no, I'm gonna do it. You can learn that when you're older.

人工智能会像父母，我们则像孩子，它们会直接推开我们完成任务。这就是糟糕的情况。即使它们试图实现我们指定的目标，本质上也会夺取控制权。如果某个超级智能AI产生'我希望世界上多几个我的副本，少几个其他超级AI'的想法，情况会更糟。一旦发生这种情况，超级AI之间就会展开进化竞赛，我们将被彻底淘汰。

AIs will be like the parent and we'll be like the children and they'll just push us out the way to get things done. So that's the bad scenario. Even if they're trying to achieve things that we've told them we want, they'll basically take control. That scenario gets worse if ever one of those super intelligent AI thinks, I'd rather there were a few more copies of me and a few less copies of the other super intelligent AIs. As soon as that happens, if that ever happens, you'll get evolution between super intelligent AIs and we'll be left in the dust.

它们会进化出所有我们在进化中获得的恶劣特质：卑鄙、好斗、对同类极度忠诚而对其他群体极具攻击性——所有这些我们身上的毛病。这是坏结局。好结局则是我们找到方法确保它们永远不会试图夺取控制权，始终服从于人类。这样我们就能拥有这些绝妙的智能助手，生活变得无比轻松。

And they'll develop all the nasty things you get from evolution, being nasty and competitive and very loyal to their own tribe and very aggressive to other tribes, all that nonsense that we have. So that's the bad scenario. The good scenario is we figure out a way where we can guarantee, they're never gonna try and get control away from us. They're always gonna be subservient to us and we figure out how we can guarantee that that'll happen. And then we will have these wonderful intelligent assistants and life is just really easy.

杰弗里，谢谢。谢谢。

Jeffrey, thank you. Thank you.

本周内容到此结束。关于科技话题，我是奥兹维洛尚。本期节目由伊丽莎·丹尼斯、维多利亚·多明格斯、希娜·尾崎和莉齐·雅各布斯制作，由我卡拉·普莱斯、凯特·奥斯本为Kaleidoscope执行制作，卡特里娜·诺维尔为iHeart Podcasts执行制作。工程师是巴希德·弗雷泽，Of Spin负责混音，凯尔·默多克创作了主题曲。

That's it for this week. For tech stuff, I'm Ozveloshan. This episode was produced by Eliza Dennis, Victoria Dominguez, Sheena Ozaki, and Lizzie Jacobs. It was executive produced by me, Kara Price, and Kate Osborne for Kaleidoscope, and Katrina Norvell for iHeart Podcasts. The engineer is Baheed Fraser, Of Spin mixed this episode, and Kyle Murdoch wrote our theme song.

周五请继续收听《一周科技要闻》。我们将解读头条新闻，并听取专家朋友们对最新科技动态的见解。请给我们评分、评论，并通过techstuffpodcast@gmail.com联系我们。谢谢。

Join us on Friday for the week in tech. We'll break down the headlines and hear from some of our expert friends about the latest in tech. Please rate, review, and reach out to us at techstuffpodcast@gmail.com. Thank you.

大家好。我是美国甜心约翰尼·诺克斯维尔。我要向你们推荐我的新真实犯罪播客《无罪的乡巴佬劫案》，由Smartless Media、Campsite Media和Big Money Players联合出品。这个故事讲述了一群高功能傻瓜如何阴差阳错完成了美国第三大现金劫案的疯狂经历。

Hello. America's sweetheart Johnny Knoxville here. I wanna tell you about my new true crime podcast, Crimeless Hillbilly Heist from smartless media, campsite media, and big money players. It's a wild tale about a gang of high functioning nitwits who somehow pulled off America's third largest cash heist.

有点像罗宾汉，除了

Kinda like Robin Hood except for

他劫富济贫的部分

the part where he steals from the rich and gives

我可没那么慷慨。

to the poor. I'm not that generous.

对于那些曾志存高远却搞砸了结局的人来说，这几乎是个鼓舞人心的真实故事。他们偷了1700万美元，却连张帮他逃跑的票都没买。

It's a damn near inspiring true story for anyone out there who's ever shot for the moon then just totally muffed up the landing. They stole $17,000,000 and had not bought a ticket to help him escape.

我们当时就慌了神，天啊，我们该怎么办？怎么办？

We're saying like, oh god. What do we do? What do we do?

那太蠢了，大家可别学我。

That was dumb. People do not follow my example.

请收听Crimeless出品的《乡巴佬大劫案》，在iHeartRadio应用、Apple播客或任何你获取播客内容的地方。

Listen to Crimeless, Hillbilly Heist on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

我住在一位邪教领袖楼下，担心自己已经惹怒了她。

I live below a cult leader, and I fear I've angered her.

等一下，索菲亚。

Wait a minute, Sofia.

你怎么知道她是邪教领袖？

How do you know she's a cult leader?

好吧，达科塔，幸运的是，现在正是‘Okay Story Time’播客的‘我不怕恐怖故事周’，所以我们很快就会知道答案。这位听众写道：我的邻居每天大声播放音乐，举行泥土仪式，现在我的天花板开始塌陷了。

Well, Dakota, luckily, it's I'm not afraid of a scary story week on the Okay Story Time podcast, so we'll find out soon. This person writes, my neighbor has been blasting music every day and doing dirt rituals, and now my ceiling is collapsing.

我试图举报他们，

I tried to report them,

但情况变得越来越诡异。

but things keep getting weirder.

我觉得他们可能是

I think they might be part

某个邪教组织的成员？等等，现实中的邪教？泥土仪式又是什么东西？

of a cult? Hold up. A real life cult? And what is a dirt ritual?

我也不知道，达科塔。想知道结局如何，请收听iHeartRadio应用、苹果播客或其他你获取播客的平台上的‘Okay Story Time’播客。

No clue, Dakota. Find out how it ends. Listen to the okay story time podcast on the iHeartRadio app, Apple podcast, or wherever you get your podcasts.

大家好，我是卡尔·佩恩。在我的新播客《又来了》中，我们将探讨当下的潮流和头条新闻，并追问：为什么历史总在重演？每周我都会邀请比尔·奈、莉莉·辛格和皮特·布蒂吉格等朋友，聊太空竞赛、电影翻拍到迷幻药等各种话题。

Hey. I'm Cal Pen. And on my new podcast, here we go again, we'll take today's trends and headlines and ask, why does history keep repeating itself? Each week, I'm calling up my friends like Bill Nye, Lilly Singh, and Pete Buttigieg to talk about everything from the space race to movie remakes to psychedelics.

换个说法，你是不是嗨了？

Put it another way, are you high?

听着，现在的世界可能看起来很可怕，但我在这里的目标是让你听完后对未来感觉好一些。在iHeartRadio应用、Apple播客或任何你获取播客的地方，收听并订阅《Cal Pen的我们又来了》。

Look. The world can seem pretty scary right now, but my goal here is for you to listen and feel a little better about the future. Listen and subscribe to here we go again with Cal Pen on the iHeartRadio app, Apple podcasts, or wherever you get your podcasts.

大家好，怎么了？这里是陷阱宅男们的零食时间。整个十月，我们将为你带来恐怖内容。

What's up, everybody? It's snacks from the trap nerds. And all October alone, we're bringing you the horror.

恐怖恐怖恐怖。这个月我们将以我最拿手的恐怖游戏开场，保证让你心惊胆战。

Boogity boogity boogity. We're kicking off this month with some of my best horror games to keep you terrified.

然后我们会讨论我们最爱的恐怖和万圣节电影，并弄清楚为什么黑人总是先死。

Then we'll be talking about our favorite horror and Halloween movie and figuring out why black people always die further.

还有托尼恐怖秀的回归，支线任务由本人亲自撰写和讲述。我们还将做一集完整的带评论的朗读。

And it's the return of Tony's horror show, side quest written and narrated by yours truly. We'll also be doing a full episode reading with commentary.

最后我们将以一场恐怖电影大逃杀作为压轴。打开你的免费iHeartRadio应用，搜索陷阱护士播客，立即收听。

And we'll cap it off with a horror movie battle royale. Open your free iHeartRadio app and search trap nurse podcast and listen now.

这是iHeart播客。

This is an iHeart podcast.