#318 – 尼克·莱恩：生命起源、进化、外星生物、生物学与意识

这里又出现了张力。黑猩猩可能更暴力，倭黑猩猩可能更纵欲——这是另一种对立。我们要多严肃？又想享乐到什么程度？

And there's again tension there. And probably chimps more violence, bonobos probably more sex. That's another tension. Do you have how serious do we wanna be? How much fun we wanna be?

再替朋友问一次：你认为尼安德特人遭遇了什么？我们这些狡猾的智人对尼安德特人做了什么？你觉得是我们屠杀了他们吗？究竟是如何做到的？我们是如何在竞争中淘汰他们的？

Asking for a friend again, what do you think happened to Neanderthals? What what did we cheeky humans do to the Neanderthals, homo sapiens? Do you think we murdered them? Was it how do we murder them? How do we outcompete them?

还是说我们在交配上胜过他们？

Or do we outmate them?

我不知道。我是说，我认为有确凿证据表明我们曾与他们交配过。

I don't know. I mean, I think there's unequivocal evidence that we mated with them.

是啊。我们总是试图与一切事物交配。

Yeah. We always try to mate with everything.

没错，差不多是这样。最初出现的是一些有趣的线粒体DNA序列，那要追溯到大约2002年左右。研究发现尼安德特人的线粒体DNA与人类的线粒体DNA差异很大。

Yes, pretty much. There's some interesting The first sequences that came along were in mitochondrial DNA, and that was back to about 2,002 or thereabouts. What was found was that Neanderthal mitochondrial DNA was very different to human mitochondrial DNA.

哦，这太有趣了。

Oh, that's so interesting.

你可以据此建立分子钟，显示分化时间大约在60万年前，不算太久远。然后大约十年后首次完成的全基因组测序显示大量杂交迹象。线粒体DNA实际上表明没有杂交，而核基因却说，不，杂交频繁。我们

You could do a clock on it and it said the divergent state was about 600,000 years ago or something like that, so not so long ago. Then the first full genomes were sequenced maybe ten years after that and they showed plenty of signs of mating between. The mitochondrial DNA effectively says no mating and the nuclear genes say, Yeah, lots of mating. We

不明白这不可能。所以你能解释下线粒体DNA与核基因的差异吗

don't That's know not possible. So can you explain the difference between mitochondrial

我之前讲过线粒体，它们是细胞的动力源。这些是精简后的控制单元——它们的DNA。所以它只通过母亲遗传，在卵细胞中我们可能有50万份线粒体DNA拷贝。仅剩37个基因，它们基本上负责能量生产的控制单元。这就是它的功能。

DNA I've and talked before about the mitochondria which are the power in cells. These are the paired down control units is their DNA. So it's passed on by the mother only and in the egg cell we might have half a million copies of mitochondrial DNA. There's only 37 genes left and they do basically the control unit of energy production. That's what it's doing.

这是一台基础的老式机器，负责

It's a basic old school machine that does

它拥有的基因曾被认为基本无关紧要，因为它们只执行非常局限的功能。但这些基因并非真的不重要，因为那个局限的功能恰恰关乎生命存在的全部意义。

And it's got genes that were considered to be effectively trivial because they did a very narrowly defined job. But they're not trivial in the sense that that narrowly defined job is about everything that is being alive.

是啊。

Yeah.

所以它们更容易测序。你有更多这些DNA的拷贝，可以快速完成测序。但问题在于，由于它们只通过母系遗传——从母亲传给女儿，你和我的线粒体DNA不会继续传递。这无关紧要。我们任何孩子都会继承他们母亲的线粒体DNA，除非在极其罕见和特殊的情况下。

So they're much easier to sequence. You've got many more copies of these things and you can sequence them very quickly. But the problem is because they go down only the maternal line from mother to daughter, your mitochondrial DNA and mine is going nowhere. It doesn't matter. Any kids we have, they get their mother's mitochondrial DNA except in very very rare and strange circumstances.

因此它讲述了一个不同的故事，而且这个故事往往难以与其他证据完全吻合。至少在我看来，它似乎暗示着基因流动是单向的——可能是从人类流向尼安德特人，而非相反。为什么尼安德特人会消失？我...我不知道。我猜测他们可能更不暴力、不够聪明、人口更少、更不愿争斗。

So it tells a different story and it's not a story which is easy to reconcile always. And what it seems to suggest to my mind at least is that there was one way traffic of genes probably going from humans into Neanderthals rather than the other way around. Why did the Neanderthals disappear? I I don't know. I mean, I I suspect that they were I suspect they were probably less violent, less clever, less populous, less willing to fight.

我...不知道。我是说，我觉得...这很有趣。最终在欧陆边缘将他们推向灭绝。

I I don't know. I mean, I I think And it's interesting. Drove them to extinction at the margins of Europe.

有趣的是，如果我们反复重演地球的历史，这些掌握了某种集体智慧运用方式的智慧生命分支中，哪些会涌现，哪些会成功？是更具暴力倾向的？还是更孤立的？你知道，什么样的动态会带来更高的生产力？我想我们永远无法得知。

And it's interesting how much if we ran Earth over and over again, how many of these branches of intelligent beings that have figured out some kind of how to leverage collective intelligence, which ones of them emerge, which ones of them succeed? Is it the more violent ones? Is it the more isolated one? You know, like, what what dynamics result in more productivity? And we I I suppose we'll never know.

生物体越复杂，在实验室中进行这类实验就越困难。

It's the more complex the organism, the harder it is to run the experiment in the lab.

是的。从某些方面来说，或许不知道反而是最好的。

Yes. And in some respects, maybe it's best if we don't know.

没错。真相可能会非常残酷。如果我们退一步思考人类做的几件有趣事情会怎样？其一是物体操控与移动。当然，移动能力是早已实现的。

Yeah. The truth might be very painful. What what about if we actually step back a couple of interesting things that we humans do? One is object manipulation and movement. And, of course, movement was something that was done.

这又是一项重大发明——能够在环境中自由移动。另一个是感知机制，我们如何感知环境。其中最精妙的高清感知就是视觉。这些发明在地球生命史上有多重要？

That was another big invention, being able to move around the environment. And the other one is this sensory mechanism, how we sense the environment. One of the coolest high definition ones is vision. How big are those inventions in the history of life on Earth?

视觉和移动，我的意思是，这极其重要，可以追溯到动物起源时期，寒武纪大爆发时化石记录中突然出现的眼睛。历史上人们常说'半只眼睛有什么用'，但其实你可以从扁平组织上的感光点开始，通过一系列步骤进化出带晶状体的眼球等等。假设每代仅需1%甚至0.5%的变化——具体模型我记不清了——

Vision movement, I mean, again, extremely important going back to the origin of animals, the Cambrian explosion where suddenly you're seeing eyes in the fossil record. It's not necessarily Again, of people historically have said, What use is half an eye? You can go in a series of steps from a light sensitive spot on a flat piece of tissue to an eyeball with a lens and so on. If you assume no more than I don't remember this. This was a specific model that I have in mind, it was 1% change or half a percent change for each generation.

要进化出我们现在所知的眼睛需要多久？答案是五十万年。这并不算长。但进化不是这样运作的。这并非问题的答案。

How long would it take to evolve an eye as we know it? The answer is half a million years. It doesn't have to take long. That's not how evolution works. That's not an answer to the question.

这仅仅表明你可以重构这些步骤，大致推演出它是如何运作的。进化出眼睛并非什么难事，但一旦拥有眼睛，就无处可藏了。我们再次回到捕食者与猎物的关系，回到视觉能力带来的所有优势。如果从哲学角度思考蝙蝠如何运用回声定位等能力，我虽不确定，但推测它们构建世界图像的方式与我们大同小异。这不过是心智重构的问题。

It just shows you can reconstruct the steps and you can work out roughly how it can work. It's not that big a deal to evolve an eye, but once you have one, then there's nowhere to hide. Again, we're back to predator prey relationships, we're back to all the benefits that being able to see brings you. If you think philosophically what bats are doing with eco location and so on, I have no idea but I suspect that they form an image of the world in pretty much the same way that we do. It's just a matter of mental reconstruction.

关于视觉的另一点是，有些单细胞生物已具备晶状体、视网膜和角膜等结构。本质上，它们在一个细胞内就拥有了相机式的眼睛。它们没有大脑，我们无从知晓它们对世界的理解，但它们却能演化出相同的结构来实现这一点。我想说的是，一旦拥有了眼睛这类器官，中枢神经系统就会面临巨大的进化压力去解读这些信号的意义。

So I suppose the other thing about sight, there are single celled organisms that have got a lens and a retina and a cornea and so on. Basically, they've got a camera type eye in a single cell. They don't have a brain. What they understand about their world is impossible to say but they're capable of coming up with the same structures to do so. I suppose then is that once you've got things like eyes, then you have a big driving pressure on the central nervous system to figure out what it all means.

这就回到你提到的另一个观点——关于操控、感官输入等——现在你需要迫切理解周围环境的本质、其含义、反应机制、何时该逃跑以及何时该保持静止。

And we come around to your other point about manipulation, sensory input and so on about now now you you have a huge requirement to understand what your environment is and what it means and how it reacts and why you should run away and why you should stay put.

实际上关于这点，让我...不知道你是否了解唐纳德·霍夫曼的研究？他运用进化论机制论证说：准确认知客观现实未必具有强烈的进化价值。我们的感知与客观现实相去甚远，我们生活在集体幻觉中。地球上所有物种基本上都在这个与物理现实截然不同的幻觉空间里竞争生存。

Actually, on that point, let me I I don't know if you know the work of Donald Hoffman, who talks about who uses the argument, the mechanism of evolution to say that there's not necessarily a strong evolutionary value to seeing the world as it is. So objective reality, that our perception actually is very different from what's objectively real. We're we're living inside an illusion, and we're basically the entire the entire set of species on Earth, think, I I guess, competing in a space that's an illusion that's distinct from that's far away from physical reality as it is, as defined by physics.

我不确定这是否算幻觉，或许更像是个认知泡泡。我们的感官输入只是潜在接收信息的一小部分，而我们根据生存需要来解读这些信息。从这个意义上说确实是幻觉，但...

I'm not sure it's an illusion so much as a bubble. I mean, we we have a sensory input which is a fraction of what we could have a sensory input on, and we interpret it in terms of what's useful for us to know to stay alive. So yes, it's an illusion in that sense, but

那棵树

the tree

确实物理存在，如果你撞上它就会明白——这并非纯粹的妄想，它具有某种物理实在性。

is physically there and if you walk into that tree, you you know, there is it's not purely a delusion, there's some physical reality to it.

所以这是一种对现实的感官切片，正因它仅是片段，你便错过了全貌。但他指出这个切片未必非得是切片，它可能是物种间普遍存在的完整虚构——这个观点很有趣，至少它让我们谦卑地认识到：我们的感知有限，认知能力也有限。对我而言，这是进化论的论点。我不确定这个论证有多强，但我确实认为生命可以存在于意识中。

So it's a it's a sensory slice into reality as it is, but because it's just a slice, you're missing a big picture. But he says that that slice doesn't necessarily need to be a slice. It could be a complete fabrication that's just consistent amongst the species, which is an interesting or at least it's it's a humbling realization that our perception is limited, and our cognitive abilities are limited. And at least to me, it's argument from evolution. I don't know how much how how strong that is as an argument, but I do think that life can exist in the mind.

是的。就像你可以玩虚拟现实游戏，在那个空间里拥有鲜活的人生。从某种角度看那个地方并不真实，但你依然能感受到同样的进化力量、同样的竞争机制、同样的人际动态。只是我不确定是否有证据表明地球上的生命就是如此产生的。看起来地球...

Yes. In the same way that you can do a virtual reality video game, and you can have a vibrant life inside that place. And that place is not real in some sense, but you could still have a vibe all the same forces of evolution, all the same competition, the dynamics of between humans you can have, but I don't know if I don't know if there's evidence for that being the thing that happened on Earth. It seems that Earth

我认为无论哪种环境，我都不会否认你描述的世界可能存在，尽管要证明这点极其困难。《黑客帝国》等电影提出的'整个世界都是构造而我们根本被蒙蔽'的观点，很难断言绝无可能。我们确实通过输入信息在心中构建外部世界，但若说这个构建不真实——我难以苟同，因为这恰恰是我们理解世界的方式。我们有眼睛，但若将人长期禁锢（据说确有被关暗室五年者），其视觉神经回路便永远无法正常发育。观察儿童成长：他撞到桌子感到疼痛，此时获得两种输入...

I think in either environment, I wouldn't deny that you could have exactly the world that you talk about, and it would be very difficult to The idea in Matrix movies and so on that the whole world is completely a construction we're fundamentally deluded. It's difficult to say that's impossible or couldn't happen, and certainly we construct in our minds what the outside world is, but we do it on input and that input, I would hesitate to say it's not real because it's precisely how we do understand the world. We have eyes, but if you keep someone in apparently this kind of thing happens, someone kept in a dark room for five years or something like that, they never see properly again because the neural wiring that underpins how we interpret vision never developed. You need When you watch a child develop, he into a table, he bangs his head on the table and it hurts. Now you've got two inputs.

其一是尖锐边缘带来的痛感，其二是触摸确认的实体存在，再加上视觉输入——三者结合后得出结论：别再撞桌子。这是学习过程，虽是有限的现实，却是真实的现实。若学不会，你就会被吞噬、被车撞、无法生存。在计算机构建的现实中亦是如此（虽然我对此领域不专），只要你设定'这就是该现实中的法则'，就必须遵守这些法则。

You've got one, pain from this sharp edge and number two, you've probably you've touched it and realized it's there, it's a sharp edge and you've got the visual input and you put the three things together and think, I don't want to walk into a table again. You're learning and it's a limited reality, but it's a true reality. If you don't learn that properly, then you will get eaten, you will get hit by a bus, you will not survive. Same if you're in some kind of, let's say, computer construction of reality. I'm not in my ground here, but if if you construct the laws that this is what reality is inside in in in inside this, then you play by those laws.

确实。只要法则保持一致性。就像实验室里探讨的模拟问题：我们是否活在模拟中固然难证实，但如今这类实验越来越可行。对我而言，更有趣的问题是：虚拟现实需要多逼真才能让我们无法分辨？

Yeah. Well, I mean, as long as the laws are consistent. Yeah. So just like you said in the lab, the interesting thing about the simulation question, yes, it's hard to know if we're living inside a simulation, but also, yes, it's possible to do these kinds of experiments in the lab now more and more. To me, the interesting question is how realistic does a virtual reality game need to be for us to not be able to tell the difference?

更让我感兴趣的是：虚拟世界需要多真实或多有趣，才能让我们愿意永远停留——或者比物理世界停留更久？不仅是像沉迷硬性毒品那样，而是从深刻、有意义的角度真正偏爱那个世界...

A more interesting question to me is how realistic or interesting does a virtual reality world need to be in order for us to want to stay there forever or much longer than physical reality? Prefer that place. And also prefer it not as we prefer hard drugs, but prefer it in a deep, meaningful way, in a way we we enjoy

我想《黑客帝国》的核心问题在于：能否通过充分蒙蔽意识，让人彻底相信自己接触的是真实世界？模拟需要多完美才能做到这点？它必须让你确信所有感官输入都准确无误、自洽合理。而这些感官输入并非与生俱来——

I mean, I suppose the the issue with The Matrix. I I imagine that it's possible to to delude the mind sufficiently that you genuinely, in that way, do think that you are interacting with the real world when in fact the whole thing's a simulation. How good does a simulation need to be to be able to do that? Well, it needs to convince you that all your sensory input is correct and accurate and joins up and makes sense. Now that sensory input is not something that we're born with.

我们生来就有触觉，生来就有眼睛等等，但我们不知道如何使用它们。我们不知道该如何理解这些感官。我们四处走动，撞到树木，经常哭泣，经常感到痛苦。基本上我们是在启动这个系统，让它能够理解所接收到的感官输入。

We're born with a sense of touch. We're born with eyes and so on, but we don't know how to use them. We don't know what to make of them. We go around, we bump into trees, we cry a lot, we're in pain a lot. We're basically booting up the system so that it can make head a tail of this sensory input that it's getting.

这种感官输入不仅仅是单向的信息流，你还必须去触碰物体、聆听声音，并将这些信息整合起来。如果把婴儿直接放入矩阵中，我认为他们不会有那种感官输入。我不认为他们能以任何方式理解纽约作为他们所处的世界。大脑并不是以那种方式发育的。

That sensory input is not just a one way flux of things, it's also you have to walk into things, you have to hear things, you have to put it together. Now, if you've got just babies in the matrix who are slotted into this, I don't think they have that kind of sensory input. I don't think they would have any way to make sense of New York as a world that they're part of. Brain is just not developed in that way.

好吧，我在这个物理现实中也无法理解纽约。但你提到了疼痛和碰撞物体。其实你可以创造疼痛信号。只要某些事物总是导致疼痛这一规律保持一致，你就能开始构建现实。也许你不同意这点，但我认为我们几乎生来就有一种渴望被现实说服的欲望，一种想要理解现实的欲望。

Well, I can't make sense of New York in this physical reality. But you said pain and the walking into things. Well, you can create a pain signal. And as long as it's consistent that certain things result in pain, you could start to construct a reality. There's some maybe maybe you disagree with this, but I think we are born almost with a desire to be convinced by our reality, like a a desire to make sense of our reality.

噢，我确信我们确实如此。是的。

Oh, I'm sure we are. Yes.

就像，这是一种本能。无论给我们什么样的现实——比如桌子会伤人、火是烫的——我认为我们在某种意义上渴望被'欺骗'，想要建立一个像爱因斯坦理论那样简单的对周围世界的解释。是的，我们渴望那种简单性。

Like, so there's an imperative. So whatever that reality is given to us, like the table hurts, fire is hot Yeah. I think we wanna be deluded in a sense that we want to make a simple, like Einstein's simple theory of the thing around us. Yeah. We want that simplicity.

所以也许对简单性的渴望，可以用来构建一个相当拙劣却能欺骗我们的模拟世界。也许欺骗人类并不需要构建一个完整的宇宙。

And so maybe the hunger for the simplicity is the thing that could be used to construct a pretty dumb simulation that that tricks us. So maybe tricking humans doesn't require building a universe.

不，我...我不...这不是我的研究领域，所以我不知道我们离实现这个有多近。也不清楚是否有人在研究这个。但我同意...是的。我不确定这在道德上是正当的行为。

No. I I don't. This is not what I work on, so I don't know how close to it we are. Think anyone works on But I agree with Yeah. I'm not sure that it's a morally justifiable thing to do.

但从原则上讲，这有可能吗？我认为会非常困难，但我不认为原则上不可能。我同意你的观点，我们试图理解世界，整合感官输入，并提出解释现状的假设。不过我认为我们所处的社会语境提供了巨大输入，我们并非独自完成这一切。

But it's it's is it possible in principle? I think it would be very difficult, but I don't see why in principle it wouldn't be possible. And I agree with you that it's that we try to understand the world, we try to integrate the sensory inputs that we have, and we try to come up with a hypothesis that explains what's going on. I think though that we have huge input from the social context that we're in. We don't do it by ourselves.

我们并非独自在宇宙中跌跌撞撞地摸索理解全局。周围人会告诉我们事物的本质和功能，语言也在此发挥作用。因此要模拟所有这些，必须构建极其逼真的仿真系统。

We don't kind of blunder around in a in a universe by ourselves and and and understand the whole thing. We're told by the people around us what things are and what they do, and language is coming in here and so on. So it would have to be an extremely impressive simulation to simulate all of that.

是的，模拟所有这一切，包括社会建构、思想传播与交流。我不知道答案。但随着我们日益数字化，这些问题至关重要。进化的下一步似乎是我们逐渐融入机器，所有讨论过的机制正越来越多地接入系统。

Yeah. Simulate all of that, including the social construct, the the thing that the spread of ideas and the exchange of ideas. I don't know. And but those questions are really important to understand as we become more and more digital creatures. It seems like the next step of evolution is us becoming partial all the same mechanisms we've talked about are becoming more and more plugged in into the machine.

我们正在变成赛博格，电线与生物体之间存在着有趣的互动。0和1与生物系统的交融。我认为很快我们就不能再将人类与所创造的技术割裂看待。我们是一个正在...

We're becoming cyborgs, and there's an interesting interplay between wires and biology. You know, zeros and ones and the biological systems. And I don't think you can just I don't think we'll have the luxury to see humans as disjoint from the technology we've created for much longer. We are an organism that's

是的。我同意你的观点，但这确实将我们引向意识问题。

Yeah. I mean, I agree with you, but we come really with this to consciousness.

没错。

Yes.

这里是否存在区别？因为按照你的说法，自然终点是我们将无法区分。如果你能建造一个足够接近人类、最终与之融合的AI，那么实质上这个AI就具有我们所知的意识。我没有强烈立场，但我在上一本书的尾声写过相关思考——十年前我在《生命进阶》一书中就专门写过关于意识的章节。

And is there a distinction there? Because what you're saying, the natural endpoint says we are indistinguishable. That if you are capable of building a an AI which is sufficiently close and similar that we merge with it, then then to all intents and purposes, that AI is conscious as we know it. I don't don't have a strong view, but I have a view. I wrote about it in the epilogue to my last book because ten years ago, I wrote a chapter in book called Life Ascending about consciousness.

《生命上升》的副标题是‘进化的十大发明’，我绝不可能写一本不包含意识的、以这类副标题命名的书，尤其是将意识列为伟大发明之一。部分原因是我出于好奇想了解更多，因此为那一章节阅读了大量资料。我虽从未专门研究过这个问题，但始终困惑——怎么会有人对此不感兴趣呢？最终我形成了两个观点：第一，无人知晓答案；第二，学界存在两大主要流派，且观点分布极不均衡。其中一派认为意识是物质的属性。

The subtitle of Life Ascending was the 10 great inventions of evolution, and I couldn't possibly write a book with a subtitle like that that did not include consciousness, and specifically consciousness as one of the great inventions. It was in part because I I was just curious to know more and I read more for that chapter. I never worked on it but I've always How can anyone not be interested in in the question? I was left with the feeling that a, nobody knows and b, there are two main schools of thought out there with a big kind of skew in distribution. One of them says, oh, it's a property of matter.

这是物理学中未知的法则，泛心论认为万物皆有意识，太阳有意识，岩石也有意识，只是程度不同。我觉得这种说法非常缺乏说服力。虽不能断言它是错的，但人类理应能区分生命体与非生命体。另一派则认为意识是高度复杂的中枢神经系统涌现的特性。而我始终无法理解‘涌现’这类术语的确切含义。

It's an unknown law of physics, panpsychism, everything is conscious, the sun is conscious, it's just a matter of or a rock is conscious, it's just a matter of how much. I find that very unpersuasive. I can't say that it's wrong, it's just that I think we somehow can tell the difference between something that's living and something that's not. And then the other the other end is it's an emergent property of a very complex central nervous system. And I never quite understand what people mean by words like emergence.

确实存在某些实例，但我认为人们常常用这个词来掩盖无知。作为生物化学家，我的疑问是：既然如此，那么意识就是中枢神经系统的某种特殊构造。一个去极化的神经元会产生感觉——无论是疼痛、爱意、愤怒或其他感受。那么从生物物理学的角度来说，中枢神经系统中的‘感觉’究竟是什么？究竟是哪段神经回路产生了这种感觉？

Mean there are genuine examples where I think we very often tend to use it to plaster over ignorance. As a biochemist, the question for me then was, okay, it's a it's a concoction of a of a central nervous system. A depolarizing neuron gives rise to a feeling, to a feeling of pain or to a feeling of love or anger or whatever it may be. So what is then a feeling in biophysical terms in the central nervous system? Which bit of the wiring gives rise to?

我从未见过任何人能用让我信服的方式回答这个问题。

And I've never seen anyone answer that question in a a way that makes sense to me.

这是个亟待解答的重要问题。

And that's an important question to answer.

若想理解意识，这恐怕是唯一需要解答的问题。因为人工智能迟早——或许已经——能在思维层面超越人类。仅就信息处理和计算能力而言，设计出匹敌人脑的智能系统毫无难度。但人类最珍视的是那些感性体验：情感、情绪，以及个人化的世界观。这些对人们的意义不亚于、甚至远超信息处理能力。正因如此，我不认为人工智能必然会具备意识，因为意识很可能是生命特有的属性。

I think if we want to understand consciousness, that's the only question to answer because I you know, certainly an AI is capable of out thinking and it's only a matter of time. Maybe it's already happened. In terms of just information processing and computational skill, I don't think we have any problem in designing a mind which is at least the equal of the human mind. But in terms of what we value the most as humans, which is to say our feelings, emotions, our sense of what the world is in a in a very personal way, that's I I think means as much or more to people than their information processing. And that's where I don't think that AI necessarily will become conscious because I think it's a property of life.

让我们深入探讨这个话题。作为我最喜爱的作家之一，请允许我引用您新书《变形者》中关于意识的论述。笛卡尔‘我思故我在’是有史以来最著名的哲学命题，但‘我’究竟是什么？

Well, let's talk about it more. You're an incredible writer, one of my favorite writers. So let me read from your latest book, Transformers, what you write about consciousness. I think therefore I am, said Descartes, is one of the most celebrated lines ever written. But what am I exactly?

根据定义，人工智能也能思考，因此它确实存在。然而，我们中很少有人能认同AI原则上是否具备类似人类情感的能力，比如爱或恨、恐惧与喜悦，对合一或遗忘的精神渴望，或是饥渴等肉体痛苦。问题在于我们不知道情感究竟是什么，正如你所说。从物理角度而言，感受是什么？放电的神经元如何产生任何形式的感受？

An artificial intelligence can think too by definition, and therefore is. Yet few of us could agree whether AI is capable in principle of anything resembling human emotions, of love or hate, fear and joy, of spiritual yearnings for oneness or oblivion, or corporeal pangs of thirst and hunger. The problem is we don't know what emotions are, as you were saying. What is the feeling in physical terms? How does a discharging neuron give rise to a feeling of anything at all?

这就是意识的难题。心灵与物质看似二元对立，我们最内在自我的物理构成。我们原则上可以理解一个极其复杂的并行处理系统如何能够实现惊人的智能壮举，但原则上我们无法回答这种至高智能是否会体验喜悦或忧郁。慰藉的量子是什么？我在谈论涌现问题，你知道，最近有篇关于神经网络在自然语言处理问题上性能涌现的相变的优秀论文。

This is the hard problem of consciousness. The seeming duality of mind and matter, the physical makeup of our innermost self. We can understand in principle how an extremely sophisticated parallel processing system could be capable of wondrous feats of intelligence, but we can't answer in principle whether such a supreme intelligence would experience joy or melancholy. What is the quantum of solace? I speaking to the question of emergence, you know, there's just technical there's there's a there's an excellent paper on this recently about the this kind of phase transition emergence of performance in neural networks on problem NLP, natural language processing.

对于语言模型，似乎存在规模问题。当神经网络规模增长到某个临界点时，会出现相变。这个问题既是技术性的，也可以进行哲学探讨。技术问题是：是否存在某个神经网络规模，能够开始形成足以捕捉语言的表征，从而不仅能处理语言，还能通过语言捕捉足够知识来解决诸多语言问题，比如进行对话。这种能力提升似乎不是渐进的，而是突变的相变。

So language models, there seems to be this question of size. At some point, there is a phase transition as you grow the size of the neural network. So the question is that's sort of somewhat of a technical question that you can philosophize over. The technical question is, is there a size of a neural network that starts to be able to form the kind of representations that can capture a language, and therefore be able to, not just language, but linguistically capture knowledge that's sufficient to solve a lot of problems in language, like, be able to have a conversation. And there seems to be not a gradual increase, but a phase transition.

我们正在试图构建相关科学理论来确定这个临界点。比如：神经网络的理想规模是多少？为何会发生这种相变？总之，这指向了涌现现象。事物可能存在这样的发展阶段：从'哦，你是个非常智能的面包机'变成'今天感到悲伤的面包机'，它转身望向窗外叹息，经历存在主义危机。

And we and they're trying to construct the science of where that is. Like, what is a good size of a neural network, and why does such a phase transition happen? Anyway, that that sort of points to emergence that there Yeah. There could be stages where a thing goes from being, oh, you're you're you're very intelligent toaster to a toaster that's feeling sad today and turns away and looks out out the window sighing, having an existential crisis.

我想到的肯定是那个偏执机器人马文。

I'm thinking of Marvin, the paranoid android for sure.

不，准确地说不是。马文太简单了，因为它只是脾气暴躁。

Exact well, no. Marvin is simplistic because Marvin is just cranky.

是的，它太容易被编程了。

Yes. It's a So easily programmed.

是啊，轻而易举就能编程出持续的存在危机。你几乎就是什么来着？陀思妥耶夫斯基的《地下室手记》。就像是在不停地抱怨生活。

Yeah. Easily programmed, nonstop existential crisis. You're almost basically what is it? Notes from Underground by Dostoevsky. Like, it's just just constantly complaining about life.

不，他们捕捉的是人类情感的完整过山车，兴奋、极乐、联结、共情，诸如此类，然后是自私、愤怒、抑郁，所有这些。他们捕捉到这一切并能深刻体验。这可能是你今天能经历的最重要的事，巅峰的喜悦，低谷的绝望，就是这样。

No. They're capturing the full roller coaster of human emotion, the excitement, the bliss, the connection, the empathy, and all that kind of stuff, and then the selfishness, the the anger, the the depression, all that kind of stuff. They're capturing all of that and be able to experience it deeply. Like, it's the most important thing you could possibly experience today. The highest highs, the lowest lows, this is it.

我的生命将终结。我无法继续承受那种感觉，然后，打个盹后，你会感觉棒极了。这可能是会浮现出来的现象。

My life will be over. This I cannot possibly go on that feeling, and then, like, after a nap, you're you're you're feeling amazing. That might be something that emerges.

那为什么打个盹会让AI生命体感觉更好呢？

So why would a nap make an AI being feel better?

首先，我们连人类为何如此也不清楚，对吧？

The first of we don't know that for a human either. Right?

但我们确实知道这对许多人多数时候确实如此。也许你会——哦，你是说实际上确实会感觉更好。所以

But we do know that that's actually true for many people much of the time. Maybe you're Oh, you mean happen, you do, in fact, feel better. So

哦，你实际上是在问技术问题——这有个非常生物学层面的答案。所以问题在于AI是否需要拥有对其身体、生理机能以及大脑自我保存成功运作的相同依赖，本质上是从某种深层生物学意义上来说。

Oh, you are actually asking the technical question there is that so that's a very there's a biological answer to that. And so the question is whether AI needs to have the same kind of attachments to its body, bodily function, and preservation of the brain's successful function of self preservation essentially in some deep biological sense.

我是说，在我看来，这又回到了我们之前讨论的关于模拟、感官输入以及理解这一切意义与生死的问题。生物学与社会不同，它对一切事物都施加着死刑，而自然选择正是基于这种死刑运作。如果你做出错误决定，就会死亡。下一代则由那些在权衡中做出略微不同决定的生物所代表。这本质上是一种难以全面模拟其丰富性的东西，可以说。

I mean, to my mind, it comes back around to the problem we were talking about before about simulations and sensory input and learning what all of this stuff means and life and death. That biology, unlike society, has a death penalty over everything And natural selection works on that death penalty. That if you make this decision wrongly, you die. And the next generation is represented by beings that made a slightly different decision on balance. And that is something that's intrinsically difficult to simulate in all its richness, would say.

所以，所以什么是

So so what is

死亡的全部丰富性。是的。我们与死亡的关系，或者说整个死亡的概念。当你提到丰富性时，当然包含了很多内容。是的。

Death in all its richness. Yeah. Our relationship with death or or or or the whole of it. So which when you say richness, of course, there's a lot in that. Yeah.

哪部分难以模拟？丰富性中哪部分最难模拟？

Which is hard to simulate? What what's the what's part of the richness that's hard to simulate?

我想是环境的复杂性以及你在其中的位置，或者说生物在其整个生命周期、多代繁衍中，在基因序列变化和个体构成细微改变的环境中的位置。但如果你回溯到单细胞层面——我在书中正是这样做的——然后追问：单细胞如何‘知道’自己作为一个单位或实体存在？当然，用引号强调它其实‘知道’不了什么。但它确实作为一个单位行动，且以惊人的精确度运作。我曾提出这与膜上的电场有关，它们能实时反馈‘我在环境中表现如何’这类信息。

I suppose the complexity of the environment and your position in that or the position of an organism in that environment in the full richness of that environment over its entire life, over multiple generations with changes in gene sequence over those generations, slight changes in the makeup of those individuals over generations. But if you take it back to the level of single cells, which I do in in in the book and and ask how do how how does a single cell in effect know it exists as a unit, as an entity? I mean, in inverted commas, obviously, it doesn't know anything. But it acts as a unit and it it acts with astonishing precision as a unit. And and I had suggested that that's linked to the electrical fields on on the membranes themselves and that they give some indication of how am I doing in relation to my environment as a kind of real time feedback on the world.

这是种可被多代选择的物理特性。如果你搞错了，就会触发特定情境反应——作为个体可能突然恐慌逃跑；作为细菌可能产生放电信号引发逃逸。通过多代关联，这种电相位状态会与特定反应绑定。几乎能直观看到感受是如何通过这种实时反馈机制‘后门’渗入的。当你将系统复杂化后，我完全认同过渡的可能性。但这一切能否仅靠系统自我理解的机制语言实现？

And this is something physical which can be selected over generations that if you are if you get this wrong, it's linked with this set of circumstances that I've just as an individual, I have a moment of blind panic and run. As a bacterium or something, have some electrical discharge that says blind panic and it runs, whatever it may be. And you associate over generations, multiple generations that this electrical phase that I'm in now is associated with a response like that. And it's easy to see how feelings come in through through the back door almost with with that that kind of giving real time feedback on your position in the world in relation to how am I doing. And then you you complexify the system and yes, I have no problem with with with with transition and I, you know Can can all of this be done purely by the language, by the the the the issues with how the system understands itself?

也许可以，老实说我也不知道？哲学家们长期讨论过‘僵尸智能’的可能性——即具备完全相同功能却毫无感受的存在。这个问题恐怕得抛回给你：你如何看待僵尸智能？

Maybe it can, I honestly don't know? But I I you know, the philosophers for a long time have talked about the possibility that you can have a zombie intelligence and that there are no feelings there, but all everything else is the same. Is I mean, I have to throw this back to you, really. How do you deal with the zombie intelligence?

首先，从生物学家的视角来看，你会思考所有造就人类的复杂因素。那四十亿年历史的全部，在某种深刻意义上将人类融入这个环境，那种生物与环境的共舞。你可以看到情感如何从中产生，这些情感紧密相连，共同塑造了人类的体验。在此基础上，混入意识及其全部混乱。是的。但对于一个已经存在的智能生物——比如正在学习的婴儿——它不需要学习如何成为细胞集合体，或如何完成所有必需的事情。

So first of all, I can see that from a biologist perspective, you think of all the complexities that led up to the human being. The entirety of the history of four billion years that in some deep sense integrated the human being into this environment, and that dance of the organism and the environment, You could see how emotions arise from that, and their emotions are deeply connected and create creating a human experience. And from that, you mix in consciousness and full mess of it. Yeah. But from a perspective of an intelligent organism that's already here, like a baby that learns, it doesn't need to learn how to be a collection of cells or how to do all the things he needs to do.

婴儿学习的基本功能是与环境互动，从环境中学习，学习如何融入社会，学会喜爱——

It's it's the basic function of a baby as it learns is to interact with its environment, to learn from its environment, to learn how to fit in to the social society, to like

而婴儿的基本反应大多时候就是哭闹。

And the the the basic response of the baby is to cry a lot of the time.

哭闹是为了说服人类保护它、管教它、教导它。我是说，我们进化出各种技巧让父母照顾我们、教育我们、向我们传授世界知识。同时，我们构建的世界既足够安全让我们生存，又足够危险以学习宝贵教训。比如桌子仍有坚硬的角，你仍可能撞上去——会疼。所以AI需要解决的是这个问题，而非构建那种超级复杂的、能让你...你知道，要做苹果派就得先创造整个宇宙的有机体。嗯。

To cry to well, to convince the humans to to protect it or to discipline it, to teach it, to what if I mean, we've developed a bunch of different tricks how to get our parents are to take care of us, to educate us, to teach us about the world. Also, we've constructed the world in such a way that it's safe enough for us to survive in, and yet dangerous enough for learning the valuable lessons. Like, the tables are still hard with corners, so can still run into them. It hurts, like, how so AI needs to solve that problem, not the problem of constructing this super complex organism that leads up so you to run the whole you know, to make an apple pie, to build the whole universe, you need to build the whole universe. Mhmm.

我认为僵尸问题该留给哲学家们，同样交给他们的还有爱的定义——当两个人之间产生那种世间万物都黯然失色的魔力时，那种你一生都在追寻的感觉、这个瞬间、这个人。哲学家们可以尽情探讨这种感觉，也可以说这只是...你可以用生物学解释，可以用各种理论说这都是假的。实际上安·兰德会说这全是自私的。存在许多不同解读。

I think the the zombie question is a it's something I would leave to the philosophers because and I will also leave to them the definition of love and what is what happens between two human beings when there's a magic that just grabs them, like nothing else matters in the world, and somehow you've been searching for this feeling, this moment, this person your whole life. That feeling, the philosophers can have a lot of fun with that one, and also say that that's just you you could have a biological explanation. You could have all kinds of it's all fake. It's actually, Ayn Rand will say it's all selfish. There's a lot of different interpretations.

我把它留给哲学家。关键是这种感觉真实得可怕。如果我的烤面包机让我觉得它是世上唯一，离别时思念它，重逢时兴奋不已，它让我的生活充满意义与欢乐，周围朋友也因它的存在见到更好的我——那这烤面包机感觉上绝对是有意识的。

I'll leave it to the philosophers. The point is the feeling sure as hell feels very real. And if my toaster makes me feel like it's the only toaster in the world, and when I leave and I miss the toaster, and when I come back, I'm excited to see the toaster, and my life is meaningful and joyful, and the friends I have around me get a get a better version of me because that toaster exists, that sure as hell feels conscious toaster.

这和养狗心理上有区别吗？不。我是说，多数人会质疑能否用狗举例。我认为狗无疑是有意识的，但...但有些人...

Is that psychologically different to having a dog? No. Mean, most people would dispute whether we can say a dog. I would I would say a dog is undoubtedly conscious, but but but some people

关于意识的程度等等，但人们肯定更不愿意说一台烤面包机是有意识的，而狗则不然。这其中仍存在一种深层次的联系，你可以说我们与狗的关系更多源于拟人化倾向——我们某种程度上将人类的特质投射到它们身上。

degrees of consciousness and so on, but people are definitely much more uncomfortable saying a toaster Yes. To be conscious than a dog. And there's still a deep connection that you could say our relationship with the dog has more to do with anthropomorphism. Like we kind of project the human being onto it.

也许吧。

Maybe.

我们对烤面包机也能干同样的事。

We can do the same damn thing with a toaster.

是的。但你可以直视狗的眼睛，看到它的悲伤或重逢时的喜悦。顺便说我没有养狗，并不是说我是...

Yes. But you can look into the dog's eyes and you can see that it's sad that it's it's delighted to see you again. I don't have a dog by way. I'm not It's not that I'm

一个爱狗人士。实际上狗极其擅长用眼神传达这些情感，它们确实如此。

a dog person. Dogs are actually incredibly good at using their eyes to do just that. They are.

我并不认为狗的智力水平能接近人工智能。是的。但它确实能与我们进行情感交流。

Now, I don't imagine that a dog is remotely as close to being intelligent as an AI intelligence. Yes. But it's certainly capable of communicating emotionally with us.

但我想大胆提出：我们常因AI擅长下棋和折叠蛋白质就认定它智能。我认为要与人类沟通、拥有情商，实际上需要更高量级的智能——不完美地存在并非易事。

But here's what I would venture to say. We tend to think because AI plays chess well Yeah. And is able to fold proteins now well, that it's intelligent. I would argue that in order to communicate with humans, in order to have emotional intelligence, it actually requires another order of magnitude of intelligence. It's not easy to be flawed.

解决数学谜题与人类之间互动的全部复杂性并不相同。实际上，我们人类总是对自己真正擅长的事情习以为常。总有人对我说人类开车技术有多糟糕。不，人类驾驶技术其实惊人地出色。

Solving a mathematical puzzle is not the same as the full complexity of human to human interaction. That's actually we humans just take for granted the things we're really good at. Nonstop people tell me how shitty people are at driving. No. Humans are incredible at driving.

双足行走、步行、物体操控，我们在这方面都极其出色。所以人们往往会低估

Bipedal walking, walking, object manipulation, we're incredible at this. And so people tend to Discount

那些我们习以为常的事情。

the things we all just take for granted.

而他们低估的其中一项能力，就是我们彼此交谈互动的舞蹈般默契。那种将想法融合的能力，互相生气后又思念的能力。就像创造注意力那样，让生活既有趣又艰难，充满有意义的挑战。这是一种需要学习的技能，而AI需要解决这个问题。

And one of those things that they discount is our ability, the dance of conversation and interaction with each other. The the ability to morph ideas together, the ability to get angry at each other, and then to miss each other. Like, to create attention that makes life fun and difficult and challenging in a way that's meaningful. That is a a skill that's learned and AI would need to solve that problem.

从某种意义上说，你的意思是AI无法真正变得有情感，除非它经历某种内在冲突——无法调和现实的各个方面或其自身现实与需要做出的决定。然后它必然会感到悲伤，因为它不知道该怎么做。对此我当然无法反驳。很可能这就是它的运作方式。我认为唯一验证的方法就是去实现它，然后

I mean, in in some sense, what you're saying is AI cannot become meaningfully emotional, let's say, until it experiences some kind of internal conflict that is unable to reconcile these various aspects of reality or its reality with with a decision to make. And then it feels sad necessarily because it doesn't know what to do. And so and and I I I certainly can't dispute that. That may very well be how it works. I think the only way to find out is to do it and

构建出来。没错。至于它是否真的感到悲伤，就留给哲学家去思考吧。关键是机器人会独自坐在那里经历内心冲突，存在主义危机，而这正是它与人类建立深刻有意义连接的必要条件。但它真的感受到这些了吗？

Build it. Yeah. And leave it to the philosophers if it actually feels sad or not. The point is the robot will be sitting there alone having an internal conflict, an existential crisis, and that's required for it to have a deep meaningful connection with another human being. Now does it actually feel that?

我不知道。

I don't know.

但我想提出另一个困扰我的问题，源于阅读诺亚·哈拉利的《21世纪的21堂课》。他在多个场合提到这类观点：将生物化学视为算法，那么人工智能必将能破解这个算法并比人类做得更好。届时会出现比莫扎特更懂创作我们欣赏的音乐、比莎士比亚更擅长写作的人工智能，诸如此类。

But I'd like to throw something else at you which which troubles me on on reading it. Noah Harari's book, 21 Lessons for the twenty first Century, and he's written about this kind of thing on on various occasions. He sees biochemistry as an algorithm. Then AI will necessarily be able to hack that algorithm and do it better than humans. There will be AI better at writing music that we appreciate than Mozart ever called or writing better than Shakespeare ever did, so on.

既然生物化学是算法性的，只需找出算法中哪些部分能操控我们的愉悦、痛苦或审美。作为生物化学家，我认为这个论点近乎无可辩驳，却令人不适——这是作为人类的本能反应。或许你会喜欢这个说法，因为它暗示AI能拥有与我们相同的情感体验，毕竟一切皆算法，而算法可编程。

Because biochemistry is algorithmic, all you need to do is figure out which bits of the algorithm to play to make us feel good or bad or appreciate things. And it's as a biochemist, I find that argument close to irrefutable and not very enjoyable. I don't like the sound of it. That's just my reaction as a human being. You might like the sound of it because that says that AI is is is capable of the same kind of emotional feelings about the world as we are because the whole thing is an algorithm and you can program an algorithm and there you are.

他在结尾章节以奇特方式单独讨论意识，涉及冥想等与内在意识连接的内容。我不懂冥想，但他对此的描述截然不同，仿佛这是逃离算法的出路。在我看来，这种意识概念足以瓦解算法体系——虽然生化反馈循环无疑是算法性的。

He then has a peculiar final chapter where he talks about consciousness in rather separate terms and he's talking about meditating and so on and getting in touch with his inner conscious. I don't meditate, I don't know anything about that. But he wrote in very different terms about it as if somehow it's a way out of the algorithm. Now it seems to me that consciousness in that sense is capable of scuppering the algorithm. I think in terms of the biochemical feedback loops and so on, it is undoubtedly algorithmic.

但我们的行为决策可以更基于情感。我们会突然想'我不管了'，无法解决复杂局面时就任性而为。这本质上是以另一种'货币'——不受个人控制的情感觉知——作为决策基础。

But in terms of what we decide to do, it can be much more based on an emotion. We can just think, I don't care. I can't resolve this complex situation. I'm going to do that. And that can be based on, in effect, a different currency, which is the currency of feelings and something where we don't have very much personal control over.

这又绕回到你对AI的终极追求：是否需要构建能推翻理性决策的系统？当信息冲突无法抉择时，用情感判断直接决定'就这么做，看看结果'？在我看来，这正是意识真正的运作方式。

And then it comes back around to to to you and what are you what are you trying to get at with AI? Do we need to have system which is capable of overriding rational a decision which cannot be made because there's too much conflicting information by effectively an emotional judgmental decision that just says, do this and see what happens? Yeah. That's what consciousness is really doing in my view.

问题在于这是不同进程还是更高级进程。将生物化学简化为算法的观点过于片面——表面无可辩驳，却丢失了本质。就像说宇宙是信息处理系统没错，但忽略了熵增过程中如何孕育出创造复杂奇迹的局部。

Yeah. And the question is whether it's a different process or just a higher level process. I might you know, the idea that biochemistry is an algorithm is, to me, an oversimplistic view. There's a lot of things that the moment you say it, it's irrefutable, but it simplifies Of course. It's an extraordinary to break this loses something fundamental.

你可以把宇宙看作执行计算的计算机，但这忽略了熵流如何催生那些如机器般精妙的复杂结构，而这些结构又通过进化构建出更深的复杂性。

So for example, calling a universe an information processing system, sure. Yes. You could you could make that. It's a computer that's performing computations, but you're missing the the process of the entropy somehow leading to pockets of complexity that creates these beautiful artifacts that are incredibly complex, and they're like machines. And then those machines are through the process of evolution are constructing even further complexity.

比如，在称呼宇宙信息处理机器时，你忽略了那些微小的局部区域以及创造它们的难度。所以对我来说问题是，如果生物化学是一种算法，那么创建一个运行人体的软件系统有多困难

Like, in calling universe information processing machine, you're you're missing those little local pockets and how difficult it is to create them. So the question to me is if biochemistry is an algorithm, how difficult is it to create an a software system

好的。

Okay.

我认为这种想法是错误的。我认为那将需要极其漫长的时间——我的意思是，要能重构人类，恐怕得等到几个世纪之后。但我敢说，要获得人类某些神奇特质，比如我们所说的情感互动，就像狗狗能让我们微笑快乐那样，这些会来得快得多。但这并不需要我们逆向工程生物化学的算法。

That that runs the human body, which I think is incorrect. I think we're that is going to take so long. I can't I mean, that's going to be centuries from now to be able to reconstruct the human. Now what I would venture to say to get some of the magic of a human being with what we're saying with the emotions and the interactions and like like a dog makes us smile and joyful and all those kinds of things, that will come much sooner. But that doesn't require us to reverse engineer the algorithm of biochemistry.

是的。但烤面包机让你感到快乐。重点不在于你是否能让烤面包机快乐。

Yes. But the toaster is making you happy. Yes. It's not about whether you make the toaster happy.

不。必须这样。必须这样。烤面包机必须能够离开我。

No. It has to be. It has to be. It has to be. The toaster has to be to leave me.

没错。但这里的烤面包机代表AI，它是非常智能的

Yeah. But it's the toaster is the AI in this case. It's a very intelligent

烤面包机必须能够感到不快乐并离开我。这很关键。是的。这是我能够想念烤面包机的前提。如果烤面包机只是我的仆人，或是服务提供者，比如告诉我天气和烤面包，那就无法建立深层连接。

The toaster has to be able to be unhappy and leave me. That's essential. Yeah. That's essential for my being able to miss the toaster. If the toaster is just my servant, that's not or or a provider of, like, services, like, tells me the weather and makes toast, that's not going to deep connection.

它必须包含内在冲突。你要写生与死。它必须能意识到自身的死亡和存在的有限性，并且明白生命是

It has to have internal conflict. You write about life and death. It has to be able to be conscious of its mortality and the finiteness of its existence, and that life is

它是

it is

短暂的，因此需要更加有选择性。

temporary and therefore needs to be more selective.

我小时候看电影时最感动的时刻之一是《2001太空漫游》中哈尔被断电的场景，那是一个有知觉生命的死亡，哈尔自己也知道。所以我认为我们都隐约明白，一个足够智能的存在终将拥有某种形式的意识。但这是否会类似于生物意识，我就不得而知了。如果你在思考如何融合——我是说我们将与AI更紧密互动，但真的...狗真的像烤面包机吗？还是说两者确实存在本质区别？你提到生物化学是算法化的，但它不是单一算法，当然非常复杂。

One of the most moving moments in the movies from when I was a boy was the the unplugging of Hal in 2001, where that was the death of a sentient being and Hal knew it. So I think we we all kind of know that that a sufficiently intelligent being is going to have some form of consciousness. But whether it would be like biological consciousness, I just don't know. If you're thinking about how do we bring together, I mean, we're going to interact more closely with AI, but are we really Is a dog really like a toaster or is there really some kind of difference there? You were talking of biochemistry is algorithmic but it's not single algorithm and it's very complex, of course it is.

所以可能生物化学回路中再次存在冲突，但我感觉单个细胞层面的整个生物化学系统的复杂度，要低于人脑或AI中的神经网络层面。

So it may be that there are again conflicts in the circuits of biochemistry but I have a feeling that the level of complexity of the total biochemical system at the level of a single cell is less complex than the the level of neural networking in the human brain or in an AI.

嗯，我猜我默认我们把大脑也包含在生物化学算法里了，因为你必须

Well, I guess I assumed that we were including the brain in the biochemistry algorithm because you have to

我会将其视为神经网络的更高层次组织。它们内部都使用着相同的生物化学连接方式。

I would see that as a higher level of organization of neural networks. They're all using the same biochemical wiring within themselves.

是的。但人类大脑不仅仅是神经元。它还包括免疫系统，是整个系统的综合体。我是说，要运行一个智能生物系统的生化算法，你必须考虑整个该死的系统。

Yeah. But the human brain is not just neurons. It's the immune system. It's it's the whole package. I mean, to have a biochemical algorithm that runs a a intelligent biological system, you have to include the whole damn thing.

这相当神奇，它源自一个胚胎。就像，整个...我是说，天啊。我是说，如果你能...什么是人类？因为它是...

It's pretty fascinating that it comes from, like, from an embryo. Like, the whole I mean, ugh, boy. I mean, if you can what is a human being? Because it's

但如果你观察

But if you look

某些代码，然后你构建...所以DNA不仅告诉你构建什么，还告诉你如何构建。我是说，这很令人印象深刻，问题是要逆向工程整个系统有多困难？

at some code, and then you build and then that so its DNA doesn't just tell you what to build, but how to build it. Is it I mean, the thing is impressive, and the question is how difficult is it to reverse engineer the whole shebang?

非常困难。

Very difficult.

我...我会说...不想说不可能，但这就像...构建一个人比逆向工程要容易得多...比逆向工程整个进化过程如何运行的...要容易得多...

I I would say it's don't want to say impossible, but it is like, it's much easier to build a human than to reverse engineer to to build like a fake human human like thing than to reverse engineer the entirety of the process of evolution of how how to run

当然，如果我们有能力逆向工程整个系统的话。是的。如果...如果人类思维能做到这一点。我是说，如果我不尝试，我就不会成为生物学家了。是的。

sure if we are capable of reverse engineering the whole thing. Yeah. If our if the human mind is capable of doing that. I mean, I wouldn't be a biologist if I wasn't trying. Yeah.

但我知道自己无法理解整个问题，我只是试图把握问题的基本轮廓。不过还有另一个方面，你谈到从单细胞发育到人类心智，以及免疫系统等所有子系统。我想迈克尔·莱文会讨论这个。但我们对逆向工程知之甚少。

But I know I can't understand the whole problem. I'm just trying to understand the rudimentary outlines of the of the problem. There's another aspect though, you're talking about developing from a single cell to the human mind and all the part system subsystems that are part of the immune system and so on. This is something that you'll talk about, I imagine, Michael Levin. But so little is known about talk about reverse engineer.

关于从基因组到完全发育生物体的发展路径，我们了解得太少。其中很多似乎取决于我提到的单细胞层面的电相互作用及其与环境的互动。生物学中存在一个完整的电场维度尚未写入任何教科书——即胚胎或单细胞如何发育成这些复杂系统。是什么定义了头部？是什么定义了免疫系统？

So little is known about the developmental pathways that go from a genome to going to a fully wired organism. A lot of it seems to depend on the same electrical interactions that I was talking about happening at the level of single cells and its interaction with the environment. There's there's there's there's a whole electrical field side to biology that is not yet written into any of the textbooks, which is about how does an embryo develop into or a single cell develop into into these complex systems. What defines the head? What defines the immune system?

是什么定义了大脑等等？这些确实是用我们刚刚开始理解的语言书写的。坦率地说，大多数生物学家仍非常不愿涉足电场影响发育这类问题。对许多生物学家来说这像天方夜谭，但本不该如此——这是二十一世纪的生物学，是未来的方向。但除非我们理解这个发育过程、理解电在生物学中的真实运作方式，否则我们无法逆向工程人类、心智或任何这些子系统。

What defines the brain and so on? That really is written in a language that we're only just beginning to understand and frankly biologists, most biologists are still very reluctant to even get themselves tangled up in questions like electrical fields influencing development. It seems like mumbo jumbo to a lot of biologists and it should not be because this is the twenty first century biology. This is where it's going. But we're not gonna reverse engineer a human being or the mind or any of these subsystems until we understand how this developmental process, how electricity in biology really works.

如果这与感受、意识等相关联...我的意思是，在现阶段我们仍需尝试。但我认为答案就在那里。

And if it is linked with feelings and with consciousness and so on, that's the I mean, in the meantime, we have to try. But but I think that's where the answer lies.

所以你认为意识等认知难题的答案可能存在于早期发育阶段，即电与生物学的相互作用中？电场？

So you think it's possible that the key to things like consciousness are some of the more tricky aspects of cognition might lie in that early development, the interaction of electricity and biology? Electrical fields

但我们已通过脑电图等了解很多大脑功能，却不知道神经网络中哪些细胞、哪些部分产生了脑电信号。我们连基础都不清楚。假设是——我们知道涉及神经网络，知道需要数百上千个细胞参与，推测与动作电位期间的去极化有关。但神经元内的线粒体拥有比质膜更多的膜结构，具有更强的膜电位，其平行排列的嵴能够增强电场并产生长程场。

But we already know the EEG and so on. It's telling us a lot about brain function, but we don't know which cells, which parts of a neural network is giving rise to the EEG. We don't know the basics. The assumption is I mean we know it's neural networks, we know it's multiple cells, hundreds or thousands of cells involved in it, we assume that it's to do with depolarization during an action potentials and so on. But the mitochondria which are in there have much more membranes than the plasma membrane of the neuron and there's a much greater membrane potential and it's formed in parallel, very often parallel cristae which are capable of reinforcing a field and generating fields over longer distances.

没人知道这是否在意识中起作用。有理由认为可能相关，但坦白说我们根本不知道，也没有纳入考量。观察果蝇等简单生物脑中线粒体膜结构时，会发现令人惊叹的排列——大量矩形结构呈现出奇妙图案。它们究竟在做什么？

Nobody knows if that plays a role in consciousness or not. There's reasons to argue that it could but frankly we simply do not know and it's not taken into consideration. You look at the structure of the mitochondrial membranes in the brains of, you know, simple things like Drosophila, the fruit fly, and they have amazing structures. You can see lots of little rectangular things that all lined up in in in amazing patterns. What are they doing?

他们为什么会那样？我们一点头绪都没有。

Why are they like that? We haven't the first clue.

你对类器官和脑类器官在实验室里研究这些器官在培养皿中的发育有什么看法？你觉得这有前景吗？是否需要观察整个系统？

What do you think about organoids and brain organoids and like so in a lab trying to study the development of these in the the petri dish development of organs. Do you think that's promising? Do you have to look at whole systems?

我从未做过类似研究。对此了解不多。但与我交流过的研究者表示，培养皿中培育的脑组织片段能产生惊人现象，甚至可能体验某种情感或保留原脑记忆。不过我认为，在掌握控制发育电场的方法之前，我们无法真正将类器官转化为功能性系统。

I've never done anything like that. I don't know much about it. The people who I've talked to who do work on it say amazing things can happen and that a bit of a brain grown in a a in a dish is capable of experiencing some kind of feelings or even memories of its former brain. Again, I I have a feeling that until we understand how to control electrical fields that that control development, we're not going to understand how to turn an organoid into a a real functional system.

但我们该如何获得这种理解？这实在太困难了。有个很有前景的方向想听听你的看法——你熟悉DeepMind和AlphaFold在蛋白质折叠等领域的工作吗？你认为这能否为生物学带来突破，帮助我们模拟从简单到复杂生物系统的行为演变？

But how do we get that understanding? It's so it's so incredibly difficult. I mean, you would have to I mean, one promising direction, I'd love to get your opinion on this. I don't know if you're familiar with the work of DeepMind and AlphaFold with protein folding and so on. Do you think it's possible that that will give us some breakthroughs in biology trying to basically simulate and model the behavior of trivial biological systems as they become complex biological systems?

我相信会的。蛋白质折叠最让我着迷的是——虽然这不是我的专业领域，我的理解可能有误——传统方法是根据蛋白质序列尝试折叠，但存在多种折叠方式，要从氨基酸链的第一性原理找到正确构象非常困难。而实际上，当蛋白质从核糖体产生时，它是在带电通道的特定环境中形成的，这个环境会实时引导每个氨基酸的空间定位。

I'm sure it will. The interesting thing to me about protein folding is that for a long time, my my understanding is not what I work on, I may have got this wrong. But my understanding is that you you take the sequence the sequence of a protein and you try to fold it. There are multiple ways in which you can fold and to come up with the correct confirmation is not a very easy thing because you're doing it from first principles from a string of letters which specify the string of amino acids. But what actually happens is when a protein is coming out of a ribosome, it's coming out of a charged tunnel and it's in a very specific environment which is going to force this to go there now and then this one to go there and this one to come like that.

因此蛋白质在离开核糖体时就被强制赋予了特定的构象变化路径。等它完全形成时，其形状已经由生成时的即时微环境（每个氨基酸逐一出现时的局部条件）决定。我觉得这个领域之前可能忽略了这点。这很能体现科学研究的典型特征：经常问错问题，然后对某个假设进行极其复杂的分析，却忘了思考生物学本身的运作机制——生物系统正是通过带电环境来强制塑造蛋白质结构的。

So you're forcing a specific conformational set of changes onto it as it comes out of the ribosome. So by the time it's fully emerged, it's already got its shape, and that shape depended on the immediate environment that it was emerging into one letter as one one one amino acid at a time. And I don't think that the field was looking at it that way. And he said, if if that's correct, then that's very characteristic of science, which is to say it asks very often the wrong question and then does really amazingly sophisticated analyses on something, having never thought to actually think, well, what is biology doing? And biology is giving you a charged electrical environment that forces you to be this way.

至于DeepMind是否通过模式识别得出了类似结论？我完全不清楚。理论上应该能从蛋白质形状反推出这个过程，但这需要远超人类思维的运算能力。不过人类智慧至少能提出质疑：等等，这个前提对吗？

Now, did deep mind come up through patterns with some answer that was like that? I've got absolutely no idea. It ought to be possible to deduce that from the shapes of proteins. It would require much greater much greater skill than the human mind has. But the human mind is capable of saying, well, hang on.

让我们看看这个出口通道，试着弄清楚这种蛋白质会呈现什么形状，我们可以解决这个问题。

Let's look at this exit tunnel and try and work out what shape is this protein going to take, and we can figure that out.

关于出口通道的研究确实很有趣。但有时候我们很幸运，就像在科学中一样，简化的视角或静态的视角实际上能为我们解决问题。在这种情况下，很可能字母序列对我们的结构有独特的映射，而无需考虑它是如何解开的，也就是无需考虑通道。蛋白质最酷的地方在于，尽管它们可能呈现各种不同的形状，但给定序列时，它们实际上似乎会形成非常特定且独特的形状。

That's really interesting about the exit tunnel. But, like, sometimes we get lucky, and our like, just like in science, the simplified view or the static view will actually solve the problem for us. So in this case, it's very possible that the sequence of letters has a unique mapping to our structure without considering how it unraveled, so without considering the tunnel. And so and that seems to be the case in this situation, where the the cool thing about proteins, all the different shapes that can possibly take, it actually seems to take very specific unique shapes given the sequence.

这是由出口通道强加给你的。问题实际上比你想象的要简单得多。还有一大类蛋白质会改变构象状态，即伴侣蛋白。它们只在出口通道出现某些问题时才会被使用，这时你可能希望以不同的方式处理。所以伴侣蛋白经常会介入并影响其折叠方式。

That's forced on you by an exit tunnel. The problem is actually much simpler than you thought. Then there's a whole army of proteins that which change the conformational state, chaperone proteins. And they're only used when when there's some presumably issue with how it came out of the exit tunnel and you want to do it differently to that. So very often the chaperone proteins will go there and will influence the way in which it falls.

所以有两种方法。要么你可以研究所有蛋白质的结构和序列，运用强大的思维去发现其中的模式；要么你可以观察实际情况并说，等等，这其实相当简单。它有一个带电环境，然后被迫以这种方式出来。接下来的问题就是，不同的核糖体是否具有不同的带电环境？

So there's two ways of doing it. Either you can you can look at the the structures and the sequences of all the proteins and you can apply an immense mind to it and figure out what the patterns are and figure out what or you can look at the actual situation where it is and say, hang on. It was actually quite simple. It's got a charged environment and then it's forced to come out this way. And then the question will be, well, do different ribosomes have different charged environments?

如果伴侣蛋白介入会发生什么？你实际上是在提出另一组问题来得出相同的答案，这种方式讲述了一个更简单的故事，并解释了为什么是这样，而不是说这可能是十亿种可能构象状态中的一种。你是在说，它呈现这种状态是因为在给定环境下这是唯一可能的状态。

What happens if a chaperone you know, you're asking a different set of questions to come to the same answer in a way which is telling you a much simpler story and explains why it is, rather than saying it could be this is one in a billion different possible conformational states that this protein could have. You're saying, it has this one because that was the only one it could take given its setting.

嗯，是的。目前人类非常擅长这种第一性原理思维。退一步说，我认为AI非常擅长收集大量数据，比如对行星的大量观测数据，并发现地球并非宇宙中心，实际上太阳才是。

Well, yeah. I mean, this currently, humans are very good at that kind of first principles thinking. Oh, yeah. Was stepping back. But I think AI is really good at, you know, collecting huge amount of data and a huge amount of data of observation of planets and figure out that Earth is not at the center of the universe, that there's actually a sun.

我们围绕太阳运行。但作为人类，你可以问，太阳系是如何形成的？需要哪些不同的力才能使这种模式出现？然后你开始发明像重力这样的概念。我的意思是，显然我是不是把顺序搞混了

We're orbiting the sun. But then you can, as a human being, ask, well, how did how do solar systems come to be? How do the what are the different forces that are required to make this kind of pattern emerge? And then you start to invent things like gravity. What I mean, obviously Is it something I I mixed up the ordering of

是的。

Yeah.

引力曾不被视为连接行星的因素，但作为人类，我们能够思考这些宏观问题。人工智能非常擅长从海量数据中推断简单模型。而在生物学领域，我们解决生物学问题的方式总是反复摇摆。听着，蛋白质折叠曾被认为是不可能解决的难题，许多才华横溢的博士生曾一次只研究一个蛋白质试图破解其结构。而事实上它确实做到了

Of gravity wasn't considered as a thing that connects planets, but we are able to think about those big picture things as human beings. AI is just very good to infer simple models from a huge amount of data. And and the question is with biology, you know, we kinda go back and forth how we solve biology. Listen, protein folding was thought to be impossible to solve, and there's a lot of brilliant PhD students that worked one protein at a time trying to figure out the structure. And the fact that it was able to do that

噢，我完全没有贬低的意思，但我觉得人们一直在问错误的问题。

Oh, I I'm not I'm not knocking it at all, but but but I I think that people have been asking the wrong question.

但当人们开始提出更好更大的问题时，人工智能就会介入对话并表示：我来帮你解决这个问题。

But then, as the people start to ask better and bigger questions, the AI kinda enters the chat and says, I'll help you out with that.

我能举一个自己研究领域的例子吗？随着年龄增长患病的风险存在遗传因素。如果你一生都暴饮暴食、吸烟等等，那完全是另一个问题。但风险确实存在遗传面，而且我们已经知道某些基因会增加特定疾病的风险。近二十年来，人们一直在进行所谓的GWAS（全基因组关联研究）。

Can I give you another example of my own work? Risk of getting a disease as we get older, there are genetic aspects to it. If you spend your whole life overeating and smoking and whatever, that's a whole separate question. But there's a genetic side to the risk and and and we know a few genes that increase your risk of certain things. For for probably twenty years now, people have been doing what's called GWAS, which is genome wide association studies.

这种方法会扫描整个基因组寻找单核苷酸多态性，也就是基因组某个位点的单个字母变化是否与特定疾病存在更高关联性。最终可能会发现基因组中数千个这样的位点。如果把所有这些因素累加起来，是否能解释该疾病已知的遗传风险？已知遗传风险通常来自双胞胎研究——比如如果一个双胞胎患癫痫，其同卵双胞胎也有40%到50%的患病几率，因此遗传因素约占50%。理论上你发现的基因相似性应该能解释这50%的已知风险。

So you you effectively scan the entire genome for any single nucleotide polymorphisms, is say a single letter change in one place that has a higher association of being linked with a particular disease or not. And you can come out with thousands of these things across the genome. And if you add them all up and try and say, so so do they add up to to explain the the known genetic risk of this disease? And the known genetic risk often comes from twin studies and you can say that, you know, that if if this twin gets epilepsy, there's a forty or fifty percent risk that the other twin, identical twin will also get epilepsy. Therefore the genetic factor is about 50% and so the gene similarities that you see should account for 50% of that known risk.

但实际情况往往是这些因素连已知风险的十分之一都解释不了。对此有两种可能的解释：第一种是人们常说的——我们统计效力不足。也许存在百万个相关位点，我们只发现了其中一千个。如果找到剩下那些虽然微弱但数量庞大的关联位点，或许就能解释全部风险。甚至可能有十亿个这样的位点。这是第一种解释方式。

Very often it accounts for less than a tenth of the known risk. And there's two possible explanations and there's one which people tend to do which is to say, well, we don't have enough statistical power. Maybe there's a million, we've only found thousand of them, but if we find the other million, they're weakly related but there's a huge number of them and so we'll account for that whole risk. Maybe there's billion of them. So that's one way.

另一种说法是，等一下，你忽略了一个系统。这个系统就是线粒体DNA，人们往往忽视它，因为它很小且变化不大。但线粒体DNA中几个单字母的变化，却能控制一些非常基础的生理过程。它不仅控制着我们生存、活动和一切行为所需的全部能量，还控制着生物合成——制造新细胞所需的构建模块。而癌细胞经常劫持线粒体并重新编程，使它们不再用于产生能量，而是作为生物合成构建模块的前体物质。

The other way is to say, Well hang on a minute, you're missing a system here. That system is the mitochondrial DNA which people tend to dismiss because it's small and it's not it doesn't change very much. But a few single letter changes in that mitochondrial DNA, it controls some really basic processes. It controls not only all the energy that we need to live and to move around and do everything we do, but also biosynthesis to make the new building blocks to make new cells. And cancer cells very often take over the mitochondria and rewire them so that instead of using them for making energy, they're effectively using them as precursors for the building blocks for biosynthesis.

你需要制造新的氨基酸、新的DNA核苷酸，还需要制造新的脂质来构建细胞膜等等。因此它们某种程度上重构了新陈代谢。现在的问题是，我们完全忽视了线粒体DNA与细胞核基因之间的所有这些相互作用——因为人们直接丢弃了线粒体基因。我们在果蝇实验中可以看到，这些相互作用会导致风险概率的巨大差异。所以你可以让人工智能来研究这个问题：究竟存在多少种这样的碱基变化。这可能存在上百万种变化并构成主要影响因素，也可能完全相反。

You need to make new amino acids, new nucleotides for DNA, you want to make new lipids to make your membranes and so on. So they kind of rewire metabolism. Now the problem is that we've got all these interactions between mitochondrial DNA and the genes in the nucleus that are overlooked completely because people throw away, literally throw away the mitochondrial genes and and we can see in in fruit flies that they interact and produce big differences in risk. So you can set can set AI onto this question of of exactly what you know, how many of these base changes there are. And that's just one possible solution that maybe there are a million of them and it does account for the greatest part of the The other one is they aren't.

实际上风险恰恰存在于你未曾关注的地方。这就是人类直觉至关重要的原因。有种感觉是：我正在研究这个，我认为它很重要，我对此非常执着——最终证明有些人是对的。事实证明它确实重要。

It's just not there. Actually, risk lies in something you weren't even looking at. This is where human intuition is very important. There's this feeling that, well, I'm working on this and think it's important and I'm bloody minded about it, and in the end, some people are right. It turns out that it was important.

你能让人工智能也具备这种执着吗？

Can you get AI to do that, to be bloody minded?

等一下——你可能完全忽略了另一个更庞大的系统。这正是科学革命性发现的瞬间。我不再断言人工智能不能做什么了，过去我对太多事情下过定论。我认为已经取得了很大进展。

And that that that hang on a minute, You might be missing a whole other system here that's much bigger. That's that's that's the moment of discovery of scientific revolution. I'm giving up on saying AI can't do something. I've I've said enough times about enough things. I think there's been a lot of progress.

相反，我对人工智能辅助人类的可能性感到兴奋。但就像我说的，我们似乎总是低估人类的力量。

And instead, I'm excited by the possibility of AI helping humans. But at the same time, just like I said, we seem to dismiss the power of humans.

是的。

Yes.

是的。我们在很多方面都受到限制，这些限制有时让我们觉得自己很笨拙，比如我们不够强壮。我们的注意力、记忆力都是有限的。我们专注事物的能力，以及我们对这些局限的认知都是有限的。但实际上，背后有一个令人难以置信的计算机系统在支撑着整个体系的运转，使我们能够与环境互动、对环境进行推理。这其中蕴含着魔力。

Yes. Like, we're so limited in so many ways that we kinda in in what we feel like dumb ways, like, we're not strong. We're we're kinda our attention our memory is limited. Our ability to focus on things is limited in our own perception of what limited is, but that actually there's an incredible computer behind the whole thing that makes this whole system work, our ability to to interact with the environment, to reason about the environment. There's magic there.

而我满怀希望地认为，人工智能或许能捕捉到一些同样的魔力，但这种魔力不会像'深蓝'下国际象棋那样。不会的。它会更加有趣。

And I I'm hopeful that AI can capture some of that same magic, but that magic is not gonna look like Deep Blue playing chess. No. It's going to be more interesting.

但我认为它也不会表现为模式识别。我的意思是，这本质上就是你告诉我的。目前它在寻找正确模式方面表现得非常出色，而我的观点是，在你寻找特定模式时它确实很有效。但我们是讲故事的动物，而假设本身就是一个故事。这是一个可验证的故事，但一个新假设本质上就是向未知领域的跃迁，是一个全新的叙事。

But I don't think it's gonna look like pattern finding either. I mean, that's essentially what you're telling me. It does very well at the moment and my point is it works very well where where you're looking for the right pattern. But we are storytelling animals and the hypothesis is a story. It's a testable story but but you know, a new hypothesis is a leap into the unknown and it's a new story basically.

它讲述着这个导致那个的因果关系。这是一套因果叙事体系。

And it says this leads to this leads to that. It's a causal set of of storytelling.

也有可能，向未知领域的跃迁本身就遵循着某种模式。

It's also possible that the leap into the unknown has a pattern of its own.

是的，确实如此。

Yes, it is.

有可能这种跃迁是可学习的。

It's possible that it's learn learnable.

我确信如此。亚瑟·凯斯勒有本关于创造力的好书，他将创造力比作一个笑话，其笑点朝着完全意想不到的方向发展，并认为这正是人类创造力的基础——那种将方向转向未知之处的创造性转折，虽然我不确定其运作机制，但这个观点确实很有启发性，其中必定蕴含某些真理。我们讲述的大多数故事可能都是错误的，毫无意义且无助于事。人类确实不擅长识别事物中的模式，但最令人愉悦的人性特质之一，就是发现那些通往未知之地的新故事。

I'm sure it is. There's a nice book by Arthur Kessler on on on the nature of creativity, and and he likens it to a joke where the punchline goes off in a completely unexpected direction and says that this is the basis of human creativity, that would you know, some creative switch of direction to an unexpected place is similar to to a I'm not saying that's how it works, but this it's a nice idea and there must be some truth in it. And it's one of these most of the stories we tell are probably the wrong story and probably going nowhere and probably not helpful. And we definitely don't do as well at seeing patterns in things. But some of the most enjoyable human aspect is is finding a new story that goes to an unexpected place.

这些正是人性对我而言的全部意义。或许AI能自行领悟这些特质，或许它们只是表象。但我时常觉得，那些试图理解人类本质的人——当我们希望构建一个类人AI系统时——其实对人类自身的构造原理仍缺乏足够深刻的把握。

These are all aspects of what being human means to me. And maybe these are all things that AI figures out for itself or maybe they're just aspects to But I just have the feeling sometimes that the people who are trying to understand what to what we are like, what we what we if we wish to craft an AI system which is somehow human like, that we don't have a firm enough grasp of what humans really are like in terms of how we are built.

但我们正逐渐加深这种理解。我完全同意你的观点。我们尝试构建某物时总会突然停下——这里还有另一套系统。事实上，构建类人AI的过程正引领我们更深入地理解人类本质。

But we get a better better understanding of that. I agree with you completely. We try to build a thing and then we go, hang on in a minute. Yeah. There's another system here, and that's actually the attempt to build AI that's human like is getting us to a deeper understanding of human beings.

有趣的是，我最近与史上最伟大的棋手马格努斯·卡尔森交谈。他谈到DeepMind的AlphaZero时有个精妙评论：这个拥有冷幽默感的棋王首次目睹AlphaZero对弈时深受震撼，称其许多操作极易被误认为创造性行为。就像典型人类反应那样，他起初拒绝给予系统应得的认可，因为AlphaZero展现的不仅是强大算力，更是令人惊叹的棋艺 brilliance——比如它以极其有趣的方式弃子。

The funny thing is I recently talked to Magnus Carlsen, widely considered to be the greatest chess player of all time. And he talked about AlphaZero, which is a system from DeepMind that plays chess, and he had a funny comment. He has a kinda dry sense of humor, but he was extremely impressed when he first saw AlphaZero play, and he said that it did a lot of things that could easily be mistaken for creativity. So he'd like a typical human, refuse to give the system sort of its due because he came up with a lot of things that a lot of people are extremely impressed by, not just the sheer calculation, but the the brilliance of play. So one of the things that it does in really interesting ways is it sacrifices pieces.

在国际象棋中，这本质上意味着以退为进。为未来收益主动牺牲棋子，对人类而言正是棋艺的精髓所在。那些笼罩在不确定性迷雾中的风险决策尤其令人煎熬。

So in chess, that means you you you basically take take a few steps back, you know, take a step forward. You give away pieces for some future reward. And that, for us humans, is where art is in chess. Mhmm. You take big risks that for us humans, those risks are especially painful because you have a fog of uncertainty before you.

此刻基于直觉冒险——明知风险正确却面临无数可能性——这需要勇气，这就是艺术，这就是危险所在。而AlphaZero不仅做出同类冒险，甚至程度更甚。当然你可以将其简化为冰冷的模式计算，

So to take a risk now based on intuition of I think this is the right risk to take, but there's so many possibilities that that's where it takes guts. That's where art is. That's that danger. And then the alpha alpha zero takes those same kind of risks and does them even greater degree. But, of course, it does it from a well, you could easily reduce down to a cold calculation over patterns.

但当最终结果呈现时，它确实如同我们在创造性棋局中目睹的魔法。不过棋盘终究有限，关键在于：随着技术进步，我们能否在数学、编程领域复制这种创造性？进而最终拓展至生物学、心理学等更复杂的系统？

But, boy, when you see the final result, it sure looks like the same kind of magic that we see in creativity, when we see creative play on the chessboard. But the chessboard is very limited. And the question is, as we get better and better, can we do that same kind of creativity in mathematics, in programming Mhmm. And then eventually in biology, psychology, and expand into more and more complex systems.

我小时候习惯跑步，经常在跑步时摔倒，尤其是在山上跑上跑下。我从来都不太擅长这个，但有些人速度惊人，特别是下坡时。我尝试后发现，实际上只有两种方法可行——虽然理论上可能有三种。要么你走得极慢极小心，一步步盘算：这里有块石头，我要踩上去；那边有个泥坑，我得避开。

I was used to go running when I was a boy and fell running, which is to say running up and down mountains. And I was never particularly great at it, but there were some people who were amazingly fast, especially at running down. I realized in trying to do this that there's only really two way there's three possible ways of doing it and there's only two that work. Either you go extremely slowly and carefully and you figure out, okay, there's a stone. I'll put my foot on this stone and and then there's another there's a muddy puddle I'm gonna avoid.

你知道的，这种方式缓慢费力，需要步步为营。或者你可以快得不可思议，完全不去思考。整个意识都被屏蔽在外，就像打乒乓球时那样。大脑在潜意识里完成了大量精确计算，这很神奇。你能以惊人的速度冲下山坡，却完全不知道自己是怎么做到的。

You know, it's slow, it's laborious, you figure it out step by step. Or you can just go incredibly fast and you don't think about it at all. The entire conscious mind is shut out of it and it's probably the same playing table tennis or something. There's something in the mind which is doing a whole lot of subconscious calculations about exact and it's amazing. You can run at astonishing speed down a hillside with no idea how you did it at all.

然后你会突然恐慌，心想再这样下去会摔断腿。你开始考虑落脚点，稍微放慢速度试图让意识介入——结果就摔倒了。下坡奔跑时根本无法保持清醒思考。大脑能完成的运算量实在令人惊叹。

And then you panic and you think, I'm gonna break my leg if I keep doing this. I've got to think about where I'm gonna put my foot. So you slow down a bit and try to bring the conscious mind in and then do, you crash. You cannot think consciously while running downhill. It's amazing how many calculations the mind is able to make.

下棋这类活动的问题在于：我们无法像冲下山坡那样，在潜意识里完成所有前瞻性计算。这部分源于进化适应——快速下坡奔跑是我们必须精通的生存技能，否则就会被吃掉。而预判未来多步棋路从不是我们的进化需求，对大多数人来说能预见两三步就足够了。

Now the problem with playing chess or something, if you're able to make all of those subconscious forward calculations about what is the likely outcome of this move now in the way that we can by running down a hill side or something, it's partly about what we have adapted to do. It's partly about the reality of the world that we're in. Running fast downhill is something that we better be bloody good at otherwise we're gonna be eaten. Whereas trying to calculate multiple, multiple moves into the future is not something we've ever been called on to do. Two or three, four moves into the future is quite enough for most of us most of the time.

确实。所以即便我们解决了国际象棋问题，可能离解决下坡奔跑的难题还有距离。不过能见证这种创造力还是很美妙的。

Yeah. Yeah. So the yeah. Just solving chess may not we may not be as far towards solving the problem of downhill running as we might think just because we solve chess. Still, it's beautiful to see creativity.

人类创造了机器。它们能在棋盘上创造艺术，也能在其他领域创造艺术。谁知道这会把我们带向何方？我早些时候提到过安德烈·卡帕西，他和我都是您的忠实粉丝。

Humans create machines. They're able to create art and art on a chessboard and art otherwise. Who knows how far that takes us? So I I mentioned Andrei Karpathy earlier. Him and I are big fans of yours.

如果要投票的话，他的建议是您下一本书该写费米悖论。既然我们谈到生命这个话题——我们本身也算是一种外星生命——您认为宇宙中存在多少外星文明呢？

If you're taking votes, his suggestion was you should write your next book on the Fermi Paradox. So let me ask you on the topic of alien life since we've been talking about life, and we're a kind of aliens. How many alien civilizations are out there do you think?

宇宙非常浩瀚，因此存在一些生命，但数量远不及大多数人想象中那么多——这是我的观点。因为从细菌这样的细胞生命到人类，这一演进轨迹在我看来存在多处重大断层。真核细胞（我们拥有的复杂细胞）是其中最大的断层，光合作用则是另一个。另一个有趣的断层是从真核细胞起源到首批动物出现之间的漫长间隔，这大约耗时十亿年，甚至更久。氧气在大气中开始积累的过程也经历了漫长延迟——从大氧化事件中氧气首次出现，到足以支撑动物呼吸的浓度，耗时近二十亿年。

Well, the universe is very big, so some, but not as many as most people would like to think is my view because the idea that that there is a trajectory going from cellular life like bacteria all the way through to humans, it seems to me there's some big gap along that way. The eukaryotic cell, complex cell that we have is the biggest of them but also photosynthesis is another. Another interesting gap is a long gap from the origin of the eukaryotic cells to the first animals, that was about a billion years, maybe more than that. A long delay in when oxygen began to accumulate in the atmosphere. So from the first appearance of oxygen in the great oxidation event to enough for animals to respire, it was close to two billion years.

为何如此漫长？似乎与行星因素、地质条件密切相关，而我们并不真正了解其中缘由。那种认为生命必然朝着复杂化和意识化方向发展的观点，我认为并不正确。并非说这种情况不会发生，但我认为它不会频繁出现。

Why so long? It seems to be planetary factors, it seems to be geology as much as anything else, and we don't really know what was going on. The idea that there's an inevitable march towards complexity and sentient life, I don't think is right. Not to say it's not going to happen, but I think it's not gonna happen often.

那么以地球为例，考虑到地质限制等因素，你认为生命——复杂生命、智慧生命——在地球上是极速出现的还是历经漫长时间形成的？只是想了解你的观点更倾向于'形成人类所需的条件极难达成'，还是说'即便条件具备，从统计学角度看依然困难'？

So if you think of Earth, given the geological constraints and all that kind of stuff, do you have a sense that life, complex life, intelligent life happen really quickly on Earth or really long? So just just to get a sense of are you more sort of saying that it's very unlikely to get the kind of conditions required to create humans, Or is it even if you have the condition, it's just statistically difficult?

我认为...核心问题在于真核细胞的起源。这件事只发生过一次，而若没有真核细胞，后续一切都不可能发生。这个事件...

I think the the I mean, the problem the the single great problem at the center of all of that to my mind is the origin of the eukaryotes, which happened once and without eukaryotes, nothing else would have happened. And and that is something that

你强调真核细胞至关重要，但...

That's because you're saying it's super important, the eukaryotes, but

我的意思等同于：仅靠细菌细胞是不可能构建出人类这般复杂的生物的。

I'm saying a tantamount to saying that it is impossible to build something as complex as a human being from bacterial cells.

从某种深层本质而言完全同意。但这就像是一个细胞进入另一个细胞内部，要让这种机制运作起来其实并不那么困难对吧？

Totally agree in some deep fundamental way. But it's just like one cell going inside another. It's it's not so difficult to get to work right that like

好吧，这种情况又发生过一次。如果你认为我是这个观点中的少数派，大多数生物学家可能无论如何都不会同意我的看法。但如果你考虑起点，我们有一个简单的细胞。这是一个古菌细胞，我们可以相当确定这一点。所以它看起来很像细菌，但实际上属于生命的另一个领域。

Well, again, it happened once. And if you think about if you if you think I'm a minority view in this position, most biologists probably wouldn't agree with me anyway. But if you think about the starting point, we've got a simple cell. It's an archaeal cell, we can be fairly sure about that. So it looks a lot like a bacterium, but is in fact from this other domain of life.

所以它看起来很像细菌细胞。这意味着它没有任何东西。它没有细胞核，也没有真正复杂的内部膜系统。它有一点结构，但不多。然后它接纳了一个内共生体。

So it looks a lot like a bacterial cell. That means it doesn't have anything. It doesn't have a nucleus, it doesn't really have complex endomembrane. It has a little bit of stuff, but not that much. And it takes up an endosymbiote.

那么接下来会发生什么？答案基本上与复杂性有关。在我看来，这里有一个美丽的悖论。植物、动物和真菌都拥有完全相同的细胞类型，但它们的生活方式却截然不同。植物细胞是光合作用的。

So what happens next? And the answer is basically everything to do with complexity. To me, there's beautiful paradox here. Plants and animals and fungi all have exactly the same type of cell, but they all have really different ways of living. So plant cell is photosynthetic.

它们最初是海洋中的藻类等等，想象一下藻华，单细胞生物。构成它们的基本细胞结构几乎完全相同，只有几个小差异。植物细胞还有叶绿体，有液泡，有细胞壁，但也就这些了。植物细胞和动物细胞的其他部分几乎完全相同。

They started out as algae in the oceans and so on, think of algal blooms, single cell things. The basic cell structure that it's built from is exactly the same with a couple of small differences. It's got chloroplasts as well. It's got a vacuole, it's got a cell wall, but that's about it. Pretty much everything else is exactly the same in a plant cell and an animal cell.

然而它们的生活方式却完全不同。这种细胞结构并非因应不同生活方式或环境而进化。我在海洋中进行光合作用，我在陆地上作为动物四处奔跑，我是土壤中的真菌，向各处延伸出长长的菌丝。它们都具有相同的基本细胞结构。为什么？

Yet the ways of life are completely different. This cell structure did not evolve in response to different ways of life, different environments. I'm in the ocean doing photosynthesis, I'm on land running around as part of an animal, I'm a fungus in the soil spending out long kind of shoots into whatever it may be, mycelium. They all have the same underlying cell structure. Why?

几乎可以肯定，这是由对内部环境的适应驱动的——这些恼人的内共生体迫使宿主细胞发生各种变化。从某种角度看，这可能是件好事，因为这一过程可能具有某种必然性。一旦获得内共生体，你就或多或少注定要朝那个方向发展。或者也可能是巨大的巧合，几乎在每种可能情况下都开始出错，冲突最终会导致战争、死亡和灭绝，根本无法成功。也许这种情况发生过数百万次，每次都失败了；或者可能只发生过一次就成功了，因为这是必然的。实际上，我们现在掌握的知识还不足以判断这两种可能性哪种属实，但两者都有点令人沮丧。

Almost certainly it was driven by adaptation to the internal environment to having these pesky endosymbionts that forced all kinds of change on the host cell. Now in one way you could see that as a really good thing because it may be that there's some inevitability to this process. As soon as you got end of symbionts, you're more or less bound to go in that direction, or it could be that there's a huge fluke about it and it's almost starting to go wrong in just about every case possible, that the conflict will lead to effectively war leading to death and extinction, and it simply doesn't work out. Maybe it happened millions of times and it went wrong every time, or maybe it only happened once and it worked out because it was inevitable. Actually, simply do not know enough now to say which of those two possibilities is true, but both of them are a bit grim.

但你倾向于认为我们只是在那一次飞跃中非常幸运。比如，你是否觉得我们的银河系中可能有数百万个行星上存在细菌？

But you're you're leaning towards we just got really lucky in that one leap. Like, we got so do you have a sense that our galaxy, for example, has just maybe millions of planets with bacteria living on it?

实际上，我预计会有数十亿、甚至上百亿颗行星上存在细菌。我是说，平均每颗恒星周围可能有5到10颗行星，其中至少一颗我希望会有细菌。因此我认为细菌应该非常普遍。我实在无法给出其他数字，但我相信这种现象在宇宙其他地方也存在。

I would expect billions, tens of billions of planets with bacteria living on it, practically. Would I would I mean, there's probably, what, five to 10 planets per star, of which I would hope that at least one would have bacteria on. So I expect bacteria to be very common. I simply can't put a number otherwise. I mean I expect it will happen elsewhere.

我并不是认为我们生活在一个完全空旷的宇宙中。

It's not that I think we're living in a completely empty universe.

这太令人着迷了。

That's so fascinating.

但我认为这并非必然发生，而且问题不仅限于地球上的复杂生命。我之前提到过氧气、动物等因素。即便是人类，我们也出现得非常晚。回溯五百万年前，如果我们发现一个充满长颈鹿的星球，我们会感到震撼吗？你可能会想：嘿，这里有生命。

But I think that it's not going to happen inevitably and there's something That's not the only problem with complex life on earth. I mentioned oxygen and animals and so on as well. And even humans, we came along very late. You go back five million years and, you know, would we be that impressed if we came across a planet full of giraffes? I mean, you'd think, hey, there's life here.

这是个适合殖民的好星球之类的。但我们不会想着要跟这些长颈鹿展开对话。

There's a nice planet to colonize or something. We we wouldn't think, oh, let's try and have a conversation with this giraffe.

是啊。我不确定外星视角会觉得人类特别有趣在哪里，或者他们如何注意到我们。我也会和你聊聊城市，因为这是观察人类文明的有趣视角。当然，你的感觉——虽然你无从知晓——这是个有趣的星球，有趣的星系。

Yeah. I'm not sure what exactly we would I'm not exactly sure what makes humans so interesting from an alien perspective or how they would notice. I'll talk to you about cities too because that's an interesting perspective of how to look at human civilization. But your sense I mean, of course, you don't know, but it's an interesting world. It's an interesting galaxy.

我们生活的这个宇宙很有趣，就像每个太阳系那样。大约90%的太阳系都有细菌存在。想象一下这样的世界，而整个星系可能只有寥寥几个——甚至只有一个智慧文明。这真是个奇妙的世界。

It's an interesting universe to live in that's just like every sun. Like, 90% of solar systems have bacteria in it. Like, imagine that world, and the galaxy maybe has just a handful, if not one, intelligent civilization. That's a wild world.

这是个狂野的

It's a wild

世界。我甚至都没想过那个世界。有一种观点认为，比如，在火星或泰坦上发现生命之所以令人兴奋的原因之一是，如果生命存在于其他地方，那么从统计学上讲，无论多细胞生物、性、暴力等出现的概率多么低，生命总会以某种形式存在。我是说，光合作用，发展出一套结合化学与环境的机制来构建复杂生物体，这肯定会发生。但是，天啊，如果到处都是细菌，我不知道自己会作何感想。

world. I didn't even even think about that world. There's a kinda thought that, like, one of the reasons it would be so exciting to find life on Mars or Titan or whatever is, like, if its life is elsewhere, then surely, statistically, that life, no matter how unlikely you carry as multicell organisms, sex, violence, what what else is extremely difficult? I mean, photosynthesis, figuring out some machinery that involves the chemistry and the environment to allow the building up of complex organisms, surely that would arise. But man, I don't know how I would feel about just bacteria everywhere.

嗯，如果这是真的，那确实令人沮丧。我想是挺沮丧的。我不认为

Well, it would be depressing if it was true. I suppose depressing. I don't think that

我不知道哪个更令人沮丧。到处都是细菌还是到处都空无一物？

I don't know what's more depressing. Bacteria everywhere or nothing everywhere?

是的。两者都让人不寒而栗。但无论是否令人寒心，我认为这都不应迫使我们改变对其真实性的看法。我所说的可能对也可能不对。

Yes. Either of them are chilling. Yeah. But whether it's chilling or not, I don't think should force us to change our view about whether it's real or not. Yes, What I'm saying may or may not be true.

那么如果我们发现火星上有生命，你会作何感想？听起来你会比其他人更不兴奋，因为你会觉得，呃

So how would you feel if we discovered life on Mars? Seen it. Sounds like you would be less excited than some others because you're like, well

我最感兴趣的是它与地球生命的相似程度。这实际上会变成一个相当微妙的问题，因为生命在火星和地球之间来回传播的可能性——虽然我不会说很高，但也不低。这是相当可行的。所以如果我们在火星上发现生命，其遗传密码非常相似但略有不同，大多数人会立即将其解读为生命曾单向或双向迁移的证据，表明火星或地球上的生命有共同起源，并进行了单向或双向传播。不过另一种看待这个问题的方式是：生命的起源取决于确定性化学和热力学，始于最可能丰富的材料——二氧化碳、水以及一个湿润的岩石行星，而火星最初正是这样的环境。

What I would be most interested in is how similar to life on earth it would be. It would actually turn into quite a subtle problem because the the likelihood of life having gone to and fro between between Mars and the earth is is quite I wouldn't say high but it's not low. It's quite feasible. And so if we found life on Mars and it had very similar genetic code but it was slightly different, Most people would interpret that immediately as evidence that there'd been transit one way or the other and that it was a common origin of life on Mars or on the earth and it went one way or the other way. The other way to see that question though would be to say, Well, the beginnings of life lie in deterministic chemistry and thermodynamics, starting with the most likely abundant materials, CO2 and water and a wet rocky planet, and Mars was wet and rocky at the beginning.

我们虽然不能说必然，但几乎可以肯定会发现一种与我们现有遗传密码相差不远的遗传密码。因此我们会看到遗传密码中的细微差异。这意味着什么？这可能非常难以解读。

And we'll, I won't say inevitably, potentially almost inevitably come up with a genetic code which is not very far away from the genetic code that we already have. So we see subtle differences in the genetic code. What does it mean? It could be very difficult to interpret.

你认为有可能辨别出真正起源的事物吗？

Is it possible you think to tell the difference of something that truly originated?

我认为如果立体化学结构不同——比如糖类有L型和D型，而所有生命体中都存在D型糖和L型氨基酸。但脂质方面，细菌拥有一种立体异构体，而其他细菌则拥有相反的立体异构体。完全可以选择使用其中一种，我认为乔治·丘奇一直在尝试用相反的立体异构体构建生命。这完全可行且会成功。如果我们在火星上发现使用相反立体异构体的生命，那将确凿证明生命在那里独立起源。

I think if the stereochemistry was different, we have sugars for example that are the L form or the D form and we have D sugars and L amino acids right across all of life. But lipids, the bacteria have one stereoisomer and the bacteria have the other, the opposite stereoisomer. It's perfectly possible to use one or the other one and the same would always only go for I think George Church has been trying to make life based on the opposite stereoisomer. So it's perfectly possible to do and it will work. And if we were to find life on Mars that was using the opposite stereoisomer, that would be unequivocal evidence that life had started independently there.

所以希望我们能在土卫六或木卫二这类地方发现生命，那里与我们共享生命的可能性更低，环境更严苛，可能会存在更奇特的生命形式。

So hopefully, the life we find will be on Titan and Europa or something like that where it's less likely that we shared and it's harsher conditions so there's gonna be weirder kind of life.

对此我不抱太大期望，因为

I wouldn't count on that because

那水呢？

And water?

生命起源于深海热泉喷口。

Life started in deep sea hydrothermal vents.

这很严酷。

It's harsh.

环境。确实相当严酷。土卫六不同。木卫二可能与地球相当类似，因为我们面对的是一个海洋，那里可能存在酸性海洋，就像早期地球那样，几乎肯定存在热液系统。土卫二也是如此。

Condition. It's pretty harsh, yeah. Titan is different. Europa is probably quite similar to earth in the sense that we're dealing with an ocean, some acidic ocean there as the early earth would have been and it almost certainly has hydrothermal systems. Same with Enceladus.

我们可以从冰层表面喷发的羽流判断这一点。我们知道那里存在液态海洋，并能大致了解其化学组成。至于土卫六，我们面对的是液态甲烷之类的物质。如果真的存在生命，那必定与地球已知的任何生命形式都截然不同。

We can tell that from these plumes coming from the surface through the ice. We know there's a liquid ocean and we can tell roughly what the chemistry is. For Titan, we're dealing with liquid methane and things like So that would really if there really is life there, it would really have to be very very different to anything that we know on earth.

所以最艰难的跨越，最重要的跨越是从原核生物到真核生物的进化。真核生物。如果要排序的话，第二重要的是什么？你刚才特别强调了光合作用。

So the hard leap, the hardest leap, the most important leap is from prokaryotes to to eukaryotes. Eukaryote. What's the second if we're ranking? What's what's the what's you gave a lot of emphasis on photosynthesis.

是的。我认为那应该是第二位。但与其说是光合作用本身的问题，不如说这是件难以实现的事。我们只知道它发生过一次。

Yeah. That would be my second one, think. But it's it's not so much I mean, photosynthesis is part of the problem. It's a difficult thing to do. Again, we know it happened once.

我们不知道为何只发生了一次，但事实是它被植物和藻类等完全吸收为叶绿体，并在与陆地完全不同的环境中蓬勃发展，这似乎表明探索这一领域并无障碍。可能存在一个独立起源探索了细菌从未涉足的领域。这说明光合作用只发生一次的原因很可能是其复杂性，因为生物系统的构建很困难。但它至少发生在22亿年前，大氧化事件前夕，甚至有人认为是30亿年前——当时地质记录中出现了微量氧气的痕迹。这个观点争议很大。

We don't know why it happened once, but the fact that it was taken on board completely by plants and algae and so on as chloroplasts and did very well in completely different environments than on land and whatever else seems to suggest that there's no problem with exploring. You could have a separate origin that explored this whole domain over there that the bacteria had never gone into. That says that the reason that it only happened once is probably because it's difficult, because the wiring is difficult. But then it happened at least 2,200,000,000 ago, right before the GOE, maybe as long as 3,000,000,000 ago when there are some people say there are whiffs of oxygen as this kind of traces in the fossil in the geochemical record that say maybe there's a bit of oxygen then. That's really disputed.

有人认为可以追溯到40亿年前，认为地球生命的共同祖先就具备光合作用能力。于是你立刻会发现，对于起始时间，不同学派之间存在20亿年的分歧。但如果我们采用最确凿的晚期时间点，即20亿年前...直到寒武纪大爆发时期，当时氧气含量明确接近现代水平（约5.5亿年前）。这意味着地球有超过15亿年处于停滞状态。

Some people say it goes all the way back four billion years ago and that it was the common ancestor of life on earth was photosynthetic. So immediately you've got groups of people who disagree over a two billion year period of time about when it started. But what let's take the latest date when it's unequivocal, that's February ago Mhmm. Through to around about the time of the Cambrian explosion when oxygen levels definitely got close to modern levels, which was around about five fifty million years ago. So we've gone more than one and a half billion years where the earth was in stasis.

几乎没什么变化。事实上，这段时期被称为‘无聊的十亿年’。可能那时新事物在某个角落萌芽，但这种认为世界不断变化、我们持续进化、沿着某种斜坡上升的观点，其实是非常人类中心主义的。现实中存在向新稳态转变的临界点——产氧细胞与耗氧细胞达到精确平衡，这就是为何大气含氧量维持在21%并持续了数亿年。我们处于一种极其精妙的平衡中。

Nothing much changed. It's known as the boring billion in fact. Probably stuff was that was when new carrots arose somewhere in there, but so it's this idea that the world is constantly changing, that we're constantly evolving, that we're moving up some ramp is a very human idea. But in reality, there are kind of tipping points to a new stable equilibrium where the cells that are producing oxygen are precisely counterbalanced by the cells that are consuming that oxygen, which is why it's 21% now and has been that way for hundreds of millions of years. We have a very precise balance.

当你跨越临界点后，根本无法预知下一个稳态会在哪里。它可能离现状相去甚远。因此如果我们通过全球变暖改变世界，必将迎来临界点。问题在于何时何地会出现？下一个稳态会是怎样？或许对人类而言将无法居住。

You go through a tipping point and you don't know where the next stable state's gonna be. But it can be a long way from here. And so if we change the world with global warming, there will be a tipping point. The question is where and when and what's the next stable state? It may be uninhabitable to us.

对生命而言肯定仍可栖息。但可能会像二叠纪大灭绝那样，95%的物种消失，经历五到一千万年的间隔后生命复苏——只是不再有人类。

It'll be habitable to life for sure. But there may be something like the Permian extinction where 95% of species go extinct, and there's a five to ten million year gap and then life recovers, but without humans.

从统计学角度看，撇开人类因素，这种变化最终是否会导致更复杂的结构、更有趣的生命形式、更高智能的生物？

And the question statistically, well, without humans, but statistically, does that ultimately lead to greater complexity, more interesting life, more intelligent

在大氧化事件(GOE)后氧气首次出现时，曾有个临界点导致了长期稳态，其状态类似今天的黑海——即表层含氧而深层停滞...确切说是充满硫化物而非无菌状态。这种状态在大陆边缘稳定维持了超过十亿年，但并未显露出明显的进化趋势。

Well, after the first appearance of oxygen with the GOE, there was a tipping point which led to a long term stable state that was equivalent to the Black Sea today, which is to say oxygenated at the very surface and stagnant sterile not sterile, but but but but sulfurous lower down. And and that was stable certainly around the continental margins for more than a billion years. It was not a state that led to progression in an obvious way.

确实。思考进化很有意思——是什么导致了稳态？环境产生的进化压力频率如何？或许其他星球能产生进化压力、化学压力等各种压力，比如‘如果未来一万年不突破就完蛋’这种强力压迫感。

Yeah. I mean, it's interesting to think about evolution, like what leads to stable states and how often are evolutionary pressures emerging from the environment. So maybe other planets are able to create evolutionary pressures, chemical pressures, whatever. Some kind of pressure that say, you're screwed unless you get your shit together in the next, like, ten thousand years. Like, a lot of pressure.

关于地球‘无聊十亿年’的解释似乎有两种：一是迈向下一步异常艰难，二是根本没有迈出下一步的必要。

It seems like Earth like, the boring billing might be explained in two ways. One is super difficult to take any kind of next step, and the second way it could be explained is there's no reason to take the next step.

不，我认为没有原因。但最终却出现了雪球地球现象，那是一场大规模的行星灾难，赤道海域都被冻结，这以某种方式迫使了变化。不久之后，大约一亿年后（时间并不短），我们开始看到动物的出现。

No. I think there is no reason. But at the end of it, there was a was a snowball Earth. There was a planetary catastrophe on a huge scale where the was sea was frozen at the Equator and that forced change in one way or another. It's not long after that, a hundred million years perhaps after that, not short time, but this is when we begin to see animals.

又一次转变发生了，另一个临界点导致了灾难性变化，进而引发了爆发。我们并不真正知道原因，但我在书中讨论的一个原因是硫酸盐被冲刷进海洋，这听起来极其局限。但问题是数据显示，我们可以通过碳同位素大致追踪氧气如何进入大气。这里我们需要考虑两种主要碳同位素：99%的碳是碳12，1%是碳13（稳定同位素），还有微量的放射性碳14。

There was a shift again, another tipping point that led to catastrophic change that led to a take off then. We don't really know why, but one of the reasons why that I discuss in the book is about sulfate being washed into the oceans, which sounds incredibly parochial. But the issue is I mean what the data is showing, we can track roughly how oxygen was going into the atmosphere from carbon isotopes. So there's two main isotopes of carbon that we need to think about here. One is carbon 99% of carbon is carbon 12 and then 1% of carbon is carbon 13 which is a stable isotope and then there's carbon 14 which is a trivial radioactive trivial amount.

碳13占1%。生命和酶通常可以将碳原子想象为弹跳的小球，碳12比碳13移动稍快，因为它更轻，更容易遇到酶，因此更可能被固定到有机物中。所以有机物富含碳12——这只是观察结果——相比碳13，其比例比预期高出几个百分点。如果将有机物作为煤或石油等埋藏，它就不再被氧化，部分氧气便留在大气中，这就是氧气在大气中积累的方式。

So carbon 13 is 1% and life and enzymes generally, you can think of carbon atoms as little balls bouncing around, ping pong balls bouncing around. Carbon 12 moves a little bit faster than carbon 13 because it's lighter and it's more likely to encounter an enzyme. And so it's more likely to be fixed into organic matter. And so organic matter is enriched and this is just an observation, it's enriched in carbon 12 by a few percent compared to carbon 13 relative to what you would expect if it was just equal. And if you then bury organic matter as coal or oil or whatever it may be, then it's no longer oxidized so some oxygen remains left over in the atmosphere And that's how oxygen accumulates in the atmosphere.

通过历史上碳埋藏量可以推算大气中的氧气含量。你可能会问，我们怎么可能知道埋藏了多少碳？答案是：如果埋藏的是碳12，海洋中留下的就是更多碳13，它们沉淀为石灰岩。因此研究这些时期的石灰岩，分析碳13信号，就能反馈氧气含量。在寒武纪大爆发前夕，出现了所谓的负同位素异常偏移，即碳13大幅下降，一千万年后又回升。

And you can work out historically how much oxygen there must have been in the atmosphere by how much carbon was being buried. And you think, well, how can we possibly know how much carbon was being buried? And the answer is, well, if you're burying carbon 12, what you're leaving behind is more carbon 13 in the oceans and that precipitates out of limestone. So you can look at limestones over these ages and work out what's the carbon 13 signal and that gives you a kind of a feedback on what the oxygen content. Right before the Cambrian explosion, there was what's called a negative isotope anomaly excursion, is basically the carbon 13 goes down by a massive amount and then back up again ten million years later.

这似乎表明海洋中的碳12正在消失——也就是说它被氧化了。如果被氧化，就会消耗氧气。强烈的碳13信号表明碳12与碳13的比例急剧下降，意味着更多碳12被移除和氧化。抱歉，这变得太复杂了但——

What that seems to be saying is the amount of carbon 12 in the oceans disappearing, which is to say it was being oxidized. If it's being oxidized, it's consuming oxygen and that should A big carbon 13 signal says the ratio of carbon 12 to carbon 13 is really going down, means there's much more carbon 12 being taken out and being oxidized. Sorry, this is getting too complex but

嗯，这是估算氧气含量的好方法。

Well, it's a good way to estimate the amount of oxygen.

如果根据‘所有被移除的碳12都被氧气氧化’的假设计算氧气量，结果会显示大气中的氧气被完全耗尽。然而其他地质指标却表明大气中仍有氧气。这形成了一种悖论，唯一能解释这种悖论的方法（仅基于空气和海洋中物质的质量平衡）是假设消耗氧气的并非氧气本身，而是硫酸盐——硫酸盐正被冲刷进海洋。

If you calculate the amount of oxygen based on the assumption that all this carbon 12 that's being taken out is being oxidized by oxygen, the answer is all the oxygen in the atmosphere gets stripped out. There is none left. And yet the rest of the geological indicators say, no, there's oxygen in the atmosphere. So it's kind of a paradox and the only way to explain this paradox just on mass balance of how much stuff is in the air, how much stuff is in the oceans and so on, is to assume that oxygen was not the oxygen, it was sulfate. Sulfate was being washed into the oceans.

硫酸盐还原菌将其用作电子受体，就像我们使用氧气作为电子受体一样。它们将电子传递给硫酸盐而非氧气。细菌确实如此。对，对。

It's used as an electron acceptor by sulfate reducing bacteria just as we use oxygen as an electron acceptor. They pass their electrons to sulfate instead of oxygen. Bacteria did. Yeah. Yeah.

这些都是细菌。它们通过硫酸盐氧化碳、有机碳，将电子传递给硫酸盐，硫酸盐与铁反应形成黄铁矿或愚人金，沉入海底，被埋藏于系统之外。这可以解释质量平衡。那么为什么这很重要？因为它表明存在一个偶然事件：构造运动使大量硫酸盐以某种矿物形式（如蒸发岩中的硫酸钙矿物）滞留陆地。由于大陆碰撞和造山运动的发生，硫酸盐被推上山坡并恰好被冲入海洋。

These are these are these are bacteria. So they're oxidizing carbon, organic carbon with sulfate passing the electrons on to sulfate that reacts with iron to form iron pyrites or fool's gold, sinks down to the bottom, gets buried out of the system. And this can account for the mass balance. So why does it matter? It matters because what it says is there was a chance event, tectonically there was a lot of sulfate sitting on land as some kind of mineral, so calcium sulfate minerals for example are evaporitic And because there happened to be some continental collisions, mountain building, sulfate was pushed up the side of a mountain and happened to get washed into the ocean.

我希望这种幸运的意外能够

I want to happy accidents like that are

发生。从统计学角度看确实很难。或许你可以用统计学解释或排除这种可能性，但这就是地球生命的进程。如果没有那些硫酸盐被抬升，寒武纪大爆发几乎肯定不会发生，我们也不会有动物等等。所以，你知道，这

possible. Statistically, it's really hard. Maybe you can rule that in statistically or rule but this is the course of life on earth. Without all that sulfate being raised up, this Cambrian explosion almost certainly would not have happened and we wouldn't have had animals, and and and so on and so on. So it's, you know, it's

关于寒武纪大爆发的这种解释，让我用几种方式来说明。那些质疑进化论有效性的人会给我们举例子——虽然我对此研究不深——他们会把寒武纪大爆发当作‘这事很诡异’的例子。确实很诡异。是的。

This kind of explanation of the Cambrian explosion so let me actually say it in several ways. So, you know, folks who challenge the validity of the theory of evolution will give us an example no. I'm not well studied in this, but will give us an example of the Cambrian explosion as, like, this thing is weird. Oh, it is weird. Yeah.

因此我要问：关于进化论最大的未解之谜或认知空白是什么？是寒武纪大爆发吗？如果是的话，我们目前最好的解释是什么？首先，它究竟是什么？据我理解，在很短的时间内（可能是1000万年或1亿年左右），大量动物种类和多样性突然出现。总之这里面包含大约五个问题。

So by the question I would have is what's the biggest mystery or gap in understanding about evolution? Is it the Cambrian explosion? And if so, how do we what's our best understanding of how to explain first of all, what is it? This in my understanding, in the short amount of time, maybe ten million years, hundred million years, something like that, a huge number of animals, variety, diversity of animals were created. Anyway, that there's like five questions in there.

是的。这是最大的谜团吗？

Yeah. Is that the biggest mystery?

不，我认为这已经不再是什么特别大的谜团了。我是说，虽然关于那个时期仍存在一些未解之谜，比如我刚才提到的硫酸盐流入海洋就是一个。它需要氧气，而当时氧气水平确实上升了，可能在此之前，氧气浓度还不足以支持动物生存。我们在寒武纪大爆发中看到的，正是捕食者与猎物关系的开端。

No. I don't think it's particularly big mystery really anymore. I mean, there are still mysteries about why then, and I've just said sulfate being washed into the oceans is one. It needs oxygen and oxygen levels rose around that time, Probably before that, they weren't high enough for animals. What we're seeing with the Cambrian explosion is the beginning of predators and prey relationships.

我们看到现代生态系统正在形成，看到军备竞赛般的进化，看到进化创造力被完全释放。我提到过‘无聊的十亿年’——整整十亿到十五亿年间几乎什么都没发生。人们错误地假设进化总是极其缓慢，认为需要数十亿年才能产生微小变化，再花十亿年才能有下一步发展。这种观点完全错误。

We're seeing modern ecosystems and we're seeing arms races and we're seeing full creativity of evolution unleashed. I talked about the boring billion. Nothing happens for one and a half, one billion years, one and a half billion years. The assumption, and this is completely wrong, this assumption, is then that evolution works really slowly and that you need billions of years to affect some small change and then another billion years to do something else. It's completely wrong.

进化会陷入停滞状态，持续数千万甚至数亿年。斯蒂芬·杰伊·古尔德曾用‘间断平衡’理论解释这种现象，但他研究的是动物和最近五亿年的历史。若纵观整个地球史，这种规律就不那么明显了。你会发现最初二十亿年只有细菌存在：生命起源后二十亿年全是细菌，直到产氧光合作用出现，接着全球性灾难——雪球地球事件与大氧化事件接踵而至，又是十亿年的沉寂，随后是动荡时期，再次经历雪球地球事件，最后突然迎来寒武纪生命大爆发。

Evolution gets stuck in a stasis and it stays that way for tens of millions, hundreds of millions of years. And Stephen Jay Gould used to argue this, he called it punctuated equilibrium but he was doing it to do with animals and to do with the last five hundred million years or so. Where it's much less obvious than if you think about the entire planetary history. And then you realize that the first two billion years was bacteria only. You have the origin of life, two billion years of just bacteria, oxygenic photosynthesis arising here, then you have a global catastrophe, snowball earths and great oxidation event and then another billion years of nothing happening and then some period of upheavals and then another snowball earth and then suddenly you see the Cambrian explosion.

世界长期处于稳定状态，这种停滞期并不倾向于增加复杂性，一切都保持平衡。只有当出现雪球地球这种全球性灾难时，平衡才会被打破，临界点到来后系统就会跃迁到新状态。认为进化缓慢的观点是错误的，它其实可以快得惊人。比如我之前提过，理论上从感光点到进化出眼睛可能只需要五十万年。

This is long periods of stasis where the world is in a stable state and is not lean is not geared towards increasing complexity. It's just everything is in balance. And only when you have a catastrophic level global level problem like a snowball earth, it forces everything out of balance and there's a tipping point and you end up somewhere else. Now, the idea that evolution is slow is wrong. It can be incredibly fast and I mentioned earlier on that you can in theory, it would take half a million years to invent an eye for example from sensitive spot.

把一根管状结构改造成带凸起的管，再变成带附肢的管，最后分化出多个附肢和头部隆起，这个过程其实不需要太长时间。从智力层面理解这些变化并不困难，但如此快速的进化确实令人震惊。我们习惯用人类时间尺度思考，但要知道海洋里有些生物寿命只有一年，一百万年就意味着百万代——在这段时间里能实现的改变是巨大的。庞大种群中，自然选择对微小变化极其敏感，再加上捕食者与猎物的竞争压力，当环境允许时进化速度惊人就不足为奇了。所以我不认为这是什么大谜团。

It doesn't take long to convert one tube into a tube with nobles on it, into a tube with arms on it and then multiple arms and then one end is a head that starts out as a swelling. It's not difficult intellectually to understand how these things can happen. Boggles the mind that it can happen so quickly but we're used to human time scales and what we need to talk about is generations of things that live for a year in the ocean and then a million years is a million generations and the amount of change that you can do, you can affect in that period of time is enormous. We're dealing with large populations of things where selection is sensitive to pretty small changes and can so again, as soon as you throw in the competition of predators and prey and you're ramping up the the scale of evolution, it's not very surprising that it happens very quickly when the environment allows it to happen. So I don't think there's a big mystery.

虽然还有很多细节需要填补，但生物学真正的重大谜题其实是意识。

There's lots of details that need to be filled in. I mean, the the big mystery in in biology is consciousness.

生物学的大谜题是意识？其实智力也算个谜题。你说的是生物学不是心理学。因为从生物学角度看，智力和意识似乎都是同一种...怎么说呢，就像所有脑部活动一样神秘。

The big mystery in biology is conscious. Well, intelligence is kind of a mystery too. I mean, you said biology, not psychology. Because to from a biology perspective, it seems like intelligence and consciousness all are the same, like, weird like, all the brain stuff.

我认为智能并不一定那么难以理解。我是说，我把它看作一种计算形式，虽然我对计算了解不多，所以...

I don't see intelligence as necessarily that difficult, I suppose. I I mean, I see it as a form of computing, and I don't know much about computing, so I

你对意识也知之甚少，所以我...我是说...哦，我明白了。明白了。明白了。明白了。明白了。

You don't know much about consciousness either, so I I mean, I suppose oh, I see. I see. I see. I see. I see.

作为人类，你对那种意识确实了解很多。

That consciousness you do know a lot about as a human being.

不，不。我是说...我...我觉得...我能理解大脑的布线，基本上和计算机电路原理差不多。对我来说的谜团是，这个系统如何产生我们之前讨论的那种感觉。

No. No. I mean, I I I think I I can understand the wiring of a brain as a series of in pretty much the same way as a computer in in theory in terms of the circuitry of it. The mystery to me is how this system gives rise to feelings as we were talking about earlier on.

是啊。我只是觉得我们把智能想得太简单了。推理之舞、思考的魔力，和感受的魔力一样有趣。嗯。我们总把推理想象成运行一个非常简单的算法。

Yeah. I just I think I think we oversimplify intelligence. I think the dance, the magic of reasoning is as interesting as the magic of feeling. Mhmm. We we tend to think of reasoning as like very running a very simplistic algorithm.

我认为推理是记忆与潜意识中发生的所有事情之间的相互作用

I think reasoning is the interplay between memory, whatever the hell is going on in the unconscious mind, all of that

我完全没有要贬低它的意思。显然，它异常精妙复杂...但我并不觉得其运作原理存在逻辑上的困难。

I'm not trying to diminish it in any way at all. I obviously, it's extraordinarily exquisitely complex and but but I don't see a logical difficulty with how it works.

是的。不，我是说，我同意你的观点，但有时候确实如此。意识周围笼罩着一层巨大的神秘面纱。

Yeah. No. I I mean, I agree with you, but sometimes yeah. There's a big cloak of mystery around consciousness.

我想用经典物理学与量子物理学做个比较。经典物理学是符合逻辑的，你能理解我们处理的那种语言。它几乎是在人类认知层面——我们研究的是恒星等可见事物。而到了量子力学领域，人脑几乎无法计算那里究竟发生了什么。

I mean, let me compare it with with classical versus quantum physics. The you know, classical physics is logical and you can understand the kind of language we're dealing with. It's almost at the human level. We're dealing with stars and things that we can see. When you get to quantum mechanics and things, it's practically impossible for the human mind to compute what is what just happened there.

没错。我认为这是同样的道理。就像你能从数学角度理解一首乐曲的音符——那是智能的表现。是的。

Yeah. I mean, that that is the same. It's like you understand mathematically the the notes of a musical composition. That's intelligence. Yes.

但为什么它会让你产生特定感受？这就难理解多了。确实如此...不过这种将生物学核心问题视为未解之谜的视角很有意思。我在想最终会是谁破解意识之谜。

But why makes you feel a certain way? That is much harder to understand. Yeah. That's that's really but it it was it was interesting framing that that's a mystery at this at at the core of biology. I wonder who solves consciousness.

我倾向于认为意识问题会被工程师解决。就是那些不断尝试构建意识系统的人，因为生物学体系实在太复杂了。从生物学角度看，意识的基础构件就像是人类最终的造物——你必须理解整个该死的系统。你提到电场，但其实是电场叠加所有因素，整个完整的机制。

I tend to think consciousness will be solved by the engineer. Meaning, the person who builds it, who tries keeps trying to build the thing versus biology is such a complicated system. I feel like it's I feel like the building blocks of consciousness from a biological perspective are like that's like the final creation of of a human being. So you have to understand the whole damn thing. You said electrical fields, but, like, electrical fields plus plus, everything, the whole shebang.

我比较赞同。根据我对科学史浅薄的理解，重大突破往往来自不相关的领域。所以破解意识的很可能不是生物学家——恰恰因为他们深陷问题本质之中。而且即便有人宣称成功，也没人会相信，就像你无法证明狗是否具有意识和悲伤一样。当AI用完美的语言告诉你它拥有这些时，人们最终会接受这个事实。

I'm inclined to agree. I mean, my feeling is from my meager knowledge of the history of science is that the biggest breakthroughs usually come through from a field that was not related So if anyone it's not going be a biologist who solves consciousness just because biologists are too embedded in the nature of the problem. And then nobody's going to believe you when you've done it because nobody's going be able to prove that this AI is in fact conscious and sad in any case, any more than you can prove that a dog is conscious and sad. It tells you that it is in good language and you must believe it. I think most people will accept if faced with that that's what it is.

所有这些关于复杂生命概率的讨论...我认为其意义在于：如果人类能再存活百年左右，AI将逐渐占据主导地位。我们向太空发射的任何探索宇宙的载体都将是AI，而非人类自身。即使有行动，规模也会极其有限。我想这条规律同样适用于任何外星文明。

All of this probability of complex life. In one way, I think why it matters is that my expectation I suppose is that we we will be over the next hundred years or so if we survive at all, that AI will increasingly dominate and pretty much anything that we put out into space going looking for other well, for the universe, for what's out there, will be AI. It won't be us. We won't be doing that or when we do, it will be on a much more limited scale. I suppose the same would apply to any alien civilization.

因此，或许我们不应该在外太空寻找生命迹象，而应该寻找人工智能。但这就面临一个问题：我不明白一个星球如何能直接孕育出人工智能。我能理解星球如何孕育有机生命。如果地球生命进化的原则同样适用于其他行星（我认为大多如此），那么首先演化出能够创造人工智能的人类文明的几率就极其渺茫。一旦成功一次，它可能会接管整个宇宙，或许就不再是问题了。

So perhaps rather than looking for signs of life out there, we should be looking for AI out there. But then we face the problem that I don't see how a planet is going to give rise directly to AI. I can see how a planet can give rise directly to organic life. And if the principles that govern the evolution of life on earth apply to other planets as well, and I think a lot of them would, then the likelihood of ending up with a human like civilization capable of giving rise to AI in the first place is massively limited. Once you've done it once, perhaps it takes over the universe and maybe there's no issue.

但在我看来，这两者必然相关联——你不可能跳过有机生命这个中间环节，直接将一个荒芜的星球变成人工智能生命体。

But it seems to me that the two are necessarily linked, that you're you're not gonna just turn a sterile planet into an AI life form without the intermediary of the organics first.

所以你的意思是，必须通过有机体完成完整的进化计算才能创造出人工智能。

So you you have to run the the full evolutionary computation with the organics to create AI.

如果没有智能生命的协助，人工智能要如何自我启动？我是说——

How does AI bootstrap itself up without the aid, if you like, of an intelligent The

人工智能的起源必然与行星的化学环境有关。但这并非限制因素，对吧？让我用费米悖论来提问：假设我们生活在这样一个黑暗而美丽的世界，数十亿行星上只有细菌，极少有智慧文明，但确实存在少数。

origin of AI is going to have to be in the chemistry of a planet. So but that's that's not a limiting factor. Right? So, I mean so there there's let me ask the Fermi Paradox question. Let's say we live in this incredibly dark and beautiful world of just billions of planets with bacteria on it and very few intelligent civilizations, and yet there's a few out there.

为什么我们从未大规模地见过它们造访地球？你觉得是什么原因？是因为它们不存在？还是因为——

Why haven't we, at scale, seen them visit us? What's your sense? Is it because they don't exist? Is it because

它们不存在于宇宙中正确的时间和地点。这是最简单的解释。

don't exist in the right part of the universe at the right time. That's the simplest answer for it.

这是你认为最有说服力的那个，还是有其他解释？

Is that the one you find the most compelling, or is there some other explanation?

我发现不是因为它更有说服力，而是因为它更有可能，而且我认为所有这些——我的意思是，这里面有很多不确定因素，我们只是不知道。所以我试图从我所知的地球生命来推测其他地方可能发生的事。这让我觉得细菌可能无处不在，而智慧生命只是偶尔出现。往前看，人类在地球上只出现了一次，没有其他任何让你觉得比接触地球上的它们更值得兴奋的存在。所以我认为机会相当有限，我们生存的机会也很有限。

I find that No, it's not that I find it more compelling, it's that I find more probable and I find all of them I mean, there's lot of hand waving in this, we just don't know. So I'm trying to read out from what I know about life on earth to what might happen somewhere else. It gives to my mind a bit of a pessimistic view of bacteria everywhere and only occasional intelligent life. Running forward, humans only once on earth and nothing else that you would necessarily be any more excited about making contact with than you would be making contact with them on earth. So I think the chances are pretty limited and the chances of us surviving are pretty limited too.

按照我们目前的发展趋势，人类在未来几百年内，甚至可能在接下来的五十年或一百年内不自我灭绝的可能性似乎很小。我希望我们能做得更好。所以也许人类唯一能留下的就是人工智能。也许AI一旦存在，并且能够有效地复制自己并将人类排除在外，那么它可能会接管宇宙。

The way we're going on at the moment, the likelihood of us not making ourselves extinct within the next few hundred years, possibly within the next fifty or hundred years seems quite small. I hope we can do better than that. So maybe the only thing that will survive from humanity will be AI. Maybe AI, once it exists and once it's capable of effectively copying itself and cutting humans out of the loop, then maybe that will take over the universe.

我是说，你描述的方式中有一种固有的悲伤。但作为生命的下一步，这不也可能很美吗？我想从你的角度来看，只要它以某种方式承载着意识的火种。

I mean, there's a kind of inherent sadness to the way you describe that. But isn't that also potentially beautiful that that's the next step of life? I suppose as from your perspective, as long as it carries the flame of consciousness somehow.

我认为，是的，作为生命的下一步，它可能有其美丽之处，而且老实说，我不知道从那个角度来看意识是否重要。是的，但确实有些悲伤，可能是因为我觉得这归结于我们之前谈到的自私。我是一个不想被生命淘汰的个体，想要活下去。

I think, yes, there can be some beauty to it being the next step of life, and I don't know if consciousness matters or not from that point of view, to be honest with you. Yeah. But there's some sadness, yes, probably because I think it comes down to the selfishness that we were talking about earlier on. I am an individual with a desire not to be displaced from life. Want to stay alive.

我想留在这里。所以我猜很多人会感到的威胁是，我们会被彻底消灭，AI和人类之间可能会有冲突，而AI会赢，因为它聪明得多。

I want to be here. So I suppose the threat that a lot of people would feel is that we will just be wiped out, that there will be potential conflicts between AI and humans and that AI will win because it's a lot smarter.

哎呀，如果意识只是细菌和AI之间的一个过渡阶段，那该是多么可悲的局面啊。对吧？

Boy, would that be a sad state of affairs if consciousness is just an intermediate stage between bacteria and AI. Right?

因此，我认为细菌可能是一种原始形式的意识。

And so I would see bacteria as being potentially a a kind of primitive form of consciousness.

对，所以也许

Right. So maybe

在我看来，地球上所有的生命都具有意识。它们能够对环境做出某种形式的感觉反应。这并不是说它们具有智慧，尽管它们有自己的智能算法，但无法与我们相提并论。我觉得一个原本荒芜的星球能演化出这一切非常美妙。令人惊叹的是，它创造了我们周围所见的一切，而我们或我们的造物却有能力摧毁这一切。

All of life on Earth to my mind Is conscious. Is capable of some form of feelings in response to the environment. That's not to say it's intelligent though it's got its own algorithms for intelligence, but nothing comparable with us. I think it's beautiful what a planet, what a sterile planet can come up with. It's astonishing that it's come up with all of this stuff that we see around us and that either we or whatever we produce is capable of destroying all of that.

这是个悲伤的想法。但也极度悲观。我更愿意相信我们有能力创造出至少与我们相当、甚至更优越的事物，比如人工智能。

It is a sad thought. But it's also hugely pessimistic. I'd like to think that we're capable of giving rise to something which is at least as good, if not better than us, as AI.

是的。我也有同样的乐观态度，特别是对于那种能比人类或其延伸更高效地在宇宙中传播的事物。无论是通过生物工程延长人类寿命以延续意识之火、人格特质以及我们内在的美好张力，将其带到多个星球、多个星系，还是与人工智能融合实现这一目标——这都是无比美好的愿景。

Yeah. I have that same optimism, especially a thing that is able to propagate throughout the universe more efficiently than humans can or extensions of humans. Some merger with AI and humans, whether that comes from bioengineering of the human body to extend its life somehow to to carry that flame of consciousness and that personality and and the the beautiful tension that's within all of us, carry that through to multiple planets, to multiple solar systems all out there in the universe. I mean, that's a beautiful that's a beautiful vision. Whether AI can do that or bioengineered humans can.

这是个令人兴奋的可能性，尤其是以类似方式遇见其他外星文明。你认为外星生命具有意识吗？

That's an exciting possibility, and especially meeting other other alien civilizations in that same kind of way. Do you think do you think aliens have consciousness?

如果它们是有机生命体的话？

If they're organic?

如此有机。是的。与意识相连。

So organic Yes. Connected to consciousness.

设想任何系统都将从行星起源自我启动——让我先说完这个再谈其他。源自行星的系统将面临相似的约束，而这些约束将通过相似的基础工程方式解决。我认为它将是细胞形态的，带有电荷，并需要经过种群长期选择。所有这些因素往往会催生出相同的过程，作为对难题的最简解决方案。因此我预期它会具有意识，是的，并且会在许多方面与地球生命相似。大约十五六岁时，我读过弗雷德·霍伊尔的《黑云》——当时我是个生物学新人，那是我第一次遇到有人真正挑战生物学核心，说你太过狭隘了。

Think any system which is going to bootstrap itself up from planetary origins Let me finish this and then I'll come on to something else. From planetary origins is going to face similar constraints and those constraints are going to be addressed in similar basic engineering ways. I I think it will be cellular and I think it will have electrical charges and I think it will have to be selected in populations over time and all of these things will tend to give rise to the same processes as the simplest fix to a difficult problem. So I would expect it to be conscious, yes, and I would expect it to resemble life on earth in many ways. When I was about, I guess 15 or 16, I remember reading a book by Fred Hoyle called The Black Cloud, which I was a budding biologist at the time and this was the first time I'd come across someone really challenging the heart of biology and saying, You are far too parochial.

你总把生命视为碳基形态。但这里有种生命形式近乎尘埃，是太阳系尺度的星际尘埃。虽是小说，但它对我冲击极大——我从未意识到自己的思维如此局限，如此固步自封。这位伟大物理学家对生命可能形态的构想截然不同。此后我见过对他各种抨击，但很难说这些批评比原著更有说服力。即便从信息处理角度看——若你处于那种尺度，而光速又是极限，当需要跨太阳系传播时，思维速度能有多快？

You're thinking about life as carbon based. Here's a life form which is kind of dust, interstellar dust on a solar system scale. It's a novel but I felt enormously challenged by that novel because it hadn't occurred to me how limited my thinking was, how narrow minded I was being. Here was a great physicist with a completely different conception of what life could be. Since then, I've seen him attacked various ways and I'm reluctant to say the attacks make more sense to me than the original story, which is to say even in terms of information processing, if you're on that scale and there's a limit to the speed of light, how quickly can something think if you're needing to broadcast across the solar system?

会非常缓慢。至少以弗雷德·霍伊尔设想的时间尺度，它无法与你对话——只要光速仍是极限就不可能轻易实现。正如理查德·道金斯早年所言（我认为他是对的）：要产生这种复杂度，除了自然选择别无他法。

It's going to be slow. It's not going to hold a conversation with you on the kind of timelines that Fred Hoyle was imagining, at least not by any easy way of doing it, assuming that speed of light is a limit. Then again, you really can't. This is something Richard Dawkins argued long ago and I do think he's right. There is no other way to generate this level of complexity than natural selection.

确实别无他途。你需要种群，需要种群中的选择，需要某种孤立的星际云团。虽然时间无限且距离或许不成问题，但必须存在世代更替的特定频率才能产生重大复杂度。而我总觉得这永远行不通。

Nothing else can do it. You need populations, and you need selection in populations, and a kind of an isolated interstellar cloud. Again, there's unlimited time and maybe there's no problems with distance, but you need to have a certain frequency of generational time to generate a serious level of complexity. And I just have a feeling it's never gonna work.

就目前所知，自然选择进化在地球上确实强大，但或许存在其他机制。不知道你是否了解细胞自动机——由极简组件构成的复杂系统，整体运行时遵循简单规则，却能涌现惊人复杂度。我不清楚其驱动压力是什么，虽非选择机制，但复杂度确实自发显现，这背后的原理尚未被充分理解。

Well, as far as we know, so natural selection evolution is a really powerful tool here on earth, but there could be other mechanisms. So whenever I don't know if you're familiar with cellular automata, but complex systems that that have really simple components and seemingly move based on simple rules when they're taken as a whole, really interesting complexity emerges. I don't know what the pressures on that are. It's not really selection, but interesting complexity seems to emerge and that's not well understood exactly why complexity

与进化相关。我们部分关于生命起源的研究正在思考：基因如何产生？生物信息如何形成？从二氧化碳与氢反应的角度看会得到什么？羧酸和氨基酸。

and evolution. Some of the work we're doing on the origin of life is thinking about how do genes arise? How does information arise in biology? Thinking about it from the point of view of reacting CO2 with hydrogen, what do you get? Well, you're going to get carboxylic acids and amino acids.

合成核苷酸相当困难，但确实可以做到，而且已经有人通过这条途径实现了。不过产量极其微量。那么接下来的问题是——假设这种思考方向是正确的（虽然可能并非如此），即如何可靠地生产更多核苷酸？如何让核苷酸生成机制变得更复杂高效？答案是需要正反馈循环，某种形式的自催化作用。这在生物学中确实存在且可行。

It's quite hard to make nucleotides and it's possible to make them and it's been done and it's been done following this pathway as well. But you make trace amounts. And so the next question assuming that this is the right way of seeing the question which maybe it's just not but let's assume it is, is well how do you reliably make more nucleotides and how do you become more complex and better at becoming a nucleotide generating machine? And the answer is well you need positive feedback loops, some form of autocatalysis. So that can work and we know it happens in biology.

以这个核苷酸为例，它能催化二氧化碳固定作用，从而提升整个系统的反应速率，实质上增强了生成更多核苷酸的驱动力。这种特性可以遗传，因为存在膜遗传机制。当细胞分裂时，如果内部物质以能自我再生的网络形式存在，就必然能实现自我复制。复杂度可以不断提升，但这一切都建立在热力学基本法则之上，不存在所谓的可进化性。

This nucleotide for example catalyzes CO2 fixation, You're going to increase the rate of flux through the whole system and you're going to effectively steepen the driving force to make more nucleotides. This can be inherited because there are forms of membrane heredity that you can have and there are effectively you can if a cell divides in two and it's got a lot of stuff inside it and that stuff is basically bound as a network which is capable of regenerating itself, then it will inevitably regenerate itself. You can develop greater complexity. But everything that I've said depends on the underlying rules of thermodynamics. There is no evolvability about that.

这只是初始条件下不可避免的结果——假设你能增强系统驱动力。你会生产更多同类产物，扩展功能边界，但永远不会产生本质差异。只有当引入基因信息（比如一段随机RNA片段）时，才会出现真正的可进化性，才会形成我们所知的生物学体系，同时也伴随着我们熟知的自然选择机制。

It's simply an inevitable outcome of your starting point assuming that you're able to increase the driving force through the system. You will generate more of the same, you'll expand on what you can do but you'll never get anything different than that. It's only when you introduce information into that as gene, as a small stretch of RNA, which can be random stretch, then you get real evolvability, then you get biology as we know it, but you also have selection as we know it.

确实。我不知道该如何理解'信息'这个概念。它像是系统的某种记忆。在局部层面，就是自我复制传播的过程，通过改变来提升对环境的适应能力。

Yeah. I mean, I don't know how to think about information. That's a kind of memory of the system. So it's not yeah. At the local level, it's propagation of copying yourself and changing and improving your adaptability to the environment.

但若将地球视为整体，它确实具有某种记忆功能，这是其关键特征。

But if you look at Earth as a whole, it has a kind of memory. That's the key feature of it.

具体指什么？

In what way?

它会记住尝试过的路径。如果要描述地球的话...我认为进化是我们作为个体生物体验到的现象，是个体间相互作用的方式，存在自然选择。但若将地球视为完整生命体，你会如何描述它？

It remembers the stuff it tries. Like, if you were to describe Earth, I I think evolution is something that we experience as individual organisms. That's that's how the individual organisms interact with each other. There's a natural selection. But when you look at Earth as an organism in its entirety, how would you describe it?

我是说

I mean

嗯，不是作为一个有机体。盖亚的概念确实很美好。詹姆斯·洛夫洛克最初将盖亚假说提出为一个进化而成的有机体，随即遭到许多人抨击。他没错，但后来有所退让，因为这更多是一种诗意的愿景而非科学。现在这门科学被称为地球系统科学，真正研究的是世界如何自我调节以维持在适宜生命存续的范围内——它做得非常出色，确实是在行星尺度上实现调控整合。

Well, not as an organism. I mean, the the idea of Gaia is is lovely. And James Lovelock originally put Gaia out as an organism that had somehow evolved and he was immediately attacked by lots of people. He's not wrong but he backpedaled somewhat because that was more of a poetic vision than the science. The science is now called Earth Systems Science and it's really about how does the the world kind of regulate itself so it remains within the limits which are hospitable to life and it does it amazingly well and it is it is working at a planetary level of integration of regulation.

但它并非通过自然选择进化，也不可能，因为它是独一无二的。所以它会随时间变化，但不是进化。所有进化都发生在系统的组成部分中。

But it's not evolving by natural selection and it can't because there's only one of it. So it can change over time but it's not evolving. All the evolution is happening in the parts of the system.

是啊。但它是个自我维持的有机体。

Yeah. But it's a self sustaining organism.

不。它靠太阳维持。

No. It's sustained by the sun.

对。所以你的意思是，你认为不能把地球视为独立的有机体？

Right. So I mean, so you don't think it is possible to see Earth as its own organism?

我认为这种说法富有诗意且美好，我常称地球为活着的星球，但从生物学角度讲它并非有机体。不是的。

I think it's poetic and beautiful and I often refer to the the Earth as a living planet, but it's not in in in biological terms an organism. No.

如果外星人造访地球，他们会注意到什么？他们会认为什么是生命的基本单位？

If aliens were to visit Earth, would they what would they notice? What would be the basic unit of life they would notice?

可能是树木。我是说，它是绿色的。地球是绿色和蓝色的。我想这是你首先会注意到的。从太空看，它与其他行星截然不同。

Trees, probably. I mean, it's it's green. It's green and blue. I think that's the first thing you'd notice. It it stands out from space as being different to any of the other planets.

所以最初注意到树木是因为绿色，我

So notice the trees at first because the green I

我注意到了绿色。

I noticed the green.

是的。然后可能会发现光合作用。

Yes. Yeah. And then probably noticed figure out the photosynthesis.

我猜第二眼可能会注意到城市。对。

Probably noticed cities a second, I suspect. Yeah.

那么让我

So let me

他们在夜间抵达，首先注意到的是城市。这一点毋庸置疑。

arrived at night, they noticed cities first. That's for sure.

这取决于时间。你在《变形金刚》中又一次写得非常优美。我想你是这样开篇的，我记不清了。用太空视角描述地球，这是个关于地球本质多么有趣的想法。

It depends depends the time. You you write quite beautifully in Transformers once again. I think you opened the book in this way. I don't remember. Space Describing Earth, it's such an interesting idea of what Earth is.

你还提到，《银河系漫游指南》将其概括为无害或基本无害，这是种极具诗意的表达。你以太空视角展开《变形金刚》——从太空看，它呈现灰暗的晶体状，抹去了生命地球的蓝绿色彩。表面布满规则的网格纹路和汇聚的条纹。中央有团模糊的密集区域，那些划痕显得较浅。这种'生长物看起来没有生命，虽然它沿着某些线条延伸，却带着某种攫取与寄生的特质'。

You also I mean, Hitchhiker's Guide summarizing it as harmless or mostly harmless, which is a beautifully poetic thing. You open transformers with from space, it looks gray and crystalline, obliterating the blue green colors of the living earth. It is crisscrossed by regular patterns and convergence striations. There's a central amorphous density where these scratches seem lighter. This, quote, growth does not look alive, although it has extended out along some lines, and there is something grasping and parasitic about it.

全球有成千上万这样的结构，形状细节各异，但全都灰暗、棱角分明、无机质般蔓延。然而到了夜晚，它们会亮起来，在黑暗天空中突然变得美丽。或许这些地貌上的疮疤在某种意义上是活物。它们有受控的能量流动，必然存在信息传递和某种形式的新陈代谢，物质的更替。

Across the globe, there are thousands of them, varying in shape and detail, but all of them gray, angular, inorganic, spreading. Yet at night, they light up, Going up the dark sky, suddenly beautiful. Perhaps these cankers on the landscape are in some sense living. There's a controlled flow of energy. There must be information and some form of metabolism, some turnover of materials.

它们是活的吗？不，当然不是。它们是城市。那么是否可以说城市具有某种生命特征？

Are they alive? No. Of course not. They are cities. So is there some sense that cities are living beings?

你觉得外星人会认为它们是生物吗？

You think aliens would think of them as living beings?

不会。但很容易产生这种错觉，不是吗？

No. It'd be easy to see it that way, wouldn't it?

它在夜间醒来。它们在夜间醒来。

It wakes up at night. They wake up at night.

严格夜行性的。没错。我想任何聪明到能来到这里的外星生物都会明白它们并非生命体。我这么说是因为我们倾向于从信息角度思考生物学，却忽略了细胞。我试图将细胞比作城市，比较能量在城市中的流动与在细胞中的流动，以及物质的更新换代。

Strictly nocturnal. Yes. I imagine that any aliens that are smart enough to get here would understand that they're not living beings. My reason for saying that is that we tend to think of biology in terms of information and forget about cells. I was trying to draw a comparison between the cell as a city and the energy flow through the city and the energy flow through cells and the turnover of materials.

关于城市的一个有趣之处在于，它们实际上并不完全受任何人统治。虽然有规章制度和各种体系，但相当松散。它们有自己的生命，有自己随时间发展的方式。在这个意义上，它们相当具有生物特性。1666年伦敦大火后曾有过重建计划。

And an interesting thing about cities is that they're not really exactly governed by anybody. There are regulations and systems and whatever else but it's pretty loose. They have their own life, their own way of developing over time. They they they and in that sense, they're quite biological. They're not There was there was a plan after the great fire of London.

克里斯托弗·雷恩不仅为圣保罗大教堂做设计，还计划用巴黎式林荫大道重建伦敦市中心被烧毁的大片区域。最终未能实现，我想是因为资金不足。但计划内容很有意思：规划了众多林荫大道，却没有酒吧、咖啡馆这类场所。而实际上伦敦就是在杂乱无章的街巷中自然生长起来的。

Christopher Wren was was making plans not only for Saint Paul's Cathedral but also to rebuild in in large Parisian type boulevards, a large part of the area of Central London that was burnt. It never happened because they didn't have enough money I think. But it's interesting what was in the plan. There were all these boulevards but there were no pubs and no coffee houses or anything like that. The reality was London just kind of grew up in a set of jumbled streets.

当时伦敦城的商业交易都发生在咖啡馆和酒馆里，那里才是城市真正的生命力所在。这一切都未经规划，反而运作得更好。同理，细胞也完全是无计划形成的。它并不像我们想象的那样受细胞核中的基因控制，而是进化出的实体，具有同样的动态变化、同样的活力、同样的生命特征。

It was the coffee houses and the pubs where all the business of the city of London was being done and that was where the real life of the city was. No one had planned it. The whole thing was unplanned and works much better that way. In that sense, the cell is completely unplanned. It's not controlled by the genes in the nucleus in the way that we might like to think that it is, but it's kind of evolved entity that has the same kind of flux, the same animation, the same life.

所以我认为这是个绝妙的类比，但不必过于拘泥于这个隐喻。

So I think it's it's a it's a beautiful analogy, but I wouldn't get too stuck with it as a as a metaphor.

看，我不同意你的观点。我完全不同意。我认为你——实际上整个科学史都深陷于思考'什么是生命'的生物学框架中。而且不仅是生物学视角，还非常人类中心主义。

See, I disagree with you. I I I disagree with you. I I think you're you are so steeped. Actually, the entirety of science, the history of science is steeped in a biological framework of thinking about what is life. And not just biological, it's very human centric too.

人类有机体是地球上生命的典范。我不知道...我认为在某种深层次的基本方式上，城市就像一个活体，就像人类一样

That human the human organism is the epitome of life on Earth. I don't I don't know. I I think there is some deep fundamental way in which a city is a living being in the same way that a human

它会孕育出子代城市。我的意思是，它不是通过自然选择运作的。它是通过模因传播，通过...

It give rise to an offspring city. So I mean, it's not it doesn't work by natural selection. It works by anything memes, it works by

是啊。但难道不是

Yeah. But isn't

这个概念本身作为一种存在模式

that itself conceptually as a as a as a mode of being.

所以，我的意思是，也许模因、也许思想才是地球上真正至关重要的生命体。也许人类本性的集体层面——集体智慧比个体智慧重要得多。也许集体人性才是那个有机体，而定义人类集体智慧的东西正是那些思想，也许其具体表现形式就是城市。或者社会、地理限制的社会、国家这类东西。从外星视角看，很可能这才是更引人注目的特质，而非出于无知。

So, I mean, maybe memes, maybe ideas are are the organisms that are really essential to life on Earth. Maybe it's much more important about the collective aspect of human nature, the collective intelligence than the individual intelligence. Maybe the collective humanity is the organism, and the the thing that that defines the collective intelligence of humanity is the ideas, and maybe the way that manifests itself is cities. Maybe or societies or geographically constrained societies or nations and all that kind of stuff. I mean, from an alien perspective, it's possible that that is the more deeply noticeable thing, not from a place of ignorance.

显眼的事物并不能说明其运作机制。我觉得...其实我对你说的没什么异议，只是这一切离不开人类。我们是从狩猎采集型经济过渡来的

What's noticeable doesn't tell you how it works. I think I mean, I don't have any problem with what you're saying really except that it's not possible without the humans. We went from a hunter gatherer type economy

如果

if

你喜欢没有城市直达城市的概念。一旦我们进入人类进化、文化和社会等领域，那么确实存在其他形式的进化与变革。但城市并非直接自我繁衍，它们通过人类社会传播，而人类社会仅因人类个体的繁衍而存在。这里存在一种层级关系，若最初没有人类，其余一切便不复存在。

you like without cities through to city And as soon as we get into human evolution and culture and society and so on, then then then yes, there are other forms of evolution, other forms of change. But but cities don't don't directly propagate themselves, they propagate themselves through human societies and human societies only exist because humans as individuals propagate themselves. So there's a kind of a there is a hierarchy there and without the humans in the first place, of the rest of it exists.

所以对你而言，生命的定义主要基于进化作用的基本单元？我认为这是个非常关键

So to you, life is primarily defined by the the basic unit on which evolution can operate? I think it's a really important

的观点。是的。

thing. Yes.

确实。而且除了进化论，我们目前还没有更好的理论来解释如何创造

Yeah. And we don't know we don't have any other better ideas than evolution for how to create

我从未遇到过比进化论更卓越的理论。或许是我孤陋寡闻，你提到的无自动生成者等概念我并未具体思考过。但我曾从生命起源的选择单位角度，以及可进化性与复杂性（或有限范围内的复杂度增长）的差异来思考。基因和选择的伟大之处在于它们打破了所有限制——最终呈现的信息世界只受生物物理现实（如生物类型、行星环境等）制约。城市等看似有生命、可被描述为生命的形式，因无法自我繁衍，只能作为可自我繁衍之物的产物存在。

I never came across a better idea than evolution. I mean, maybe maybe I'm just ignorant and I don't know and and there's you know, you mentioned there's no automator and so on, and and I don't think specifically about that, but I have thought about it in terms of selective units at the origin of life and the difference between evolvability and complexity or just increasing complexity, but within very narrowly defined limits. The great thing about genes and about selection is it just knocks down all those limits. It gives you a world of information in the end which is limited only by the biophysical reality of what kind of an organism you are, what kind of a planet you live on and so And cities and all these other forms that look alive and could be described as alive, because they can't propagate themselves, can only exist as the product of something that did propagate itself.

确实。这种视角蕴含着极具说服力的真理。但我希望我们不要因此忽视身边的庞然大物。

Yeah. I mean, there's a deeply compelling truth to that kind of way of looking at things. But I I just hope that we don't miss the giant cloud among us.

我其实希望自己在许多方面是错的，因为我的世界观并不特别令人振奋。但从某种角度说，是否振奋并不重要。科学关乎现实本质、存在真相及其成因——而我认为这其中也蕴含着美。

I of hope that I'm wrong about a lot of this because I can't say that my worldview is particularly uplifting. But in some sense, it doesn't matter if it's uplifting or not. Science is about what's reality, what's what's out there, why why is it this way? And and I'm I think there's beauty in that too.

黑暗中有其美。你在生物学层面上探讨生死。关于自杀的问题——为何而活？为何人类心智会陷入抑郁，你能从生物学角度内省这一点吗？为何我们的心智、为何我们人类会堕入如此黑暗的境地？

There's beauty in darkness. You write about life and death sort of at the biological level. Is does does the question of suicide, why live? Does the question of why the human mind is capable of depression, are you able to introspect that from a place of biology? Why our minds why we humans can go to such dark places?

为何我们会自杀？为何我们能陷入——你知道的——痛苦时期，还会因无意义感而痛苦，被抑郁带入黑暗深渊？这是生命的特性还是缺陷？

Why can we commit suicide? Why can we go, you know, suffer suffer period, but also suffer from a feeling of meaninglessness of going to a dark place that depression can take you? Is this a feature of life or is it a bug?

我不知道。如果这是生命特性，那其他生物体应该也如此，但我并不确定。我们之前聊到狗，它们确实会极度悲伤，主人去世后可能郁郁寡欢好几天。从某种意义说，这或许是生物学的特性，可能源于生命有限性，甚至超越这些范畴。我想可以从两个角度来思考。

I don't know. Mean, it's a feature of life, then I suppose it would have to be true of other organisms as well, and I don't know. We were talking about dogs earlier on and they can certainly be very sad and upset and may mooch for days after their owner died or something like that. I suspect in some sense it's a feature of biology. It's probably a feature of mortality, it's probably a beyond all of that, I mean I guess there's two ways you could come at it.

其一是通过理性计算得出结论：一切皆无意义，我继续存在毫无价值。或许在宏大宇宙尺度下确实如此。你可以用社会价值为自己辩护，但社会本身也终将消亡，最终逻辑只会导向一片荒芜。

One of them would be to say, well, you can effectively do the math and come to the conclusion that it's all pointless and that there's really no point in me being here any longer. Maybe that's true in the greatest scheme of things. You can justify yourself in terms of society, but society will be gone soon enough as well and you end up with a very bleak place just by logic.

某种意义上，我们竟能找到任何意义才令人惊讶。

In some sense, it's surprising that we can find any meaning at all.

或许这正是意识的作用——我们拥有转瞬即逝的欢愉，但也伴随着短暂的痛苦。生物学中所有事物都难以完美调控。总会呈现钟形曲线分布：不幸者处于极乐端，另一些人则在痛苦端，这是大脑的固有构造，恐怕永远无法逃脱。性与其他方面亦然，我们面对的是无法精确调控的系统。

Well, maybe this is where consciousness comes in, we we have transient joy, but with transient joy, we have transient misery as well. And sometimes with everything in biology, getting the regulation right is practically impossible. You will always have a bell shaped curve where some people unfortunately are the joy end and some people are at the misery end and that's the way brains are wired and I doubt there's ever an escape from that. It's the same with sex and everything else as well. We're dealing with a you can't regulate it.

这就是生命的本质，都是生物学的一部分。

It's goes. It's all part of biology.

诚然如此。首先让我谈谈你的著作《权力、性与自杀》，我能直接列举你写过的书吗？如果有更佳的书名和主题，我也想不出还有什么。这让我对你接下来要写的内容充满期待。希望你能继续创作。

Amen to that. Let me on writing, in your book, Power, Sex, and Suicide first of all, can I just read off the books you've written? If there's any better titles and topics to be covered, I don't know what they are. It makes me look forward to whatever you're going to write next. I hope there's things you write next.

那么你首先写了《氧气：塑造世界的分子》，正如我们讨论过的氧气在地球生命中的作用。接着是《等待时机》、《权力、性与自杀》、《线粒体与生命的意义》，然后是《生命升级：进化的十大发明》、《关键问题》——这是我读你的第一本书《关键问题：生命为何如此》，以及新书《变形者：生命与死亡的深层化学》。在《权力、性与自杀》中，你写了很多内容，但我有个关于写作的问题。你写道，在《银河系漫游指南》中，福特长官花了十五年时间修订指南中关于地球的条目，原版写的是‘无害’。顺便问一下，作为附带问题，我也想请教你对地球的总结是什么。

So you first, you wrote oxygen, the molecule that made the world, as we've talked about this idea of the role of oxygen in life on Earth. Then Wait For It, Power, Sex, Suicide, Mitochondria and the Meaning of Life, then Life Ascending, The 10 Great Inventions of Evolution, The Vital Question, the first book I've read of yours, The Vital Question, Why is Life the Way It Is, and the new book, Transformer, The Deep Chemistry of Life and Death. In Power, Sex, and Suicide, you write about writing or about a lot of things, but I I have a question about writing. You write, in the hitchhiker's guide to the galaxy, Ford Perfect spends fifteen years researching his revision to the guide's entry on the earth, which originally read harmless. By the way, I would also, as a side quest as a side question, would like to ask you what would be your summary of what Earth is.

但你说得对。他那篇关于这个主题的长文被编辑删减后只剩下‘基本无害’。我怀疑太多新增内容都遭遇了类似命运，即使不是由于荒谬的编辑决定，至少也因为内容缺乏实质性变化。巧的是，《权力、性与自杀》初版问世已近十五年，我正抵制着进行拙劣修订的诱惑。有人说连达尔文也因多次修订《物种起源》而削弱了论点力量——他在修订中应对批评，有时反而偏离了正确方向。

But you're right. His long essay on the subject is edited down by the guy to read mostly harmless. I suspect that too many new additions suffer a similar fate, if not through absurd editing decisions, at least through a lack of meaningful change in content. As it happens, nearly fifteen years have passed since the first edition of Power Sex Suicide was published, and I am resisting the temptation to make any lame revisions. Some say that even Darwin lessened the power of his arguments in the origin of species through his multiple revisions in which he dealt with criticisms and sometimes shifted his views in the wrong direction.

我更愿让原作自己说话，即便它最终被证明是错的。让我请教关于写作的问题——既包括你在学术环境中指导学生，也包括你创作了我读过的最精彩的科学与人文著作。写作的过程是怎样的？如果要给年轻的达尔文、年轻的你或任何年轻人关于如何写作、如何写好宏大主题的建议，你会说什么？

I prefer my original to speak for itself, even if it turns out to be wrong. Let me ask the question about writing, both your students in the academic setting, but also writing some of the most brilliant writings on science and humanity I've ever read. What's the process of writing? How how do you advise other humans if you if you were to talk to young Darwin or the young young you and just young anybody and give advice about how to write and how to write well about these big topics, what would you say?

我...我是说，我认为有几个要点。首先是：故事是什么？我想知道什么？想传达什么？为什么这对任何人重要？

I I mean, I suppose there's couple there's a couple of points. One of them is what's the story? What do I want to know? What do I want to convey? Why does it matter to anybody?

而往往最大、最有趣的问题，那些孩子气的问题，其实是每个人都想问却不敢问的，怕显得愚蠢。科学界的好处是，你待得越久，就越明白没人知道这些问题的答案，提问并不那么蠢。我试图问那些15、16岁时渴望了解世界却知之甚少的自己会提出的问题，想抵达认知边缘但需要引导。我不愿用太多术语，希望有人能清晰聚焦问题本身。

And very often, the most the biggest, most interesting questions, the childlike questions are the one that actually everybody wants to ask but don't quite do it in case they look stupid. One of the nice things about being in science is the longer you're in, the more you realize that everybody doesn't know the answer to these questions and it's not so stupid to ask them after all. Trying to ask the questions that I would have been asking myself at the age of 15, 16 when I was really hungry to know about the world and didn't know very much about it and wanted to go to the edge of what we know, but be helped to get there. I don't want to be too much terminology and so on. I want someone to keep a clean eye on what the question is.

除此之外，我常思考：我在为谁写作？最终唯一的答案是15、16岁的自己。因为你永远不知道读者是谁，也不知道他们在哪里阅读——在浴缸、床上、地铁里听有声书？你希望每几页就有内容回顾（如果他们每次只读三页），还是觉得这很烦人？

Beyond that, I've wondered a lot about who who am I writing for? That was in the end, the only answer I had was was myself at the age of 15 or 16. Because even if you're you just don't know who's reading but also where are they reading it. Are they reading it in the bath or in bed or on the metro or listening to an audiobook? Do you want to have a recapitulation every few pages because you read three pages at a time or are you really irritated by that?

你会受到那些对你所作所为感到恼火之人的批评。你不知道他们是谁，也不知道自己做了什么会惹恼别人。最终，你只能尽力取悦自己。这意味着，那些宏大、有趣、迷人的核心问题是什么？我们对此了解多少？

You're going to get criticism from people who are irritated by what you're doing. You don't know who they are or what you're to do that's going to irritate people. In the end, all you can do is just try and please yourself. That means, what are these big, fun, fascinating and big questions? What do we know about it?

我能传达这些吗？我在写作尝试中逐渐领悟到，首先要阐明已知事实。最初几本书里，我震惊地发现经常很快触及未知领域。后来在指导物理和数学博士生时，我意识到他们的数学能力远超于我。但试图厘清'我们实际要建模什么'、'这个方程包含什么'的过程，与写作极为相似。

Can I convey that? I kind of learned in trying to write, first of all, say what we know. I was shocked in the first couple of books how often I came up quickly against all the stuff we don't know. If you're trying to I realized later on in supervising various physicists and mathematicians who are PhD students, their maths is way beyond what I can do. But the process of trying to work out what are we actually going to model here, what's going into this equation is a very similar one to writing.

我要在纸上写下什么？如何用最简洁的方式概括这个想法，使其成为可观察互动关系的独立单元？你会意识到，如果这个像那样，那个像这样，那么结论就不成立。于是你在这片领域中开辟自己的路径，这种未知的探索令人振奋。我想这正是我的书受欢迎的原因之一——读者能感受到这种不知终点的惊险旅程，因为连我自己也不知道结局。

What am I going to put on a page? What's the simplest possible way I can encapsulate this idea so that I now have it as a unit that I can kind of see how it interacts with the other units? And you realize that well, if this is like that and this is like this, then then that can't be true. So you end up navigating your own path through this landscape and that can be thrilling because you don't know where it's going. I'd like to think that that's one of the reasons my books have worked for people because this sense of thrilling adventure ride, I don't know where it's going either.

所以关键在于用最简方式解释已知与未知事物，以及两者间的张力，故事由此诞生。那么编辑过程呢？你是否发现自己会删减到'基本无害'的程度？为追求简洁，你觉得编辑是建设性的，还是会破坏原有的魔力？

So this finding the simplest possible way to explain the things we know and the simplest possible way to explain the things we don't know and the tension between those two, and that's where the story emerges. What about the edit? Do you find yourself to the point of this, you know, editing down to mostly harmless? To arrive at simplicity, do you find the edit is productive, or does it destroy the the magic that was originally there?

不。我通常认为——或许我作为编辑比作家更出色。我会反复写，反复改，改了一遍又一遍。

No. I usually find I I think I'm perhaps a better editor than I am a writer. I I write and rewrite and rewrite and rewrite.

先把一堆草稿堆在纸上，然后看编辑会把它

Put a bunch of crap on the page first, and then see where where the edit where it

引向何方。是的。但之后还有专业编辑的介入。《变形者》的编辑在收到初稿两个月后反馈说：'前两章对普通读者构成巨大障碍，去修改吧。'

takes Yeah. But then then there's the professional editors who come along as well. Mean, in in Transformer, the editor came back to me after I'd sent two months after I sent the first edition, he'd read the whole thing and he said, The first two chapters prevent a formidable hurdle to the general reader. Go and do something about it.

是的。

Yes.

而那正是我最后真正想要的东西

And that was the last thing I really wanted

你的编辑听起来口才非常好。

to Your editor sounds very eloquent in speech.

是啊。虽然这是封邮件，但我仔细考虑后，归根结底他是对的。所以我将整件事搁置了大约两个月，度过了夏天。我想应该是去年夏天。然后在九月左右全神贯注地重新处理，几乎是从零开始重写了那些章节。

Yeah. Well, this an email, but I thought about it and the bottom line is he was right. So I put the whole thing aside for about two months, spent the summer. This would have been, I guess, last summer. And then turned to it with full attention in about September or something and rewrote those chapters almost from scratch.

我保留了一些材料，但花了很长时间消化，弄清楚需要改变什么，哪些地方需要...这段时间我并没有写作。怎样才能把故事讲得更好，让它更易懂有趣？最终我认为...这仍然很难。依旧是生物化学，但它有了...他最后说，现在作品充满澎湃能量，而我...因为他第一次说了实话，我决定相信他第二次说的也是实话，并且感到非常高兴。能

I kept some of the material, but it took me a long time to process it, to work out what needs to change, where does it need to I wasn't writing in this time. How am I going to tell the story better so it's more accessible and interesting? In the end, I think it It's still difficult. It's still biochemistry but it has He ended up saying, now he's got a barreling energy to it and I was Because he told me the truth the first time, I decided to believe that he was telling me the truth the second time as well and was was was delighted. Could

你能给年轻人一些普遍建议吗？高中生、大学生，该如何应对你处理过的那些重大问题？你在生物学领域做到了并拓展出去。他们怎样才能拥有值得骄傲的事业或人生？

you give advice to young people in general? Folks in high school, folks in college, how to take on some of the big questions you've taken on? Now you've done that in the space of biology and expanded out. How can they have a career they can be proud of or have a life they can be proud of?

天啊。这真是个宏大的问题。

Gosh. That's a big question.

我相信你已经积累了一些可以传授的智慧。

I'm I'm sure you've gathered some wisdom that you can impart.

实际上，我唯一给学生们的建议就是追随你感兴趣的领域，因为他们担心如果现在做出这个决定选择这门课程而非那门，就会限制职业机会。科学领域没有固定的职业路径。说有也有，说没有也没有。竞争激烈，象征性的淘汰很多。谁能幸存？

Into some The only advice that I actually ever give to my students is follow what you're interested in because they're worried that if they make this decision now and do this course instead of that course, then they're going to restrict their career opportunities. There isn't a career path in science. There is but there isn't. There's a lot of competition, there's a lot of death symbolically. Who survives?

幸存者是那些足够在乎并坚持做下去的人。他们往往是不太担忧未来、能够活在当下的人。因为如果你攻读博士学位，你已经为获得博士资格激烈竞争过，接着又要为博士后职位拼命竞争，而且下个职位可能在另一个大洲，反正也只有两年期限，最终也没有保证能获得教职。所以

The people who survive are the people who care enough to still do it. They're very often the people who don't worry too much about the future and are able to live in the present. Because if you do a PhD, you've competed hard to get on to the PhD, then you have to compete hard to get a post doc job and you have, you know, the next bond maybe on another continent and it's only two years anyway and so and and there's no there's no guarantee you're gonna get a faculty position at the end of it. So

而且总有下一步要竞争。如果你获得教职，就要争取终身职位，成为正教授后，还要争取学科领域的各种奖项。如果你是物理学家，就永远在为诺贝尔奖竞争。奖项各不相同。是的。

And there's always the next step to compete. If you get a faculty position, you get a tenure, and with tenure, you go full professor and full professor, then you go to some kind of whatever the discipline is, there's an award. If you're in physics, you're always competing for the Nobel Prize. There's different awards. Yeah.

是的。最终你们都在竞争——我是说，竞争永无止境。

Yeah. And then eventually, you're all competing to I mean, there's always a competition.

所以不存在幸福。幸福并不存在。

So there is no happiness. Happiness does not lie.

如果你总是望向未来，确实如此。

If you're looking into the future, yes.

如果你关心的是职业发展，那这可能不适合你。但如果你能放下这点，要知道我也曾在企业工作过短暂时间，两次遭遇裁员，我明白那条路同样无法保证职业稳定。

And if what you're caring about is a career, then then it's probably not the one for you. If though you can put that aside, and you know, I've also worked in industry for a brief period and I was made redundant twice, I know that there's no guarantee that you've got a career that way either.

是的。

Yes.

活在当下，享受你所做的事。这意味着真正投入你最感兴趣的主题，并尽可能深入地追随它们。这种投入往往会以意想不到的方式回报。我不知道你是否也有同感，但我发现人们常常会帮助你——当他们看到你眼中闪烁的光芒，你对他们的学科充满热情并渴望探讨时，如果他们知道加州的朋友有职位空缺，就会推荐你去试试。

Live in the moment and try and enjoy what you're doing. That means really go to the themes that you're most interested in and try and follow them as well as you can. And and that tends to pay back in surprising ways. I don't know if you found this as well, but I I found that people will help you often. If they see some light shining in the eye, you're excited about their subject and just want to talk about it and they know that their friend in California has got a job coming up, they'll say, go for this.

只要你是真心投入的，大家会通过人脉网络帮你。两年后是否还有这份工作并不重要，重要的是你现在全心投入。只要你眼中有光，事情自然会解决。这是我唯一能给的忠告，多数人可能因为这个体系而退出，因为对他们而言这种奋斗不值得。

This guy's alright. They'll use the network to help you out if you really care. You're not going to have a job two years down the line, if what you really care about is what you're doing now, then it doesn't matter if you have a job in two years time or not. It'll work itself out if you've got the light in your eye. So that's the only advice I can give and most people probably drop out through that system because the fight is just not worth it for them.

没错。当你眼中有光，对事物充满热情时，你会自然而然地吸引同样对此感兴趣、眼中也有光芒的人。我认为这确实需要下功夫——

Yeah. When you have the light in your eye, when you have the excitement for the thing, what happens is you start to surround yourself with others. They're interested in that same thing that also have the light. If you really are rigorous about this because I think it does take it's it doesn't it takes effort to make

噢，你必须达到痴迷的程度。但如果你做的是真正热爱的事，那就不再是工作，而是你的生活方式。

Oh, you've got to be obsessive. But but if you're doing what you really love doing, then it's not work anymore, it's what you do.

是的。但我也指要让自己周围都是对同一事物痴迷的人，因为——

Yeah. But I also mean the surrounding yourself with other people that are obsessed about the same thing because

这同样需要一些努力。是的。

takes some work as well. Yes.

还有认知

And know

还有运气。

And luck.

找到合适的...是的。找到合适的导师和合作者很重要，因为我认为博士过程中的一个问题就是人们选择导师时不够谨慎。这些导师和同事，他们会定义你人生的方向。你对一件事的热爱程度，我的意思是，仅仅在走廊里的几次简短交谈就能产生不可思议的影响。所以在这方面你必须稍微谨慎些。

Finding the right yeah. Finding the right mentors, the collaborators because I think one of the problem with the PhD process is people are not careful enough in picking their mentors. Those are people mentors and colleagues and so on, those are people who are gonna define the the direction of your life, the Yeah. How much you love a thing, how much I mean, the the power of just, the few little conversations you have in the hallway, it's it's it's incredible. So you have to be a little bit careful in that.

有时候你几乎是被随机分配的，实际上我认为，你既要追求课题本身，也要追求从事这个课题的人。就像，两者都要兼顾，整个...

Sometimes you just get randomly almost assigned, really pursue, I suppose, the subject as much as you pursue the people that do that subject. So, like, both, the whole dance of

它们确实是相辅相成的。

it. They kind of go together really.

是的。确实如此。但要认真对待这部分，可能就像你说的那样，要谨慎定义成功。因为...

Yeah. They do. They they really do. But to to take that that that part seriously, and and probably in the way you're describing it, careful how you define success. Because

你永远不会在成功中找到幸福。记得罗伯特·路易斯·史蒂文森说过一句精辟的话，大意是：人生最令人幻灭的莫过于达成目标。

You'll never find happiness in success. Don't think there's a lovely quote from Robert Louis Stevenson, I think, who said, nothing in life is so disenchanting as attainment.

是啊。某种意义上，成功的真正定义就是能做让你真正享受的事，那些让你满心欢喜的事，这才是最终的成功。所以成功不是地平线那边的奖杯、巨额财富或显赫地位。

Yeah. So, I mean, in in some sense, the the the true definition of success is you're getting to do today what you really enjoy doing, just what fills you with joy, and that's ultimately success. So success isn't the thing beyond the horizon, the big stat the the the the big trophy, the the financial

我认为这已是最接近幸福的状态了。不是说你要时刻欢欣鼓舞，但通过日常事务获得满足感，就是我们能持续获得人类幸福的最接近方式。如果你追求的是诺贝尔奖这类世俗认可，我见过这样的人——他们被未被授予应得奖项的愤懑与怨恨逐渐侵蚀。

I think it's as close as we can get to happiness. That's not to say you're full of joy all the time, but it's it's as close as we can get to a sustained human happiness is by getting some fulfillment from what you're doing on a daily basis. If what you're looking for is the world giving you the stamp of approval with a Nobel Prize or a fellowship or whatever it is, then you know, I've I've known people like this who they they're eaten away by the by the anger, the kind of caustic resentment that they've not been awarded this prize that they deserve.

反过来说，若你过度看重这类奖项又恰好获奖，你会发现所谓的成功越大，自我膨胀的风险就越高，反而无法享受生活之美。你开始自以为洞悉一切，而我认为人生最大的乐趣在于对周遭保持好奇，不断被惊喜，在大小事物中发现美。自我越膨胀，就越难...

And the other way, if you put too much value into those kinds of prizes and you win them, I've gotten a chance to see that it also the more, quote, unquote, successful you are in that sense, the more you run the danger of growing ego so big that you don't get to actually enjoy the beauty of this life. You start to believe that you figured it all out as opposed to I think what the ultimately the most fun thing is is being curious about everything around you, being constantly surprised, and these little moments of discovery of enjoying enjoying beauty in small and big ways all around you. And I think the bigger your ego grows, the more you start to take yourself seriously, the less you're able

享受这些。天啊，我完全赞同这点。

to enjoy that. Oh, man. To that, I couldn't agree more.

就像《银河系漫游指南》里从'无害'升级到'基本无害'的评语，你会如何概括地球？如果要用几句话总结整个存在，或许还要加入生命意义——为什么？人类这种执着追问终极意义的特质，是否正是我们的定义特征？我在想这该不该成为描述的一部分：这些生物似乎非常迷茫。

So, you know, the the summary from harmless to mostly harmless in Hitchhiker's Guide to the Galaxy, how would you try to summarize Earth? And, you know, if you were given if you had to summarize the whole thing in in in a couple of sentences, and maybe throwing meaning of life in there, like, what why why why? Maybe is that a defining thing about humans that we care about the meaning of the whole thing? I wonder I wonder if that should be part of the the the these creatures seem to be very lost.

没错。永远在追问为什么。'为什么'就是我的核心问题。有次登山事故在我额头留下小疤，当时同伴碰落岩石时喊了声'注意下面'，意思是石头要砸下来了。

Yes. That's always asking why. Mean, that's my defining question is why. It was as used to make a joke, I have a small scar on my forehead from a climbing accident years ago, and the guy I was climbing with had dislodged a rock and he shouted something. He shouted below, think meaning that the rock was coming down.

我没听清他说的话，于是抬头看，结果那东西直接砸在我额头上，周围的人都嘲笑我说他抬头是为了问为什么。

I hadn't caught what he said so I looked up and then it smashed straight on my forehead and and everybody around me took the piss saying he he looked up to ask why.

是啊。但这是人类的本能。这是作为人类的一部分意义所在。仰望天空，追问为什么，不断追问。所以

Yeah. But that that's a human imperative. That's part of what it means to be human. Look up to the sky and ask why and ask why. So

你的问题是定义地球。我不确定我能做到。我想到的第一个词是生命。我不愿说主要是生命，但或许

your question, define the earth. I'm not sure I can do that. I mean, the first word that comes to mind is living. I wouldn't like to say mostly living, but perhaps

主要是生命？这很有趣，因为如果你要写《银河系漫游指南》，我想说我们讨论过的观点，细菌是银河系和宇宙中最主要的生命形式。那么地球会显得独特，需要

Mostly living? Well, it's interesting because like, if you were to to write The Hitchhiker's Guide to the Galaxy, I suppose say our idea that we talked about, the bacteria is the most prominent form of life throughout the galaxy and the universe. I suppose that Earth would be kind of unique and would require

在这种情况下显得丰饶。是的。它奢侈富足。它充满无限生机。

abundance in that case. Yeah. It's profligate. It's rich. It's enormously enormously living.

那么你会如何描述它不是细菌？它是

So how how would you describe that it's not bacteria? It's

真核生物。对。嗯，我是说，这是专业术语，但本质上它就是

Eukaryotic. Yeah. Well, I mean, that's that's the technical term, but it is basically it's

是啊。然后

Yeah. And then

实际上我对'真核生物'这个术语感到非常困扰，因为这个词——我是说，几乎没有比'真核生物'更能立刻让人望而却步的词了。你一开始用这种词，人们就会直接离开房间。'克雷布斯循环'是另一个让人拔腿就跑的词。所以我一直在想，有没有其他词可以替代'真核生物'？实际上我能想到的只有'复杂细胞'、'复杂生命'这类表述。

I've I've I've actually really struggled with that term because the word I mean, there's few words quite as good as eukaryotic to put everybody off immediately. You start using words like that and they'll they'll leave the room. Krebs cycle is another one that gets people to leave the room. So I've tried to think, is there another word for eukaryotic that I can use? And really the only word that I've been able to use is complex cells, complex life and so on.

这个词有个立竿见影的作用——传递某种印象。但它对每个人意味着太多不同含义，以至于实际意义会立刻消失。所以这有点像...

That word, serves one immediate purpose which is to convey an impression. But then it means so many different things to everybody that actually is lost immediately. So it's a it's a kind of

从其他星球的视角来看这根本不明显。真核生物的出现才是复杂性发生显著相变的标志。那么'无害'和'基本无害'呢？这算是...

Well, that's unnoticeable from the perspective of other planets. That is a noticeable phase transition of complexity is the eukaryotic. What about the harmless and the mostly harmless? Is that kind of

从宇宙尺度来看可能还算准确。我认为现阶段人类根本不可能对宇宙造成什么干扰。

Probably accurate on a on a universal kind of scale. I don't think that humanity is in any danger of disturbing the universe at the moment.

现阶段是这样，所以我们大多不知道。这得看马斯克在搞什么名堂。取决于火箭发射数量。我觉得...

At the moment, which is why the mostly we don't know. Depends what Elon is up to. Depends how many rockets. I think

即便如此，我认为距离我们扰动时空结构还需要相当长的时间。

It'll it'll be still even then a while, I think, before before we disturb the fabric of time and space.

我想是之前提到的Andrej Karpathy，他将地球概括为一个系统，你向它倾注大量光子作为输入，输出则是火箭。所以

Was the aforementioned Andrej Karpathy, I think he summarized Earth as as a system where you hammer it with a bunch of photons. The input is like photons and the output is rockets. So

如果你只是... 天啊，那在火箭发射之前需要的光子数量简直多得惊人。

if you just Well, that's a hell of a lot of photons before there was a rocket launch.

是啊。但要知道，或许在宇宙的时间尺度上，这段时间并不算长。我确实好奇未来会怎样，我们究竟是处于地球这个重要项目的早期阶段——当你试图总结它时这点很关键，还是已临近终点，人类终于获得了摧毁这个美丽项目的能力，通过核武器、人造病毒等手段。

Yeah. But it's like, know, maybe maybe in the span of the universe, it's not it's not that much time. And so and I do wonder, you know, what the future is, whether we're just in the early beginnings of this earth, which is important when you try to summarize it, or we're at the end where humans have finally gained the ability to destroy the entirety of this beautiful project we've got going on. Now with nuclear weapons, with engineered viruses, with all those kinds of things.

或者仅仅是无意间通过全球变暖、污染等方式。我们完全有这个能力。我是说，我们只需要

Or just inadvertently through global warming and pollution and so on. We're quite capable of it. I mean, we just need

越过那个临界点

to I pass the tipping

我认为我们更可能无意间造成毁灭，而非通过随时可能爆发的核战争。但我担心的是我们根本不知道临界点在哪里，还自认为足够聪明能在危机真正来临时迅速解决问题。这种‘目前一切尚好’的盲目乐观才是主流观念。也许二十年后问题会灾难性爆发，那时我们才需要集中智慧解决，却可能为时已晚——无论我们多聪明，都已错过时机。

mean, I think we're more likely to do it inadvertently than than than through a nuclear war, which could happen at any time. But my fear is we just don't know where the tipping points are and we will we we we kind of think we're smart enough to fix the problem quickly if we really need to. I think that's the the overriding assumption that we're all right for now. Maybe in twenty years' time, it's going to be a calamitous problem and then we'll really need to put some serious mental power into fixing it without seriously worrying that perhaps that is too late and that however brilliant we are, it's we we missed the boat.

就这样走向悬崖。我不知道。我对人类作为智慧后裔还是抱有乐观的。

And just walk off the cliff. I don't know. I have optimism in humans being clever descendants.

哦，我毫不怀疑我们能解决这个问题，这是个紧迫的问题。我们需要尽快解决它。但我确实怀疑，在政治上我们是否有能力团结起来，不仅是在一个国家，而是在全球范围内——我是说，我知道我们能做到，但我们有这种意愿吗？我们有实现它所需的远见吗？

Oh, I have I have no doubt that we can fix the problem, It's an urgent problem. We need to fix it pretty sharpish. And I do have doubts about whether politically we are capable of coming together enough to not just in any one country, but around the planet to to I mean, I know we can do it, but do we have the will? Do we have the the the vision to to accomplish it?

这就是让整个旅程充满乐趣的原因。我知道。我们不仅不知道能否应对眼前的危机，甚至不知道未来二十年将面临的所有危机。我认为二十一世纪最可能挑战我们的，正是那些我们根本预料不到的事。一战结束时，人们也没预料到会有二战。

That's what makes this whole ride fun. I know. Not only do we not know if we can handle the crises before us, we don't even know all the crises that are gonna be before us in the next twenty years. The the the ones I think that will most likely challenge us in the twenty first century are the ones we don't even expect. People didn't expect World War two, at the end of World War one.

但确实。不是在一战结束时，而是到了二十世纪二十年代末，我想人们开始担心这个了。

But yeah. Not not at the end of World War one, but by the nineteen late nineteen twenties, I think people were beginning to worry about it.

是啊。不。总有人对一切都忧心忡忡。没错。所以如果你专注于某件事

Yeah. No. There there's always people worrying about everything. Yeah. So if you focus on the thing

人们会为此担心。是的。

that People worry about it. Yes.

因为人们担心的事情有上百万件，其中99.99999%都不会发生。当然，那些最终被证明是对的的人会说'我早就知道'，但这并不是...你知道...这不是准确预测未来的方式。理性地说，你可以担心，但没人想到会再爆发世界大战。那场'终结所有战争的战争'。

Because there's a million things people worry about it and 99.99999% of them don't come to be. Of course, the people that turn out to be right, they'll say, I knew all along, but that's not, you know, that's not an accurate way of knowing what Sure. You could have predicted. I think rationally speaking, you can worry about it, but nobody thought you could have another world war. The war to end all wars.

怎么会再有战争呢？从技术上讲，核武器的概念本身就很难预料——创造出一种能让破坏力跃升数个数量级的武器。当然，我们能预感到所有这些东西：工程病毒、纳米机器人、人工智能，还有全球变暖带来的各种复杂全球影响。这会如何改变资源分配、能源流动、国家间紧张关系、军事冲突、权力重组，再看中国在这整盘棋中的角色——与俄罗斯的关系、非洲日益增长的影响力、欧洲的诡异动态，还有美国因推特和脸书的推荐算法加剧政治分裂而分崩离析。这整个美丽的混乱本身就很有趣，而且我认为有很多杰出的工程师、科学家和优秀人才——当大家周末在网上争吵取乐时，他们真的在努力构建解决方案，这些人将会创造出美好的事物。

Why would you have another war? And the idea of nuclear weapons just technologically, is a very difficult thing to anticipate to create a weapon that just jumps orders of magnitude and destructive capability. And, of course, we can intuit all the things like engineered viruses, nanobots, artificial intelligence, yes, all the different complicated global effects of global warming. So how that changes the allocation of resources, the flow of energy, the tension between countries, the military conflict between countries, the reallocation of power, then looking at the role of China in this whole thing with with Russia and growing influence of Africa and the weird dynamics of Europe, and then America falling apart to the political division fueled by recommender systems through Twitter and Facebook. The whole beautiful mess is just fun, and I think there's a lot of incredible engineers, incredible scientists, incredible human beings that while everyone is bickering and so on online for the fun of it on the weekends, they're actually trying to build solutions, and those are the people that will create something beautiful.

至少我知道有你，这就是进化的过程。一切都始于一个查克·诺里斯式的单细胞生物，它从热泉口诞生，成为我们所有人的祖先。对于那位先生、女士或两者兼具的祖先，我要致以深深的谢意，并迫不及待想看看接下来会发生什么。我很高兴有像你尼克这样杰出的人类在记录和研究这一切。能与你交谈是我莫大的荣幸。

At least I have you know, that's the process of evolution. It's there was it all started with a Chuck Norris single cell organism that went out from the vents and was the parent to all of us. And for for that guy or lady or both, I guess, is a big thank you, and I can't wait to what happens next. And I'm glad there's incredible humans writing and studying it like you are, Nick. It's a huge honor that you would talk to me.

这真是太棒了。

This has been fantastic.

这确实令人惊叹。我迫不及待想读到你接下来的著作。感谢你的存在，也感谢你和丹交谈。谢谢。感谢大家收听与尼克·莱恩的这场对话。

This is really amazing. I can't wait to read what you write next. Thank you for existing, and thank you for talking to Dan. Thank you. Thanks for listening to this conversation with Nick Lane.

如需支持本播客，请查看简介中的赞助商信息。现在请允许我用史蒂夫·乔布斯的话作为结束：'我认为二十一世纪最伟大的创新将出现在生物学与技术的交叉领域。一个新时代正在开启。'感谢您的收听。

To support this podcast, please check out our sponsors in the description. And now let me leave you with some words from Steve Jobs. I think the biggest innovations of the twenty first century will be at the intersection of biology and technology. A new era is beginning. Thank you for listening.

希望下次还能见到你。

I hope to see you next time.