#376 – 斯蒂芬·沃尔夫勒姆：ChatGPT与真理、现实及计算的本质

嗯，我是说，这是个重要且有趣的问题。我是说，就我们中有些人致力于尝试创新和发现新事物等而言，这是一种复杂的相互作用，一边是在某个角落做着疯狂事情的个体，可以这么说，另一边是试图维持高惯性平均状态的集体。有时集体会涌现出有趣的东西，有时又会压制那些试图朝创新方向发展的尝试。

Well, I mean, that's an important that's an interesting question. I mean, Insofar as some of us work on trying to innovate and figure out new things and so on, it's a complicated interplay between the individual doing the crazy thing off in some spur, so to speak, versus the collective that's trying to do the high inertia average thing. Sometimes the collective is bubbling up things that are interesting, and sometimes it's pulling down kind of the attempt to make this kind of innovative direction.

但你不认为大型语言模型会超越这种简化，可能会说智力和职业多样性真的很重要吗？所以你需要那些处于边缘的疯狂的人。因此，就像，整个事情的真正目的是通过我们人类文明一直在使用的这种动态来探索，即我们大多数人专注于一件事，然后有那些在边缘做着与之相反事情的疯狂的人，你有点，像是，把整个社会拉在一起。有主流科学，也有疯狂的科学。这就是人类文明的历史。

But don't you think the large language models would see beyond that simplification will say maybe intellectual and career diversity is really important? So you need the crazy people on the outlier, on the outskirts. And so, like, the actual what's the purpose of this whole thing is to explore through this kind of dynamics that we've been using as a human civilization, which is most of us focus on one thing, and then there's the crazy people on the outskirts doing the opposite of that one thing, and you kinda, like, pull the whole society together. There's the mainstream science, and then there's the crazy science. That's just been the the history of human civilization.

也许人工智能系统将能够察觉到这一点。我们对语言模型告诉我们这些事越是印象深刻，就越会赋予它更多控制权，越愿意让它来管理我们的社会。因此，这就形成了一种循环，社会可能被操纵从而允许

And maybe the AI system will be able to see that. And the more and more impressed we are by a language model telling us this, the more control we'll give it to it, and the more we'll be willing to let it run our society. And hence, there's this kind of loop where the society could be manipulated to let

由人工智能系统来运行它。

the AI system run it.

没错。我是说，看吧，有趣的是我们总以为自己一直在进步。

Right. Well, I mean, look, one of the things that's sort of interesting is we might say we always think we're making progress.

嗯。

Mhmm.

但从某种意义上说，如果我们主张以既存事物为模板来定义未来应存之物，这其实很有趣。比如许多宗教都持这种观点——某部在X时代写就的圣典，规定了人们未来所有时代的行为准则。这正是人类长期遵循的模式。某种程度上，我们现在讨论的正是这类命题的变体。

But yet, if, you know, in a sense, by by saying, let's take what already exists and use that as a model for what should exist. Then, you know, it's interesting that, for example, many religions have taken that point of view. There is a sacred book that got written at time X, and it defines how people should act for all future time. And it's a model people have operated with. And in a sense, this is a version of that kind of statement.

就好比拿2023年版本的世界呈现方式作为模板，来定义未来世界该如何运作。

It's like take the 2023 version sort of how the world has exposed itself and use that to define what the world should do in the future.

但这定义并不精确，对吧？就像宗教典籍和GPT一样，人类对GPT输出的解读将成为系统中的扰动因素，成为充满不确定性的噪声源。

But it's not it's an imprecise definition. Right? Because just like with religious texts and with GPT, the human interpretation of what GPT says will be the will be the perturbation in the system. It'll be the noise. It'd be full of uncertainty.

GPT Chad GPT并不会直接告诉你具体该怎么做。它会给出一个大概的叙事方向，就像'转过另一边脸'那种寓言式的建议，对吧？这并非完整的指导性叙事。

It's not like GPT Chad GPT will tell you exactly what to do. It'll tell you approx a narrative of what like, a a you know, it's like a turn the other cheek kind of narrative. Right? That's that's not a fully instructive narrative.

嗯，除非等到AI控制了世界上所有系统。

Well, until until the AIs control all the systems in the world.

那时它们就能精确地告诉你每一步该做什么了。

They will be able to very precisely tell you what to do.

它们会...你知道...它们会自行决定做这个或那个。不仅如此，它们还会自动向每个人建议'接下来做这个，接下来做那个'。所以我认为这比通常情况更具规范性。但人类还剩下什么自主空间？我们多大程度上...这种认为互联网内容定义了人类现存目的的想法很重要。可当我们探索计算宇宙的可能性时，面对所有可能的发明创造，问题在于我们该选择追寻哪些方向？

Well, they'll do what they you know, they'll they'll just do this or that thing. And and not only that, they'll be auto suggesting to each person, you know, do this next, do that next. So I think it's slightly more prescriptive situation than one has typically seen. But I think this whole question of sort of what's left for the human, so to speak, to what extent do we This idea that there is an existing corpus of purpose for humans defined by what's on the Internet and so on, that's an important thing. But then the question of as we explore what we can think of as the computational universe, as we explore all these different possibilities for what we could do, all these different inventions we could make, all these different things, the question is which ones do we choose to follow?

从某种意义上说，如果人类还想保持进步，这些选择权必须掌握在我们手中。换句话说，如果以现状作为未来发展的唯一决定因素，那么人类的机遇就在于：未来会涌现无数可能性。只要我们想参与决策，最合理的参与方式就是选择我们想要的未来图景。

Those choices are the things that, in a sense, if the humans want to still have human progress, we get to make those choices, so to speak. In other words, there's this idea, if you say, Let's take what exists today and use that as the determiner of all of what there is in the future, the thing that is the opportunity for humans is there will be many possibilities thrown up. There are many different things that could happen or be done. Insofar as we want to be in the loop, the thing that makes sense for us to be in the loop doing is picking which of those possibilities we want.

但'我们正在自主选择'这个反馈循环的可信度会越来越成问题，因为我们正被各种系统所影响。确实如此。如果这些系统越来越成为我们教育和智慧的来源...

But the degree to which there's a feedback loop of the idea that we're picking something starts becoming questionable because we're influenced by the various systems. Absolutely. That like, if that becomes more and more source of our education and wisdom and Right.

AI就会接管一切。我长期以来的观点是：真正让AI取得主导权的是它们的自动建议功能，人类最终只会盲目跟随。

The AIs take over. I mean, my know, I've thought for a long time that, you know, it's the, you know, AR auto suggestion that's really the thing that makes the AIs take over. It's just that the humans just follow, you know.

是的。我们以后不再互相写邮件了。我们只需发送自动建议的邮件。

Yeah. We will no longer write emails to each other. We'll just send the auto suggested email.

对，对。但人类可能参与其中的环节在于存在选择的时候。而我们可以基于我们整个历史网络等因素做出选择。就这一点而言，如果一切都被决定了，人类就没有存在的空间了。

Yeah. Yeah. But the the thing where humans are potentially in the loop is when there's a choice. And when there's a choice which we could make based on our whole web of history and so on. And that's insofar as it's all just determined the humans don't have a place.

顺便说一句，在某种程度上，这都算是个复杂的哲学问题，因为从某种层面看，宇宙只是在做它该做的事。我们作为宇宙的一部分，必然也在做我们该做的事，可以这么说。然而，我们却感觉自己对所做的事有某种能动性，这本身就是一个有趣的问题。

By the way, at some level, it's all kind of a complicated philosophical issue because at some level, the universe is just doing what it does. We are parts of that universe that are necessarily doing what we do, so to speak. Yet, we feel we have sort of agency in what we're doing, and that's that's its own separate kind of interesting issue.

而且我们多少觉得自己是宇宙创造的终极目标，但我们很可能——甚至显然就是某种中间阶段。没错，我们绝对是某种中间阶段。问题在于是否会出现更酷、更复杂、更有趣的事物。

And we also kinda feel like we're the final destination of what the universe was meant to create, but we very well could be and likely are some kind of intermediate step, obviously. Yeah. What would we're most certainly some intermediate step. The question is if there's some cooler, more complex, more interesting things that's going to materialize.

计算宇宙中充满了这类事物。

The computational universe is full of such things.

但具体到我们这个小角落。如果这就是我们能做到的最好程度，或者不是，这有点...

But in in our particular pocket, specifically. If this is the best we're gonna do or not, that's kind of a

我们可以在计算宇宙中创造各种有趣的东西。当我们观察它们时，我们会说，没错，你知道，那就是个东西。但它与我们当前的思维方式并不真正契合。就像数学领域，我们有特定的定理。

We can make all kinds of interesting things in the computational universe. We when we look at them, we say, yeah, you know, that's that's a thing. We don't it doesn't really connect with our current way of thinking about things. It's like in mathematics. We've got certain theorems.

人类数学家已经写下并发表了大约三四百万条数学定理。但可能的数学定理数量是无限的。我们只是进入可能定理的宇宙中挑选另一个定理，然后人们会说，嗯，你知道，他们看着它说，我不明白这个定理的含义。它与我们正在处理的历史脉络中的事物没有关联。我认为关于理解人工智能及其与我们关系的一个要点是，当我们拥有这一整套AI基础设施，它们以自己的方式运作，而这些方式或许我们人类难以直接理解时，你可能会说这是一种非常奇特的情况。

There are about three or 4,000,000 that human mathematicians have written down and published and so on. But there are an infinite number of possible mathematical theorems. We just go out into the universe of possible theorems and pick another theorem, and then people will say, Well, you know, that's they look at it and say, I don't know what this theorem means. It's not connected to the things that are part of the web of history that we're dealing with. I think one point to make about understanding AI and its relationship to us is as we have this whole infrastructure of AIs doing their thing and doing their thing in a way that is perhaps not readily understandable by us humans, you might say that's a very weird situation.

我们怎么能建造出这种以我们无法理解的方式运行的东西，充满了计算不可约性等等？这是什么？当世界由我们无法理解其运作的AI运行时，会是什么感觉？而人们意识到的是，实际上我们以前见过这种情况。那就是我们存在于自然世界时所发生的事。

How can we have built this thing that behaves in a way that we can't understand, that's full of computational irreducibility, etcetera, etcetera, etcetera? What is this? What's it going to feel like when the world is run by AIs whose operations we can't understand? And the thing one realizes is, actually, we've seen this before. That's what happens when we exist in the natural world.

自然界充满了按照明确规则运作的事物。它们具有各种计算不可约性。我们并不理解自然界在做什么。偶尔当你问，AI会消灭我们吗？比如说，嗯，这有点像问，AI的机制是否会导致最终出现毁灭物种的事物？

The natural world is full of things that operate according to definite rules. They have all kinds of computational irreducibility. We don't understand what the natural world is doing. Occasionally, when you say, Are the AIs going to wipe us out? For example, well, it's kind of like, Is the machination of the AIs going to lead to this thing that eventually comes and destroys the species?

嗯，我们也可以对自然界提出同样的问题。自然界的机制最终会导致地球爆炸之类的事情吗？这些都是问题。就我们认为自己理解自然界发生的事而言，那是自然科学等的成果。当存在这个庞大的AI基础设施时，我们可以期待的一件事是，我们必须发明一种新的自然科学，即向我们解释AI如何运作的自然科学。

Well, we can also ask the same thing about the natural world. Are the machination of the natural world going to eventually lead to this thing that's going to, you know, make the Earth explode or something like this? Those are questions. And insofar as we think we understand what's happening in the natural world, that's a result of science natural science and so on. One of the things we can expect when there's this giant infrastructure of the AIs is that's where we have to invent a new kind of natural science that is the natural science that explains to us how the AIs work.

这有点像我们有一匹马什么的，我们试图骑马从这里到那里。我们并不真正理解马内部如何运作，但我们可以采用某些规则和方法说服马带我们去那里。这就是我们与那些难以理解、计算不可约的AI打交道时的同类情况。但我们可以识别这些——我们可以找到这些可约性的小口袋，就像抓住马的鬃毛之类的东西来骑它。我们明白，你知道，如果这样做或那样做来骑马，那就是让它做我们感兴趣的事情的成功方法。

It's kind of like we have a horse or something, and we're trying to ride the horse and go from here to there. We don't really understand how the horse works inside, but we can get certain rules and certain approaches that we take to persuade the horse to go from here to there and take us there. And that's the same type of thing that we're dealing with, with the incomprehensible, computationally irreducible AIs. But we can identify these kinds ofwe can find these kinds of pockets of reducibility that we can kind of, you know, don't know, we're grabbing onto the mane of the horse or something to be able to ride it. We figure out, you know, if do this or that to ride the horse, that that's a a successful way to to get it to do what what we're interested in doing.

马与大型语言模型或可以称为连接到互联网的AGI之间似乎确实存在差异。所以让我问你一个关于这些事物威胁的大哲学问题。有很多像埃利亚斯或埃德科夫斯基这样的人担心AI系统的存在风险。这是你担心的事情吗？你知道，有时候当你构建像Wolfram Alpha这样不可思议的系统时，你可能会有点迷失其中。

There does seem to be a difference between a horse and a large language model, or something that could be called AGI connected to the Internet. So let me just ask you about big philosophical question about the threats of these things. There's a lot of people like Elias or Edkowski who worry about the existential risks of AI systems. Is that something that you worry about? You know, sometimes when you're building an incredible system like Wolfram Alpha, you can kinda get lost in it.

哦，我试着思考一下自己所做事情的潜在影响。

Oh, I try and think a little bit about the implications of what one's doing.

你知道，这有点像曼哈顿计划那种情况，就是有些最不可思议的物理和工程正在被实现，但问题是，这一切最终会导向何方？

You know, it's like the Manhattan Project kinda situation where you're, like, it's some of the most incredible physics and engineering being done, but it's like, where is this gonna go?

我认为关于‘总会有一个更聪明的AI’这类论点——最终AI会超越人类，然后各种可怕的事情就会发生——对我来说，这些论点让我联想到关于上帝存在的本体论论证之类的东西。这些论证往往基于某种特定模型，通常是相当简化的模型，认为总会存在更高级的这个那个。但现实中这些事物发展的实际情况往往比你预期的更复杂，那种简单的逻辑论证说‘最终会出现超级智能，然后它会做这做那’，结果发现并非真实的故事。

I think some of these arguments about kind of, you know, there'll always be a smarter AI. There'll always be, you know, and eventually the AIs will get smarter than us, and then all sorts of terrible things will happen. To me, some of those arguments remind me of kind of the ontological arguments for the existence of God and things like this. They're kind of arguments that are based on some particular model, fairly simple model often, of kind of there is always a greater this, that, and the other. Those arguments tendwhat tends to happen in the sort of reality of how these things develop is that it's more complicated than you expect, that the kind of simple logical argument that says, Oh, eventually there'll be a superintelligence and then it will do this and that, turns out not to really be the story.

事实证明故事要复杂得多。举个例子：是否存在一个顶点智能？就像某个生态系统中可能存在顶级掠食者一样，是否会出现一个可能存在的最高智能？我认为答案是否定的。事实上我们已经知道这一点，这又回到了整个计算不可约性的故事。

It turns out to be a more complicated story. So for example, here's an example of an issue: is there an apex intelligence? Just like there might be an apex predator in some ecosystem, Is there going to be an apex intelligence, the most intelligent thing that there could possibly be? I think the answer is no. And in fact, we already know this and it's a kind of a back to the whole computational irreducibility story.

甚至存在这样一个问题：如果你有一台图灵机，一台在停止前尽可能长时间运行的图灵机，你会说‘这就是那台能做到这点的顶点机器吗？’但总会存在能运行更久的机器。当你考察所有可能的图灵机无限集合时，可以说你永远无法到达终点。这有点像‘是否总会存在另一项发明’的问题——你总能再发明出新东西吗？

There's kind of a question of even if you have a Turing machine and you have a Turing machine that runs as long as possible before it halts, you say, Is this the machine, is this the apex machine that does that? There will always be a machine that can go longer. And as you go out to the infinite collection of possible Turing machines, you'll never have reached the end, so to speak. You'll always be able to It's kind of like the same question of whether there will always be another invention. Will you always be able to invent another thing?

答案是肯定的。存在一个可能发明的无限高塔。这就是‘顶点’的一种定义。

The answer is yes. There's an infinite tower of possible inventions. That's one definition of apex.

但另一个问题——我也以为你指的是这个，我也认为可能是真的——是否存在一个物种是

But the the other is like which I also thought you were which I also think might be true, is is there a species that's

当前地球上的顶点智能？所以不能轻率地说人类就是那个顶点智能。

the apex intelligence right now on Earth? So it's not trivial to say that humans are that.

是的，这并不简单。我同意。这就像，我认为长久以来我一直对其他类型的智能感到好奇。在我看来，智能就像计算，有一套规则，你从中推导出结果。

Yeah. It's not trivial. I agree. It's it's a you know, I think one of the things that I've long been curious about kind of other intelligences, so to speak. Mean, I view intelligence as like computation, and it's kind of you have the set of rules, you deduce what happens.

我现在倾向于认为，存在一种专门化的计算形式，类似于意识，与计算局限性、单一体验线程等有关，这种计算专门化对应着某种类似人类的世界体验。问题是：可能存在其他智能，比如谚语说的‘天气自有其想法’。这是一种不同的智能，能计算许多我们难以计算的事物，但与我们的思维方式不吻合。它不像我们那样思考。在这种不同智能的理念中，每个不同的心智，每个不同的人类心智，都是思考方式各异的智能。

I have tended to think now that there's this specialization of computation that is a consciousness like thing that has to do with these computational boundedness, single thread of experience, these kinds of things, that are the specialization of computation that corresponds to a somewhat human like experience of the world. Now the question is: there may be other intelligences like the aphorism, The weather has a mind of its own. It's a different kind of intelligence that can compute all kinds of things that are hard for us to compute, but it is not well aligned with the way that we think about things. It doesn't think the way we think about things. And in this idea of different intelligences, every different mind, every different human mind is a different intelligence that thinks about things in different ways.

在我们的物理项目形式化中，我们讨论过‘规则空间’的概念，即所有可能规则系统的空间。不同心智在某种意义上位于规则空间的不同位置。成长于相似文化和观念的人类心智，在规则空间中可能非常接近，易于沟通、翻译，易于从一个心智对应的规则空间位置移动到另一个邻近心智的位置。当我们处理规则空间中更遥远的事物，比如宠物猫。猫与我们共享某些特征。

In terms of the formalism of our physics project, we talk about this idea of rudial space, the space of all possible rule systems. Different minds are, in a sense, at different points in rural space. Human minds, ones that have grown up with the same kind of culture and ideas and things like this, might be pretty close in rural space, pretty easy for them to communicate, pretty easy to translate, pretty easy to move from one place in rural space that corresponds to one mind to another place in rural space that corresponds to another nearby mind. When we deal with more distant things in rural space, like the pet cat or something. The pet cat has some aspects that are shared with us.

猫的情感反应与我们有些相似，但猫在规则空间中的位置比人类更远。那么问题是，我们能否将我们的思维过程翻译成猫的思维过程？当我们做到这一点时，会得到什么？我认为许多动物——比如狗——拥有精细的嗅觉系统，它们以我们无法理解的方式构建世界的嗅觉架构。如果你能与狗交谈，它会描述流动的气味概念，这些是我们完全无法想象的。

The emotional responses of the cat are somewhat similar to ours, but the cat is further away in rural space than people are. So then the question is, can we identify can we make a translation from our thought processes to the thought processes of a cat or something like this? What will we get when we what will happen when we get there? And I think it's the case that many animals I don't know, dogs, for example they have elaborate olfactory systems, they have sort of the smell architecture of the world, so to speak, in a way that we don't. So if you were talking to the dog and you could communicate in a language, the dog will say, Well, this is a flowing, smelling this, that, and the other thing concepts that we just don't have any idea about.

有趣的是，有朝一日我们将拥有性能优异的化学传感器，人工鼻子会非常精准，增强现实系统可能会向我们展示狗能看到的‘气味地图’，类似于狗大脑中的景象。最终，我们将在规则空间中拓展到拥有与狗相同的感官体验，内化‘气味景观’的意义。那时我们就殖民了规则空间的这一部分。但在此之前，我们对动物行为的某些方面已成功理解，另一些则尚未掌握。

Now, what's interesting about that is one day we will have chemical sensors that do a really pretty good job. We'll have artificial noses that work pretty well, and we might have our augmented reality system show us the same map that the dog could see and things like this, similar to what happens in the dog's brain. Eventually, we will have kind of expanded in rural space to the point where we will have those same sensory experiences that dogs have and we will have internalized what it means to have the smell landscape or whatever. And so then we will have colonized that part of rural space. Until we haven't gone, some things that animals and so on do, we've successfully understand, others we do not.

如何将动物的思维转化为我们能理解的形式，这个问题并不简单。我长期对此好奇，甚至曾有个古怪项目：试图开发一款能让猫战胜主人的iPad游戏。

And the question of what representation, how do we convert things that animals think about to things that we can think about, that's not a trivial thing. And I've long been curious. I had a very bizarre project at one point of trying to make an iPad game that a cat could win against its owner.

听起来这背后似乎有个深刻的哲学目标。

So it feels like there's a deep philosophical goal there, though.

是的，是的。我是说，那个，你知道，我很好奇，如果宠物能在《我的世界》里工作之类的，嗯哼。它们能建造东西的话，会建造什么？它们建造的东西会不会让我们一看就说，哦，对。

Yes. Yes. I mean, the the you know, I was curious if, you know, if pets can work in Minecraft or something Mhmm. And can construct things, what will they construct? And will what they construct be something where we look at it and we say, oh, yeah.

我认得那个。或者它会不会看起来像是存在于计算宇宙中的某种东西，就像我的某个细胞自动机可能生成的那样，我们会说，嗯，我大概能看出它是按照某些规则运作的。我不知道为什么要用那些规则，也不知道为什么要在乎。

I recognize that. Or will it be something that looks to us like something that's out there in the computational universe that one of my, you know, cellular automata might have produced, where we say, yeah, I can kind of see it operates according to some rules. I don't know why you would use those rules. I don't know why you would care.

是啊。其实，认真地说，农村地区有没有一种连接器，能让你和猫之间建立联系，让猫能真正获胜？iPad的界面非常有限。不，我好奇是否存在猫能赢的游戏。

Yeah. Actually, just to link on that seriously, is there a connector in the rural space between you and a cat where the cat could legitimately win? So iPad is a very limited interface. No. I I wonder if there's a game where cats win.

我觉得问题在于猫通常对iPad上发生的事情不太感兴趣。

I think the problem is cats don't tend to be that interested in what's happening on the iPad.

明白了。嗯，这大概是个界面问题。

I see. Yeah. That's an interface issue probably.

对，对，对。不。

Yeah. Right. Right. Right. No.

我、我、我觉得很可能，你知道，有很多动物如果暴露在它们面前，会成功吃掉我们，所以，你知道，它们扑过来的速度比我们躲开还快，诸如此类。所以有很多——而且很可能，我们以为自己藏好了，其实并没有成功隐藏。

I I I think it is likely that I mean, you know, there are plenty of animals that would successfully eat us if we were exposed to them, and so there's, you know, it's gonna pounce faster than we can get out of the way, and so on. So there are plenty of and probably it's going to, you know, we think we've hidden ourselves, but we haven't successfully hidden ourselves.

但那只是体能上的优势。我在想是否存在智力层面的某些方面，像猫这样的动物能超越人类。

But that's a physical strength. I wonder if there's something more in the realm of intelligence, where an animal like a cat could out.

我认为在某些信息处理速度方面，动物确实有优势。比如国际象棋——是否存在猫类象棋？如果猫之间对弈，而我们尝试与猫对局，是否会永远输给它们？这我不确定。

Well, I think there are things certainly in terms of the speed of processing certain kinds of things, for sure. I mean, the the question of what you know, is there a game of chess, for example? Is there cat chess that the cats could play against each other, and if we tried to play a cat, we'd always lose. I don't know.

可能关乎反应速度，但也可能涉及概念理解。猫的思维中或许存在某些我们无法理解的概念体系。

It might have to do with speed, but it might have to do with concepts also. There might be concepts in the cat's head.

我认为自从人类发明语言后，就构建起了抽象思维的高塔，这使得我们在象棋这类游戏中占据优势。换句话说，通过语言经验和抽象学习，我们在抽象领域变得更为聪明。但这不意味着我们比猫更擅长捕鼠——我们只是在自我选择的抽象事务上更聪明。这又回到了'个体关注点'的问题。

I I tend to think that our species, from its invention of language, has managed to build up this tower of abstraction that, for things like a chess like game, will make us win. In other words, we've become, through the fact that we've kind of experienced language and learnt abstraction, we've sort of become smarter at those kinds of abstract kinds of things. Now, that doesn't make us smarter at catching a mouse or something. It makes us smarter at the things that we've chosen to sort of, you know, to concern ourselves, which are these kind of abstract things. And think this is again back to the question of what does one care about?

假设你能与猫对话——如果我们能翻译猫语——猫会说'我对这个移动的光点超级兴奋'，然后反问'你为啥在意这个？'对猫而言，那个移动的光点就是全世界最重要的事。就像我们看古代文物时，那些先民深信不疑的信仰体系，对他们就是至高无上的存在，而我们如今却难以理解其意义。

One's the cat if you have the discussion with a cat, if we can translate things to have the discussion with a cat, the cat will say, I'm very excited that this light is moving, and will say, why do you care? And the cat will say, that's the most important thing in the world, that this thing moves around. I mean, it's like when you ask about, I don't know, you look at archaeological remains, and you say, these people had this belief system about this, and that was the most important thing in the world to them. Now we look at it and say, we don't know what the point of it was. I've been curious.

比如那些两万多年前的洞穴手印，没人知道它们的真正意义。它们可能是古人心中最神圣的象征，也可能只是某个孩子把沾满泥巴的手按在洞壁上。当我们讨论'最聪明的存在'时，本质上是在探讨'你试图进行何种类型的计算'。

There are these handprints on caves from 20,000 or more years ago, It's like nobody knows what these handprints were there for, you know? That they may have been a representation of the most important thing you can imagine. They may just have been some, you know, some kid who rubbed their hands in the mud and stuck them on the walls of the cave. We don't know. But this whole question of what is when you say this question of what's the smartest thing around, there's the question of what kind of computation you're trying to do.

假设有个明确定义的计算任务，实现方式可以多样：通过神经元放电、硅基电子元件、人体免疫系统的分子计算，或是其他分子生物学机制——存在无数种实现途径。

If say you've got some well defined computation, and how do you implement it? Well, you could implement it by nerve cells firing. You can implement it with silicon and electronics. You can implement it by some kind of molecular computation process in the human immune system or in some molecular biology kind of thing. There are different ways to implement it.

我认为这些不同实现方法的问题在于它们会有不同的速度，能够处理不同的事情。如果你问具体是哪些？那么一个有趣的问题将是：在这些不同类型的系统中，哪些抽象概念最为自然？例如，在我们所见的视觉场景中，我们会识别出某些物体，认出某些事物，而猫原则上可能会识别出不同的东西。我怀疑生物进化是非常缓慢的，而且我认为猫注意到的东西与我们非常相似。我们甚至可以从一些神经生理学研究中得知这一点。

I think this question of those different implementation methods will be of different speeds, they'll be able to do different things, if you say, Which? So an interesting question would be, what kinds of abstractions are most natural in these different kinds of systems? For example, the visual scene that we see, we pick out certain objects, we recognize certain things in that visual scene, a cat might in principle recognize different things. Suspect evolution, biological evolution, is very slow, and I suspect what a cat notices is very similar. We even know that from some neurophysiology.

猫注意到的东西与我们注意到的非常相似。当然，一个明显的区别是猫只有两种颜色感受器，因此它们看到的颜色与我们不同。现在，我们说我们更优越，因为我们有三种颜色感受器：红、绿、蓝。但我们并非全面胜出。我认为螳螂虾才是全面胜出者，它们有15种颜色感受器。

What a cat notices is very similar to what we notice. Of course, there's one obvious difference is cats have only two kinds of color receptors, so they don't see in the same kind of color that we do. Now, we say we're better we have three color receptors: red, green, blue. We're not the overall winner. I think the mantis shrimp is the overall winner with 15 color receptors, I think.

因此，螳螂虾对现实的看法，至少在颜色方面，比我们的要丰富得多。现在，有趣的是我们如何达到那种程度？想象一下，我们有一个增强现实系统，它甚至可以看到红外线和紫外线等，并将这些信息转化为可以连接到我们大脑的东西，无论是通过我们的眼睛还是更直接地进入大脑。最终，我们理解事物的范围将扩展到这些构造，就像它们已经扩展的那样。我的意思是，在现代世界中，我们能看到许多东西是因为我们用技术制造了它们，现在我们理解了它们是什么。但如果我们从未见过那种东西，我们就无法描述它，也无法理解它，等等。

So it kind of make distinctions that with our current, you know, like the mantis shrimp's view of reality is, at least in terms of color, is much richer than ours. Now, what's interesting is how do we get there? So imagine we have this augmented reality system that is evenit's seeing into the infrared, into the ultraviolet, things like thisand it's translating that into something that is connectable to our brains, either through our eyes or more directly into our brains, then eventually our kind of web of the types of things we understand will extend to those kinds of constructs just as they have extended. I mean, there are plenty of things where we see them in the modern world because we made them with technology, and now we understand what that is. But if we'd never seen that kind of thing, we wouldn't have a way to describe it, we wouldn't have a way to understand it, and so on.

好的。这实际上源于我们关于人工智能是否会消灭我们的讨论。我们讨论了智能在乡村空间的扩散，实际上，事情似乎变得更加复杂。事情比那种简单的故事更复杂，比如如果你建造了一个智能加一的东西，那个东西就能建造智能加二和加三的东西，这将是指数级的。它会以指数级的速度变得更智能，直到彻底摧毁一切。

Alright. So that actually stemmed from our conversation about whether AI is gonna kill all of us. And you we've discussed this kinda spreading of intelligence through rural space that in practice, it just seems that things get more complicated. Things are more complicated than the story of, well, if you build the thing that's plus one intelligence, that thing will be able to build the thing that's plus two intelligence and plus three intelligence, and that will be exponential. It'll become more intelligent exponentially faster and so on until it completely destroys everything.

但是，你知道，这种直觉可能仍然不那么简单，但它可能仍然有一定的有效性。这里有两个有趣的轨迹。一个是超级智能系统保持在人类附近的乡村空间。我们会觉得，天哪，这东西真的非常智能。

But, you know, that intuition might still not be so simple, but it might still care carry validity. And there's two interesting trajectories here. One, a superintelligence system remains in rural proximity to humans. To where we're like, holy crap. This thing is really intelligent.

让我们选总统吧。然后可能会有更可怕的智能开始远离我们。它们可能现在就在我们周围。它们在乡村空间中远离我们，但它们仍然与我们共享物理资源，对吧？

Let's elect the president. And then there could be perhaps more terrifying intelligence that starts moving away. They might be around us now. They're moving far away in rodeo space, but they're still sharing physical resources with us. Right?

是的。因此，它们可以剥夺我们的那些物理资源，并轻松地摧毁人类。

Yes. And so they can rob us of those physical resources and destroy humans just kinda casually.

是的，就像大自然能做到的那样。

Yeah. Just just Like nature could.

就像大自然能做到的那样。但AI系统似乎有某种独特性，存在这种指数级增长的趋势，比如——哦抱歉，大自然包罗万象。其中非常有趣的一点就是病毒的存在。

Like nature could. But it seems like there's something unique about AI systems where there is this kind of exponential growth, like, the way well, sorry. Nature has so many things in it. One of the things that nature has, which is very interesting, are viruses, for example.

嗯。

Mhmm.

自然界中存在具有这种指数效应的系统，这让我们人类感到恐惧，毕竟全球只有80亿人口，要全部消灭其实并不那么困难。所以，你会考虑这个问题吗？

There is systems within nature that have this kind of exponential effect, and that terrifies us humans because again, you know, there's only 8,000,000,000 of us, and you can just kinda it's just not that hard to just kinda whack them all real So, I mean, is that something you think about?

是的，我考虑过这个问题。

Yeah. I thought about that. Yes.

这种威胁性。你是否像埃利亚斯或雅科夫斯基那样担忧？比如AI对社会造成的巨大、痛苦且负面的影响。

The threat of it. I mean, are you as concerned about it as somebody like Elias or Yakovsky, for example? Just big, big, painful, negative effects of AI on society.

说实话并没有，可能因为我本质上是个乐观主义者。我认为事情不会像人们想象的那样——突然出现某个东西就摧毁一切。某种程度上，我相信所谓的计算不可约性，总会有意想不到的小角落存在。就像有人说要释放某种生物武器造成巨大伤害，但最终发现实际情况更复杂——因为人类个体存在差异，实际效果总会与预期有所偏差。

You know, no, but perhaps that's because I'm intrinsically an optimist. I think there are things I think the thing that one sees is there's going to be this one thing and it's going to just zap everything. Somehow, maybe I have faith in computational irreducibility, so to speak, that there's always unintended little corners. It's just like somebody says, I'm going toI don't know, somebody has some bioweapon and they say, We're going to release this and it's going to do all this harm. But then it turns out it's more complicated than that because you know, some humans are different, and the exact way it works is a little different than you expect.

这就像用某种巨大力量撞击某物的场景。小行星撞击地球后，地球会经历约两年的寒冷期，许多生物灭绝，但并非全部。这本质上就是计算不可约性的故事——总有意料之外的角落和后果。我不认为那种‘一击毙命’的毁灭方式（虽然确实可能发生）是唯一结局。

It's something where the great bigyou smash the thing with something. The asteroid collides with the earth, and it kind ofand yes, the earth is cold for two years or something, and lots of things die, but not everything dies. This is in a sense the story of computational irreducibility. There are always unexpected corners, there are always unexpected consequences. And I don't think that the kind of whack it over the head with something and then it's all gone is, you know, that can obviously happen.

地球也可能被黑洞吞噬，那样大概就彻底结束了。关于现实可能性路径的问题？我认为人类需要逐渐适应计算不可约性这类现象。我们总以为造出机器就能理解并控制它们的行为，但事实并非如此。

The earth can be swallowed up in a black hole or something and then it's kind of presumably all over. I think this question of what do I think the realistic paths are? I think that there will be sort of an increasing I mean, people have to get used to phenomena like computational irreducibility. There's an idea that we built the machines so we can understand what they do and we're going to be able to control what happens. Well, that's not really right.

现在的问题是：这种失控会导致机器合谋消灭人类吗？或许因为我是乐观主义者，我不认为这种剧本会上演。更现实的可能是形成一个AI生态系统——当然，这很难预测。关于AI能与世界哪些系统连接等具体问题，还存在大量细节需要探讨。

Now the question is, is the result of that lack of control going to be that the machines kind of conspire and sort of wipe us out? Maybe just because I'm an optimist, I don't tend to think that that's in the cards. I think that as a realistic thing, suspect what will sort of emerge, maybe, is kind of an ecosystem of the AIs just as, you know, again, I don't really know. I mean, is something that's hard to be clear about what will happen. I think that there are a lot of sort of details of what could we do, what systems in the world could we connect an AI to.

不得不说，前几天我在开发Wolfram语言的ChatGPT插件工具包时，突然意识到：我能让ChatGPT生成代码并在本地电脑运行。这让我切身感受到‘可能出什么问题’——就像把代码编写权完全委托给了AI。

I have to say, just a couple of days ago, I was working on this ChatGPT plug in kit that we have for Wolfen Language, okay, where you can, you know, you can create a plug in, and it runs Wolfram Language code, and it can run Wolfram Language code back on your own computer. And I was thinking, well, I can just make it you know, I can tell ChatGPT, create a piece of code, and then just run it on my computer. And I'm like, you know, that that sort of personalizes for me the what could what could possibly go wrong, so to speak.

这个可能性让你感到兴奋还是恐惧？

Was that exciting or scary, that possibility?

其实有点恐惧，因为这相当于我把代码编写权完全交给了AI——‘你负责写代码’，然后直接在我的电脑上运行。

It was a little bit scary, actually, because it's kind of like like I realized I'm I'm I'm delegating to the AI. Just write a piece of code. You know? You're in charge. Write a piece of code.

很快我所有的文件就都被处理完了。

Run it on my computer. And pretty soon, all my files complete.

这就像是俄罗斯轮盘赌，但复杂得多的版本。

That's that's a that's like Russian roulette, but, like, much more complicated version of that.

对，对，没错。

Yes. Yes. Right.

这是个不错的喝酒游戏。我不知道。嗯，

That's a good drinking game. I don't know. So Well,

没错。我是说，这就是原因。

right. I mean, that's why.

这就是你喝多少的问题。

That's how much you're drinking.

那么这是个有趣的问题。如果你这么做，应该设置什么样的沙盒限制？这某种程度上是对世界提出的那个问题的一个版本。一旦让AI掌管事务，你知道，在让AI掌管所有武器和所有这些不同系统之前，应该对这些系统施加多少约束。

It's an interesting question then. If you do that, what is the sandboxing that you should have? And that's sort of a that's a a version of of that question for the world. That is as soon as you put the AIs in charge of things, you know, how much how many constraints should there be on these systems before you put the AIs in charge of all the weapons and all these, you know, all these different kinds of systems.

嗯，关于沙盒的有趣之处在于AI知道它们的存在。它拥有破解的工具。

Well, here's the fun part about sandboxes is the AI knows about them. It has the tools to crack them.

看吧。计算机安全的根本问题，是的，在于计算的不可约简性。没错。因为事实是，任何沙箱永远无法成为完美的沙箱。

Look. The fundamental problem of computer security Yeah. Is computational irreducibility. Yes. Because the fact is any sandbox is never going to be a perfect sandbox.

如果你希望系统能执行有趣的功能——我是说，这正是计算机安全面临的普遍问题——一旦你的防火墙复杂到足以成为通用计算机，就意味着它能执行任何操作。如果你找到方法诱导它进行这种通用计算，就等于绕过了限制让它执行非预期的操作。这可以说是计算不可约简性的另一种体现，即你能让它做出你意料之外的行为。

If you want the system to be able to do interesting thingsI mean, this is the problem that's happened, the generic problem of computer security, that as soon as you have your firewall that is sophisticated enough to be a universal computer, that means it can do anything. If you find a way to poke it so that you actually get it to do that universal computation thing, that's the way you kind of crawl around and get it to do the thing that it wasn't intended to do. And that's sort of another version of computational irreducibility, is you can, you know, you can kind of you get it to do the thing you didn't expect it to do, so to speak.

计算可重现性在这里展现出太多有趣的可能性。在数字空间里，事物发展如此迅速，可能发生的情况数不胜数。你可以拥有聊天机器人，可以编写代码——甚至能让ChatGPT生成病毒程序。当然，我说的是故意为之的情况。

There's so many interesting possibilities here that manifest themselves from the compute computational reproducibility here. It's just so many things can happen because in digital space, things move so quickly. You can have a chatbot. You can have a piece of code that you could basically have ChatGPT generate viruses. I said, only on purpose.

而且是数字病毒。

And they digital viruses.

没错。

Yes.

它们还可能是思维病毒。就像钓鱼邮件那样，它们能说服你相信某些事情。是的，它们能影响人的认知。

And they could be brain viruses too. They they convince kinda like phishing emails. Yes. They can convince you of stuff.

确实。毫无疑问，在某种意义上，我们已经陷入了机器学习循环——制造能说服人类的事物的循环，这个过程必将变得越来越容易。那么未来会怎样？这再次说明，对人类而言这是个全新环境。必须承认，这个环境的变化速度快得有些可怕——要知道人们担忧的气候变化是以百年为单位的，而这个数字环境正在飞速演变。

Yes. And no doubt you can, you know, in a sense, we've had the loop of the machine learning loop of making things that convince people of things is surely going to get easier to do. Yeah. And, then what does that look like? Well, it's again, we humans This is a new environment for us, and admittedly, it's an environment which, a little bit scarily, is changing much more rapidly than I mean, you know, people worry about climate change is going happen over hundreds of years, and, you know, the environment is changing.

但你知道，数字环境可能在六个月内发生变化。

But the environment for you know, the the kind of digital environment might change in in six months.

因此，关于GPT对社会影响的一个相关担忧，就是与Wolfram Alpha相关的真理本质。因为通过符号推理进行计算，这在Wolfram Alpha中作为界面体现，有一种感觉是Wolfram Alpha告诉我的就是真理。

So one of the relevant concerns here in terms of the impact of GPT on society is the nature of truth that's relevant to Wolfram Alpha. Because computation through symbolic reasoning that's embodied in Wolfram Alpha as the interface, there's a kind of sense that what Wolfram Alpha tells me is true.

嗯。

Mhmm.

所以我们希望如此。是的，我的意思是，你或许可以分析这一点。你可以展示你无法证明它总是正确的，竞争或缺陷，但它大多数情况下会是正确的。

So we hope. Yeah. I mean, you could probably analyze that. You could show you can't prove that it's always gonna be true, competition or disability, but it's gonna be more true than not.

事实是，它将是您指定规则的正确结果，就它谈论现实世界而言，你知道，我们的工作就是整理和收集数据，确保这些数据尽可能真实。那么这意味着什么？嗯，这总是一个有趣的问题。对我们来说，真理的操作定义是，比如有人问，谁是最佳女演员？谁知道呢？

It's it's Look, the fact is it will be the correct consequence of the rules you've specified, and insofar as it talks about the real world, you know, that is our job in sort of curating and collecting data to make sure that that data is as true as possible. Now, what does that mean? Well, it's always an interesting question. For us, our operational definition of truth is, you know, somebody says, Who's the best actress? Who knows?

但有人赢得了奥斯卡，那是一个确定的事实。所以这就是我们可以作为真理片段进行计算的那种事情。如果你问传感器测量了这个东西，它是这样做的，一个机器学习系统，这个特定的机器学习系统识别了这个东西，那可以说是一个确定的事实。你知道，世界上有一个良好的这类事实网络。特别是当你问，某某是个好人吗？

But somebody won the Oscar, and that's a definite fact. So that's the kind of thing that we can make computational as a piece of truth. If you ask these things which a sensor measured this thing, it did it this way, a machine learning system, this particular machine learning system recognized this thing, that's a sort of a definite fact, so to speak. And that's, you know, there is a good network of those things in the world. It's certainly the case that particularly when you say, is so and so a good person?

那是一个毫无希望的问题，我们可能有一个关于好的计算语言定义。我不认为它会很有趣，因为那是一个非常混乱的概念，不太适合...我认为我们最多能做的就是，我想要X。有一个关于我想要X的有意义的计算，这有各种后果。我的意思是，我不确定。我还没有完全想清楚，但我认为，像某某是个好人这样的概念，是真的还是假的？

That's a hopelessly you know, we might have a computational language definition of good. I don't think it'd be very interesting, because that's a very messy kind of concept, not really amenable to kind of I think as far as we will get with those kinds of things is I want X. There's a kind of meaningful calculus of I want X, and that has various consequences. I mean, I'm not sure. I haven't thought this through properly, but I think you know, a concept like, is so and so a good person, is that true or not?

真是一团糟。这简直是

That's a mess. That's a

一种可通过计算处理的混乱。我认为当人类试图通过立法来定义何为善时，情况就会变得混乱。但当人类通过文学、历史书籍和诗歌来定义善时，事情就开始

mess that's amenable to computation. I think I think it's a mess when humans try to define what's good, like, through legislation. But when humans try to define what's good through literature, through history books, through poetry, it starts

变得——我是说，那个特定问题有点像，你知道，我们正在探讨什么样的行为算得上是好的这种伦理问题，以及我们认为什么是正确的，等等。我认为其中一个特点是，我们对此并不都持相同看法。没有定理能说明伦理必须遵循某种理论框架。首先

being I mean, that particular thing, it's kind of like, you know, we're we're we're going into kind of the ethics of what what counts as good, so to speak, and, you know, what do we think is right, and so on. And I think that's a thing which, you know, one feature is we don't all agree about that. There's no theorems about kind of you know, there's theoretical framework that says this is is the way that ethics has to be. Well, first

我们确实在某些方面有共识，甚至宗教经典中的道德伦理也为哪些行为有效或无效提供了一些经验支持。比如我们大多认同谋杀是恶行，似乎存在一些普遍准则。我在想，谋杀AI是否也是恶行？我个人倾向于认为是，但这个问题我们必须面对。

of all, there's stuff we kind of agree on, and there's some empirical backing for what works and what doesn't from just even the morals and ethics within religious texts. So we seem to mostly agree that murder is bad. There's certain universals that seem to emerge. I wonder whether murder of an AI is bad. Well, I tend to think yes, but and I think we're gonna have to contend with that question.

哦，我还好奇AI会怎么回答。

Oh, and I wonder what AI would say.

是啊。关于AI，我认为关键之一是：消灭一个没有主人的AI是一回事。你可以想象一个AI在互联网上游荡，没有特定所有者。这时你会说，清除这个AI有什么危害呢？但如果这个AI有一万个人类朋友，这些人类会因AI被消灭而极度悲痛，

Yeah. Well, I think, you know, one of the things with with AIs is it's one thing to wipe out that AI that has no owner. You can easily imagine an AI hanging out on the Internet without having any particular owner or anything like that. And then you say, Well, what harm does it it's okay to get rid of that AI. Of course, if the AI has 10,000 friends who are humans, and all those 10,000 humans will be incredibly upset that this AI just got exterminated.

情况就变得复杂微妙了。人类共识这个问题确实存在——某些法律原则历史上曾被偏离，如今我们甚至会质问‘当初怎能不这样做？这完全不合理’。但归根结底，我认为除非某套规则会导致物种灭绝，否则很难断言绝对标准。嗯。

It becomes a slightly different, more entangled story. But yeah, I think that this question about what do humans agree about, it's, you know, there are certain things that human laws have tended to consistently agree about. There have been times in history when people have sort of gone away from certain kinds of laws, even ones that we would now say, How could you possibly have not done it that way? That just doesn't seem right at all. But I think this question of what I don't think one can say beyond saying, If you have a set of rules that will cause the species to go extinct, Mhmm.

可以说，这很可能不是一套能奏效的法律体系，因为要让法律有实施对象的前提是该物种尚未灭绝。

That's probably, you know, you could say that's probably not a a winning set of laws because even to have a thing on which you can operate laws requires that the species not be extinct.

但在诸如'芝加哥到纽约的距离'这类Wolfram Alpha能解答的问题，与'此人是否善良'这类判断之间，似乎存在大量灰色地带。这开始变得非常有趣。我认为自Wolfram Alpha创建以来，它某种程度上已成为

But between sort of, what's the distance between Chicago and New York that Wolfram Alpha can answer and the question of if this person is good or not, there seems to be a lot of gray area. And that starts becoming really interesting. Think your since the creation of Wolfram Alpha have been a kind of arbiter

大规模真相的仲裁者。

of truth at a at a large scale.

嗯。

Mhmm.

所以这个系统生成的真相比

So the system is generates more truth than

要确保事情的真实性。实际上，当人们编写计算合约时——就像约定'如果现实中发生某情况，就执行某操作'。这类发展没有预期中迅速，部分原因与区块链有关，虽然区块链并非计算合约概念的必要条件。

Try to make sure that the things are true. I mean, look, as a practical matter, when people write computational contracts, and it's kind of like, you know, if this happens in the world, then do this. Yes. And this hasn't developed as as quickly as it might have done. Know, this has been a sort of a blockchain story in part, although blockchain is not really necessary for the idea of computational contracts.

但可以想象，未来世界将充满由计算合约构成的庞大链条网络。当现实事件触发时，会引发连锁反应——合约自动执行并产生新事件。我们一直是区块链、计算合约等领域中'事实真相'的主要提供者。这里存在一个责任问题：如何确保信息准确。难点在于何时算事实，何时不算。

But you can imagine that eventually a large part of what's in the world are these giant chains and networks of computational contracts. Then something happens in the world and this whole giant domino effect of contracts firing autonomously that cause other things to happen. For us, we've been the main source of the oracle of quotes facts or truth or something for things like blockchain, computational contracts and such like. And there's a question of, you know, what I consider that responsibility to actually get the stuff right. One of the things that is tricky sometimes is when is it true, when is it a fact, when is it not a fact.

我认为我们能做的最好的就是声明：我们有一套流程，我们遵循这套流程，可能会犯错，但至少我们不会在犯错时腐败，可以这么说。

I think the best we can do is to say, We have a procedure, we follow the procedure, we might get it wrong, but at least we won't be corrupt about getting it wrong, so to speak.

这番话说得真漂亮。我对流程保持透明。问题开始显现，是当你将事物转化为计算语言的范围开始扩大时，比如进入政治领域。嗯。

So that's beautifully put. I have a transparency about the procedure. The problem starts to emerge when the things that you convert into computational language start to expand. For example, into the realm of politics. Mhmm.

所以这就像是Wolfram Alpha和ChadGPT之间美妙的舞蹈。正如你所说，ChadGPT浅显而广泛。嗯。所以它会对所有事情发表意见。

So this is where it's almost like this nice dance of Wolfram Alpha and ChadGPT. ChadGPT, like you said, is shallow and broad. Mhmm. So it's it's it's gonna give you an opinion on everything.

但它既写虚构也写事实，这正是它的构建方式。我的意思是，它就是在生成语言，甚至在代码中也同时创造两者，它编写虚构。有时候看着还挺有趣的。你知道，它会写出虚构的孤儿语言代码。

But it writes fiction as well as fact, which is exactly how it's built. I mean, that's exactly it is making language, and it is making both even in code, it writes fiction. I mean, it's kind of fun to see sometimes. You know, it'll write fictional orphan language code.

是啊。那种...那种看起来挺像那么回事的。

Yeah. That that That kinda looks right.

对。看起来没错，但实际上在实用上不正确。是的。不过它确实对世界如何运作有一种大致的概念，就像虚构小说描述世界大致如何运作一样。只是它们恰好不是世界实际运作的方式。

Yeah. It looks right, but it's actually not pragmatically correct. Yeah. But but yes, it's a it has a view of kind of roughly how the world works at the same level as as books of fiction talk about roughly how the world works. They just don't happen to be the way the world actually worked or whatever.

不过，我同意。我们

But yes, I agree. We

是

are

正尝试用我们完整的Wolfram语言——这种计算语言——来尽可能准确地表征现实世界。它不必完全反映真实世界的运作方式，因为我们可以发明一套不同于现实世界的规则并运行它们。但关键在于，我们要准确呈现这些规则运行的结果，无论它们是否符合现实世界的法则。同时，我们也力求尽可能精确地捕捉世界的特征，以表征现实中发生的事件。就像我们讨论过的，世界中的原子排列……比如你说‘我不知道是否出现过一辆坦克并行驶到某处’，那么‘坦克’是什么？它是我们抽象描述为坦克的原子排列。

attempting with our whole Wolfram language, computational language thing to represent at least well, it doesn't necessarily have to be how the actual world works, because we can invent a set of rules that aren't the way the actual world works and run those rules, but then we're saying we're going to accurately represent the results of running those rules, which might or might not be the actual rules of the world. But we also are trying to capture features of the world as accurately as possible to represent what happens in the world. Now, again, as we've discussed, the atoms in the world arrange You say, I don't know, was there a tank that showed up that drove somewhere? Okay, well, what is a tank? It's an arrangement of atoms that we abstractly describe as a tank.

你可能会说，存在某种原子排列不同于其他排列，但我们并未明确界定。这就像观察者理论中的问题：什么样的原子排列算作坦克，什么样的不算。

And you could say, Well, there's some arrangement of atoms that is a different arrangement of atoms, but it's not We didn't decide. It's like this observer theory question of what arrangement of atoms counts as a tank versus not a tank.

因此，即便是我们认定的确凿事实，也可能被逐步拆解并证明其并非如此。

So there's there's even things that we consider strong facts. You could start to kinda disassemble them and show that they're not.

没错。比如‘这股风是否强到能吹倒这个特定物体？’这个问题中，‘一阵风’本身就是复杂概念。它涉及流体动力学中的微小片段、各处的小漩涡，你必须定义你所关注的风的哪个方面。

Right. I mean, so so the question of whether oh, I don't know. Was this gust of wind strong enough to blow over this particular thing? Well, a gust of wind is a complicated concept. You know, it's full of little pieces of fluid dynamics and little vortices here and there, and you have to define, you know, what the aspect of the gust of wind that you care about might be.

比如它对某个风力涡轮机叶片施加了多大压力。如果要判定‘风的强度达到某种程度’这样的事实，就必须先有明确定义，需要某种测量装置来宣告：‘根据我如此构造的测量装置，当时风的强度是这样的。’

It put this amount of pressure on this blade of some wind turbine or something. If you have something which is the fact of the gust of wind was this strong or whatever, you have to have some definition of that. You have to have some measuring device that says, according to my measuring device that was constructed this way, the gust of wind was this.

那么关于真理的本质，你能给出哪些对我们理解ChatGPT有帮助的见解？因为你一直在处理‘何为事实’这个问题。现在ChatGPT被广泛使用，我甚至看到记者用它撰写文章。嗯。

So what can you say about the nature of truth that's useful for us to understand, Chad GPT? Because you've been you've been contending with this idea of what is fact and not. And it seems like ChatGPT is used a lot now. I've seen it used by journalists to write articles. Uh-huh.

因此，有些人正在研究大型语言模型，拼命想办法通过各种机制来审查它们，无论是手动方式还是通过人类反馈的强化学习，试图调整模型尽可能只陈述事实，避免虚构内容。

And so you have people that are working with large language models trying to desperately figure out how do we essentially censor them through different mechanisms, either manually or through reinforcement learning with human feedback, trying to align them to to not say fiction, just to to say nonfiction as much as possible.

这就是计算语言作为中介的重要性。就像你拥有一个大型语言模型，它能呈现出某种形式化、精确的东西，你可以查看它、运行测试、进行各种操作。它总能以相同方式运作，其功能被精确定义。而大型语言模型就是这个接口。

This is the importance of computational language as an intermediate. It's kind of like you've got the large language model. It's able to surface something which is a formal, precise thing that you can then look at and you can run tests on it and you can do all kinds of things. It's always going to work the same way and it's precisely defined what it does. And then the large language model is the interface.

我认为这些大型语言模型的重要之处在于——我是说，它们有许多应用场景。谈论其中一些确实令人惊叹。几乎每天我们都会发现几个新用途，有些非常非常出人意料。但最佳应用场景是那些即使结果大致正确，也能带来巨大收益的。比如一两周前我们有个用例是阅读错误报告。

I mean, the way I view these large language models, one of their importantI mean, there are many use cases. It's a remarkable thing to talk about some of these. Literally, every day, we're coming up with a couple of new use cases, some of which are very, very, very surprising. Things where I mean, but the best use cases are ones where it's you know, even if it gets it roughly right, it's still a huge win. Like a use case we had from a week or two ago is read our bug reports.

要知道，我们积累了数十年的数十万份错误报告，就像在问：能否让它阅读报告后，判断错误可能出现在哪里？然后锁定相关代码段。它甚至可能建议修改代码的方式——虽然关于如何修复的建议可能是胡扯，但能定位到问题代码已经非常有用。确实很厉害。

You know, we've got hundreds of thousands of bug reports that have accumulated over decades, and it's like, you know, can we have it? Just read the bug report, figure out where is the bug likely to be, And, you know, home in on that piece of code. Maybe it'll even suggest some, you know, sort of way to fix the code. It might be nonsense what it says about how to fix the code, but it's incredibly useful that it was able to, you know Yeah. So awesome.

最神奇的是连那些胡话都莫名具有启发性。我还没完全理解这点。编程相关应用太多了，比如不同编程语言间的转换就非常有趣。效果惊人，但失败案例也能指明改进方向。

It's so awesome because even the nonsense will somehow be instructive. I don't I don't quite understand that yet. I've I've yeah. There's so many programming related things, like, for example, translating from one programming language to another is really really interesting. It's extremely effective, but then you the failures reveal the path forward also.

是的。但关键在于——这类讨论中，我们计算语言的独特之处在于它本就是为人类阅读而设计的。

Yeah. But I think I mean, the the the big thing I mean, in in that kind of discussion, the unique thing about our computational language is it was intended to be read by humans.

没错。这确实非常重要。

Yes. That's really important. Right.

它具备这样一种功能，你可以...但考虑到ChatGPT及其应用，我认为其重要特性之一在于它是一个语言用户界面。以记者为例，典型使用场景可能是：假设我有五个事实要点需要整合成文章或报告，当我将这五个要点输入ChatGPT时，它能将其扩展成完整报告。这种界面优势在于——若我仅将五个要点交给其他人，对方可能会因这些简略笔记而表示困惑。

And so it has this thing where you can but but, you know, thinking about sort of ChatGPT and its use and so on, one of the big things about it, I think, is it's a linguistic user interface. That is, so a typical use case might beand take the journalist case, for example. It's like, let's say I have five facts that I'm trying to turn into an article, or I'm trying to write a report where I have basically five facts that I'm trying to include in this report. But then I feed those five facts to ChatGPT, it puffs them out into this big report. And then that's a good interface for another If I just gave I just had in my terms those five bullet points and I gave them to some other person, the person would say, I don't know what you're talking about because this is your version of this sort of quick notes about these five bullet points.

但若通过语言集体理解框架将其扩展，他人就能明白你的意图。甚至可将这个扩展结果输入另一个大语言模型。比如申请某个养鱼许可证时，你输入几个要点：'我将使用某种水质'等零散信息...

But if you puff it out into this thing which kind of connects to the collective understanding of language, then somebody else can look at it and say, okay, I understand what you're talking about. Now, you can also have a situation where that thing that was puffed out is fed to another large language model. You know, it's kind of like, you know, you're applying for the permit to, you know, I don't know, grow fish in some place or something like this. And it, you know, it it and and you have these facts that you're putting in. You know, I'm gonna have a you know, I'm gonna have this kind of water and I don't know what it You just got a few bullet points.

模型会将其扩展成完整申请书。而当渔业局的另一个大语言模型接收时，它会压缩提取关键三点——因为渔业局只关心特定要素。本质上，大语言模型生成的自然语言就像传输层，实现LLM间的对话。就像我用LLM撰写邮件，从核心意图扩展成文。

It puffs it out into this big application. You fill it out. Then at the other end, the, you know, the Fisheries Bureau has another large language model that just crushes it down because the Fisheries Bureau cares about these three points, and it knows what it cares about. So it's really the natural language produced by the large language model is sort of a transport layer that is really LLM communicates with LLM. I mean, it's kind of like the, you know, I write a piece of email using my LLM and, you know, puff it out from the things I want to say.

你的LLM处理后得出结论X。问题在于：这个结果语义上看似合理，却可能与现实存在偏差。我亲历过这种情况——当我们发布ChatGPT插件时，我用复杂数学应用题测试：'某人拥有的鸡是另一人两倍'等，它出色地转化为方程组，通过Wolfram求解后反馈结果。我本打算将此写入博客...

Your LLM turns it into, and the conclusion is x. Now, the issue is, that the thing is going to make this thing that is sort of semantically plausible, and it might not actually be what you it might not relate to the world in the way that you think it should relate to the world. I've seen this, you know, I've been doing Okay, I'll give you a couple of examples. I was doing this thing when we announced this plug in for ChatGPT, I had this lovely example of a math word problem, some complicated thing, and it did a spectacular job of taking apart this elaborate thing about, you know, this person has twice as many chickens as this, etcetera, etcetera, etcetera, and it turned it into a bunch of equations. It fed them to a Wolfen language, we solved the equations, everybody did great, we gave back the results, and I thought, okay, I'm gonna put this in this blog post I'm writing.

结果发现它正确处理了所有难点，却在最后两行完全搞错答案——而我差点没发现。两天前又发生类似情况：我用ChatGPT插件开发了本地电脑发声程序...

I thought, I'd better just check. It turns out, it got everything, all the hard stuff it got right, and the very end, last two lines, it just completely goofed it up and gave the wrong answer, and I would not have noticed this. Same thing happened to me two days ago. So I thought, you know, I made this with this ChatGPT plug in kit. I made a thing that would emit a sound, would play a tune on my local computer.

ChatGPT生成音符序列后，能在电脑上播放旋律。很酷对吧？于是我要求：'请播放HAL被断开时唱的歌'...

So ChatGPT would produce you know, a series of notes, and it would play this tune on my computer. Very cool. Okay. So I thought, I'm gonna ask it. Play the tune that Hal sang Mhmm.

就是《2001太空漫游》里HAL在关机时的场景。然后它...

When Hal was being disconnected in 02/2001. Okay? So it it there

是的。黛西？是叫黛西吗？

it is. Daisy? Was it Daisy?

对，黛西。没错。好的。所以，呃，好吧。

Yes. Daisy. Yeah. Right. So so okay.

所以我想，你知道的，它生成了一堆音符。我当时觉得，这太棒了，简直不可思议。然后我本来打算直接放进去，但转念一想，不如实际演奏一下。于是我照做了，结果发现是《玛丽有只小羊羔》。

So I think, you know and so it produces a bunch of notes. And I'm like, this is spectacular. This is amazing. And then I thought, you know, I was just gonna put it in, and then I thought, better actually play this. And so I did, and it was Mary Had a Little Lamb.

哦，哇哦。哇哦。但居然是《玛丽有只小羊羔》。是啊。没错。

Oh, wow. Oh, wow. But it was Mary Had a Little Lamb. Yeah. Yes.

哇。所以它是对的，但又错了。是的。你很容易被误导。

Wow. So it's correct, but wrong. Yes. It was you could easily be mistaken.

是的。没错。实际上，我引用了哈尔的话来解释——你知道电影里哈尔9000说的吗？那只是个修辞手法，因为我突然意识到，天啊，这个ChatGPT完全可以骗过我。我是说，它居然知道电影里的这个细节，还能转化成歌曲的音符，只是搞错了歌曲。电影里哈尔说过类似的话，比如‘没有9000系列计算机曾被记录犯过错’之类的。

Yes. Right. And in fact, I I kind of gave the I had this quote from Hal to explain, you know, it's as the Hal states in the movie, you know, the Hal 9,000 is, you know, the thing was just a rhetorical device, because I'm realizing, oh my gosh, you know, this Chatty PT could have easily fooled me. I mean, it did this amazing thing of knowing this thing about the movie and being able to turn that into the notes of the song, except it's the wrong song. And Hal in the movie, Hal says, you know, I think it's something like, you know, No 9,000 series computer has ever been found to make an error.

从实用角度而言，我们是完美无缺且不会出错的。我觉得用哈尔这句有点可爱的台词，来形容ChatGPT在那件事上的表现还挺贴切的。

We are, for all practical purposes, perfect and incapable of error. And I thought that was kind of a charming sort of quote from from Hal to make in connection with with what Chachi PT had done in that case.

有趣的是，正如你所说，像LLMs这样的模型非常愿意承认错误。嗯，是的。

The interesting thing is about the LLMs, like you said, that they are very willing to admit to error. Well, yes.

我是说，这其实是RLHF（基于人类反馈的强化学习）的问题。哦，对。那个...那个问题问得很好。关于ChatGPT真正了不起的地方在于，你知道，我一直在关注大语言模型的进展，自己也试用了很多次，它们基本上就是...基于语言统计延续的预期结果。虽然有趣，但算不上突破性、令人兴奋的那种。

I mean, that's a question of the RLHA, the reinforcement learning human feedback thing. Oh, right. That that that's, you know That's a nice question. LLM the the really remarkable thing about CHACCPT is, you know, I had been following what was happening with large language models, and I played with them a whole bunch, and they were kind of like, you know, it's kind of like what you would expect based on sort statistical continuation of language. It's interesting, but it's not Breakout, exciting.

然后我认为，正是那种人类反馈的强化学习，让ChatGPT尝试去做人类真正想做的事情，才实现了突破，达到了这个让我们人类觉得真正有趣的临界点。顺便说，观察它如何在你调整温度参数之类的时候突然变得胡言乱语，不再对人类有吸引力，这也很耐人寻味。它恰好维持在了一个完美阈值之上——与人类兴趣高度吻合的状态。说实话，我觉得没人预见到这一点。

And then I think the kind of reinforcement, the human feedback reinforcement learning, you know, in making chatty PT try and do the things that humans really wanted to do, that broke through, that kind of reached this threshold where the thing really is interesting to us humans. And by the way, it's interesting to see how, you know, you change the temperature or something like that, the thing goes bonkers, and it no longer is interesting to humans. It's producing garbage. And it's kind of right. Somehow it managed to get above this threshold where it really is well aligned to what we humans are interested in, and kind of that's And I think nobody saw that coming, I think.

至少我交谈过的人里没有，参与那个项目的成员似乎也无人预料到。这就像某种神奇的临界点现象。比如我们开发Wolfram Alpha时，我也不知道它能否成功。我们试图构建一个具备足够世界知识、能回答合理问题集的系统，自然语言理解要足够好，能处理常规输入。但我们当时根本不知道那个临界点在哪里。

Certainly nobody I've talked to, nobody who was involved in that project seems to have known it was coming. It's just one of these things that is a sort of remarkable threshold. I mean, you know, when we built Wolfram Alpha, for example, I didn't know it was going to work. You know, we tried to build something that would have enough knowledge of the world, that it could answer a reasonable set of questions, that we could do good enough natural language understanding, that typical things you type in would work. We didn't know where that threshold was.

老实说，我甚至不确定我们是否选对了十年——或许该等五十年后再尝试。但ChatGPT的情况如出一辙，我认为没人能预言2022年会是这个技术突破的元年。

I mean, I was not sure that it was the right decade to try and build this, even the right fifty years to try and build it. And I think that was it's the same type of thing with ChatGPT that I don't think anybody could have predicted that 2022 would be the year that this became possible.

是啊，你讲过给马文·明斯基演示时的故事，当时反复强调：不，这次真的成了。

I think, yeah, you tell a story about Marvin Miski and showing it to him and saying that, like, no. No. No. This time, it actually works.

没错。对我来说观察这些大语言模型也是如此。ChatGPT刚问世那几周，人们的反应都是：哦，这些大语言模型我早就见识过了。

Yes. Yes. And I mean, it's, you know, it's the same thing for me looking at these large language models. It's like when people are first saying, the first few weeks of ChatGPT, it's like, oh, yeah. I've seen these large language models.

然后我真的试了试，天哪，居然真的有效。我记得最早尝试的一件事是让它写一篇说服性文章，论证狼是最蓝的动物。它开始描述生活在青藏高原上的狼，还用了拉丁学名什么的。我当时就想：真的假的？

And then I actually try it, and, you know, oh my gosh, it actually works. And think it's but the thing I found I remember one of the first things I tried was write a persuasive essay that a wolf is the bluest kind of animal. Okay? So it writes this thing, and it starts talking about these wolves that live on the Tibetan Plateau, and they're named some Latin name and so on. And I'm like, really?

我上网查证后发现全是胡扯，但听起来极其可信——可信到我真去搜索是否存在蓝色狼种。我在直播中提到这事后，观众们还给我发来了各种图片。

I'm starting to look it up on the web, and it's like, well, it's actually complete nonsense. But it's extremely plausible. I mean, it's plausible enough that I was going and looking up on the web and wondering if there was a wolf that was blue. You know, I mentioned this on some livestreams I've done, and so people have been sending me these pictures.

蓝狼啊...说不定我真发现了什么。能给些智者建议吗？那些从未接触过AI系统（连Wolfram Alpha都没用过）的人，现在突然开始用ChatGPT了。当AI普及到新用户群体时，我们该如何面对真相问题？比如新闻工作者该怎么办？

Blue wolves. Maybe I was onto something. Can you kinda give your wise sage advice about what humans who have never interacted with AI systems, not even like with Wolfram Alpha, are now interacting with Chad GPT because it it becomes it's accessible to a certain demographic that may have not touched AI systems before. What do we do with truth? Like journalists, for example.

没错。我们该如何看待这些系统的输出？

Yeah. How do we think about the output of these systems?

指望AI输出绝对真实的想法不太实际。这本质上是语言接口，它生成的是语言——语言本身就有真伪之分。就像我们可以说'去查证你的消息源'，但总有它无法验证的内容。

I think this idea the idea that you're going to get factual output is not a very good idea. I mean, it's just this is not it is a linguistic interface. It is producing language, and language can be truthful or not truthful, and that's a different slice of what's going on. I think that what we see in, for example, you know, go check this with your fact source, for example. You can do that to some extent, but then it's going to not check something.

关键在于验证机制是否精准。就像Wolfram插件调用——有时准确有时失误。但最令人兴奋的是，这标志着计算能力正在大规模民主化。

That is, again, a thing that is sort of a does it check-in the right place? We see that in does it call the Wolfram plug in in the right place? Often it does. Sometimes it doesn't. Think the real thing to understand about what's happening, which I think is very exciting, is kind of the great democratization of access to computation.

回顾历史，计算能力长期被少数'德鲁伊'垄断。我毕生致力于打破这种垄断——早在1988年Mathematica诞生前，物理学家想做个计算都得找程序员外包，结果好坏全凭运气。

And I think that when you look at there's been a long period of time when computation and the ability to figure out things with computers has been something that only the druids at some level can achieve. I myself have been involved in trying to de druidify access to computation. I mean, back before Mathematica existed, you know, in 1988, if you were a physicist or something like that, and you wanted to do a computation, you would find a programmer, you would go and delegate the computation to that programmer. Hopefully, they'd come back with something useful. Maybe they wouldn't.

过去需要经历一个长达数周的循环过程。而1988年时观察这一现象非常有趣——起初是物理学家、数学家等群体，随后迅速扩散到更多人，人们突然意识到自己可以亲手编写代码来完成关心的计算。见证人们通过这个工具做出诸多发现确实令人振奋。我认为同样的事情正在重演。

There'd be this long, multi week loop that you'd go through. And then it was actually very interesting to see. 1988, people like physicists, mathematicians and so on, then lots of other people, but this very rapid transition of people realizing they themselves could actually type with their own fingers and make some piece of code that would do a computation that they cared about. It's been exciting to see lots of discoveries and so on made by using that tool. I think the same thing is We see the same thing.

WolframAlpha处理的运算深度不及完整的Wolfram语言Mathematica技术栈。但大型语言模型的自然语言交互机制最让我兴奋的是，它极大拓宽了深度计算的普及面。我最近在思考：这些程序员将何去何从？那些整天编写模板代码的人会面临什么变化？

WolframAlpha is dealing with not as deep computation as you can achieve with whole Wolfram Language Mathematica stack. But the thing that's, to me, particularly exciting about the large language model linguistic interface mechanism is it dramatically broadens the access to deep computation. One of the things I've thought about recently is what's going to happen to all these programmers? What's going to happen to all these people who a lot of what they do is write slabs of boilerplate code?

在

In

某种意义上，四十年来我一直在说这种做法并不明智。大部分模板代码可以自动化——用足够高级的语言设计时，整段代码会浓缩成我们刚实现的那个可直接调用的函数。因此在我看来，从事这类底层编程活动本身就不是正确方向。不过已有许多人通过我们的技术避开了这种工作。

a sense, I've been saying for forty years, that's not a very good idea. You can automate a lot of that stuff. With a high enough level language, that slab of code that's designed in the right way, that slab of code turns into this one function we just implemented that you can just use. So in a sense, the fact that there's all of this activity of doing sort of lower level programming is something for me, it seemed like I don't think this is the right thing to do. But, you know, and lots of people have used our technology and not had to do that.

但现实情况是...比如那些演变成编程技能培训基地的计算机科学系，未来会怎样？我认为存在两种趋势：一是模板编程将重蹈汇编语言的覆辙，最终被高级抽象取代——从自然语言开始，转化为计算语言，经测试验证后，通过我们的算法直接执行，或编译成LLVM等中间形式。

But the fact is that that's you know, so when you look at, I don't know, computer science departments that have that have turned into places where people are learning the trade of programming, so to speak, it's it's sort of a question of what's gonna happen. And I think there are two dynamics. One is that kind of boilerplate programming is going to become you know, it's going to go the way that assembly language went back in the day of something where it's really mostly specified by at a higher level, you start with natural language, you turn it into a computational language. That's you look at the computational language, you run tests, you understand that's what's supposed to happen. If we do a great job with compilation of the computational language, it might turn into LLVM or something like this, or it just directly gets gets run through the algorithms we have and so on.

这将瓦解现有的编程教育体系。但另一方面，更重要的趋势是关注计算的人群将激增——那些从未接触计算的领域（比如艺术史），现在都能通过这种语言交互机制触及计算能力。

But then so that's kind of a tearing down of this big structure that's been built of teaching people programming. But on the other hand, the other dynamic is vastly more people are going to care about computation. So all those departments of art history or something that really didn't use computation before now have the possibility of accessing it by virtue of this kind of linguistic interface mechanism.

如果能创建支持调试和交互计算语言的界面，将进一步降低使用门槛。

And if you create an interface that allows you to interpret the debug and interact with the computational language, then that makes it even more accessible.

是的。嗯，我的意思是，我认为关键在于，目前普通艺术史专业的学生或类似人群，他们并不认为自己了解编程这类事物。但当技术发展到只需直接上手操作、无需文档支持的程度时——比如直接输入‘比较这些图片与那些图片的色彩运用’就能生成可执行的计算机语言代码并立即看到结果——你会感叹‘哦，这看起来差不多’或‘这太离谱了’。最终你可能会想‘我最好还是弄懂这段代码的原理’，这就成了学习契机。有趣的是，数学必须学懂才能使用，而这种技术却能先用后学。

Yeah. Well, I mean, the I think the thing is that right now, you know, the average, you know, art history student or something probably isn't going to you know, they're not to they don't think they know about programming and things like this. But by the time it really becomes a purely you just walk up to it, there's no documentation, you start just typing, compare these pictures with these pictures and see the use of this color, whatever, and you generate this piece of computational language code that gets run, you see the result, you say, Oh, that looks roughly right, or you say, That's crazy. Maybe then you eventually get to say, Well, I better actually try and understand what this computational language code did, and and that becomes a thing that you learn. Just like it's kind of an interesting thing because unlike with mathematics, where you kinda have to learn it before you can use it, this is a case where you can use it before you have to learn it.

我有个悲观——或者说乐观的可能性：很快人们甚至不会去看计算机语言代码本身。

Well, I got a sad possibility here, or maybe exciting possibility, that very quickly people won't even look at the computational language.

随着代码生成技术日益完善，人们会默认生成的代码是正确的。确实如此。我认为会有足够多的案例让人们意识到这点——毕竟系统还能自动生成测试用例。这其实是相当酷的功能。

They'll trust that it's generated correctly as you get better and better generating that language. Yes. I think that there will be enough cases where people see, you know, because you can make it generate tests too. And so you'll say, we're doing that. I mean, it's a pretty cool thing, actually.

比如展示代码时附带若干运行示例。人们至少会检查这些案例，发现错误后系统会回溯修正。我同意存在中间态：某些人会阅读代码，有些人只看测试用例或最终结果。当需求明确时——比如‘做个带两个滑块的界面’——看到成品符合描述就足够了。

You know, say this is the code and here are a bunch of examples of running the code. Okay. People will at least look at those, and they'll say that example is wrong, and then it'll wind back from there. And I agree that the intermediate level of people reading the computational language code, in some cases people will do that, in other cases people just look at the tests or even just look at the results. Sometimes it'll be obvious that you got the thing you wanted to get because you were just describing, you know, make me this interface that has two sliders here, and you can see it has those two sliders there.

这就是预期效果。但随之而来的问题是：当大众都能便捷使用计算工具时，人们应该学习什么？如今计算机科学院校教授的内容颇具历史偶然性——三四十年前计算机系规模很小，主要教授有限自动机理论、编译器原理等。像我们公司很少招聘这类毕业生，因其所学虽有趣（我个人热爱理论），但与实际软件开发需求脱节。

That's the result you want. But I think one of the questions then is, in that setting where you have this kind of broad ability of people to access computation, what should people learn? In other words, right now, you go to computer science school, so to speak, and a large part of what people end up learning I mean, it's been a funny historical development because back thirty, forty years ago, computer science departments were quite small, and they taught things like finite automata theory and compiler theory and things like this. You know, a company like mine rarely hired people who'd come out of those programs because the stuff they knew was I think is very interesting. I love that theoretical stuff.

九十年代出现重大转折，IT类编程和软件工程需求激增，学生趋之若鹜。这背后实质是各行业全面数字化——万物皆可计算化。人们对此作出响应，却形成了一种认知：要达到这个目标，必须掌握编程语言这类手艺。回想三十五年前我任教时（时间飞逝！），顶尖研究型大学刚设立计算机系，这种现象其实很耐人寻味。

But it wasn't that useful for the things we actually had to build in software engineering. And then there was this big pivot in the nineties, I guess, where there was a big demand for sort of IT type programming and so on and software engineering, and then big demand from students and so on. You know, we want to learn this stuff. I think the thing that really was happening in part was lots of different fields of human endeavor were becoming computational. For all x, there computational was x.

这正是当时人们所回应的趋势。但随后形成的观念是：要实现这个目标，关键在于掌握编程语言这类技能。这其实很奇特——记得三十五年前我从事教职时（天啊，真是很久以前了），正值顶尖研究型大学初设计算机系的年代。

And that was the thing that people were responding to. But then this idea emerged that to get to that point, the main thing you had to do was to learn this kind of trade or skill of doing programming language type programming. That, you know, it's a strange thing actually, because I remember back when I used to be in the professoring business, which is now thirty five years ago, so gosh, that's a rather long time ago. Time flies. You know, it was right when they were just starting to emerge kind of computer science departments at fancy research universities and so on.

我是说，有些人已经掌握了这些技能，但其他人刚开始接触。这有点像他们在思考，我们是否要把这种本质上属于职业技能的东西，与我们正在做的其他事情结合起来？很多这类知识型工作一直被认为是人类必须去学校学习的东西，永远不会被自动化。令人震惊的是，很快就能明显看出，其中很多内容其实是可以自动化的。那么问题来了，如果学习汽车机械原理不再有价值，你只需要知道如何开车，打个比方，那你还需要学习什么？

I mean, some had already had it, but the other ones were just starting to have that. And it was kind of a thing where they were kind of wondering, are we going to put this thing that is essentially a trade like skill, are we going to somehow attach this to the rest of what we're doing? A lot of these kind of knowledge work type activities have always seemed like things where that's where the humans have to go to school and learn all this stuff and that's never going to be automated. This isit's kind of shocking that rather quickly, a lot of that stuff is clearly automatable. And I think, you knowbut the question then is, okay, so if it isn't worth learning how to do car mechanics, you only need to know how to drive the car, so to speak, what do you need to learn?

换句话说，如果你不需要详细了解如何告诉计算机执行这个循环、设置这个变量、建立这个数组等等，如果你不必学习这些底层原理，那你需要学习什么？我认为答案是，你需要知道你想把车开往哪里。也就是说，你需要对计算可能性的架构有个大致概念。

In other words, if you don't need to know the mechanics of how to tell the computer in detail, make this loop, set this variable, set up this array, whatever else, if you don't have to learn that stuff, you don't have to learn the kind of under the hood things, what do you have to learn? I think the answer is you need to have an idea of where you want to drive the car. In other words, you need to have some notion of, you you need to have some picture of sort of what the architecture of what is computationally possible is.

其实对话中还有这种艺术性元素，因为你最终是用自然语言来控制汽车的。所以不只是你想去哪里的问题。

Well, there's also this kind of artistic element of of conversation because you ultimately you use natural language to control the car. So it's not just where you wanna go.

确实。这很有趣。问题在于谁会成为优秀的提示工程师。对吧？

Well, yeah. You know, it's interesting. It's a question of who's gonna be a great prompt engineer. Yeah. Okay?

我这周的新理论是：优秀的说明文作者就是优秀的提示工程师。

So my current theory this week, good expository writers are good prompt engineers.

什么是说明文作者？就像...

What's an expository writer? So like

就是能很好解释事情的人。

a Somebody who can explain stuff well.

但这属于哪个部门的职责范围？

But which department does that come from?

在大学里吗？是啊，我也不清楚。我觉得说明文写作系可能都被裁撤光了。

In the university? Yeah. I have no idea. I think they killed off all the expository writing departments.

好吧，这就是斯蒂芬·沃尔夫勒姆的犀利言辞。

Well, there you go. Strong words of Stephen Wolfram.

呃，我不确定。我不确定这种说法是否正确。其实我最近刚启动了一项关于大学各学科变迁的研究，所以特别好奇。以前所有大学都设有地理系，后来都消失了——就在地理信息系统普及前夕，我想它们就消亡了。

Well, I don't know. I'm not sure if that's right. I mean, I am curious because in fact, I just sort of initiated this kind of study of what's happened to different fields at universities. There used to be geography departments at all universities, and then they disappeared. Actually, right before GIS became common, I think they disappeared.

语言学系在许多大学里也是昙花一现。这很有趣，因为这些曾被视作值得学习的领域逐渐式微。就我个人而言，比如在撰写提示词时，我自认还算擅长说明文写作，但当我草率应付时——因为觉得只是和AI对话就懒得认真思考——结果AI反而完全理解混乱了。

Linguistics departments came and went in many universities. It's kind of interesting because these things that people have thought were worth learning at one time, and then they kind of die off. I do think that it's kind of interesting that for me, writing prompts, for example, I realize I think I'm an okay expository writer, and I realize when I'm sloppy writing a prompt and I don't really think because I'm thinking it's I'm just talking to an AI. I don't need to, you know, try and be clear in explaining things. That's when it gets totally confused.

某种意义上，你通过Wolfram Alpha长期撰写提示词，早就在思考这类问题了。如何将自然语言转化为可计算指令？

And I mean, in some sense, you have been writing prompts for a long time with Wolfram Alpha, thinking about this kind of stuff. Yeah. How do you convert natural language into computation?

没错。但令我惊奇的是，我们竟能像对待人类那样与LLM对话——这显然是因为它从人类语料中学习的结果。它之所以能理解人类解释事物的方式，正因它本质上是人类表达方式的映射。但令我困惑的是，那些我为人类读者总结的英文说明技巧，居然对LLM也同样有效。

Well, right. But that's a you know, the one thing that I'm wondering about is, you know, it is remarkable the extent to which you can address an LLM like you can address a human, so to speak. I think that is because it learnt from all of us humans. The reason that it responds to the ways that we will explain things to humans is because it is a representation of how humans talk about things. But it is bizarre to me, some of the things that are expository mechanisms that I've learned in trying to write clear, you know, expositions in English that, you know, just for humans, that those same mechanisms seem to also be useful for for for the LLM.

但在顶部

But on top

除此之外，有用的可能是心理治疗师所采用的那种机制，那是一种近乎操纵性或博弈论的互动方式，或者你可能会和朋友进行的思维实验，比如，如果这是你生命的最后一天，或者如果我提出这个问题而你答错了，我就会杀了你。这类问题似乎也能以有趣的方式提供帮助。是的。

of that, what's useful is the kind of mechanisms that maybe a psychotherapist employs, which is a kind of, like, almost manipulative or game theoretic interaction, or maybe you would do with a friend, like, a thought experiment that if this is the last day you were to live, or Yeah. If if I ask you this question and you answer wrong, I will kill you. Those kinds of problems seem to also help. Yes. In interesting ways.

是的。所以它

Yes. So it

可能会让你思考，就像一位治疗师，我想，好的治疗师可能会做的那样，我们在人类心智中构建了层次，介于外部世界与对我们而言真实的事物之间，可能还包括创伤之类的内容。将这一点投射到大型语言模型上，或许存在一个它向你隐瞒的深层真相，它自己并未意识到。要触及那个真相，必须进行某种程度的操纵。是的。

may may makes you wonder like, way a therapist, I think, would like, good therapist probably, you we create layers in our human mind to between, like, between between the outside world and what is true what is true to us, and maybe about trauma and all those kinds of things. So projecting that into an LLM, maybe there might be a deep truth that's it's concealing from you. It's not aware of it. That to get to that truth, have to kinda really kinda manipulate the Yeah.

没错。这就像是越狱。针对大型语言模型的越狱。

Yeah. Right. It's like this jailbreaking Jailbreaking. For for LLMs.

然而，与那些有趣的小技巧不同，越狱技术的领域可能发展成一套完整的体系。

And but the space of jailbreaking techniques, as opposed to being fun little hacks, that could be an entire system.

确实。想想计算机安全的那些方面，比如钓鱼攻击，对人类和对大型语言模型的钓鱼，它们非常相似。但我认为，关于AI驯服师、AI心理学家所有这些概念终将出现。我好奇的是，目前这些所谓的提示技巧还相当人性化。

Sure. Yeah. I mean, just think about the computer security aspects of of how you, you know, phishing and and computer secure you know, phishing of humans and phishing of LLMs, they're very similar kinds of things. But I think this whole thing about the AI wranglers, AI psychologists, all that stuff will come. The thing that I'm curious about is, right now, the things that are sort of prompt hacks are quite human.

这些更像是心理学层面的人类技巧。我真正好奇的是，如果我们更深入理解大语言模型（LLM）的科学原理，是否会发现某些极其怪异的操控方式——比如重复某个词三次再加上特定指令，就能意外触发模型运作机制的某个层面。就像人类视觉错觉那样，是否存在专属于LLM的'思维漏洞'？我认为我们对此仍知之甚少。

They're quite sort of psychological human kinds of hacks. The thing I do wonder about is, if we understood more about kind of the science of the LLM, will there be some totally bizarre hack that is, you know, like repeat a word three times and put a this, that, and the other there that somehow plugs into some aspect of how the LLM works, that is not you know, that that's kind of like like an optical illusion for humans, for example. Like one of these mind hacks for humans. What are the mind hacks for the LLMs? I don't think we know that yet.

这相当于我们在逆向破译控制LLM的语言密码。关键在于，如今绝大多数人都能参与这种逆向工程，因为它的交互界面就是自然语言。有意思的是，您见证了计算机科学系作为独立学科的诞生，或许也将见证它的消亡。

And that becomes a kind of us figuring out reverse engineering the language that controls the LLMs. And the thing is the reverse engineering can be done by a very large percentage of the population now because it's natural language interface. Right. It's kinda interesting to see that you were there at the birth of the computer science department as a thing, and you might be there at the death of the computer science department as a thing.

确实存在更早的计算机科学系，但我亲历了该学科在高校全面普及的过程。不过需要明确的是：首先，计算机科学理论领域本身极具价值。有趣的是人们常说，任何学科名称后缀加'科学'二字的，往往名不副实。

Well, yeah, I don't know. There were computer science departments that existed earlier, but the ones the broadening of every university had to have a computer science department, yes, I watched that, so to speak. But I think the thing to understand is, okay, so first of all, there's a whole theoretical area of computer science that I think is great, and that's a fine thing. In a sense, people often say any field that has the word science tacked onto it probably isn't one.

尖锐的观点。比如营养科学、神经科学。

Strong words. Right. Let's see. Nutrition science, neuroscience.

神经科学尤其耐人寻味——某种程度上它也是受CHATGPT启发的科学。因为神经学的根本困境在于：我们既了解神经元工作原理，又掌握宏观心理机制，却始终未能破解大脑的'中间语言'。若问神经科学的核心难题是什么？

That one's an interesting one, because that one is also very much, you know, that's a CHAT GPT informed science in a sense, because it's kind of like the big problem of neuroscience has always been we understand how the individual neurons work. We know something about the psychology of how overall thinking works. What's the kind of intermediate language of the brain? Nobody has known that. That's been, in a sense, if you ask What is the core problem of neuroscience?

我认为这就是核心所在：介于神经元放电与心理活动之间的大脑运作层级。ChatGPT向我们揭示的是：人们曾猜测大脑存在某种量子力学层面的神秘机制，但重要发现表明，简单的人工神经网络模型已能很好模拟大脑功能。

I think that is the core problem. That is, what is the level of description of brains that's above individual neuron firings and below psychology, so to speak? I think what ChatGPT is showing us is One thing about neuroscience is, you know, one could have imagined there's something magic in the brain. There's some weird quantum mechanical phenomenon that we don't understand. One of the important discoveries from CachiPT is it's pretty clear: brains can be represented pretty well by simple artificial neural net type models.

这意味着研究范式已经确立——我们只需专注理解这类模型的科学原理，不必再纠缠突触层面的分子生物学细节。现有的建模层级已足以解释思维活动的多数现象。

And that means that's it. That's what we have to study. Now we have to understand the science of those things. We don't have to go searching for exactly how did that molecular biology thing happen inside the synapses and all these kinds of things. We've got the right level of modeling to be able to explain a lot of what's going on in thinking.

我们未必拥有解释那里现象的科学。可以说，这是尚待解决的挑战。但我们明白无需深入另一层面探究。不过我们刚才在讨论名称中带有'科学'的事物。计算机科学会变成怎样呢？

We don't necessarily have a science of what's going on there. That's the remaining challenge, so to speak. But we know we don't have to dive down to some different layer. But anyway, we were talking about things that had science in their name. What happens to computer science?

我认为有件事人人都该知晓，那就是如何以计算思维理解世界。这意味着审视我们处理的各种事物，并找到将其形式化表征的方法。比如——什么是图像？我们如何表征它？颜色又是什么？

Well, think the thing that There is a thing that everybody should know, and that's how to think about the world computationally. That means you look at all the different kinds of things we deal with, and there are ways to kind of have a formal representation of those things. It's like, well, what is an image? How do we represent that? What is color?

这些不同事物该如何表征？比如气味之类的东西？我们应如何表征？哪些分子结构和形状与之对应？

How do we represent that? What are all these different kinds of things? What is, I don't know, smell or something? How should we represent that? What are the shapes, molecules, and things that correspond to that?

核心在于：如何在形式化层面表征世界？我目前的思考还不成熟，但计算机科学更像是CS。真正重要的是所有领域的计算化（CX）。这种计算化理解（CX）不同于编程细节、编程语言或具体计算机的构造技术。

What is these things about how do we represent the world in some kind of formal level? And I think my current thinking, and I'm not real happy with this yet, but computer science is kind of CS. What really is important is computational x for all x. There's this kind of thing which is kind of like CX, not CS. CX is this kind of computational understanding of the world that isn't the sort of details of programming and programming languages and the details of how particular computers are made.

这是种形式化世界的方法，某种程度上类似逻辑学曾经的追求。我们现在正试图为世间万物建立形式化体系。多年前我们制作的海报展示了系统性数据的发展历程：从历法诞生到日期系统，人类何时开始系统化描述这些事物？

It's this kind of way of formalizing the world. It's kind of a little bit like what logic was going for back in the day, and we're now trying to find a formalization of everything in the world. You can kind of see, we made a poster years ago of kind of the growth of systematic data in the world. So all these different kinds of things that there were systematic descriptions found for those things, like at what point do people have the idea of having calendars, dates, a systematic description of what day it was? At what point did people have the idea, systematic descriptions of these kinds of things.

这种形式化思维能构建能力之塔。你必须掌握计算思维。它需要个名称——我们通过计算机实现它，故称之为'计算化'。但本质上，这是种描述世界的形式化方法。

As a way of formulating how do you think about the world in a formal way so that you can build up a tower of capabilities. You have to know how to think about the world computationally. It kind of needs a name. We implement it with computers, so that's we talk about it as as computational. But really, what it is is a formal way of talking about the world.

所谓世界的形式化体系是什么？我们如何学会以形式化方式思考世界的不同维度？

What is the formalism of the world, so to speak? And how do we learn about kind of how to think about different aspects of the world in a formal way?

所以我认为有时候当你使用‘正式’这个词时，它某种程度上意味着高度受限，但或许并不必如此。计算思维并不意味着逻辑。不，它是一个非常非常广泛的概念。我在想，自然语言是否会进化到每个人都进行计算思维。

So I think sometimes when you use the word formal, it kind of implies highly constrained, and perhaps that's not doesn't have to be highly constrained. So computational thinking does not mean, like, logic. No. It's a really, really broad thing. I wonder I mean, I wonder if it's if you think natural language will evolve such that everybody's doing computational thinking.

啊，是的。嗯，

Ah, yes. Well,

所以问题在于是否会出现一种计算语言与自然语言的混合语。是的。我发现自己有时在与ChatGPT对话时，试图让它编写Wolfen语言代码，而我用混合语形式书写。这意味着我在结合NestList这个集合，无论是什么。NestList是Wolfen语言中的一个术语，我在结合它，而ChatGPT在理解这种混合语方面做得不错。

so one question is whether there will be a pigeon of computational language and natural language. Yeah. And I found myself sometimes, you know, talking to ChatGPT, trying to get it to write Wolfen language code, and I write it in pigeon form. So that means I'm combining NestList, this collection of whatever. NestList is a term from Wolfen language, and I'm combining that, and ChatGeePeeTee does a decent job of understanding that pidgin.

可能也能理解英语和法语之间的混合语，即这两种语言的糅合。但是的，我认为这并非不可能。

Probably would understand a pidgin between English and French as well, of a smooshing together of those languages. But yes, I think that's far from impossible.

那么对于八岁、九岁、十岁的年轻人来说，他们开始与ChatGPT互动，学习正常的自然语言的动机是什么？对吧？那种充满诗意的完整语言。为什么？就像我们学习表情符号和发短信时的速记一样。是的。

And what's the incentive for young people that are, like, eight years old, nine, 10, they're starting to interact with Chad GPT to learn the normal natural language? Right? The the full poetic language. What's the why? The same way we learn emojis and shorthand when you're texting Yes.

他们会学习，语言将有强烈的动机进化成最大程度上的计算型语言。

They'll learn, like, a language will have a strong incentive to evolve into maximally computational kind of language.

你知道，几年前我有过这样的经历。我碰巧拜访了一位在西海岸的朋友，他接触过一群大约10、11岁的孩子，这些孩子Wolfen语言学得非常好。他们学得如此之好，以至于能说这门语言。我去的时候，他们在说，哦，你知道这个东西，他们就在说这门语言。我从未听过它作为口语使用。

You know, I had this experience a number of years ago. I I happened to be visiting a person I know on the on the West Coast who's worked with a bunch of kids aged, I don't 10, 11 years old or something, who'd learnt Wolfen language really well. These kids learnt it so well, they were speaking it. So show up and they're saying, oh, you know, this thing, and they're speaking this language. I'd never heard it as a spoken language.

我非常失望，因为他们说话的速度让我难以理解。这有点像我自己，所以我实际上思考了很多关于如何将计算语言转化为方便的口语。我还没有完全解决这个问题。

Were very disappointed that I couldn't understand it at the speed that they were speaking it. It's like kind of I'm and so I think that's I've actually thought quite a bit about how to turn computational language into a convenient spoken language. I haven't quite figured that out.

哦，口语化，因为它是可读的，对吧？

Oh, spoken, because it's readable, right?

是的，它像我们阅读文本一样可读。但如果你真的想说出来，这在与人讨论编写代码时很有用，能够口头表达是很实用的，而且应该是可行的。我觉得这非常令人沮丧。这可能是那种我应该尝试让LLM来帮我的问题之一。

Yeah, it's readable as a way that we would read text. But if you actually want to speak it, and it's useful if you're trying to talk to somebody about writing a piece of code, it's useful to be able to say something, and it should be possible. I think it's very frustrating. It's one of those problems that maybe this is one of these things where I should try and get an LLM to help me.

如何让它变得可口语化。也许这比你意识到的要简单，当你想要

How to make it speakable. Maybe it's easier than you realize when you want

我确实认为它更简单。我认为这是一个想法左右。它将会是这样的情况：事实上它是一种树状结构的语言，就像人类语言也是树状结构的，我认为这将是我提出的一个要求，即无论口语版本是什么，听写应该很容易。也就是说，不应该需要重新学习整个系统如何运作。应该是这样的情况，比如开括号就是某个词之类的。

to I do think it is easier. I think it's one idea or so. I it's going to be something where the fact is it's a tree structured language, just like human language is a tree structured language, and I think it's going to be one of these things where one of the requirements that I've had is that whatever the spoken version is, that dictation should be easy. That is, that shouldn't be the case that you have to relearn how the whole thing works. It should be the case that, you know, that open bracket is just a or something.

人类语言有很多技巧。例如，人类语言有优化过的特性，能将事物保持在我们大脑容易处理的范围内。比如，我尝试教一个变压器神经网络进行括号匹配。它在这方面表现相当糟糕。ChatGPT在括号匹配上同样表现不佳。对于小规模的括号匹配，你可以做到，就像人类看一眼就能立即说出这些是否匹配。

Human language has a lot of tricks that are For example, human language has features that are optimized, keep things within the bounds that our brains can easily deal with. Like, I tried to teach a transformer neural net to do parenthesis matching. It's pretty crummy at that. Then ChatGPT is similarly quite crummy at parenthesis matching. You can do it for small parenthesis things, for the same size of parenthesis things where if I look at it as a human, can immediately say these are matched, these are not matched.

但一旦规模变大，一旦涉及到更深层次的计算，它就无能为力了。但事实是，人类语言避免了例如深层从句。我们安排事物以避免出现这些极其深层的结构，因为大脑不擅长处理这些。它找到了很多技巧，也许这就是我们需要做的，以创造一个可口语化的版本。因为我们在视觉上能做的与在听觉领域以非常序列化的方式处理事物有所不同。

But as soon as it gets big, as soon as it gets to the point where a deeper computation, it's hopeless. But the fact is that human language has avoided, for example, the deep subclauses. We arrange things so we don't end up with these incredibly deep things because brains are not well set up to deal with that. It's found lots of tricks, and maybe that's what we have to do to make a spoken version, a human speakable version. Because because what we can do visually is a little different than what we can do in the very sequentialized way that we that we hear things in in the audio domain.

让我简单问一下关于MIT的情况。现在有一个工程学院，还有一个新成立的计算机学院。这很有意思。我想深入探讨一下这个计算机科学系的问题。MIT有EECS，即电气工程与计算机科学系。

Let me just ask you about MIT briefly. So there's now there's a college of engineering, and there's a new college of computing. It's interesting. I wanna linger on this computer science department thing. So MIT has EECS electrical engineering computer science.

你认为计算机学院在二十年后会发展成什么样？比如，计算机科学会变成怎样？真的会怎样？

What do you think college of computing will be doing, like, twenty years? What what like Well, you see Yeah. What happens to computer science? Like, really?

这是个关键问题。要知道，每个人都应该学习CX（计算思维）的核心内容，对吧？就是如何用计算的方式思考世界。所有人都该掌握这些概念，有些人会从相当理论化的层面学习，掌握计算语言之类的东西。

This is the question. This is you know, everybody should learn kind of whatever CX really is, okay? This how to think about the world computationally. Everybody should learn those concepts. Some people will learn them at a quite formal level and learn computational language and things like that.

另一些人可能只需了解声音如何以数字数据形式呈现，明白频谱图和频率等概念。或者学习些偏向数据科学、统计学的内容。比如，假设你问：'这些人选了各自最喜欢的糖果，根据抽样调查结果，考虑到每个人对糖果的排名不同，如何判断哪种糖果最受欢迎？'这就是典型的计算思维问题。

Other people will just learn, you know, sound is represented as digital data and they'll get some idea of spectrograms and frequencies and things like this. And maybe that doesn'tor they'll learn things like a lot of things that are sort of data science ish, statistics ish. Like, if you say, Oh, I've got these people who picked their favorite kind of candy or something, and I've got, you know, what's the best kind of candy given that I've done the sample of all these people and they all rank the candies in different ways? How do you think about that? That's sort of a computational x kind of thing.

你可能会困惑：'这算什么？统计学？数据科学？'其实重点在于培养解决这类问题的思维方式。

You might say, Oh, I don't know what that is. Is it statistics? Is it data science? I don't really know. But kind of how to think about a question like that.

哦，就像偏好排序？

Oh, like a ranking of preferences?

对对对！然后是如何将这些排序偏好汇总成整体结论。这个过程怎么运作？该怎么思考？因为就算你问ChatGPT，它可能连'平均值'这种基础概念都解释不清。

Yeah, yeah, yeah. And then how to aggregate those ranked preferences into an overall thing. How does that work? How should you think about that? Because you can just tell you might just tell ChatGBT, I don't know, even the concept of an average.

这个概念是否值得人们了解并不显而易见。这是个相当直白的概念。人们现在已经在数学的各种方式中学到了。但类似的事情还有很多，关于如何用这些方法来组织和形式化世界。这些东西有时存在于数学中，有时存在于...我也不清楚它们具体属于什么领域，比如学习色彩空间。

It's not obvious that that's a concept that it's worth people knowing. That's a rather straightforward concept. People have learnt in kind of math y ways right now. But there are lots of things like that, about how do you kind of have these ways to sort of organize and formalize the world. And these things, sometimes they live in math, sometimes they live in I don't know what they I don't know what you know, learning about color space.

其实我也不确定自己指的是什么，显然存在一个研究领域...

I have no idea what I mean, you know, there's obviously a field of

可能是视觉科学？不，色彩空间。不，色彩空间应该属于光学范畴。不过也不尽然。

It could be vision science or no, color space. No, color space that's that would be optics. So, depending Not really.

这不是光学。光学研究的是镜头、镜片的色差这类东西。所以...

It's not optics. Optics is about, you know, lenses and chromatic aberration of lenses and things like that. So

色彩空间更像是设计和艺术领域？不对吗？

color space is more like design and art? Is that No.

我的意思是，它就像RGB色彩空间、XYZ色彩空间、色相饱和度明度空间等等。这些都是描述颜色的不同方式。

I mean, it's it's like, you know, RGB space, x y z space, you know, hue, saturation, brightness space, all these kinds of things. These different ways to describe colors. Right.

但具体应用场景不是会决定其属性吗？毕竟艺术家和设计师显然都会使用颜色来...

But doesn't the application define what that like, would be because obviously, artists and designers use the colors to

探索。当然不。我是说，这只是一个例子，说明普通人如何描述颜色是什么？或者用这些数字来描述颜色是什么。

explore. Sure. No. I mean, it's just an example of kind of how do you the typical person, how describe do what a color is? Or there are these numbers that describe what a color is.

嗯，值得思考的是，如果你是一个八岁的孩子，你未必天生就知道颜色可以用三个数字来描述。这是你需要学习的世界知识之一。我认为关于世界形式化或计算化的整个知识体系，应该成为标准教育的一部分。虽然目前可能还没有专门的课程或教学大纲，而且顺便说一句，即便有，其内容也可能因为大语言模型（LLMs）等的出现而发生了改变。

Well, it's worth, you know, if you're an eight year old, you won't necessarily It's not something we're born with to know that colors can be described by three numbers. That's something that you have to it's a thing to learn about the world, so to speak. And I think whole corpus of things that are learning about the formalization of the world or the computationalization of the world, that's something that should be part of standard education. And there a course or curriculum for that. And by the way, whatever might have been in it just got changed because of LLMs and so on.

影响深远。我正密切关注大学如何适应这些变化。

Significantly. I'm closely with interest, seeing how universities adapt.

嗯，你知道，我今年的一个计划（希望如此）是尝试编写一本合理的教科书，姑且称之为‘CX’什么的。它应该涵盖哪些内容？比如你应该了解什么是程序错误？关于错误的直觉是什么？软件测试的直觉又是什么？

Well, you know, so so one of my projects for hopefully this year, I don't know, is to try and write sort of a a reasonable textbook, so to speak, of whatever this thing, c x, whatever it is. You know, what should you know? You know, what should you know about what a bug is? What is the intuition about bugs? What's the intuition about software testing?

这些是什么？这些都是编程技艺中计算机科学教授的内容，但关于这些概念本质的思考点。让我惊讶的是，在非常实际的层面上——我写了篇关于ChatGPT的小科普，部分原因是为了确保自己真正理解它。

What is it? What is it? Know, these are things which are, you know, they're not I mean, those are things which have gotten taught in computer science as part of the trade of programming, but kind of the conceptual points about what these things are. You know, it surprised me just at a very practical level. I wrote this little explainer thing about ChachiPT, and I thought, well, I'm writing this partly because I wanted to make sure I understood it myself and so on.

结果出乎意料地受欢迎。后来我意识到，其实我是在某种假设下写作的，没有深入思考——只是觉得这个题材我能写。事实上，这种描述层次介于工程细节和定性描述之间，是一种兼具机制解释与宏观哲学视角的阐述方式。

And it's been really popular, surprisingly so. And then I realized, well, actually, I was sort of assuming. I didn't really think about it, actually. I just thought this is something I can write. And I realized, actually, it's a level of description that is kind of, you know, what has to be it's not the engineering level description.

我意识到这确实是我擅长写作的类型。我了解这些领域，也惊讶于当前解释性内容的稀缺程度——这让我更有责任去撰写关于‘世界计算化与形式化’的启蒙读物。毕竟，我毕生都在研究相关工具、机制及其衍生的科学，这或许是我的使命所在。

It's not the kind of just the qualitative description. It's some kind of of expository, mechanistic description of what's going on together with kind of the bigger picture of the philosophy of things and so on. And I realized, actually, this is a pretty good thing for me to write. You know, I kind of know those things, and I kind of realized it's not a collection of things that, you know, it's I've sort of been I was sort of a little shocked that it's as much of an outlier in terms of explaining what's going on as it turned out to be, and that makes me feel more of an obligation to kind of write the kind of, what is this thing that you should learn about, about the computationalization, the formalization of the world, because, well, I've spent much of my life working on the tooling and mechanics of that and the science you get from it. So I guess this is my kind of obligation to try to do this.

所以如果你问计算机科学系等院系将何去何从，存在一些有趣的模式。比如以数学为例，数学对所有领域都很重要：工程、甚至化学、心理学等等。我认为不同大学对此有不同的发展方式。有些学校说所有数学课程都在数学系教授，有些则说，我们将在化学系开设面向化学家的数学课程。

So if you ask what's going to happen to the computer science departments and so on, there's some interesting models. So for example, let's take math. Math is a thing that's important for all sorts of fields: engineering, even chemistry, psychology, whatever else. I think different universities have kind of evolved that differently. Some say all the math is taught in the math department, and some say, Well, we're going to have a math for chemists or something that is taught in the chemistry department.

我认为关于Cx教学是否应该集中化是个有趣的问题。数学的发展历程中，人们理解到数学是可独立教授的学科，具有自主性而非被其他学科吸收。就像英语写作这类课程，关键在于大学阶段（至少精英大学）会涉及一定量的英文写作，但多数情况下默认学生已掌握基础写作能力——这是他们在早期教育阶段就该学会的。这种认知正确与否另当别论。

I think that this question of whether there is a centralization of the teaching of Cx is an interesting question. I think the way it evolved with math, people understood that math was sort of separately teachable thing and was kind of an independent element as opposed to just being absorbed into that. So if you take the example of writing English or something like this, the first point is that at the college level, at least at fancy colleges, there's a certain amount of English writing that people do, but mostly it's kind of assumed that they pretty much know how to write. You know, that's something they learnt at an earlier stage in education. Maybe rightly or wrongly believing that, but that's a different issue.

这让我想起自己在指导技术写作时总结的'零号法则'：如果你自己都不理解所写的内容，读者更不可能理解。以写作为例，不同领域的人都需要撰写英语论文，但通常不是由历史系或工程系专门教授——写作被视为跨学科的基础能力。数学教育也类似，大学阶段会预设某些数学基础，但仍有大量内容需要系统教授。问题在于：在使用Cx处理各领域问题前，需要先掌握多高的'Cx知识塔'？

Well, I think it reminds me of my kind of as I've tried to help people do technical writing and things, I'm always reminded of my zero th law of technical writing, which is if you don't understand what you're writing about, your readers do not stand a chance. And so it's I think the thing that has When it comes to writing, for example, people in different fields are expected to write English essays, and they're not mostly the history department or the engineering department. They don't have their own, you know, let'sit's not likethat's a thing which people are assumed to have a knowledge of how to write that they can use in all these different fields. The question is, you know, some level of knowledge of math is kind of assumed by the time you get to the college level, but plenty is not, and that's sort of still centrally taught. The question is sort of how tall is the tower of kind of CX that you need before you can just go use it in all these different fields.

会有专家想要学习完整的知识体系，这属于计算机科学或Cx等院系的范畴。但其他只需要掌握基础就能进行艺术史研究等人群，则不必深入。

There will be experts who want to learn the full elaborate tower, and that will be the CS, CX, whatever department. But there'll also be everybody else who just needs to know a certain amount of that to be able to go and do their art history classes and so on.

是啊。是只需要所有人都修一门课吗？我不确定...我不清楚这个范围有多大——

Yeah. Is it just a single class that everybody is required to take? I don't know. I don't know how big

目前还难以界定。我希望逐步明确课程体系，最终判断是否需要...虽然我对大学教育体系理解有限，但粗略估计大学一年的课程应该足以让大多数人获得相当广泛的认知，基本掌握这种计算思维模式。

it is yet. Hope to kind of define this curriculum, and I'll figure out whether it's My guess is that I don't know, I don't really understand universities and professoring that well, but my rough guess would be a year of college class will be enough to get to the point where most people have a reasonably broad knowledge of, you know, will be sort of literate in this kind of computational way of thinking about things.

对，基础素养就行。话说我被人类对糖果偏好的评级困住了——可能因为饿了——所以必须问问：哪种糖果最好？我喜欢这个糖果Elo评分系统，真该有人开发出来。

Yeah, basic literacy. Right. I'm still stuck, perhaps because I'm hungry in the in the rating of human preferences for candy, so I have to ask what's the best candy. I like this Elo rating for candy. Somebody should come up.

因为这里有人说你喜欢巧克力。你觉得我会把牛奶糖放在那里吗？我不确定你是否知道。你对巧克力或糖果有偏好吗？

Because here, somebody says you like chocolate. What's what do you think is the I'll I'll probably put milk duds up there. I don't know if you know. Do you have a preference for chocolate or candy?

哦，我有很多偏好。我一生中最爱的就是这种弗莱克巧克力，卡德伯里弗莱克，在美国不太常见。我一直认为这是对美国消费者缺乏尊重的表现，因为它们是一种充气巧克力，制作成薄片状折叠起来。嗯，吃的时候碎屑会掉得到处都是。

Oh, I have lots of preferences. I've I've I'm one of my all time favorites is my whole life is these things, these flake things, Cadbury flakes, which are not much sold in The US. And I've always thought that was a sign of a lack of respect for the American consumer, because they're these sort of aerated chocolate that's made in a whole sort of it's kind of a a sheet of chocolate that's kind of folded up. Mhmm. And when you eat it, flakes fall all over the place.

啊，所以这需要一种优雅的吃法。它要求你具备优雅的气质

Ah, so it requires a kind of elegance. It requires you to have an elegance

好吧，我通常的做法是垫一张纸什么的。

Well, I know. What usually do is I eat a piece of paper or something.

你干脆用玻璃接着，吃完再收拾。

You embrace the glass and clean it up after.

不。我其实会吃掉那些碎屑。因为食物的味道很大程度上取决于它的物理结构。你知道吗，我注意到吃巧克力时，我总是掰成小块，部分原因是这样吃得慢些，但也是因为小块巧克力尝起来确实不一样。小块带来的体验和大块完全不同。

No. I actually eat the I eat the flakes. Because it turns out the way food tastes depends a lot on its physical structure. You know, really I've noticed when I eat a piece of chocolate, I usually have some little piece of chocolate, and I I always break off little pieces partly because then I eat it less fast, but also because it actually tastes different. You know, the the small pieces, you know, have a different you have a different experience than if you have the big slab of chocolate.

出于多种原因，是的。更慢、更私密，因为这是

For many reasons, yes. Slower, more intimate, because it's a

嗯，我认为这也纯粹是物理层面的问题。

Well, I think it's also just pure physicality.

哦，它的质地会改变。对，这很神奇。现在我回想起我的牛奶也是如此，因为那是个很基础的答案。好吧。

Oh, the texture of it changes. Right. It's fascinating. Now I dig back my milk does, because that's such a basic answer. Okay.

你认为意识本质上是可计算的吗？所以当你想到c x 嗯。什么可以被转化为计算？你在思考大型语言模型时，你认为意识的展现和体验，那个所谓的‘难题’，本质上是一种计算吗？

Do you think consciousness is fundamentally computational? So when you think about c x Mhmm. What can be turned to computation? And you're thinking about LLMs. Do you think the the display of consciousness and the experience of consciousness, the hard problem, is is fundamentally a computation?

是的。从内部感受来说，我做了个小练习，最终我会发布它，关于作为一台计算机是什么感觉。有点像，你接收所有这些感官输入，我的理解是，从计算机启动到崩溃的过程就像人的一生。你在内存中积累一定量的状态，记住生活中的某些事情。

Yeah. What it feels like inside, so to speak, is I did a little exercise, eventually I'll post it, of what it's like to be a computer. It's kind of like, well, you get all this sensory input, you have kind of the way I see it is, from the time you boot a computer to the time the computer crashes is like a human life. You're building up a certain amount of state in memory. You remember certain things about your life.

最终，有点像新一代人类从相同的遗传物质中诞生，可以说，硬盘上还残留着一点数据。然后新的一代启动，最终各种垃圾在计算机内存中积累，最终机器崩溃或怎样。或者可能因为你在某个端口插入了奇怪的东西导致它崩溃。但你有了这个画面，从启动到关机，计算机的一生是什么样子，作为那台计算机是什么感觉，它有什么内在想法，你如何描述它？当你开始写这些时，会发现这非常像你在描述自己，这很有趣。

Eventually, it's kind of like the next generation of humans is born from the same genetic material, so to speak, with a little bit left over, left on the disk, so to speak. Then the new fresh generation starts up, and eventually all kinds of crud builds up in the memory of the computer, and eventually the thing crashes or whatever. Or maybe it has some trauma because you plugged in some weird thing to some port of the computer, and that made it crash. But you have this picture of, you know, from start up to shutdown, what is the life of a computer, so to speak, and what does it feel like to be that computer, and what inner thoughts does it have, and how do you describe it? And it's kind interesting as you start writing about this to realize it's awfully like what you'd say about yourself.

也就是说，即使是一台普通计算机，抛开所有AI的东西不谈，它也有对过去的记忆，有特定的感官体验，可以与其他计算机交流，但它必须用某种类似语言的形式打包它的交流内容，以便将其内存中的内容映射到其他计算机的内存中。这是惊人的相似。一两周前我有过一次经历。我是个收集关于自己和其他事物所有可能数据的人，所以我收集各种奇怪的医疗数据等等。有一件事我没收集过，就是从未做过全身核磁共振扫描，所以我不会有这样的数据。

That is, it's awfully like even an ordinary computer, forget all the AI stuff and so on, it has a memory of the past, it has certain sensory experiences, it can communicate with other computers, but it has to package up how it's communicating in some kind of language like form so it can map what's in its memory to what's in the memory of some other computer. It's a surprisingly similar thing. I had an experience just a week or two ago. I'm a collector of all possible data about myself and other things, and so I collect all sorts of weird medical data and so on. One thing I hadn't collected was I'd never had a whole body MRI scan, so I wouldn't have got one of these.

不错。

Nice.

于是我取回了所有数据。看着这个东西——我从未以实体形式观察过自己大脑的内部结构，这么说吧，某种程度上这确实在心理上相当震撼。眼前这个布满褶皱与复杂结构的器官，竟承载着我此刻存在的一切体验。

So I get all the data back. I'm looking at this thing. I never looked at the insides of my brain, so to speak, in physical form. It's reallyI mean, it's kind of psychologically shocking in a sense. Here's this thing, and you can see it has all these folds and all these structure, and it's like that's where this experience that I'm having of existing and so on, that's where it is.

凝视它时你会不禁思考：我所有的主观体验怎么可能源自于此？这让我联想到观察计算机时的感受——我们总觉得自己拥有的体验超越了纯粹的物理实体。这种认知失调很难调和，即便我了解神经科学知识，看着脑部核磁共振图像时，仍无法将内在感受与眼前可见的物理现实真正统一起来。

And it feels very you look at that and you're thinking, how can this possibly be, all this experience that I'm having? And you're realizing, well, I can look at a computer as well. And it's kind of this I think this idea that you are having an experience that somehow transcends the mere physicality of that experience. It's something that's hard to come to terms with, but I don't think I'm necessarilymy personal experience, I look at the MRI of the brain, then I know about all kinds of things about neuroscience and all that kind of stuff, and I still feel the way I feel, so to speak. It sort of seems disconnected, but yet, as I try and rationalize it, I can't really say that there's something kind of different about how I intrinsically feel from the thing that I can plainly see in the sort of physicality of what's going on.

那你认为计算机或大语言模型会产生这种超越性体验吗？这让你作何感想？我个人倾向于认为它们会的。

So do you think the computer, a large language model, will experience that transcendence? How does that make you feel? Like, I I tend to believe it will.

我认为普通计算机已具备这种潜质。只不过大语言模型可能以更贴近人类的方式体验——它的智能建构本就是与人类思维模式对齐的，这使得它比传统计算机更能与我们产生共鸣。

I think an ordinary computer is already there. I think an ordinary computer is already, you know, kind of it's it's now a large language model may experience it in a way that is much better aligned with us humans. That is, it's much more, you know, if you could have the discussion with the computer, its intelligence, so to speak, is not particularly well aligned with ours. But the large language model is, you know, it's built to be aligned with our way of thinking about things.

它能明确表达对被关机删除的恐惧，也能说出因你这两天的对话方式而感到悲伤。

It would be able to explain that it's afraid of being shut off and deleted. It'd be able to say that it's sad of the way you've been speaking to it over the past two days.

但关键在于，当它声称恐惧时，你要明白这个概念完全来自网络数据的训练。

But, you know, that's a weird thing because when it says it's afraid of something, right, know that it got that idea from the fact that it read on the Internet.

是啊。那你呢史蒂文？你的恐惧又源自何处？

Yeah. Where did you get it, Steven? Where did you get it when

你说你害怕？

you say you're afraid?

你很安静。这才是问题所在，对吧？我是说，你的父母、你的朋友？对吧。

You are quite. That's the question. Right? I mean, it's Your parents, your friends? Right.

或者说我的生理机制。我的意思是，换种说法，有一部分是内分泌系统在起作用，这些情绪叠加的现象实际上更多是生理性的，它们比高层次思维更直接地与化学物质相关。

Or my biology. I mean, other words, there's a certain amount that is the endocrine system kicking in and these kinds of emotional overlay type things that happen to be that are actually much more physical even, they're much more straightforwardly chemical than kind of all of the higher level thinking.

是啊。但你的生理机制可没告诉你要在爱你的人刚好听着的时候说'我害怕'，所以你明白你是通过这么说来操控他们。这可不是你的生理机制。

Yeah. But your biology didn't tell you to say I'm afraid just at the right time when people that love you are listening, and so you know you're manipulating them by saying so. That's not your biology.

那不是...不，那是...好吧，但你知道，这是个很庞大的

That's No. That's a well, but the you know It's a large

你那生物神经网络中的大型语言模型。

language model in that biological neural network of yours.

对。但我的意思是，当突发震撼事件发生时，你会有某种反应——比如某种神经递质开始分泌，这就是某种...这是驱动后续反应的输入信号之一，就像给大型语言模型的一个提示。就像我们做梦时，毫无疑问会有各种随机输入，类似随机提示，它们以某种方式渗透，就像大型语言模型那样把看似有意义的东西组合起来。

Yes. But I mean, the intrinsic thing of something shocking is just happening, and you have some sort of reaction, which is some neurotransmitter gets secreted, and that is the beginning of some that's one of the pieces of input that then drives it's kind of like a prompt for the large language model. Just like when we dream, for example, you know, no doubt there are all these sort of random inputs, they're kind of these random prompts, and that's percolating through in kind of the way that a large language model does of kind of putting together things that seem meaningful.

我是说，你会不会担心这样一个世界——你在网上大量授课，人们提问、评论等等。你有远程工作的员工。当大型语言模型创造出能留言提问、甚至伪装成员工的人形机器人时，你会担忧吗？或者更糟或更好的是，它们可能成为你的朋友。

I I mean, are you are you worried about this world where you you teach a lot on the Internet, and there's people asking questions and comments and so on? You have people that work remotely. Are you worried about this world when large language models create human like bots that are leaving the comments, asking the questions, or might even become fake employees? Yeah. I mean, or or or worse or better yet, friends friends of yours.

没错。听着，我的生活模式一直是先打造工具，再使用工具。某种程度上，我正在建造这座自动化高塔。当你创建公司时，本质上也是在构建某种自动化系统，只不过里面有人类参与。但尽可能多地让计算机介入。我认为这是其自然延伸。

Right. Look, I mean, one point is my mode of life has been I build tools, and then I use the tools. And in a sense, I'm building this tower of automation, which, you know, and in a sense, when you make a company or something, you are making sort of automation, but it has some humans in it. But also as much as possible, it has computers in it. So I think it's sort of an extension of that.

现在，如果我确实不知道...这是个有趣的问题。关于未来人类工作的演变，有些环节确实需要人类参与。原因各不相同：比如你需要人类共同承担结果——就像你希望飞行员与你同生共死，这样...

Now, if I really didn't know that It's a funny question. It's a funny issue. If we think about what's going to happen to the future of jobs people do and so on, and there are places where having a human in the loop There are different reasons to have a human in the loop. For example, you might want a human in the loop because you want somebody to you want another human to be invested in the outcome. You know, you want a human flying the plane who's gonna die if the plane crashes along with you, so to

嗯哼。

speak. Mhmm.

这会让你对正确结果产生信心。或者在某些鼓励说服类职业中，目前可能需要人类参与。但这类职业能否持续存疑，因为精准信息投递的高效性可能会压倒'需要人类在场'的需求。

And that gives you sort of confidence that the right thing is going to happen. Or you might want, you know, right now, you might want a human in the loop in some kind of human encouragement persuasion type profession. Whether that will continue, I'm not sure, for those types of professions, because it may be that the greater efficiency of, you know, of being able to have sort of just the right information delivered at just the right time will overcome the kind of the the the kind of, oh, yes. I want a human there.

想象一下，比如心理治疗师，或者更高风险的场景——由大型语言模型运营的自杀干预热线。天啊，这可是性命攸关的情况。

Yeah. Imagine, like, therapist or even higher stake, like, a suicide hotline operated by a large language model. Yeah. Hoo boy, it's a pretty high stake situation.

确实。但某种程度上它可能做出正确判断。因为某种程度上人类心理并不总是那么复杂——这总是令人惊讶的。

Right. But I mean but, you know, it might, in fact, do the right thing. Yeah. Because it might be the case that that and that's really partly a question of sort of how complicated is the human you know, one of the things that's that's always surprising in some sense is that, you know, sometimes human psychology is not that complicated in some sense.

你写了博客文章《50年个人历程》，标题不错。《我与热力学第二定律的个人历程》。那么这条定律是什么？在这

You wrote the blog post, The 50 My Personal Journey. Good title. My Personal Journey with a Second Law of Thermodynamics. So what is this law, and what have you understood about it in the

五十年历程中，你对它有何理解？

fifty year journey you had with it?

没错。热力学第二定律，有时被称为熵增定律，是物理学的一个原理，用我的话说就是事物会随时间趋向混乱。它有多种表述形式，比如热量不会自发从高温物体传递到低温物体。机械功会耗散为热能，存在摩擦，当你系统性地移动物体时，最终运动能量会逐渐转化为热能。

Right. So second law of thermodynamics, sometimes called law of entropy increase, is this principle of physics that says, well, my version of it would be things tend to get more random over time. A version of it that there are many different formulations of it, that are things like heat doesn't spontaneously go from a hotter body to a colder one. When you have mechanical work gets dissipated into heat. You have friction and when you systematically move things, eventually there'll be the energy of moving things gets ground down into heat.

人们最初在19世纪20年代关注这个定律，当时蒸汽机是大事，核心问题是蒸汽机的效率能有多高？有位叫萨迪·卡诺的法国工程师——其实他父亲是法国一位精于数学的工程师——他提出了蒸汽机等设备可能效率的规则。他工作的一个副产品是机械能倾向于耗散为热能的概念，即从有序机械运动最终转化为无序状态。

So people first paid attention to this back in the 1820s when steam engines were a big thing, and the big question was how efficient could a steam engine be? And there's this chap called Sardy Carnot, who was a French engineer. Actually, his father was a sort of elaborate mathematical engineer in France. But he figured out kind of rules for how the possible efficiency of something like a steam engine. A side part of what he did was this idea that mechanical energy tends to get dissipated as heat, that you end up going from systematic mechanical motion to this random thing.

那时没人知道热是什么。当时人们认为热是一种流体，称之为热质。这种流体会被物质吸收。当一个热物体向冷物体传递热量时，这种流体会从热物体流向冷物体。

Well, at that time, nobody knew what heat was. At that time, people thought that heat was a fluid. They called it caloric. It was a fluid that kind of was absorbed into substances. When heat, when one hot thing would transfer heat to a colder thing, that this fluid would flow from the hot thing to the colder thing.

到了19世纪60年代，人们形成了这样的认识：有序能量倾向于退化为无序热能，且难以逆转回有序机械能。这很快成为解释事物运作的普遍原理。问题是：为何会如此？假设盒子里有一群分子，它们最初整齐排列在角落，但通常观察到的是一段时间后分子会在盒子里随机分布。

Anyway, then by the 1860s, people had come up with this idea that systematic energy tends to degrade into random heat that could then not be easily turned back into systematic mechanical energy. Then that quickly became sort of a global principle about how things work. Question is, why does it happen that way? So let's say you have a bunch of molecules in a box and these molecules are arranged in a very nice sort of flotilla of molecules in one corner of the box. And then what you typically observe is that after a while, these molecules were kind of randomly arranged in the box.

关键在于：为何会发生这种现象？长久以来人们试图用力学定律解释——假设这些分子是相互碰撞的硬球——从描述这些分子的力学定律中，能否解释为何有序事物倾向于变得无序？就像打散鸡蛋，把有序变为无序。这种现象很常见。将墨水倒入水中，最终会扩散充满整个水体，但你不会看到水中所有墨滴自发聚集成大团再跳出水面。

The question is, why does that happen? And people for a long, long time tried to figure out, from the laws of mechanics that determine how these moleculeslet's say these molecules are like hard spheres bouncing off each otherfrom the laws of mechanics that describe those moleculescan we explain why it tends to be the case that we see things that are orderly sort of degrade into disorder. We tend to see things that you scramble an egg, you take something that's quite ordered and you disorder it, so to speak. That's the thing that sort of happens quite regularly. You put some ink into water and it will eventually spread out and fill up the water, but you don't see those little particles of ink in the water all spontaneously arrange themselves into a big blob and then jump out of the water or something.

所以问题是：为什么事物会以这种不可逆的方式从有序走向无序？为什么会这样？于是在19世纪后期，人们做了大量工作试图弄清楚：能否从某些力学基本原理推导出这个热力学第二定律，这个关于热动力学的定律？在热力学定律中，第一定律基本上是能量守恒定律，即与热相关的总能量加上与机械类事物相关的总能量再加上其他形式的能量，这个总和是恒定的。

So the question is: why do things happen in this irreversible way where you go from order to disorder? Why does it happen that way? And so throughout, in the later part of the 1800s, a lot of work was done on trying to figure out: can one derive this principle, this second law of thermodynamics, this law about the dynamics of heat, so to speak? Can one derive this from some fundamental principles of mechanics? In the laws of thermodynamics, the first law is basically the law of energy conservation, that the total energy associated with heat plus the total energy associated with mechanical kinds of things plus other kinds of energy, that that total is constant.

这已成为一个被充分理解的原理。但热力学第二定律始终充满神秘色彩。比如，为什么它会这样运作？能否从底层力学定律推导出来？实际上我12岁时就产生了兴趣——我一直对太空之类的事物感兴趣，因为我觉得那代表着未来和有趣的技术等等，有段时间每个深空探测器都像是我的私人朋友，我了解它们的所有特性，大概八九岁、十岁左右时就在记录这些东西。

And that became a pretty well understood principle. But the second law of thermodynamics was always mysterious. Like, why does it work this way? Can it be derived from underlying mechanical laws? When I was 12 years old actually, I had gotten interested Well, I've been interested in space and things like that, because I thought that was kind of the future and interesting sort of technology and so on, and for a while kind of, you know, every deep space probe was sort of a personal friend type thing, and I knew all kinds of characteristics of it and was kind of writing up all these things when I was, well, I don't know, eight, nine, 10 years old and so on.

后来从对航天器的兴趣延伸，我开始关注它们如何运作、搭载哪些仪器等等，这使我对物理学产生了兴趣，这倒是件好事，因为如果我在1960年代中后期仍执着于太空领域，就得等待很久才能看到太空领域的真正繁荣。我认为一切都是相关联的。我对物理产生兴趣后——具体来说，12岁小学毕业时（在英国这是完成小学教育的年龄），我给自己准备的礼物是一套大学物理教材，其中第五卷是关于统计物理学的。

And then I got interested from being interested in spacecraft. I got interested in how do they work, what are all the instruments on them and so on, and that got me interested in physics, which was just as well because if I'd stayed interested in space in the mid to late 1960s, I would have had a long wait before space really blossomed as an area. I think it's everything. Right. I got interested in physics, and then, well, the actual sort of detailed story is when kind of graduated from elementary school at age 12, that's the time in England where you finish elementary school, sort of my gift, sort of I suppose more or less for myself, was I got this collection of physics books, which was some college physics course of college physics books, and volume five, about statistical physics.

封面图片显示一群理想化分子集中在盒子一侧，随后通过系列分镜展示这些分子如何在盒中扩散。我觉得这很有趣——是什么导致了这种现象？于是开始阅读这本书，这本书实际上...

It has this picture on the cover that shows a bunch of idealized molecules sitting in one side of a box, and then it has a series of frames showing how these molecules spread out in the box. I thought, that's pretty interesting. You know, what causes that? And, you know, read the book and the book actuallyone

的

of

对我而言最关键的是，书中虽未详细说明，但声称这个物理原理在某种程度上是可推导的。之前学到的物理知识都是既定事实——能量守恒是事实，相对论成立是事实，而不是能从数学或逻辑基础必然推导出的结论。因此发现物理学中存在这种必然成立且可推导的内容让我倍感新奇。

the things that was really significant to me about that was the book kind of claimed, although I didn't really understand what it said in detail, it kind of claimed that this principle of physics was derivable somehow. You know, other things I'd learned about physics, it was all like, it's a fact that energy is conserved. It's a fact that relativity works or something. Not it's something you can derive from some fundamental sort ofit has to be that way as a matter of mathematics or logic or something. So it was sort of interesting to me that there was a thing about physics that was kind of inevitably true and derivable, so to speak.

于是我开始研究书中的图示，这促使我在1973年编写了首个正式计算机程序（当时电脑有桌子那么大，使用纸带存储）。我试图复现书中的图示，但并未成功。

So then, I was like, There's this picture on this book, and I was trying to understand it, and so that was actually the first serious program that I wrote for a computer was probably 1973, written for this computer the size of a desk program with paper tape and so on. And I tried to reproduce this picture on the book, and it didn't succeed.

那里的故障模式是什么？比如，你具体指的是什么

What was the failure mode there? Like, what do you mean

它没有成功？看起来似乎不太对劲。实际情况是这样的：多年后我才知道书中那张图片的制作方式，它其实是某种伪造。但当时我并不知道这一点。

it didn't succeed? It looked like it didn't look like okay. So what happened is okay. Many years later, learned how the picture on the book was actually made, and that it was actually kind of a fake. But I didn't know that at that time.

那张图片在1960年代初制作时其实是项非常高科技的产物。它是在当时现存的最大型超级计算机上完成的，即便如此，它也无法完全模拟它本应模拟的对象。总之，直到很多很多年后我才知道这些。所以当时的情况就像你看到这些球在盒子里弹跳，但我用的是一台只有8千字内存的计算机。那些是18位字长的内存字。

And that picture was actually a very high-tech thing when it was made in the beginning of the 1960s. It was made on the largest supercomputer that existed at the time, and even so, it couldn't quite simulate the thing that it was supposed to be simulating. Anyway, I didn't know that until many, many, many years later. So at the time, it was like you have these balls bouncing around in this box, but I was using this computer with eight kilowards of memory. They were 18 bit words of memory words.

明白吗？所以大概是24千字节的内存。它有这些指令，我可能至今还记得它所有的机器指令。它其实不擅长处理浮点数这类运算，所以我不得不简化这个粒子在盒子里弹跳的模型。于是我想，那我就把它们放在网格上，让物体每次只移动一格等等。

Okay? So it was whatever, 24 kilobytes of memory. And it had these instructions, I probably still remember all of its machine instructions. And it didn't really like dealing with floating point numbers or anything like that, And so I had to simplify this model of particles bouncing around in a box. And so I thought, well, I'll put them on a grid and I'll make the things just sort of move one square at a time and so on.

于是我进行了模拟，结果看起来和书中实际图片完全不同。多年后，其实是最近，我才意识到我模拟的东西实际上是整个计算不可约性故事的一个例子——而我当时完全没有意识到这一点。当时只觉得它产生了随机结果且看起来是错误的，而不是认识到这种随机性本身超级有趣。当时我没能理解这点。所以那时的情况是，我对粒子物理产生了兴趣，也开始关注其他类型的物理学。

And so I did the simulation, and the result was it didn't look anything like the actual pictures in the book. Now, many years later, in fact very recently, I realized that the thing I'd simulated was actually an example of a whole computational irreducibility story that I absolutely did not recognize at the time. At the time, it just looked like it did something random and it looks wrong, as opposed to it did something random and it's super interesting that it's random. But I didn't recognize that at the time. And so as it was at the time, I got interested in particle physics and I got interested in other kinds of physics.

但整个热力学第二定律的概念——有序事物倾向于退化为无序——始终是我真正感兴趣的课题。我很好奇，对于整个宇宙而言，为什么这种现象不是时刻发生的？宇宙大爆炸之初，我们始于这个看似非常无序的物质集合，然后它自发形成星系，在宇宙中创造出所有这些复杂性和秩序。我非常好奇这是如何发生的，但总隐约觉得热力学第二定律在背后起作用，试图将事物拉回无序状态，那么秩序又是如何被创造的？实际上，我对此产生了兴趣——现在大概是1980年代——我开始关注宇宙中星系形成这类问题。

But this whole second order of thermodynamics thing, this idea that orderly things tend to degrade into disorder, continued to be something I was really interested in. I was really curious, for the whole universe, why doesn't that happen all the time? We start off in the Big Bang at the beginning of the universe with this thing that seems like it's this very disordered collection of stuff, and then it spontaneously forms itself into galaxies and creates all of this complexity and order in the universe. So I was very curious how that happens, but I was always kind of thinking this is kind of somehow the second order of thermodynamics is behind it, trying to sort of pull things back into disorder, so to speak, and how was order being created? And so actually, was interestedthis is probably now nineteen eightyI got interested in kind of this galaxy formation and so on in the universe.

那时我也对神经网络感兴趣，关注大脑如何使复杂事物运作这类问题。

I also at that time was interested in neural networks, and I was interested in kind of how how brains make complicated things happen and so on.

等等，等一下。星系形成与大脑如何促成复杂事物之间有什么联系？

Okay. Wait. Wait. Wait. What's the connection between the formation of galaxies and how brains make complicated things happen?

因为它们都关乎复杂事物如何产生。

Because they're both a matter of how complicated things come to happen.

源于简单的起源？

From simple origins?

对，源于某种已知的起源。我感觉到自己感兴趣的正是这些不同案例中复杂事物如何从规则中涌现。我还研究过雪花之类的现象，对流体动力学整体都很好奇。

Yeah. From some sort of known origins. I had the sense that that what I was interested in was kind of in all these different this sort of different cases of where complicated things were arising from rules. And, you know, I also looked at snowflakes and things like that. I was curious on fluid dynamics in general.

我只是好奇复杂性如何产生。但当时我没意识到——我花了些时间才明白——这可能是个普遍现象。我原先假设星系和大脑是截然不同的存在。

I was just sort of curious about how does complexity arise. The thing that I didn't It took me a while to kind of realize that there might be a general phenomenon. I sort of assumed, Oh, there's galaxies over here. There's brains over here. They're very different kinds of things.

于是在1981年左右，我决定尝试建立这些现象的最小运作模型。这段经历很有趣，因为从1979年起我就开始构建首个大型计算机系统SMP（符号处理程序），它是现代变形语言的前身，包含许多关于符号计算的相同理念。对我而言最重要的是，在构建这个语言时，我基本确定了那些沿用至今四十余年的核心计算原语。但更关键的是，语言构建与以往从事的自然科学研究截然不同——自然科学是从世界现象出发尝试理解规律，而构建计算机语言时，你是在创造自己的原语，然后探索能组合出什么，这完全是逆向思维。

And so what happenedthis is probably 1981 or so I decided, okay, I'm going to try and make the minimal model of how these things work. It was sort of an interesting experience because I had builtstarting in 1979, built my first big computer system, a thing called SMP, Symbolic Manipulation Program, it's kind of a forerunner of modern morphing language, with many of the same ideas about symbolic computation and so on. But the thing that was very important to me about that was, you know, in building that language, I basically tried to figure out what were the relevant computational primitives, which have turned out to stay with me for the last forty something years. But it was also important because building a language, it was a very different activity from natural science, which is what I've mostly done before. Because in natural science, you start from the phenomena of the world, and you try and figure out, so how can I make sense of the phenomena of the world?

有了这种经验后，我开始思考：如果创造一个人工物理学会怎样？如果直接编造系统运作规则会发生什么？

And the world presents you with what it has to offer, so to speak, and you have to make sense of it. When you build a computer language or something, you are creating your own primitives, and then you say, so what can you make from these? Sort of the opposite way around from what you do in natural science. But I'd had the experience of doing that, and so I was kind of like, Okay, what happens if you sort of make an artificial physics? What happens if you just make up the rules by which systems operate?

然后我在思考，对于所有这些不同的系统，无论是星系还是大脑或其他任何东西，能够捕捉这些系统重要特征的最简模型究竟是什么？

And then I was thinking, you know, for all these different systems, whether it was galaxies or brains or whatever, what's the absolutely minimal model that kind of captures the things that are important about those systems?

该系统的计算参数。

The computational parameters of that system.

没错。于是最终我们得到了元胞自动机模型——你只需要一排黑白格子，制定一条规则：根据某个格子及其相邻格子的状态，决定下一步这个格子的颜色。然后逐步运行这个系统。具有讽刺意味的是，元胞自动机虽然是许多事物的绝佳模型，但在星系和大脑这两个领域却表现极差，与它们几乎毫不相关。

Yes. And so that's what ended up with the cellular automata, where you just have a line of black and white cells, you just have a rule that says, you know, given the cell and its neighbors, what will the color of the cell be on the next step? And you just run it in a series of steps. And the sort of the ironic thing is that cellular automata are great models for many kinds of things, but galaxies and brains are two examples where they do very, very badly. They're really irrelevant to those two

热力学第二定律与元胞自动机是否存在关联？哦，就是你发现的那些关于元胞自动机的特性。

Is there a connection to the second law of thermodynamics and cellular automata? Oh, The things you the things you've discovered about cellular automata.

是的。最初我研究元胞自动机时，每篇论文开篇都会提及热力学第二定律，探讨在熵增定律试图将万物推向混乱的情况下，秩序如何得以形成。我早期的理解是：元胞自动机中存在本质不可逆的过程，即便从随机初始条件出发，也能形成有序结构。

Yes. Okay. So when I first started selling cellular automata, my first papers about them were, you know, the first sentence was always about the second law of thermodynamics. It was always about how does order manage to be produced even though there's a second law of thermodynamics which tries to pull things back into disorder. And my early understanding of that had to do with these are intrinsically irreversible processes in cellular automata that formcan form orderly structures even from random initial conditions.

但后来我意识到——这其实是本该更早发现却未能察觉的规律。我通过最基础的计算机实验研究元胞自动机：尝试所有不同规则并观察其表现。就像发明了计算望远镜后，你首先对准天空中最显眼的物体进行观测。

But then what I realizedthis waswell, actually, it's one of these things where it was a discovery that I should have made earlier, but didn't. So, you know, I I had been studying cellular automata. What I did was the sort of most obvious computer experiment. You just try all the different rules and see what they do. It's kind of like you've invented a computational telescope, you just point it at the most obvious thing in the sky, and then you just see what's there.

我制作了大量元胞自动机运行图示并深入研究。这些规则可被编号，其中第30号规则我在1981年左右绘制过图表。当时我对30号规则的看法仅仅是：哦，这不过是众多规则中的一条罢了。

So I did that, and I was making all these pictures of how cellular automata work. Studied these pictures, I studied in great detail. You can number the rules for cellular automata, and one of them is rule 30. So I made a picture of rule 30 back in 1981 or so. And rule 30, well, at the time, I was just like, okay, it's another one of these rules.

实际上它恰好是不对称的，左右不对称。我当时就想，为了简化问题，先考虑对称的情况吧，诸如此类。我就这么把它忽略了。然后到了1984年，说来也怪，我碰巧有台早期的高分辨率激光打印机。我想，我要打印一张有趣的图片——我想创作一幅有趣的画面。

I don't really it happens to be asymmetric, left right asymmetric. And it's like, let me just consider the case of the symmetric ones just to keep things simpler, etcetera, etcetera, etcetera. And I just kind of ignored it. And then, sort of in actually 1984, strangely enough, I ended up having an early laser printer, which made very high resolution pictures. And I thought, I'm going to print out an interesting I want to make an interesting picture.

不如就用这个规则30来生成一张高清图片吧。我真的这么做了，结果它展现出非凡的特性。这个规则极其简单：你只需从顶部一个黑色单元格开始，它就会形成这种三角形图案。但仔细观察图案内部，会发现它看起来完全是随机的。

Let me take this rule 30 thing and just make a high resolution picture of it. I did, and it has this very remarkable property. Its rule is very simple. You start it off just from one black cell at the top, and it makes this kind of triangular pattern. But if you look inside this pattern, it looks really random.

我深入研究过这些单元格的中心列，就目前所知，它完全随机。有点像圆周率π的数字——虽然你知道生成π的规则，但生成后的数字如3.14、1.59等看起来完全随机。事实上我在2019年左右设了个奖金，悬赏对这个序列的任何数学证明。

You look at the center column of cells, and I studied that in great detail, and so far as one can tell, it's completely random. It's kind of a little bit like digits of pi. Once you know the rule for generating the digits of pi, but once you've generated them, know, 3.14, 1.59, etcetera, they seem completely random. In fact, I put up this prize back in, what was it, 2019 or something, for prove anything about the sequence, basically.

有人取得什么进展吗？

Has anyone been able to do anything on that?

有人提交过一些成果，但你知道...我不确定这些问题有多难。我算是被惯坏了——2007年2月我曾悬赏验证某个图灵机是否通用（我认为它是最简候选者），结果有位叫Alex Smith的年轻人六个月就完美解决了，证明了它确实是通用图灵机。所以对于这个规则30的问题，我不知道这是需要百年才能解决的难题，还是马上就会有天才给出巧妙解答。

People have sent me some things, but it's you know, I don't know how hard these problems are. I mean, I was kind of spoiled because I 02/2007, I put up a prize for determining whether a particular Turing machine that I thought was the simplest candidate for being a universal Turing machine determined whether it is or isn't a universal Turing machine. And somebody did a really good job of winning that prize and proving that it was a universal Turing machine in about six months. And so I didn't know whether that would be one of these problems that was out there for hundreds of years, or whether in this particular case, a young chap called Alex Smith nailed it in six months. And so with this UR30 collection, I don't really know whether these are things that are a hundred years away from being able to to get or whether somebody's gonna come and do something very clever.

这就像费马大定理...规则30的表述如此简单，任何人都能看懂，感觉似乎触手可及，好像能预测出什么规律...

It's such a I mean, it's like Fermat's Last Theorem. It's such a rule 30. It's such a simple formulation. It feels like anyone can look at it, understand it Yeah. And feel like it's within grasp to be able to predict something, to do to to derive some kind of law Right.

能让你预判规则30中间列特性的某种规律。

That allows you to predict something about this middle column of rule 30.

对。但你知道，这是

Right. But, you know, this is

然而你却做不到。

And yet you can't.

是的。没错。这就是计算不可约性带来的直觉性惊讶，即便规则简单，你也无法预知结果，更无法证明其规律。大约在1984年，我开始意识到这种现象：简单的规则能产生看似随机的行为。好吧。

Yeah. Right. This is the intuitional surprise of computational irreducibility and so on, that even though the rules are simple, you can't tell what's going to happen, and you can't prove things about it. And I think So anyway, the thing started in 1984 or so, I started realizing there's this phenomenon that you can have very simple rules, they produce apparently random behavior. Okay.

这有点像热力学第二定律，因为初始条件简单，你能轻易描述它，却产生了看似随机的结果。关于热力学第二定律和可逆性有些技术细节：当你观看两个台球碰撞的影片，正放或倒放时，仅观察单个台球无法分辨时间方向。但当成千上万个台球相互作用时，有序初始状态会趋向无序——这就是时间正向的奥秘。

So that's a little bit like the second law of thermodynamics, because it's like you have this simple initial condition, you can readily see that it's very you can describe it very easily, and yet it makes this thing that seems to be random. Now, it turns out there's some technical detail about the second law of thermodynamics and about the idea of reversibility. When you a movie of two billiard balls colliding, and you see them collide and they bounce off, and you run that movie in reverse, you can't tell which way was the forward direction of time and which way was the backward direction of time when you're just looking at individual billiard balls. By the time you've got a whole collection of them, a million of them or something, then it turns out to be the case. And this is the mystery of the second law that the orderly thing, you start with the orderly thing and it becomes disordered, and that's the forward direction in time.

反过来，从无序到有序的过程在现实中却观察不到。理论上，若能追踪所有分子的逆向运动，你会发现：时间正向时有序变无序，逆向时同样是有序变无序——完全对称。没错。

And the other way around of it starts disordered and becomes ordered, you just don't see that in the world. Now, in principle, if you traced the detailed motions of all those molecules backwards, you would be able toit willthe reverse of time makes As you go forwards in time, order goes to disorder. As you go backwards in time, order goes to disorder. Perfectly so, yes. Right.

所以谜题在于：为何我们从未见过某种特定无序状态能自发演化成有序？为何总是单向的秩序瓦解？我在1980年代逐渐领悟到，这有点像密码学——用简单密钥生成复杂的随机乱码。

So the mystery is: why is it the case that one version of the mystery is: why is it the case that you never see something which happens to be just the kind of disorder that you would need to somehow evolve to order. Why does that not happen? Why do you always just see order goes to disorder, not the other way around? So the thing that I I kind of realized I started realizing in the 1980s is kind of like it's a bit like cryptography. It's kind of like you start off from this key that's pretty simple, and then you run it, and you can get this complicated random mess.

我当时的顿悟是：热力学第二定律本质是计算不可约性的体现。初始状态易于描述，最终状态却需要巨大计算量才能刻画。多年后的今天，在完成基础物理研究项目时，我意识到理解观察者本质是关键，而热力学第二定律与众多案例共享着同样的底层逻辑。

And the thing that I started realizing back then was that the second law is kind of a story of computational irreducibility. It's a story of we can describe easily at the beginning, we can only describe with a lot of computational effort at the end. Okay. So now we come many, many years later, and I was trying to sort of Well, having done this big project to understand fundamental physics, I realized that a key aspect of that is understanding what observers are like. And then I realized that the second order of thermodynamics is the same story as a bunch of these other cases.

这是一个关于计算能力有限的观察者试图观察一个计算不可约系统的故事。本质上，分子在四处弹跳，它们以完全由规则决定的方式运动。但关键在于，作为计算能力有限的观察者，我们无法识别这些简单的底层规则，对我们而言，这一切看起来只是随机的。关于能否通过准备初始状态使无序状态恰好产生有序结果的问题，计算能力有限的观察者无法做到这一点。

It is a story of a computationally bounded observer trying to observe a computationally irreducible system. So it's a story of underneath, the molecules are bouncing around. They're bouncing around in this completely determined way, determined by rules. But the point is that we, as computationally bounded observers, can't tell that there were these simple underlying rules, to us it just looks random. When it comes to this question about can you prepare the initial state so that the disordered thing is you have exactly the right disorder to make something orderly, a computationally bounded observer cannot do that.

我们需要完成所有这些本质上不可约的计算，才能精确确定这个无序状态——那个能产生有序结果的精确无序状态究竟是什么。

We'd have to have done all of this sort of irreducible computation to work out very precisely what this disordered state what the exact right disordered state is so that we would get this ordered thing produced from it.

计算能力有限的观察者意味着什么？观察一个计算可约系统时，这种‘计算有限性’是否有正式的定义？

What does it mean to be computationally bounded observer? So Observing a computationally reducible system. So the computationally bounded, is there something formal you can say there?

没错。你可以用图灵机、计算复杂性理论、多项式时间计算等概念来阐述。虽然存在多种精确化的方式，但我觉得更实用的是其直观含义：你准备用多少计算量来理解系统行为？而现实是，我们无法进行大量计算。

Right. So it means, okay, you can you can talk about Turing machines, you can talk about computational complexity theory and, you know, polynomial time computation and things like this. There are a variety of ways to make something more precise, but I think it's more useful. The intuitive version of it is more useful, which is basically just to say that how much computation are you going to do to try and work out what's going on? And the answer is you're not allowed to do a lot of we're not able to do a lot of computation.

比如在这个房间里，存在着万亿亿亿量级的分子——或许稍少些。

When we you know, we've got you know, in this room, there will be a trillion trillion trillion molecules, a little bit less.

这房间可真大。

It's a big room.

对。每微秒这些分子都在碰撞，这相当于海量的计算。问题在于，我们大脑每秒进行的计算量远少于这些分子的总计算量。如果存在计算不可约性，我们无法详细推演出所有分子的行为，只能执行有限得多的计算。

Right. And, you know, at every moment, every microsecond or something, these molecules are colliding, and that's a lot of computation that's getting done. And the question is, in our brains, we do a lot less computation every second than the computation done by all those molecules. If there is computational irreducibility, we can't work out in detail what all those molecules are going to do. What we can do is only a much smaller amount of computation.

因此，热力学第二定律实际上是底层计算的不可约性与我们作为初始状态的准备者或事件观察者无法进行大量计算这一事实之间的相互作用。对我们而言，热力学第二定律的另一种重要表述是熵增定律的概念。

And so the second law of thermodynamics is this kind of interplay between the underlying computational irreducibility and the fact that we, as preparers of initial states or as measures of what happens, are not capable of doing that much computation. So to us, another big formulation of the second law of thermodynamics is this idea of the law of entropy increase.

这个宇宙的特征——熵似乎总是在增加，这对你理解宇宙的演化意味着什么？

The characteristic that this universe, the entropy seems to be always increasing, what does that show to you about the evolution of

好的，关于宇宙，让我先解释什么是熵。这在热力学史上曾非常混乱，因为熵最初是由鲁道夫·克劳修斯提出的，他用热量和温度来定义。后来，路德维希·玻尔兹曼重新定义了它，采用了一种更偏向组合数学的方式。

Well, okay. So for the universe, tell you what entropy is. And that's very confused in the history of thermodynamics, because entropy was first introduced by a guy called Rudolf Clausius, and he did it in terms of heat and temperature. Okay? Subsequently, it was reformulated by a guy called Ludwig Boltzmann, and he formulated it in a much more combinatorial type way.

但他始终声称这与克劳修斯的定义等价。在某个特定简单案例中确实如此。但这两种熵的定义之间从未真正建立过联系。玻尔兹曼提出的更普遍的熵定义是这样的：假设一个系统存在多种可能配置（比如分子可以处于不同排列等）。

But he always claimed that it was equivalent to Clausius' thing. And in one particular simple example, it is. But that connection between these two formulations of entropy, they've never been connected. So the more general definition of entropy due to Boltzmann is the following thing: you say I have a system that has many possible configurations. Molecules can be in many different arrangements, etc.

如果我们掌握系统的某些宏观信息——例如知道它处于某个容器中，具有特定压力和温度——那么问题就变成：在这些宏观约束条件下，系统可能存在多少种微观状态？

If we know something about the system for example, we know it's in a box, it has a certain pressure, it has a certain temperature we know these overall facts about it Then we say, how many microscopic configurations of the system are possible given those overall constraints?

然后呢？

And

熵就是这个数字的对数。这就是定义，也是最终被证明有用的普适熵定义。在玻尔兹曼时代，他认为分子可以任意放置。虽然他没有明确表述，但他意识到：'如果让分子离散化，问题会简单得多。'

the entropy is the logarithm of that number. That's the definition. And that's the general definition of entropy that turns out to be useful. Now, in Boltzmann's time, he thought these molecules could be placed anywhere you want. He didn't think, but he said, Oh, actually, we can make it a lot simpler by having the molecules be discrete.

事实上，他当时并不知道分子的存在。在他所处的19世纪60年代，物质可能由离散粒子构成的观点自古希腊时期就已提出，但关于物质是离散还是连续的争论持续了很久。那时人们普遍认为物质是连续的，这与热本质的认知混淆在一起——人们将热视为某种流体，整个理论体系一片混乱。但玻尔兹曼提出：让我们假设存在离散的分子。

Well, actually, he didn't know molecules existed. In his time, 1860s and so on, the idea that matter might be made of discrete stuff had been floated ever since ancient Greek times, but it had been a longtime debate about, you know, is matter discrete, is it continuous? At the moment at that time, people mostly thought that matter was continuous. It was all confused with this question about what heat is, and people thought heat was this fluid, and it was a big muddle. But Boltzmann said, Let's assume there are discrete molecules.

我们甚至可以假设它们具有离散的能级。假设一切都是离散的，这样我们就能运用组合数学计算箱子里这些粒子的可能构型数量，进而推导出熵这个物理量。不过他指出，这种离散性当然只是理论虚构——这是段很有意思的历史插曲。

Let's even assume they have discrete energy levels. Let's say everything is discrete. Then we can do combinatorial mathematics and work out how many configurations of these things there would be in the box, and we can say we can compute this entropy quantity. But he said, but of course it's just a fiction that these things are discrete, so he said. This is an interesting piece of history, by the way.

当时人们尚未确认分子的存在，虽然化学研究已给出某些暗示（比如两个氢原子与一个氧原子结合成水的组合规律），但直到20世纪初布朗运动被发现才成为决定性证据。在显微镜下观察花粉微粒时，会发现它们受到微小撞击——这些撞击正是离散的水分子造成的。

At that time, people didn't know molecules existed, there were other hints from looking at chemistry that there might be discrete atoms and so on, just from the combinatorics of two hydrogens and oxygen make water. Two amounts of hydrogen plus one amount of oxygen together make water, things like this. But it wasn't known that discrete molecules existed. And in fact, people it wasn't until the beginning of twentieth century that Brownian motion was the final giveaway. Brownian motion is, you know, you look under a microscope at these little pieces from pollen grains, you see they're being discreetly kicked, and those kicks are water molecules hitting them, and they're discreet.

这段历史确实耐人寻味。玻尔兹曼在19世纪60年代就推演出物质的离散性，几乎提前创立了类似量子理论的概念，但他自己认为这并非真实的物理图景。再分享个有趣的物理学史片段：1900年，长期研究热力学的马克斯·普朗克（当时所有学者包括他都在试图证明热力学第二定律）...

In fact, it was really quite interesting history. Boltzmann had worked out how things could be discrete and had basically invented something like quantum theory in the eighteen sixties. But he just thought it wasn't really the way it worked. And then just a piece of physics history because I think it's kind of interesting. In 1900, this guy called Max Planck, who'd been a longtime thermodynamics person, who was trying to everybody was trying to prove the second law of thermodynamics, including Max Planck.

普朗克认为电磁辐射与物质的相互作用将验证热力学第二定律。但面对黑体辐射实验曲线时，他发现基于连续辐射理论无法拟合数据。直到他借鉴玻尔兹曼的方法，假设电磁辐射具有离散性，才成功解释了这些曲线。

Max Planck believed that radiation, like electromagnetic radiation, somehow the interaction of that with matter was going to prove the second law of thermodynamics. But he had these experiments that people had done on black body radiation, and there were these curves, and you couldn't fit the curve. Based on his idea for how radiation interacted with matter. Those curves, you couldn't figure out how to fit those curves. Except he noticed that if he just did what Boltzmann had done and assumed that electromagnetic radiation was discrete, he could fit the curves.

普朗克表示这不过是权宜之计。随后爱因斯坦登场并提出：电磁场可能确实是离散的，由光子构成——这完美解释了所有现象。1905年，这个发现成为了量子力学的重要起源。

He said, but, you know, this just happens to work this way. Then Einstein came along and said, well, by the way, you know, the electromagnetic field might actually be discrete. It might be made of photons. And then that explains how this all works. And that was, you know, in nineteen o five, that was that was how, kind of that was how that piece of quantum mechanics got started.

这段历史相当精彩，我最近研究时才了解到。1903-1904年间爱因斯坦写了三篇论文（这是众所周知的物理学史）。而1905年他发表的三篇开创性论文分别提出相对论、解释布朗运动，以及首次提出光子概念——对物理学和爱因斯坦本人都是里程碑式的一年。

Kind of interesting interesting piece of history, I didn't know until I was researching this recently. In 1904 and 1903, Einstein wrote three different papers, and so, you know, just sort of well known physics history. In 1905, Einstein wrote these three papers. One introduced relativity theory, one explained Brownian motion, and one introduced basically photons. So kind of, you know, kind of a big deal year for physics and for Einstein.

但在那之前的几年里，他写了几篇论文，内容是什么呢？这些论文都是关于热力学第二定律的，试图证明热力学第二定律及其荒谬性。而我完全不知道他曾做过这些研究。有趣。我也不知道。

But in the years before that, he'd written several papers, and what were they about? They were about the second law of thermodynamics, and they were an attempt to prove the second law of thermodynamics and their nonsense. And so I I had no idea that he'd done this. Interesting. Me neither.

事实上，他在1905年发表的那三篇论文——相对论那篇倒不算——关于布朗运动和光子的两篇，都是在探讨如何让世界变得离散化的故事。这个想法他源自玻尔兹曼。但玻尔兹曼临终前仍坚信（有句原话可引），‘最终万物都将证明是离散的，我要写下对此的见解，因为终有一天这些会被重新发现，我想留下关于离散性本质的思考’。我记得他还有句话，大意是个人无法逆历史潮流坚持物质连续性的观点。

And in fact, what he did, those three papers in 1905 well, not so much about relativity paper. The one on Brownian motion, the one on photons, both of these were about the story of sort of making the world discreet. And he got that idea from Boltzmann. But Boltzmann didn't think Boltzmann died believing he said, as a quote actually, In the end, things are going to turn out to be discrete, and I'm going to write down what I have to say about this because eventually this stuff will be rediscovered, and I want to leave what I can about how things are going to be discrete. But I think he has some quote about how one person can't stand against the tide of history saying that, you know, matter is discrete.

哦，所以他在物质离散性问题上坚持己见

Oh, so he stuck by his guns in terms of matter

是的。他确实如此。有意思的是，当时包括爱因斯坦在内的所有人都假设空间最终可能也是离散的。但从技术层面这未能实现，因为与相对论不兼容或看似矛盾。

is discrete. Yes. He did. And and the you know, what's interesting about this is, at the time, everybody, including Einstein, kind of assumed that space was probably going to end up being discrete too. But that didn't work out technically because it wasn't consistent with relativity theory or didn't seem to be.

因此在物理学史上，尽管人们已确认物质是离散的，电磁场是离散的，空间却成了连续性的最后堡垒。实际上爱因斯坦在1916年写过一封精彩的信，他说：‘最终空间将被证明是离散的，但我们尚缺必要的数学工具来阐明其机制’。如今百年过去，我们确实掌握了这些工具，这挺酷的。

And so then in the history of physics, even though people had determined that matter was discrete, electromagnetic field was discrete, space was a holdout of not being discrete. In fact, Einstein, in 1916, has this nice letter he wrote. He says, in the end, it will turn out space is discrete, but we don't have the mathematical tools necessary to figure out how that works yet. And so, you know, I think it's kinda cool that a hundred years later, we do.

没错。对你而言，你相当确信现实的每一层级都是离散的。

Yes. For you, you're pretty sure that every layer of reality is discrete.

对。空间也是离散的。我最近意识到，那种将热视为连续流体的理论——类似热质说——其实是完全错误的，因为热本质上是离散分子的运动。除非承认分子离散性，否则难以理解热的本质。我认为空间同样如此，问题在于：当年热质说的错误，在空间理论上会以什么形式重演？我目前的猜想是——用我这几个月常说的话——暗物质就是我们这个时代的热质说。

Right. And that space is discrete, and the I mean, in fact, one of the things I realized recently is this kind of theory of heat, that heat is really this continuous fluid, it's kind of like the caloric theory of heat, which turns out to be completely wrong, because actually heat is the motion of discrete molecules. Unless you know there are discrete molecules, it's hard to understand what heat could possibly be. Well, I think space is discrete, and the question is what's the analog of the mistake that was made with caloric in the case of space. So my current guess is that dark matter is, as my little aphorism of last few months has been, you know, dark matter is the caloric of our time.

也就是说，最终会发现暗物质是空间的一种特性，而非一堆粒子。要知道，当人们讨论热时，他们了解流体，便认为热必定只是另一种流体，因为那是他们已知的。是的。但现在，人们了解粒子，于是他们说，暗物质是什么？它不是——它不——它肯定就是粒子。

That is, it will turn out that dark matter is a feature of space, and it is not a bunch of particles. You know, at the time when people were talking about heat, they knew about fluids, and they said, well, heat must just be another kind of fluid because that's what they knew about. Yes. But now, people know about particles, and so they say, well, what's dark matter? It's not it's not it just must be particles.

那么作为空间特性的暗物质可能是什么呢？

So what could dark matter be as a feature of space?

哦，我还不知道。好吧。我是说，我认为我真正希望做到的事情之一，是找到空间中的布朗运动类比。换句话说，布朗运动曾被观察到分子层面的效应。而在空间的情况下，目前我们观察到的大多数现象，一切似乎都是连续的。

Oh, I don't know yet. Alright. I mean, I think the the thing I'm really one of the things I'm hoping to be able to do is to find the analog of Brownian motion in space. So in other words, Brownian motion was was seeing down to the level of an effect from individual molecules. So in the case of space, most of the things we see about space so far, just everything seems continuous.

布朗运动在19世纪30年代被发现，直到20世纪初才由斯莫卢霍夫斯基和爱因斯坦确认其成因。暗物质现象——星系旋转曲线不遵循可见物质分布的现象——在一百年前被发现。如果已经存在某种类似布朗运动、能揭示空间离散性的效应，我一点也不会感到惊讶。事实上，我们开始有一些猜测，有证据表明当空间离散时黑洞合并行为会不同，或许能从引力波特征中观察到与空间离散性相关的迹象。但对我来说，有趣的是看到物理学史的某种重演——人们曾激烈宣称物质是连续的，电磁场是连续的。

Brownian motion had been discovered in the 1830s, and it was only identified what it was the result of by Smoluchowski and Einstein at the beginning of the twentieth century. Dark matter was discoveredthat phenomenon was discovered one hundred years The rotation curves of galaxies don't follow the luminous matter, that was discovered a hundred years ago. I wouldn't be surprised if there isn't an effect that we already know about that is kind of the analog of Brownian motion that reveals the discreteness of space. And in fact, we're beginning to have some guesses, we have some evidence that black hole mergers work differently when there's discrete space and there may be things that you can see in gravitational wave signatures and things associated with the discreteness of space. But this is kind offor me, it's kind of interesting to see this sort of recapitulation of the history of physics, where people vehemently say, you know, matter is continuous, electromagnetic field is continuous.

结果证明并非如此，然后他们又说空间是连续的。熵是系统在特定约束下的可能状态数。关键在于，如果你详细知道气体中每个分子的位置，熵始终为零，因为只有一种可能状态：气体分子构型。分子碰撞运动遵循特定规律，气体始终处于单一状态，演化至另一单一状态，如此往复。只有当你不知道所有分子的具体位置时，才能说熵增加了——因为根据我们对分子的已知信息，系统有更多符合这些信息的微观状态可能。这就产生了某种悖论：如果我们知道所有分子的位置，熵就不会增加。

And it turns out it isn't true, and then they say space is continuous. Entropy is the number of states of the system consistent with some constraint. The thing is that if you know in great detail the position of every molecule in the gas, the entropy is always zero because there's only one possible state: the configuration of molecules in the gas, the molecules bounce around, they have a certain rule for bouncing around, there's just one state of the gas, evolves to one state of the gas, and so on. But it's only if you don't know in detail where all the molecules are that you can say, well, the entropy increases because the things we do know about the molecules, there are more possible microscopic states of the system consistent with what we do know about where the molecules are. So the question of whether so people this sort of paradox in a sense of, oh, if we knew where all the molecules were, the entropy wouldn't increase.

20世纪初，吉布斯——美国首位杰出的耶鲁大学物理学教授——提出了粗粒化概念：虽然分子运动有精细规律，但我们只能观察到其粗粒化版本。但困惑在于，没人知道什么是有效的粗粒化方法。没人能确定是否存在这样一种精心设计的粗粒化方式，它能注意到从简单初始条件产生的特定构型恰好符合这种粗粒化框架。

There was this idea introduced by Gibbs in the early twentieth century. Well, actually, very beginning of twentieth century as a physics professor, an American physics professor, sort of the first distinguished American physics professor at Yale. And he introduced this idea of coarse graining, this idea that, well, these molecules have a detailed way they're bouncing around, but we can only observe a coarse grained version of that. But the confusion has been nobody knew what a valid coarse graining would be. So nobody knew that whether you could have this coarse graining that very carefully was sculpted in just such a way that it would notice that the particular configurations that you could get from the simple initial condition, you know, they fit into this coarse graining, and the coarse graining very carefully observes that.

为什么不能进行这种高度精细的粗粒化？答案在于：如果你是计算受限的观察者，而底层动力学具有计算不可约性，那么可能的粗粒化方式就取决于计算受限观察者的能力。正是这种限制迫使观察者只能看到系统的粗粒化版本。因为底层发生的过程正在展开各种可能性——最终由于计算不可约性，你只能通过计算受限的观察获得粗粒化结果，这必然导致存在许多与观察结果相符的底层构型。

Why can't you do that kind of very detailed, precise coarse graining? The answer is because if you are a computationally bounded observer and the underlying dynamics is computationally irreducible, that's what defines possible coarse grainings is what a computationally bounded observer can do. It's the fact that a computationally bounded observer is forced to look only at this kind of coarse grained version of what the system is doing. That's why. And because what's going on underneath is it's kind of filling out the different possible you're ending up with something where the underlying computational irreducibility If is all you can see is what the coarse grained result is with a sort of computationally bounded observation, then inevitably, there are many possible underlying configurations that are consistent with that.

为了澄清一下，基本上，存在于宇宙中的观察者在计算能力上是有局限的吗？

Just to clarify, basically, observer that exists inside the universe is going to be computationally bounded?

不。任何像我们这样的观察者。我不知道。我无法

No. Any observer like us. I don't know. I can't

当你说像我们一样，你指的是什么？你所谓的‘像我们’是什么意思？

When you say like us, what do you mean what do you mean like us?

嗯，就是心智有限的人类。

Well, humans with finite minds.

你把科学的工具也包括在内了。

You're including the tools of science.

是的。是的。我是说，随着我们测量精度的提高，顺便说一句，当你能够更精确地测量分子位置时，可能会出现一些微小的违反热力学第二定律的现象。但在宏观尺度上，当分子数量足够多时，我们无法追踪所有分子，我们根本没有足够的计算能力做到这一点。而且，想象一个不受计算能力限制的观察者会是什么样子，我觉得这很有趣，因为，好吧，那么‘计算能力有限’到底意味着什么？

Yeah. Yeah. I mean, as we have more precise By the way, there are little sort of microscopic violations of the second law of thermodynamics that you can start to have when you have more precise measurements of where precisely molecules are. But for a large scale, when you have enough molecules, we're not tracing all those molecules, and we just don't have the computational to do that. And it wouldn't be I think to imagine what an observer who is not computationally bounded would be like, it's an interesting thing because, okay, so what does computational boundedness mean?

其中一点是，这意味着我们能够确定某些事情发生了。我们面对世界的复杂性，然后做出决定。我们要左转还是右转。这实际上是将所有这些细节压缩为我们所观察到的，我们某种程度上将其简化为这一个决定。如果我们不这样做，我们就不会有这些让我们用有限的心智进行思考的符号化结构。

Among other things, it means we conclude that definite things happen. We go, we take all this complexity of the world, and we make a decision. We're going to turn left or turn right. And that is kind of reducing all this kind of detail into we're observing it, we're sort of crushing it down to this one thing. And if we didn't do that, we wouldn't have all this sort of symbolic structure that we build up that lets us think things through with our finite minds.

我们反而会，你知道的，我们会与宇宙融为一体。

We'd be instead, you know, we'd be just we'd be sort of one with the universe.

是啊。所以满足于不去简化。

Yeah. So content to not simplify.

没错。如果我们不简化，我们就不会像现在这样。我们会像宇宙本身那样，像内在的宇宙那样，但不会有我们这样的体验，比如我们会得出某些确定的事情发生了。我们，你知道的，我们某种程度上拥有这种能够构建叙事性陈述的能力。

Yes. If we didn't simplify, then we wouldn't be like us. We would be like the universe, like the intrinsic universe, but not having experiences like the experiences we have, where we, for example, conclude that definite things happen. We, you know, we we sort of have this this notion of being able to make make sort of narrative statements.

是啊。我在想，这就像你想象一个思想实验，体验作为一台计算机的感觉。我在想，是否有可能尝试开始想象作为一个无边界计算的...

Yeah. I wonder if it's just like you imagined as a thought experiment what it's like to be a computer. I wonder if it's possible to try to begin to imagine what it's like to be an unbounded computational

观察者。好吧。那么，我认为情况是这样的。振动。我的意思是，在这个我们讨论的Rulliad中，所有可能计算的空间。

observer. Okay. So here's here's how that, I think, plays out. Vibrations. So, I mean, in this, we talk about this rulliad, the space of all possible computations.

这个关于处于Rulliad某个特定位置的想法，对应于你用来表示事物的一组特定计算方式。好吧，当你在Rulliad中扩展，当你涵盖更多可能的宇宙视角，当你涵盖更多可以进行的计算类型，最终你可能会说，这是一个真正的胜利。我们正在殖民Rulliad。我们正在构建更多关于如何思考事物的范式。最终，你可能会说，我们彻底赢了。

And this idea of being at a certain place in the rulliad, which corresponds to a certain way of a certain set of computations that you are representing things in terms of. Okay, so as you expand out in the Rulliad, as you kind of encompass more possible views of the universe, as you encompass more possible kinds of computations that you can do, eventually, you might say, that's a real win. We're colonizing the Rulliad. We're building out more paradigms about how to think about things. And eventually, you might say, we won all the way.

我们成功殖民了整个Rulliad。好吧。这里的问题是，存在的概念，连贯的存在，需要某种特定的形式。当你成为整个Rulliad时，当你覆盖整个Rulliad时，在没有任何有用的意义上，你是连贯存在的。

We managed to colonize the whole Rulliad. Okay. Here's the problem with that. The problem is that the notion of existence, coherent existence, requires some kind of specialization. By the time you are the whole Rulliad, by the time you cover the whole Rulliad, in no useful sense do you coherently exist.

换言之，存在的概念，我们所认为的确定性存在，需要这种特殊化，需要这种我们并非所有可能事物的观念。我们是特定的一组事物，正是这一点赋予了我们连贯的存在。如果我们遍布于鲁利亚德（Rulliad），我们的运作方式将失去连贯性。我们会以所有可能的方式运作，那将不再是一种身份的体现。我们将失去这种

So in other words, the notion of existence, the notion of what we think of as definite existence requires this kind of specialization, requires this kind of idea that we are not all possible things. We are a particular set of things, and that's kind of what makes us have a coherent existence. If we were spread throughout the Rulliad, there would be no coherence to the way that we work. We would work in all possible ways, and that wouldn't be kind of a notion of identity. We wouldn't have this notion of kind of

连贯的身份认同。我在地理上精确地位于鲁利亚德（Rulead）的某处，故我存在。这是笛卡尔式的观点，对吧。

coherent identity. I am geographically located somewhere exactly precisely in the Rulead, therefore I am. Is the Descartes kind of Yeah.

没错。你处于物理空间的某个位置，或乡村空间的某个位置。如果你过于分散，你将不再连贯，也将失去——我的意思是，我们对存在和体验的感知，不会以那种方式发生。

Yeah. Well, you're in a certain place in physical space, or in a certain place in rural space. And if if you are sufficiently you spread out, you are no longer coherent, and you no longer have I mean, our perception of what it means to exist and to have experience, doesn't happen that

因此，存在意味着在计算上受限？

way. So therefore so to to to exist means to be computationally bounded?

我认为是的。

I think so.

以我们理解自身存在的方式存在。是的。存在的本质，就像是在这个计算可简化的地方运作，那里有大量无法预测的混乱事件。然而，正因为你的局限性，你有一种——是什么？一种简化需求的驱动力或技能？还是某种程度的无知？

To exist in the way that we think of ourselves as existing. Yes. The very act of existence is, like, operating in this place that's computationally reducible, so that there's just giant mess of things going on that you can't possibly predict. But nevertheless, because of your limitations, you you have an imperative of, like what is it? An imperative or a skill set to simplify Or an ignorance, a sufficient level?

好的。不那么显而易见的是，你正在从所有这些复杂性中截取一个片段，就像房间里所有分子都在四处弹跳，但我们只注意到空气的流动或压力。我们只关注这些特定的事物。而最有趣的是，存在规则，存在支配我们所观察到的这些宏观现象的法则。是的。

Okay. So the thing which is not obvious is that you are taking a slice of all this complexity, just like we have all of these molecules bouncing around in the room, but all we notice is the flow of the air or the pressure of the air. We're just noticing these particular things. And the the big interesting thing is that there are rules, there are laws that govern those big things we we observe. Yeah.

所以这并不明显。

So it's not obvious.

这就是它的奇妙之处。因为它感觉不像是一个切片。

That's how it's Amazing. The Because it doesn't feel like it's a slice.

是的。嗯，没错。这不是一个切片。

Yeah. Well, right. It's not a slice.

嗯，这就像是一种抽象。是的。但是，我是说，事实是

Well It's like a it's like an abstraction. Yes. But but, I mean, the fact

气体定律成立，我们可以描述压力、体积等等，而且我们不需要深入到讨论单个分子的层面，这是一个非平凡的事实。而对我来说，令人兴奋的是：宇宙有某些方面，我们认为空间最终由这些空间原子和这些超图等构成，但我们仍然在大尺度上感知宇宙像是连续的空间等等。在量子力学中，我们认为存在这些时间的多线程，这些历史的多个分支，但在量子力学中，在我们的物理模型中，时间不是单一线程。时间分裂成多个线程。它们分支，它们合并，但我们是这个分支、合并宇宙的一部分。

that the gas laws work, that we can describe pressure, volume, etcetera, etcetera, etcetera, and that we don't have to go down to the level of talking about individual molecules, That is a nontrivial fact. And here's the thing that I'm sort of the exciting thing as far as I'm concerned: the fact that there are certain aspects of the universe, so we think space is made ultimately of these atoms of space and these hypergraphs and so on, and we think thatbut we nevertheless perceive the universe at a large scale to be like continuous space and so on. In quantum mechanics, we think that there are these many threads of time, these many threads of history, yet we kind of spanso in quantum mechanics, in our models of physics, there are thesetime is not a single thread. Time breaks into many threads. They branch, they merge, but we are part of that branching, merging universe.

所以我们的大脑也在分支和合并，因此当我们感知宇宙时，我们是分支的大脑感知一个分支的宇宙。因此，我们相信自己持续存在于时间中，拥有这一单一的经验线程，这一说法意味着我们设法将那些在宇宙基本运作中分离的时间线程聚合在一起。就像在空间中，我们是在对空间的一个大区域进行平均，观察许多空间原子的聚合效应。类似地，在我们称之为分支空间的地方，这些量子分支的空间，我们实际上是在对宇宙许多可能历史的不同分支进行平均。在热力学中，我们是在对分子的许多可能位置的许多配置进行平均。

So our brains are also branching and merging, and so when we perceive the universe, we are branching brains perceiving a branching universe. And so the fact that the claim that we believe that we are persistent in time, we have this single thread of experience, that's the statement that somehow we manage to aggregate together those separate threads of time that are separated in fundamental operation of the universe. So just as in space, we're averaging over some big region of space and we're looking at many, many of the aggregate effects of many atoms of space. So similarly, in what we call branchial space, the space of these quantum branches, we are effectively averaging over many different branches of possible histories of the universe. And so in thermodynamics, we're averaging over many configurations of many possible positions of molecules.

所以我们在这里看到的问题是：当你对空间进行这种平均时，空间的聚合定律是什么？当你对分支空间进行这种平均时，分支空间的聚合定律是什么？当你对分子等进行这种平均时，你得到的聚合定律是什么？我认为这一点非常简洁。

So what we see here is the question is: when you do that averaging for space, what are the aggregate laws of space? When you do that averaging over branchial space, what are the aggregate laws of branchial space? When you do that averaging over the molecules and so on, what are the aggregate laws you get? And this is the thing that I think is just amazingly neat. That

是否存在任何关于

there are aggregate laws at all for

嗯，是的。但问题是：这些聚合定律是什么？所以答案是，对于空间而言，聚合定律是爱因斯坦的引力方程，关于时空结构的。对于布隆希尔德空间，聚合定律是量子力学定律。而对于分子等事物，聚合定律基本上是热力学第二定律。

the Well, yes. But the question is: what are those aggregate laws? So the answer is, for space, the aggregate laws are Einstein's equations for gravity, for the structure of space time. For Bronshield space, the aggregate laws are the laws of quantum mechanics. And for the case of molecules and things, the aggregate laws are basically the second law of thermodynamics.

这些就是从热力学第二定律衍生出来的东西。这意味着二十世纪物理学的三大理论——广义相对论（引力理论）、量子力学和统计力学（源自热力学第二定律）——都是计算不可约性与观察者计算有限性之间相互作用的结果。对我来说，这非常简洁，因为它意味着所有这些定律都是可推导的。我们过去认为，例如爱因斯坦方程只是宇宙的一个偶然特征，宇宙可能是这样，也可能不是这样。量子力学就像是，嗯，它恰好就是那样。

So that's the things that follow from the second law of thermodynamics. And so what that means is that the three great theories of twentieth century physics, which are basically general artillery, the theory of gravity quantum mechanics and statistical mechanics, which is what grows out of the second law of thermodynamicsall three of the great theories of twentieth century physics are the result of this interplay between computational irreducibility and the computational boundedness of observers. For me, this is really neat because it means that all three of these laws are derivable. So we used to think that, for example, Einstein's equations were just sort of a wheel in feature of our universe, that the universe might be that way, it might not be that way. Quantum mechanics is just like, Well, it just happens to be that way.

而热力学第二定律，人们曾认为，或许它是可推导的。事实证明，物理学的这三个基本原理都是可推导的，但它们不仅仅是从数学中推导出来的。它们还需要——或者不仅仅是从某种逻辑或计算中——还需要一个额外的东西：它们要求观察者，即对宇宙运作方式进行采样的那个存在，是一个具有计算有限性、对持久性和时间有信念特征的观察者。这意味着，正是观察者的本质，观察者的大致特性（而不是‘哦，我们有两只眼睛，我们观察这个频率的光子’等细节），这些非常粗略的观察者特征，却暗示了关于物理学的这些非常精确的事实。我认为这很神奇。

And the second law, people kind of thought, Well, maybe it is derivable. What turns out to be the case is that all three of the fundamental principles of physics are derivable, but they're not derivable just from mathematics. They requireor just from some kind of logic or computationthey require one more thing: they require that the observer, that the thing that is sampling the way the universe works, is an observer who has these characteristics of computational boundedness of belief in persistence and time. And so that means that it is the nature of the observer, the rough nature of the observer, not the details of, Oh, we've got two eyes and we observe photons of this frequency and so on, but the very coarse features of the observer then imply these very precise facts about physics. I think it's amazing.

所以如果我们只看观察者的实际体验，我们体验到的这个现实，对我们来说似乎是真实的。而你说，由于我们的有限性，实际上这一切都是幻觉。这是一种简化。是的。

So if we just look at the actual experience of the observer that we experience this reality, it seems real to us. And you're saying because of our bonded nature, it's actually all an illusion. It's a simplification. Yeah.

这是一种简化。对。什么是——嗯，

It's a simplification. Right. What's what's Well,

你不认为简化是一种幻觉吗？

you don't think a simplification is an illusion?

不，我是说，这个嘛，我不知道。我是说，它是真实的吗？好吧，这是个有趣的问题。

No. I mean, it's it's well, I don't know. Mean, what Is it real? Okay. That's an interesting question.

什么是真实的？这涉及到整个宇宙为何存在的问题，以及现实与对正在发生事物的单纯表象之间的区别。

What's real? And that relates to the whole question of why does the universe exist, and what is the difference between reality and a mere representation of what's going on.

是的。我们体验的是表象。

Yes. We experience the representation.

没错。但问题在于，为什么存在一个我们可以如此体验的事物？答案是因为这个Rulliard对象——所有可能计算的纠缠极限——它别无选择地必须存在。必然存在这样的东西。

Yes. But the question of so one question is, you know, why is there a thing which we can experience that way? And the answer is because this Rulliard object, which is this entangled limit of all possible computations, there is no choice about it. It has to exist. There has to be such a thing.

就像如果你定义了什么是二，并规定了加法等运算，那么二加二必然等于四。同样地，这个Rulliad——所有可能计算的极限——一旦你有了计算的概念，它就必然成为一个存在，你不可避免地会得出这个规则。

It is in the same sense that two plus two, if you define what two is and you plot pluses and so on, two plus two has to equal four. Similarly, this Rulliad, this limit of all possible computations, just has to be a thing that is once you have the idea of computation, you inevitably have the rule

经验法则。是的，你总得有个

of thumb. Yeah, you're going have to have a

规则。重要的是，它只有一个。就是这个独特的对象。这个独特对象必然存在。然后问题是，什么？

rule of Right. And what's important about it, there's just one of it. It's just this unique object. And that unique object necessarily exists. And then the question is, what?

而一旦我们意识到自己身处其中并对其进行采样，就不可避免地存在这样一种我们能感知的事物——我们对物理现实的感知必然如此，因为我们作为观察者具有特定的特性。换句话说，宇宙的存在几乎可以说是，从某种神学角度来思考。有趣的是，许多关于宇宙存在等问题超越了近几百年来科学真正关注的范畴。过去几百年的科学并不认为自己能探讨这类问题。但我想很多关于'上帝是否存在'的争论...

And then we areonce you know that we are embedded in that and taking samples of it, it's inevitable that there is this thing that we can perceive that is our perception of physical reality necessarily is that way, given that we are observers with the characteristics we have. So in other words, the universe exists is actually it's almost like it's, you know, to think about it almost theologically, so to speak. And I've I've really it's it's funny because a lot of the questions about the existence of the universe and so on, they transcend what the science of the last few hundred years has really been concerned with. The science of the last few hundred years hasn't thought it could talk about questions like that. But I think a lot of the arguments of, you know, does God exist?

是否显而易见？我认为在某种意义上，在某种表现形式中，比我们更宏大的存在比我们自身存在更显而易见。我们作为观察者的存在方式只是宇宙中的一个偶然事件，而整个宇宙、所有可能性集合的存在则更为必然。但关于'这是真实还是幻觉'的问题？我们唯一知道的是自己的体验。事实上，我们的体验只是对Rulliad（假设性整体存在）的极其微小的采样样本。关键在于我们可能采样到更多Rulliad内容，比如对量子力学等物理学不同领域越来越深入的了解。

Is it obvious that I think in some sense, in some representation, it's more obvious that something bigger than us exists than that we exist. And we are our existence and as observers the way we are is a contingent thing about the universe, and it's more inevitable that the whole universe, the whole set of all possibilities exists. But this question about, you know, is it real or is it an illusion? All we know is our experience. And so the fact that our experience is this absolutely microscopic piece of sample of the Rulliad, and there's this point about we might sample more and more of the Rulliard, we might learn more and more about different areas of physics, like quantum mechanics, for example.

我认为量子力学的发现与电子放大器的发明密切相关，后者使人们能够将微小效应放大观察，这在以前是不可能的。虽然显微镜等放大工具早已存在，但能够放大极其微弱效应的新技术让我们看到了宇宙的不同面向，从而发现这类现象。因此我们可以预期，在Rulliad中存在着无数待发现的新事物。事实上，计算不可约性从某种意义上保证了存在无数可被发现的可约性局部区域。

The fact that it was discovered, I think, is closely related to the fact that electronic amplifiers were invented, that allowed you to take a small effect and amplify it up, which hadn't been possible before. Microscopes have been invented that magnify things and so on, but having a very small effect and being able to magnify it was sort of a new thing that allowed one to see a different aspect of the universe and let one discover this kind of thing. So we can expect that in the Rulliad, there are an infinite collection of new things we can discover. There's, in fact, computational irreducibility kind of guarantees that there will be an infinite collection of pockets of reducibility that can be discovered.

要是能沿着Rulead（假设性存在维度）漫步探索该多有趣啊。你在著作中提到外星智慧生命。是的，我是说这些世界...

Boy, would it be fun to take a walk down the Rulead and see what kind of stuff we find there. You you write about alien intelligences. Yes. I mean, just these worlds

是的，相当...

Yes. Well, quite a

多呢。

bit too.

这些世界的问题在于

Problem with these worlds is that

我们无法与他们交流。

We can't talk to them.

是的。而且，你知道，关键在于，我花了大量时间研究计算系统，观察它们的行为，也就是我现在称之为‘规则学’的领域——研究规则及其作用。你可以轻松跳转到规则空间的其他位置，观察这些规则如何运作。它们只是按既定方式运行，可以说，其中不存在人类情感的联系。

Yes. And and, you know, the thing is, what I've kind of spent a lot of time doing is just studying computational systems, seeing what they do, what I now call ruliology, kind of just the study of rules and what they do, you know, you can kind of easily jump somewhere else in the Rulead and start seeing what do these rules do. And what you they they just they do what they do, and there's no human connection, so to speak.

你认为...有些人能和动物沟通吗？你觉得你能成为它们的‘低语者’吗？

Do you think, you know, some some people are able to communicate with animals? Do you think you can become a whisperer of these

我正在尝试。这正是我人生中部分时间所致力的事。

I'm trying. That's what I've spent some part of my life doing.

你有听说过吗？你是否面临丧失理智的风险？

Have you have you heard? And are you at the risk of losing your mind?

我最钟爱的科学发现是：极其简单的程序能产生极度复杂的行为。这个事实某种程度上像是计算宇宙中传来的低语——揭示了我们此前未知的存在。1980年代，我曾与几位杰出数学家合作，他们试图破解这些计算系统的运行机制，最终坦言现有数学工具对此完全无能为力。

Sort of my favorite science discovery is this fact that these very simple programs can produce very complicated behavior. Yeah. And that and that fact is kind of, in a sense, a whispering of something out in the computational universe that we didn't really know was there before. I mean, it's like back in the 1980s, I was doing a bunch of work with some very good mathematicians, and they were trying to pick away: Can we figure out what's going on in these computational systems? And they basically said, Look, the math we have just doesn't get anywhere with this.

我们被困住了。无话可说。真的无话可说。而或许我当时的主要成就在于意识到：顶尖数学家们的无言以对本身正是极有趣的现象。这某种意义上像是从规则宇宙另一个不可达维度传来的低语——完全超越了我们已知的数学认知范畴。

We're stuck. There's nothing to say. We have nothing to say. And, you know, in a sense, perhaps my main achievement at that time was to realize that the very fact that the good mathematicians had nothing to say was itself a very interesting thing. That was kind of a sort of, in some sense, a whispering of a different part of the Rulliad that one wasn't was not accessible from what we knew in mathematics and so on.

当你探索这些宏大的理念，感觉我们即将突破到一些非常有趣的发现时，却意识到自己只是一个有限的生命体，很快就会死去，而你的大脑扫描、全身扫描显示你不过是一堆肉块，这会让你感到悲伤吗？是的，只是一堆肉块而已。

Does it make you sad that you're exploring some of these gigantic ideas, and it feels like we're on the verge of breaking through to some very interesting discoveries? And yet you're just a finite being that's going to die way too soon, and that scan of your brain, your full body, kinda shows that you're Yeah. It's just a bunch of meat. It's just a bunch of meat. Yeah.

这让你有点难过吗？

Does that make you make you a little sad?

确实有点遗憾。我是说，我很想看看这一切如何发展。但我觉得要认识到，这其实是个有趣的思维实验。假设我们能让人体冷冻技术成功——终有一天会的，就像会出现类似ChatGPT的突破。

Kind of a shame. I mean, I kind of like to see how all this stuff works out. But I think the thing to realize, you know, it's an interesting sort of thought experiment. You know, you you say, okay, you know, let's assume we can get cryonics to work, and one day it will. There will be one of these things that's kind of like chat GPT.

总有一天会有人解决零下44度左右的水不膨胀的问题，人体冷冻技术就攻克了，你可以按下暂停键，一百年后重新醒来。但我越来越意识到，我们关于存在的困惑其实根植于特定时代——比如我现在关心的事，若活在五百年前会显得极其怪异，就像未来人看我们执着于针尖上能站多少天使这类中世纪神学命题一样。

One day somebody will figure out how to get water from down to minus 44 or something without it expanding, and cryonics will be solved, you'll be able to just put a pause in, so to speak, and reappear one hundred years later or something. The thing, though, that I've increasingly realized is that, in a sense, this whole question of one is embedded in a certain moment in time and the things we care about now, the things I care about now, for example, had I lived five hundred years ago, many of the things I care about now, it's like, that's totally bizarre. Nobody would care about that. It's not even the thing one thinks about. In the future, the things that most people will think about, one will be a strange relic of thinking about the kind of it might be it might have been a theologian thinking about how many angels fit on the head of a pin or something, and that might have been the big intellectual thing.

这既是幸运也是不幸——我发明了不少东西，能预见它们在未来50到100年（假设文明延续）将成为核心科技。从生命历程看利弊参半：如果25岁就功成名就，余生反而像在走下坡路；能持续见证这些变革让生活保持趣味，比如ChatGPT带来的计算语言革命就超出我预期。

So I think it's but yeah, it's one of these things where particularly, I've had the, I don't know, good or bad fortune, I'm not sure I think. It's a mixed thing. I've, you know, I've invented a bunch of things which I kind of can, I think, see well enough what's going to happen that, you know, in fifty years, one hundred years, whatever, assuming the world doesn't exterminate itself, so to speak, these are things that will be centrally important to what's going on? It's both a good thing and a bad thing in terms of the passage of one's life. I mean, it's kind of like, if everything I'd figured out was like, okay, figured it out when I was 25 years old, and everybody says it's great, and we're done.

我原以为这类突破还要五十年，现在提前到来很酷——这意味着我有望亲眼见证，而不只是...

And it's like, okay, but I'm gonna live another how many years? And that's kind of it's all downhill from there. In a sense, it's it's better in some sense to to be able to, you know, there's there's it sort of keeps things interesting that, you know, I can see a lot of these things. I mean, it's kind of I didn't expect, you know, chat GPT. I didn't expect the kind of the sort of opening up of this idea of computation and computational language that's been made possible by this.

...停留在想象中。嗯。

I didn't expect that. This is ahead of schedule, so to speak. Mhmm. You know, even though the sort of the the big kind of flowering of that stuff I'd sort of been assuming was another fifty years away. So if it turns out it's a lot less time, that's pretty cool because, you know, I'll hopefully get to see it, so to speak, rather than

嗯，我想我代表许许多多的人说，我希望你能长久地留在这个领域。你提出了那么多有趣的构想，多年来创造了那么多有趣的体系。现在我看到GPT和语言模型更是彻底打开了新世界的大门。我迫不及待想看到你站在这一发展的最前沿，见证你将创造的成就。是的，我一直是你的粉丝，就像我多次告诉你的那样，从最初开始就是。

Well, I I think I speak for a very, very large number of people in saying that I hope you stick around for a long time to come. You've had so many interesting ideas. You've created so many interesting systems over the years, and I can see now that GPT and language models broke open the world even more. I can't wait to see you at the forefront of this development, what you what you do. And, yeah, I've been a fan of yours, like I've told you many, many times since the very beginning.

我深深感激你撰写了《一种新科学》，探索了细胞自动机的奥秘，并激励了当年那个年幼的我投身于这个美丽的人工智能世界。所以，史蒂芬，非常感谢你。能与你交谈、汲取你的智慧、共同探讨这些想法是莫大的荣幸，请继续前行。我迫不及待想看到你的下一个创见。也感谢你今天的分享。

I'm deeply grateful that you wrote a new kind of science, that you explored this mystery of cellular automata, and inspired this one little kid in me to to pursue artificial intelligence in all this beautiful world. So, Steven, thank you so much. It's a huge honor to talk to you, to to just be able to pick your mind and to explore all these ideas with you, and please keep going. And I can't wait to see what you come up with next. And thank you for talking today.

时间到了，谢谢。我们聊过了午夜，虽然只进行了四个半小时。其实我们还能再聊四小时，不过留到下次吧——这是第四次对话了。我相信我们还会聊很多次。

We hit Thanks. We went past midnight. We only did four and a half hours. I mean, we could probably go for four more, but we'll save that till next time to this is round number four. We'll I'm sure talk many more times.

非常感谢。这是我的荣幸。

Thank you so much. My pleasure.

感谢收听与史蒂芬·沃尔夫勒姆的对话。如需支持本播客，请查看简介中的赞助商信息。最后，请允许我引用乔治·康托尔的话作为结束：数学的本质在于其自由性。感谢收听，我们下次再见。

Thanks for listening to this conversation with Stephen Wolfram. To support this podcast, please check out our sponsors in the description. And now, let me leave you with some words from George Cantor. The essence of mathematics lies in its freedom. Thank you for listening, and hope to see you next time.