330 | 佩特·托恩伯格谈（错误）信息的动态传播

大家好，欢迎收听《心智景观》播客。我是主持人肖恩·卡罗尔。社会科学界有个概念叫'物理学嫉妒'，经济学尤其容易受此影响。这可不是什么好事。

Hello, everyone, and welcome to the Mindscape Podcast. I'm your host, Sean Carroll. There's a idea in social science circles called physics envy. Economics especially is susceptible to this idea. It's not supposed to be a good thing.

虽然不该真的嫉妒物理学，但社会科学确实很难。人性复杂多变，变量繁多。物理学能通过大幅简化问题取得巨大进展，部分原因在于其研究对象的基本构成——即便是量子力学、宇宙学、相对论这些听起来高深的内容——其实并没有太多动态组成部分。

You're not actually supposed to be envious of physics, but social science is hard. People are messy. There's a lot of variables going on. Physics is able to make enormous progress by simplifying things a great deal. In part, that's because the fundamental ingredients that we study, even though it's like quantum mechanics and cosmology and relativity and other things that sound way out, but there aren't a lot of moving parts.

我们研究的基础对象足够简单，用较少变量就能描述，并能通过少量变量分离出系统中的所有关键现象。正因如此，物理学能取得惊人进展：可以证明定理，进行精确到多位小数的实验验证。

The basic things that we're looking at sufficiently simple. You can describe them using relatively few variables, and you can isolate all the interesting things that are going on in these systems with small numbers of variables. As a result of this, you can make tremendous progress. You can prove theorems. You can do experiments that test your theories to many, many decimal places.

这确实充满乐趣，人们难免心生羡慕。但这种'物理学化'倾向是种病症，至少应该避免在社会科学研究中过度模仿物理学。做社会科学时必须承认，其复杂性无法像物理学忽略空气阻力或摩擦力那样被抽象掉。不过常听《心智景观》的听众都知道，我认为在某些情境下，物理学式思维对社会科学确实能有所启发。

It's a lot of fun. Of course, people would be envious of this. But it's a disease or at least something to be avoided to therefore try to make your social scientific research too much like physics. When you do social science, you should admit that there are complications there that cannot be abstracted away in the same way that we abstract away air resistance or friction when we're doing physics. Nevertheless, I'm sure that everyone who listens to Mindscape on a regular basis knows, I do think that there are contexts in which physics like reasoning can be helpful or even interesting in the social science.

在某些领域，物理学式的推理和概念移植到社会科学中可能非常有用且有趣。比如均衡概念、涌现现象，或是大量无意识个体互动产生的集体行为——这些都是物理学家经常思考，又与社会科学密切相关的问题。今天嘉宾彼得·索恩伯格是计算社会科学教授，他自述有物理学背景，这就不难理解了。

There can be contexts in which physics type of reasoning and concepts borrowed from physics can be really useful, very interesting in the social scientific contexts. Ideas like equilibrium, ideas of emergence in general, ideas of what is collective behavior like when it arises from the sort of mindless, non directed interaction of many small things. These are things that physicists think about all the time and are very, very relevant to the social sciences. So today's guest is Petter Thornburg, who, is a professor of computational social science. I promise I didn't know this, but he admits on the podcast that he actually has a physics background, so this makes some sense.

他运用我们近期讨论过的基于主体的模型（与唐·法默等学者合作），通过构建简单主体（行为模式固定或带有随机性）或复杂主体组成的模型来研究社会系统。比如他曾用大语言模型模拟社交媒体中的人类互动，进而探究：模型能否产生现实世界观察到的现象？

But he uses models, agent based models that we've talked about recently, with Don Farmer and others as ways to study the behavior of social systems. Can you make a little model where the individual pieces are either simple agents that always act in some way, or maybe there's a little bit of stochasticity in there, or maybe they're even very complicated. We'll talk about an example where Better used LLMs, large language models, to model human interactions in social media landscapes. And then you can ask, what is the robust behavior? Do you get things that we observe in the real world?

会出现两极分化吗？影响力会集中在特定人群而非均匀分布吗？这些现象好坏如何？若对社交媒体进行干预能否改善状况？问题全在算法，还是个体参与者的偏好使然？

Do you get polarization? Do you get sort of an accumulation of influence in certain people rather than having it be completely uniform? Is this good or bad? If you intervene in certain ways in the social medium, can you make things better? Is it all the algorithm's fault, or is it just the preferences of the individual actors?

所有这类问题都可以用这种略带态度的方式处理，但要放在社会科学背景下。我不想过多透露彼得的研究结果，但它们并不乐观。对于那些希望社交媒体能发挥作用的人来说，这些结果并不令人鼓舞。某些自然形成的动态趋势似乎会将事情推向糟糕的方向，加剧两极分化和回音室效应等现象。无论我们喜欢与否，如果事实如此，那就是现实。

All of these kinds of questions can be addressed with this sort of slightly attitude, but put into a social science context. I don't wanna give away too much about Petter's results, but they're not great. They're not very, encouraging for those of us who want social media to work. There are certain natural dynamics that seem to come up that drive things in a bad direction, that drive polarization and echo chambers and things like that. It doesn't matter whether we like it or not, if that's what happens.

了解这种现象如何发生以及在何种条件下发生，有望帮助我们改进社交媒体、媒体或整个信息生态系统，使其不仅服务于民主目标，也能让我们学习有趣的新事物、享受美好时光、成为更好的人，并通过多种方式与他人建立原本不可能的联系。我们希望在保留这些美妙技术优点的同时避免其弊端，我认为这类研究能帮助我们找到实现这一目标的方法。那么让我们开始吧。彼得·索恩伯格，欢迎来到《心灵景观》播客。

Knowing how it happens and that it happens under certain circumstances, hopefully, will be helpful in making social media or media or information ecosystems in general more functioning for the purposes of democracy, but also for the purposes of just learning fun new things and having a good time and being good human beings in in connecting with our peeps in various different ways and, building connections that otherwise wouldn't have been possible. So we wanna keep the good aspects of these wonderful technologies without being subject to the bad aspects, and I think this kind of study helps us figure out how to do this. So let's go. Peter Thornburg, welcome to the Mindscape Podcast.

谢谢邀请我参加。作为这个播客的长期听众和粉丝，能来真是莫大的荣幸。真的太棒了。

Thank you for having me on. It's a it's a real treat as a as a longtime listener and fan of the podcast. It's really great.

好的，非常好。既然你是老听众，应该知道流程。我们先做些铺垫。在你的一本书里有句引人深思的话：'权力拥有认识论'。

Okay. Very good. Well, then you know how it's gonna go. Let's just start setting some stage a little bit. In one of your books, you have this provocative line, power has an epistemology.

还记得这句话吗？出自《像平台一样观察》。对于那些日常不常使用'认识论'这类术语的听众，你能解释下这句话的含义吗？

Do you remember that one? That's from seeing like a platform. So what does that what does that mean to those people in the audience who might not use words like epistemology every day?

是的，这是个很好的切入点。我的很多研究都基于将社会视为复杂系统的理解。虽然我早年有物理学背景——不过那是很久以前的事了，所以别考我——但我主要采用复杂系统的视角。

Yeah. So it's it's a it's a good starting point, I think. So a lot of my research is kind of informed by this understanding of society being a complex system. So I kind of like, I actually have a kind of physics background myself, but like long time ago, so don't quiz me on that. But, and come very much from the kind of complex systems perspective.

在博士期间，我专注于运用这种视角，尝试通过复杂性科学和计算方法来理解社会。这本书的核心思想就是：随着复杂性概念在社会科学、城市规划乃至数字平台如何塑造社会的讨论中被广泛运用，我们试图厘清这究竟意味着什么。社会科学中，复杂性通常被理解为自下而上的系统——不像汽车或航天器这类复杂系统可以拆解成发动机等部件并轻松理解其组装逻辑。

And in my PhD, I was kind of focusing on taking that perspective to try to see how we can understand society using the methods of complexity science and using computational methods. And so this book is basically centered around this idea, okay, because we're seeing this notions of complexity becoming used throughout the social sciences, but also in kind of urban planning, and also in kind of when we talk about digital platforms and how they're shaping society. And so we try to kind of understand what does that actually mean. And so, I mean, the social sciences, the way the complexity tends to be understood is, like this bottom up system. So often you have like, there's complicated systems like a car or a spaceship that you can kind of take apart, you can decompose them, you have like an engine, and it's quite easy to figure out how they fit together.

你可以通过拆解系统来大致理解它。但另一方面，在这条界限的另一边，存在着复杂系统。比如蚁群、鸟群的流动等等。这些系统一旦被拆解——比如从蚁群中单独取出一只蚂蚁，你可以无限观察它的行为，但这几乎无法揭示蚁群如何运作，因为蚁群的智慧源于大量蚂蚁的互动，对吧？没错。

You can kind of understand the system by taking it apart. But then on the other hand, on the other side of the kind of this line, have complex systems. So like ant colonies or flux of birds or whatever. And these systems, if you take them apart, like you take out an individual ant from an ant colony, you can observe its behavior as much as you want, but it won't tell you very much about how an ant colony functions because the kind of intelligence of the ant colony emerges through the interaction of a lot of ants, right? Right.

这本书源于一个观察：我们越来越多地通过这种复杂视角来讨论社会。它与新型民主的概念交织在一起——显然这些思想很大程度上源自物理学和计算机科学领域。但当这些概念进入社会领域时，它们发生了显著变化，并开始具有政治意涵。我们追踪的正是这些意涵。最终在社会科学中，这演变成了认识论的问题。

And so the book kind of stems from the observation that we've shifted to more and more talking about society through this complex lens. And that it's kind of intertwined with this notion of new forms of democracy because obviously a lot of those idea come from kind of from your field, from physics and from computer science. But as they're entering, they're changing quite a lot as they're entering into the social world And they also begin to have kind of political implications. And so those are the kind of implications that we follow. And ultimately in the social sciences, this becomes a question of kind of epistemology.

这就引出了核心问题：我们如何构想社会的本质？从六十年代开始，我们社会科学家所称的福特主义或工业现代性时期，人们习惯将社会视为机器，视为复杂系统。我们认为，这种认知方式源于建立在大型工业、大规模生产基础上的社会，它催生了大众社会，也催生了'我们能像设计机器那样设计和规划社会'的野心。

So the question of like, kind of how do we envision what society is? And we went from, you know, in the sixties, and what we as social scientists refer to as Fordism or industrial modernity, where we tended to see society as a machine, as a complicated system. So basically, that way of seeing and way of understanding society stemmed, we argue, from a society built on large industry, large mass production, and it led to kind of a mass society. And it led to the kind of ambition of like, we can design, we can plan society, we can build it as if we're building a machine.

所以这个福特主义指的是亨利·福特和他的装配线？

So this is Fordism as in Henry Ford and his assembly lines.

正是。福特不仅创造了一种生产方式，更创造了一种消费模式。社会学家研究了这种工业理念如何渗透到社会，塑造了学校运作、公司运营的方式，广义上成为社会组织形式。本书观察到的是：我们已经从机械认识论转向了以复杂系统定义的时代，现在我们把社会比作群体或自组织系统，使用的是更具有机特质的隐喻。

Exactly. So that's, that image of the factory that Ford produced, because Ford, he didn't only produce a kind of way of producing, but also a way of consuming. And social scientists have looked at how that kind of idea of the industry, how it's kind of leaked into society and shaped how schools functions, how companies function, know, like that it became a kind of way of organizing society, broadly speaking. And so what we are kind of observing in this book is the fact that we've moved from that kind of machine epistemology into an era that's defined by kind of complex systems where we're talking about society as kind of swarms or as self organizing. Know, we're using these different metaphors that stem from, 're much more organic.

这些不再是机械隐喻。某种程度上我们也放弃了由国家设计社会以产生特定结果的野心。这种乌托邦式的、改善世界的希望已被我们抛弃。取而代之的是一种假设：系统的自然结果会天然更好——这就是这种意识形态的潜在前提。

They're not the kind of machinery. And we've also kind of in some ways abandoned the ambition of having the state design society to produce certain outcomes. So this kind of utopia, this hope of improving the world has we've kind of given up on that. And instead and there is this idea that now that it's implying that the outcomes of systems, like the bottom out outcomes of systems somehow will be inherently better. That's the kind of underlying assumption of this, what you might call ideology.

这正是我们要质疑的。嗯。因为我们要论证这并非自然现象。仍有权力形式在塑造社会，只是它们更隐蔽，通过塑造我们的互动方式（比如算法）来运作。

And so that's kind of what we're interrogating and questioning. Mhmm. Because we would argue that this isn't just something that's natural. There are still forms of power that are shaping society. They're just much less visible, and they're operating through by shaping kind of how we interact, by things like algorithms.

因此它们常常变得难以察觉，但仍能对社会产生巨大的结构性影响。比如通过调整互动规则——这正是平台在做的事，对吧？这可能带来非常重要的宏观影响。其中依然存在权力，只是这种权力具有新的认识论基础。

And so they often become kind of difficult to see, but they can still have very large structural outcomes on society. So for instance, it's by kind of fiddling with the rules of interaction, which is basically what platforms are doing, right? And that can have really important large scale outcomes. And there is still power there. It's just power that has new epistemology.

这真的非常非常有帮助，谢谢。我读了您书的开头但没看完，所以这就是我们做播客的原因——可以直接向您提问。换句话说，让我试着重新表述，看看我是否理解正确。

So that that's very, very helpful. Thank you. I mean, read the beginning of your book, but I didn't get through the whole thing. So this is this is why we have the podcast so I can just ask you questions. So in other words, let me try to rephrase it and see if I'm understanding.

我们曾幻想过设计自上而下的组织体系，无论是骨子里非常资本主义的亨利·福特试图管理工厂，还是政府的中央计划。虽然可以论证让事物自然达到平衡更有效率——无论是热力学平衡还是自由市场平衡——但您指出这或许确实是计算最优解的更好方式，却仍摆脱不了支配与权力的结构。

You know, we might have had this dream of planning an organization where it was top down and either Henry Ford, who I guess was very capitalist at heart, but still, you know, he was trying to organize his factories or, you know, central planning from a government. And there are various arguments you can make that simply letting things come to an equilibrium, is more efficient, whether it's a a sort of thermodynamic equilibrium or an economic, free market equilibrium or whatever. And and you're making the point that maybe so. Maybe that's sort of a a better way of calculating some optimum, but it still carries with it its structures of domination and power.

确实如此。即便达到平衡时，那些结果仍由初始条件决定。就像某些学术观点建立经济模型后，发现某些人变得极度贫困而少数人极度富有时，他们就说这是系统必然结果而不应改变。

For sure. I mean, there is this, you know, even emergent outcomes, even when we arrive at an equilibrium, it's still like those outcomes are still defined by the conditions that we came in with. Right? And so, I mean, it's like these arguments that you sometimes hear from certain parts of academia where it's they build some model of the economy and they find that certain people become very poor or like a large majority become very poor and certain people become very rich and they say, well, it's an inevitable outcome of this system. We shouldn't try to change it.

但这根本不成立，对吧？因为本可以制定不同规则产生不同结果。如果系统产生的后果有问题并伤害很多人，或许就该重新审视现行规则。那种认为市场是自然形成、不受国家权力影响的观念，其实早被质疑大半个世纪了——市场本质上就是国家权力的建构。

But that's just not It doesn't follow, right? Yeah, it doesn't follow because you could have produced other rules that would produce other outcomes and if your outcomes that are produced by your system are problematic and are harming a lot of people, then maybe you should reconsider the rules that you're operating under. And also with the notion that like that the this idea that the market would somehow be natural, and not, you know, like, that it would not be shaped by state power has been, you know, questioned for Of course. The better part of a hundred years that actually the market is very much a construction of state power.

我记得思考不同概率分布时的顿悟。纯数学角度讲，无论是均匀分布、幂律分布还是钟形曲线，当被问及哪种分布能使熵最大化时，答案是在不同约束条件下它们都能做到。

I remember the epiphany I had when I was thinking about different probability distributions. This is purely a math statement, but still it's it's, you know, important. Like, we have, uniform probability distributions, power laws, bell curves, whatever. And you might ask, well, which one maximizes the entropy? And the answer is they all do subject to different constraints.

关键在于宏观上对系统施加的约束条件，没有什么是真正必然的。

Like, it's it's all about the constraints you put on the system macroscopically. Nothing is there really inevitably.

是的。我认为这，我的意思是，这确实在某种程度上是我关注的核心。而且我认为它也是我许多论文和研究的核心，即我们从一个空间型社会转变而来，如果我们作为物理学家，可能会将其视为一种晶格结构。对吧，我们最近的邻居。

Yeah. And I think this, I mean, it's a really, this is very much at the core of what I am interested in to a certain degree. And I think it's also very much at the core of many of my papers and much of my research on this, which is this, the fact that as we've moved from a kind of a spatial society, we moved from a society that we could, if we were a physicist, one might think of it as a kind of lattice. Right. Our nearest neighbors.

没错，正是如此。然后我们转向了以网络结构为特征的数字社会。这些结构产生，或者说关联着不同类型的分布。作为社会科学家，一切都呈钟形曲线。对吧？

Yeah, exactly. And then we move to a digital society, which is characterized by network structures. And those produce, like they are associated with different types of distributions. As social scientists, everything is a bell curve. Right?

是的。但作为计算社会科学家，一切都遵循幂律分布。

Yeah. But as the computational social scientist, every everything is power laws.

一切都遵循幂律分布。没错。

Everything is power laws. Right.

这不仅关乎我们如何研究这些系统以及需要做出哪些假设。还在于某些人在数字网络结构中变得非常强大，而大多数人则无能为力，他们得不到关注或重要资源。这对社会来说是个深刻的问题，也是这些网络或结构的属性。

And that's not just like a a question of how we study these systems and what assumptions we need to make. It's also that certain people, when we have digital network structures become very powerful. And most people are powerless and they don't get attention or they don't get resources that are important. And that is profoundly problematic for society and they are attributes of these networks or the structures.

或许这是个不错的切入点，你刚才正好为我们做了铺垫。托马斯·谢林关于社会自组织的经典研究及其隔离模型。你有一个更新版本，但不如先告诉我们谢林原版模型是怎样的？

I mean, maybe a good inroad here, and I think you just gave us a good segue to it, but there's this classic work on self organization in society by Thomas Schelling and his, segregation model. So you have an updated version of that, but why don't you tell us what what Schelling's version was?

当然。因为他的研究我记得早在1969年就发表了，至今仍被频繁引用。我教学时也一直在用，因为我认为它至今仍是最好的模型——不仅是最早的。某种程度上我们计算社会科学家已经达到了巅峰。

Sure. Because his his I think it was published in 1969 already. And it's still kind of pointed to I I use it all the time when I'm teaching because I think it's It's still somehow the best model. It's not just the first. Also we peaked as computational social scientists.

简单来说，谢林的理论背景是这样的：他在大学食堂里走动时环顾四周，发现地理学家都坐在一张桌子旁，社会学家则在另一张桌子，这让他觉得奇怪。他联想到城市中也存在类似现象，于是试图探究为何隔离会成为各类系统中的普遍结果。为此，他拿出了一个棋盘——当时甚至还没有电脑模拟——在棋盘上摆放硬币，设想这个系统是一个网格，主体随机分布其上。

Basically, Schelling, think the backstory is that he was walking through his cafeteria in the university and was looking around and he saw like, it's weird that all the geographers are sitting in one table and sociologists are at another table and then he was like, Yeah, and the same thing with the city. He was trying to figure out why is segregation such a common outcome across systems? And so what he did, he actually took a checkers board. So he wasn't actually, it wasn't even a computer simulation back then. And then he had coins on the checkerboards and he was, okay, so let's imagine that the system is lattice and agents are randomly distributed on this checkerboard.

每个主体遵循一条简单规则：如果周围邻居中异类比例过高（因为存在两种硬币类型），他们就会移动到网格上随机可用的位置。可以设定这样的规则：当70%邻居是异类时他们完全满意，但如果达到90%，他们就会觉得'虽然我包容，但这未免太多了'。

And each agent, they follow a simple rule. So if more than a very high number of their neighbors, they're like kind of more neighbors are of a different type than themselves, because there are two types of coins, then they move to a random available space on the lattice. And you can have the rule of something like that. They're completely satisfied if 70% are of a different type than themselves. But if it's like 90%, they're like, yeah, okay, you know, I'm I'm tolerant, but like, come on.

没错。他们希望存在一定程度的隔离。

Right. They want some neighbors to be segregation.

他们希望部分邻居与自己相似，但并不强求大多数邻居必须如此。

They want some neighbors to be like themselves, but they don't insist that most of them be like themselves necessarily.

正是如此。当他们移动到新位置后会发生什么？你可能会预期，如果他们对80%异类邻居感到满意，系统最终会稳定在接近五五开的80%比例。

Exactly. Exactly. And then so they move to a rental space. And and so what happens? You would kind of expect that if they're happy with like 80% being of a different type, you would expect the system to settle on maybe something like 80%, pretty much fiftyfifty.

你不会预期出现高度隔离。但实际结果却是几乎完全隔离，即使设定很高的容忍阈值。原因何在？本质上这是连锁效应：一个人搬离社区后，社区变得更单一，促使其他邻居也相继搬离，形成连锁反应。

You wouldn't expect very high levels of segregation to emerge. But what actually happens is that you get almost complete segregation, even with very kind of high thresholds. And why is that? Well, basically you get this kind of cascade effects. So one person is leaving the neighborhood, the neighborhood is left more segregated, his neighbors also move and you get this kind of cascade.

这个模型告诉我们：融合状态极其不稳定。系统总会向某一端倾斜，任何社区最终都会倒向单一类型。这始终是我最钟爱的模型，在我看来，它对我们理解数字世界同样具有启示意义。

And so basically as what the model is telling us is that the integrated state is just very unstable. And so the system tends to tip over, like any neighborhood tends to tip over to one color or the other. And so to me looking, I mean, it's always been my favorite model. Yeah. And to me that seemed to tell us something also about the digital world.

所以我感兴趣的是，能否将这种现象推广到其他非空间性的互动结构中。于是我研究了各种论坛和平台，因为在观察社交媒体平台时，过去二十年一直存在争论——我们经常看到回音室效应，对吧？经常看到空间变得高度同质化。关于这是否由算法（比如过滤算法）驱动，长期存在争论。这就是帕里斯在其2011年著作《过滤泡》中提出的概念。

So I was interested in like, can we generalize this to other type of interaction structures that are non spatial. And so I look at kind of forums and different platforms because basically, when looking at platforms, when looking at social media, there's been like for the last twenty years of debate because we often see echo chambers, right? We often see spaces being very homogenous. Then there's been a long standing debate about whether that is driven by the algorithms, like the filtering algorithms. So there's this notion of Paris's notion of from his 2011 book, The Filter Bubble.

这种观点认为算法创造了茧房，只向我们展示我们已认同的内容。特别是近年来，越来越多观点认为事实并非如此——实际上是我们主动寻求隔离。我们不愿意接触其他观点。是的，我们不想遇到与我们意见相左的人。

So this idea that the algorithms create this cocoon, they only show us content that we already agreed to. And then especially in recent years, there's been more and more arguments that that's not true, that in fact we want to be segregated. We don't want to We don't want to be exposed to other ideas. Yeah. We don't want to encounter someone to disagree with us.

这就是两种立场之间的争论。在我看来，可能两者都不是答案。在这篇简短论文中，我实现了谢林模型，但在某种程度上将其迁移到线上场景。不同于传统网格结构，我设置了可以代表不同subreddit或网站的群体。

And so that's been the kind of the debate between those two positions. And to me it seemed like that might, you know, maybe it's neither. Right. So basically what I do in this paper, and it's a very simple and short paper, I implement the shelling model, but I move it online to a certain degree. So I look instead of having like a lattice, I look at, we have different groups that can represent kind of subreddits, it can represent websites.

智能体被随机分配位置——这与谢林模型非常相似，尽可能保持简单。每轮中它们与组内随机对象互动，规则与谢林模型相同：只要互动对象中有一定比例同类，它们就会保持满意状态。

And the agents are randomly located, so it's again very similar to the Schelling model. So it's as simple as possible. So agents are randomly located to these groups, and then in each round they interact with some random people in their group. And then it's the same rule as in the Schelling model. So they're happy with as long as at least some of their interactors are of their same type.

如果同类数量极少或没有，它们就会随机迁移到其他群体。根据这个设定，你大概能猜到结论——但我发现这种社区结构中的谢林隔离效应反而更强，更容易产生隔离。这对当前争论具有有趣启示：既非完全由过滤泡算法驱动，也非参与者主观意愿，而是这种社交互动结构的自然结果。研究还得出了一些反直觉的发现。

And if there's no one of their same type, or maybe a very few percentage, they move to a random other group. And I mean, with this background, you can kind of guess what the conclusion is, but what I find is that actually the shelling segregation effect is even stronger in this kind of communities, that they're even more prone to segregate. Of course, that has kind of interesting implications for this debate, because it's not necessarily either that this is driven by filter bubbles, that it's driven by algorithms, nor is it in the interest of anyone. It's just something that follows from having social interaction structured in these ways. And there's also some kind of counterintuitive results from this.

例如实际上，过滤算法和过滤泡的存在反而降低了隔离程度。因为如果算法总是向你推送认同你观点的内容，或确保信息中总有支持者，你的迁移倾向就会减弱。

So for instance, actually having filtering algorithms, having a filter bubble, it actually reduces segregation. Because if you have a filtering algorithm that always shows you someone who agrees with you, or whatever you're shown some messages, always get someone who agrees with you put into those messages, you will become less prone to move.

迁移倾向。明白了。

To moving. Okay.

在系统层面，这将减少隔离的程度。因此，系统在这些条件下更有可能保持稳定。

For the system level, it will reduce the amount of segregation. So the system will be much likely to likelier to be stable under those conditions.

所以

So

嗯，就是这样。

And so this is yeah.

在谢林模型中，它实际上是一个棋盘格局。甚至在我的书《大局观》里，我也稍微提到了谢林模型。所以它本质上是指你的最近邻关系，这在讨论城市中的种族隔离之类的问题时很有意义。那么你的意思是把它放在一个网络上，基本上是这样吗？也就是存在不同的节点可以相互影响？还是说它比这更具动态性？

In shelling, it's literally a a checkerboard. Even in my book, The Big Picture, I talked about the shelling model a little bit. And so it's literally your nearest neighbors, which makes sense if you're talking about racial segregation in a city or something like that. And so you're saying you're putting it on a network, basically, right, where there's different nodes that you can hear? Or is it more dynamical than that?

是像不同的空间结构那样，还是会随时间变化的某种东西？

Is it just like a different spatial structure or something that changes with time?

所以在这个案例中我关注的是群体，更像是子版块。也就是说你加入一个社区后，就会接触到该社区内的随机人群。你也可以在网络结构上运行它，但就这种意义而言，网络结构不太容易出现这种谢林式的动态，因为你需要某种传递性，明白吗？需要类似这样的机制：如果你离开社区，社区会变得更加隔离，这会增加其他人迁移的可能性。你会得到这种阈值效应，一个人的行为会触发另一个人的行为。

So I focus on groups in this case, so more like subreddits. So it's more like you join a community and then you're exposed to random people within that community. You can also run it on networks, but in this sense, the network structure are less prone to this emerging as kind of shelling dynamic because you would have to You need a kind of transitivity, you You know? Need something like if you're leaving the community, the community becomes more segregated and that increases the chance of someone else moving. You get this kind of threshold effect where one person triggers another person.

没错。而在网络中，要让这种情况发生，必须做出非常强的假设。

Right. And in network, have to make really strong assumptions for that to be the case.

哦，好吧。这个

Oh, okay. This

如果你对某人感到厌烦，直接取消关注就行，但这不会改变你朋友们的社交网络。所以他们不会更

If you get annoyed with someone, you just unfollow them, but that doesn't change your, you know, your friends' networks. So they're not gonna be more

可能。所以在你做的事情里得不到那种积极的反馈？

likely. So there's not the positive feedback you get in your in what you did?

至少不是这种谢林式的反馈。

At least not this kind of shelling feedback.

好的。顺便说一句，我的印象是，谢林当时提出的城市种族隔离解释并不需要来自高层的种族主义政策，比如红线歧视之类的。这完全只是个人偏好。但事实上，社会科学家研究发现，现实城市隔离的真正原因恰恰是来自高层的种族主义强制推行。

Good. Okay. And and by the way, my impression is that, you know, Schelling was offering an explanation for urban segregation that did not require, like, you know, racism handed down from on high via redlining or whatever. It was all just individual preferences. But in fact, when the social scientists have gone to look at it, the reason why real cities are segregated is in fact because of racism from on high, forcing it to happen.

是的。我认为这是个非常重要的观点，某种程度上也挺讽刺的。我在地理系做过四年博士后（虽然现在不在了），很快发现唯一不了解也不参与谢林隔离模型的社会科学家就是地理学者，因为这个模型从根本上就与他们的思维方式相冲突——我非常赞同这点。某种程度上这很有趣，我认为这与认识论问题相关，毕竟这是托马斯·谢林的隔离模型。

Yeah. No. I think this is a it's a really important point, and it's it's quite funny in some ways. I've been in geography for not anymore, but I was in as postdoc in the geography department for about four years and quickly realized that the only social scientists that do not really know or engage with the Shaolin model of segregation are the geographers, because it just seems fundamentally incompatible with that way of thinking, and I'm very much in agreement. It is quite interesting to a certain degree, and I think it connects to the question of epistemology because it's kind of the Thomas Schelling segregation model.

它深刻揭示了隔离的动态机制，但要把这种洞见与现有城市隔离研究（如你所说主要指向结构性种族主义、红线政策）进行对话确实很困难。不过某种程度上，两者都是对的，对吧？

It gives a very deep insight into the dynamics of segregation, but it's also really hard to bring that insight into dialogue with the existing literature on segregation in cities that, as you say very much point to, kind of structural racism, redlining. But to a certain degree, I mean, both are true. Right?

是的，正是如此。

Yeah. Exactly.

只是很难让这些理论相互对话。

It's just difficult to make these theories kind of speak to each other.

嗯，我一直以来的观点是，谢林模型非常擅长解释你一开始提出的问题，也就是人们在食堂里的座位选择。对吧？就像，虽然没有硬性规定说运动员和书呆子必须分坐两边，但由于这些偏好，他们总是会自然分开。

Well, my my line has always been that the Schelling model is really good at explaining exactly what you started with, which is where people sit in the cafeteria. Right? Like, there there's not rules like, you know, the the jocks the nerds have to sit on different sides, but they always do because of exactly these preferences.

是的，不，我认为这是个很好的观点，而且比起用它来思考城市隔离问题，这个例子可能也没那么具有争议性。

Yeah. No. I think that's a that's a good point, and it's probably also less of a a provocative example than using it to think about urban segregation.

所以这种认为我们会调整自己的社交网络或社交媒体使用习惯，使其更倾向于我们想听的那部分人群的观点，在现实世界中适用于哪些场景呢？我们是在谈论Twitter、YouTube、TikTok还是Facebook之类的平台？

And so this idea that, you know, we do change our social social network or social media usage to be just a little bit more within a set of people that we wanna hear, like, where does this apply in the real world? Is this are we thinking of Twitter or YouTube or TikTok or or Facebook or what?

没错。社会科学界对此其实已有长期争论，关于回声室效应普遍与否的问题。这至今仍是个非常激烈的辩论。但我想说的是，有迹象表明存在大量相对隔离的社区。比如观察大多数政治相关的subreddit，它们往往倾向于某一方的立场。

Yeah. So there's this has basically been a a a long debate in the social sciences, this kind of question of how pervasive echo chambers are or not. And it's it's still a very, very heated kind of debate. So but what I would say is basically that there are suggestions that there is quite a lot communities that are relatively segregated. And so looking at, for instance, most subreddits are if they are political, they tend to be towards one side or the other.

不过我认为这很有趣——Twitter历史上就是个反例。长期以来它都相当包容，能同时容纳不同政治立场。有趣的是，Twitter曾长期作为社会科学研究社交媒体平台的典型样本，因为它是少数能获取大量数据的平台之一（或者说曾经是）。因此大量计算社会科学研究都以Twitter为例来探讨社交媒体现象。可以说Twitter与其他平台截然不同，因为它确实汇集了所有政治立场，其特点更多表现为冲突性辩论。

But I do think it's an interesting I mean, Twitter has historically been a good example of the opposite. It was for a long time quite inclusive in the sense of having both political sides. And it's interesting because Twitter has kind of functioned as the kind of model of organism for social science research for looking at platforms, because it's been one of the few platforms where we can actually get a lot of data, or we could. And so, a lot of kind of computational social science research has looked at Twitter and used it as a kind of way of speaking about social media. I would say that Twitter is like, it was a very different platform from everything else because it actually had all political sides, and it was characterized much more by kind of conflictual debate.

但如果你观察较小的社区，观点上确实往往更加分化。这在某种程度上是个问题，因为政治理论家在讨论实现有效政治话语和协商所需条件时，其中一条就是我们需要多元化的观点。不能只有单一的政治立场——这道理我想相当明显。

But if you look at kind of smaller communities, do tend to be much more segregated in terms of opinion. And that can be a problem in the sense that if political theorists, when they talk about what conditions need to be fulfilled for us to have a kind of functioning political discourse, functioning deliberation, one of those conditions are that we need to have kind of diversity of opinions. We can't just have, like, political side, which is, I mean, pretty obvious, I guess.

我正想就此提问：如果人们在社交媒体上只与同类人互动，情况有多糟糕？我猜理想的政治结构不该如此。但要知道，大多数社交媒体用户并非为了扮演理想政治角色，他们只是来与朋友交流并获得认同。

I guess I was gonna ask a question about that. How bad is it if people on social media interact with people who are like them? Like, I I can imagine maybe a utopian political structure wouldn't be like that. But, you know, most people on social media are not there to be utopian political actors. They're there to talk to their friends and and be reinforced.

这真的有那么糟糕吗？

Is that so terrible?

我认为在很多情况下，这甚至可能非常有益。社交媒体改变社会的关键方式之一，就是让我们能够与世界各地的人建立联系。对许多群体（特别是少数群体）而言，比如当你作为LGBT成员在偏远村庄长大时，这种连接已被证明对心理健康和生活体验大有裨益。但就政治影响而言，其效果未必总是积极的。

So I mean, it's a I would say that I I think in a lot of cases, it can be even very beneficial. I mean, in some way, one of the ways that social media transformed society, one of the key ways was this possibility that we couldn't connect with anyone from all over the world. And so for a lot of communities, especially minorities, if you're LGBT and you grow up in a small village somewhere and you don't have anyone connect to, it's been shown that it's very beneficial for your mental health and for your experience, for your lived experience. At the same time, the way that it affects politics is not always as beneficial. Right.

因为显然，如果你所属的少数群体恰好是新纳粹主义的极端形式，这种聚集同样会产生类似效果——它使这些人能够形成共同体意识，从孤立个体转变为自信的政治群体，这在激进化方面可能相当危险。

Because obviously, if the minority that you belong to happen to be kind of some extremist form of neo Nazism, it seems to have similar kind of consequences for those communities, because it allows them to come together and form a kind of shared sense of community. And it transforms them from being someone isolated to a kind of confident political community. And that can be quite dangerous in terms of radicalization.

这是否意味着——我的印象是这种现象源于社交媒体等新技术让小众群体得以聚集？有些群体可能只是编织爱好者，另一些则是新纳粹分子。但有没有数据支持这点？相比1960年代，我们现在是否真的看到更多这类小群体的活跃迹象？

And is this mean, my impression is that it is something that comes from newfangled technology, social media, things like that, the ability of these smaller groups to come together. Like, some of them are just gonna be people who like to crochet, and others are gonna be neo Nazis. Right? And but is there is there data that backs that up in the in the sense that have we seen more viability of these small groups than we did in the nineteen sixties or whatever?

这类社会变迁其实很难评估。毕竟我们只有一个社会样本，难以比较没有社交媒体和数字媒介的社会形态。但可以确定的是，我们确实目睹了政治暴力事件的增加。

It's I mean, it's very difficult to kind of look at those kind of changes. Right? Because we unfortunately, we only have one society. We It's hard to compare how society would look different without social media, without digital media. But what we can say is that we've seen a kind of increase in political violence.

我们已经在许多国家目睹了一种民主倒退的现象。各种政治极端主义运动正进入政治主流。虽然很难断言这与社交媒体有直接关联，但很明显，在我们当前社会中，这种现象与社交及数字媒体密不可分。在我之前的著作《仇恨的亲密社群》中，我们就研究了其中一个线上社区。

We've seen a kind of democratic backsliding in a lot of countries. And we've seen the kind of political extremist movements entering into the political mainstream. Right. And whether or not that is costly linked to social media is very difficult to say, but it is very clear that it is in our current society very much entangled with social and digital media. And I've looked, so in my previous book, Intimate Communities of Hate, we look at one of these online communities.

基本上，我们试图通过深入研究'风暴前线'这个美国历史悠久的纳粹社区来回答这个问题。

Basically, try to to answer this question by going in-depth and looking at the Stormfront community, which is a it's a very old kind of Nazi community in The US. Right.

他们早在社交媒体出现前就存在了，对吧？

They predate social media, right?

没错，可以追溯到1995年左右。最棒的是，如果你能绕过他们的反爬虫安全措施，就能获取这二十多年来所有的对话数据。

Yeah. So basically it goes back to like '95. Okay. And the nice thing about it is that you like all of the data, all of the conversations over this, you know, long period of time is all available online, if you're able to scrape it, and bypass their various securities from preventing you from scraping it. So we have all of the data and all of the conversations over this, you know, twenty plus year period.

这让我们能够观察用户如何因参与该社区而改变。通过自然语言处理和文本分析，我们可以看到个体在互动中语言模式的变化，以及他们自我认知的转变轨迹。研究显示，这本质上是一个关于社群身份重塑的过程。比如新成员最初使用'我'自称，但逐渐会改用'我们'或'SF'（风暴前线缩写），这标志着他们开始将自己视为集体的一部分。

And so that allows us to kind of look at how the users are changed by interacting with this community. So we can kind of use natural language processing and various forms of text analysis, and kind of see of how individuals, when they interact in this community, how does it change their language and different markers of like how they perceive themselves and so on. And the kind of image that we come out with is much more, it's very much a kind of a question of a community formation of changing identity and so on. And so just an example, we can kind of see how when they first come in, they use I and my and speak of themselves. But then over time they start saying we or S F, you know, for Stormfront, because they start, you know, that's kind of a marker of them starting to think of themselves as part of a collective and as part of something larger.

这非常有趣，因为不同于我们刚才讨论的研究中人们只是寻找志同道合者，这里展现的是互动反馈对个体身份的塑造作用——人们在这些交流中改变着自我认同。

So that's very interesting because it's not just about people, in the study we were just talking about, you're talking about associating with people who are like minded. And here you're talking about the feedback acting on yourself. And, you know, the individual people are sort of changing their identities in response to those interactions.

确实如此。这明显是个双向反馈过程：既有群体隔离，又有身份转变。以风暴前线为例，2008年奥巴马当选后几天，新用户激增。从他们的言论中能看到情感上的困惑与焦虑——他们难以理解一个能言善辩的黑人成为总统的世界，这冲击了他们的自我认知体系。

For sure. I mean, I think it's very clearly, you know, kind of a feedback process, right, where we have a kind of segregation mixed with a kind of changing of identities. And so we can kind of see, you know, in the Stormfront case, especially after the two thousand and eight Obama election is that the few days after the election, there was just a huge surge of users coming in. So new people joining and looking at what they're saying, can see this kind of emotional kind of confusion and anxiety and they're trying to, they somehow feel confused about this world that they're living in where a black person who's articulate can become precedent. And what does that mean for their self identity and how can they make sense of this?

然后这个社群功能有点像情感倾诉疗法，让他们能找到新的叙事方式，某种程度上化解这种情感焦虑，将其从被动的焦虑转化为主动的愤怒或愤慨。他们最终会形成这些荒谬的叙事，比如关于犹太人的，或者白人至上种族之类的论调。这整个过程本质上是在身份认同、情感和自我叙事层面运作的。

And then the community function is kind of emotional talk therapy that allows them to find new narratives and kind of resolve this emotional anxiety and turn it into kind of from something passive like anxiety to something active like anger or outrage. And they come out with these narratives about, know, these absurd narratives about the Jews and, you know, like that actually the whites are the superior race. It was just, you know, whatever that happened. And, you know, so it's very much a kind of process that is on the level of, like, identity and emotion and kind of self narratives.

这非常有趣，因为一切都能回溯到《心景》播客的第一期节目。我采访了社会心理学家卡罗尔·塔夫里斯，她提出了‘选择金字塔’理论——想象两个人在做某个选择时原本五五开，比如决定穿哪双运动鞋之类。但一旦他们做出不同选择，就会开始自我合理化这个选择。

It's very interesting because it all cycles back with the very first podcast episode of Mindscape. I interviewed Carol Tavris, who's a social psychologist, And she has this idea of the pyramid of choice when if you imagine two people who are basically fifty fifty as to how they could make some certain choice. Right? You know, what what sneakers to wear or whatever. But once they make the choice, if they make it in different they start justifying that choice to themselves.

尽管在坍缩为特定选项的波函数之前它们本质上无法区分，但它们最终会相距甚远。

And they end up very far apart, even though they were essentially indistinguishable before they collapse their wave function on that particular option.

是的，不，这说得通。所以我的意思是，归根结底，这其实是人们互动的结构性环境如何导致这类结果的问题。确实如此。

Yeah. No, that makes sense. And so, I mean, ultimately, it's kind of the question of the the the kind of structural context in which people are interacting that can produce this kind of outcomes. And so yeah.

我能询问一下你的模型或谢林模型吗？作为一个物理学家，我不禁联想到伊辛模型，它们有些相似但不完全相同。在伊辛模型中，晶格上的自旋会相互作用，我们通过引入温度来设定翻转概率。对吧？自旋有一定的概率会发生翻转等等。

Can I ask about either your model or the Schelling model? There seems to be, like, as the physicist in me thinks of the Ising model, which is similar but not exactly the same, where you have spins that are interacting on a lattice. And the thing we do there is we introduce probabilities by having a temperature. Right? There's some chance that the spin is going to flip and whatever.

而在谢林模型中，概率体现在当人们决定搬迁时，目标位置是随机的。但关于是否搬迁的选择本身并非随机。对吧？这完全取决于他们周围不同类型邻居的数量。所以有人研究过这种设定吗？

And in the Schelling model, there's a probability because when the person decides to move, where they move to is random. But the choice about whether to move or not is not random. Right? That's just determined by how many neighbors they have of each kind. So have people done that?

有没有人尝试引入搬迁概率而非确定性规则，观察这是否会改变模型结果？

Have people introduced a probability of moving rather than a certainty and and seen if that changes anything?

好问题。我我真的不知道。好吧。但我知道你很感兴趣。有人

Good question. I I honestly don't know. Okay. But I I know you're interested. Someone

应该去做。那边在听的人应该

should do that. Someone listening out there should be

我们该做点什么。

something for us to do.

对。对。完全正确。

Yeah. Yeah. Absolutely.

作者就是它了。

Author it is.

然后好吧。所以你后来做了另一项研究，也是最近才发表的，用了大语言模型。这里呢，嗯，我让你来讲吧，但核心思路是：不同于之前那些网格上无意识的点互相交互，这次你们实际上是让小型智能体在社交媒体环境中互相交谈并做出选择。那么结果如何？

And then okay. So you then did a different study, which came out also very recently using large language models. And and here, well, I'll let you tell the story, but the idea is rather than just having these mindless dots on a on a grid or whatever that are interacting with each other, you literally had little agents talking to each other and making choices in a social media context. So how did that go?

是的。基本上，这个研究部分是为了回应社会科学家们长期以来的批评——至少很多社会科学家对基于智能体建模的批评——就是这些基于规则的智能体并不能很好地代表人类行为的全貌，我觉得这个批评很合理。在很多情况下，比如谢林隔离模型，这种简洁性非常有用，能帮助我们揭示某些本质上是结构性的涌现现象。但在其他场景下，这种简化也确实存在局限对吧？比如观察社交媒体上的政治讨论，就是个例子——这些更复杂的行为模式确实会产生重要影响，不是吗？

Yeah. So so basically, the aim of this is to address in part this, like, kind of long standing criticism from social scientists, or at least from a lot of social scientists when it comes to agent based modeling, which is that these rule based agents are just not very good representations of the full spectrum of human behavior, which is, I think, fair enough. And I think, I mean, in a lot of cases, like the shelling segregation model, the simplicity is very useful and it allows us to throw light on some emergent phenomenon that is ultimately structural. But in other contexts, it is also limiting, right? So looking at social media, like politics on social media, for instance, it's an example where these richer behaviors, they also can really matter, right?

确实，我们无法真正将文化因素与结构因素割裂开来。我们需要将它们视为相互交织、相互作用的整体。这就是背景情况，因为我在这里感兴趣的，或者说我们共同关注的是这个问题——毕竟我们花了大约二十年时间批评社交媒体，指出其问题及与各种负面后果的关联。但现在越来越多人开始思考：我们能否更建设性一些？

Right. We cannot really separate the cultural from the structural. We need to look at them as kind of intertwined and as interacting. And so that's kind of the background because what I'm interested in here or what we're interested in is this question because we basically, we spent twenty years or something criticizing social media and pointing to the problems and are linked to various problematic outcomes. But now there's more and more kind of interest in like, can we be a little bit more constructive?

我们能否真正采取行动？因为归根结底，如果社交媒体能塑造一种由愤怒驱动、激进化的政治形态，它也应该能够塑造一种亲社会的政治形式，产生更健康的政治和社会结果。这就是基本理念。那么我们该如何研究这个问题？

Can we actually do something about this? Because ultimately, if social media can shape a politics that is outrage driven and radicalizing, it should also be able to shape a form of politics that is pro social, that has healthier political and social outcomes. Yeah. So that's the kind of idea. And how do we study that?

使用观察数据效果不佳。因此基于建模的方法会非常有益。我们正在使用基于智能体的模型，不再采用规则遵循者，而是使用大语言模型。它们充当了人类行为的替代品。

Well, using observational data doesn't really work. So that's a kind of modeling approach can be really beneficial. And so we're using agent based models, having instead of these rule followers, we're using large language model. And they work as kind of stand ins for for humans.

或许我可以快速打断一下，能否请您简要介绍一下基于智能体模型的概念？比如它与什么模型相对立？哪些模型不属于基于智能体模型？这类模型主要用于哪些领域？

Maybe if I if I could interrupt you just quickly, I mean, maybe give a little bit of background onto the concept of an agent based model, like, as opposed to what? What is what kind of models are not agent based and what are agent based and and what is that used for?

当然。在社会科学领域，我们传统上研究社会世界的方式——回到最初讨论——更多将其视为复杂系统。我们通常将社会看作变量间的相互作用，这在很多情况下很有效。但当你思考社会世界中这种复杂层面时，比如智能体间的互动导致意外结果，那些传统的基于变量的方法就完全失效了。比如你如何用变量研究鸟群的集群飞行现象？

Sure. I mean, so in the social sciences, the way that we have traditionally approached the social world is in in to link back to where we started very much as a kind of complicated system. So we tend to think of society as kind of variables interacting, which in a lot of cases can work really well. But if you're thinking of this kind of complex aspects of the social world, where you have interaction between agents and then leading to unexpected outcomes, those traditional kind of variable based approaches just don't work at all. Like how would you, like using used variables, how would you like study the murmuration, you know, like of- birds?

这根本不可能。因此基于智能体的模型采用自下而上的建模方法。举个简单例子就是谢林模型：智能体代表个体，遵循简单规则，然后你观察结果。这种方法让你能将个体微观行为与系统层面的结果联系起来，而这些结果往往与你预设的规则大相径庭。

It just wouldn't be possible. And so agent based models is kind of using this kind of bottom up modeling approach where we And so one example would just be the shelling model, right? Like you have agents that are individuals, they follow simple rules and then you look at the outcomes. And that allows you to think together the kind of micro behavior of individuals with system level outcomes that can often be kind of unexpected from the rules that you're coming in with.

单个智能体本身不需要非常复杂对吗？

The individual agents need not be very complex themselves?

传统上它们并非如此。它们历来只是简单的规则遵循者。比如谢林阈值规则，或者优化规则。但要让智能体在推理或语言生成方面模仿人类行为，传统方法根本不可能实现。对吧？

Traditionally, they haven't been. So they traditionally have been kind of simple rule followers. You have like, know, like the shelling threshold rule, or you have maybe an optimization rule. But basically building agents that would kind of mimic human behavior in terms of reasoning or language production, it just becomes impossible traditionally. Right?

因为这确实极其复杂。是的。你需要构建一种推理智能体，而这在几年前我们还不具备。但随着ChatGPT的出现和大语言模型的兴起，人们开始高度关注能否用这些模型作为基于智能体的模型的一部分来模拟社会行为。这就是我们正在做的尝试，同时也希望借此贡献某种社会科学理论，增进我们对社交媒体及其动态的理解。

Like it would just be extremely complicated. Yeah. You would have to build a kind of a reasoning agent, which we just haven't had up until a few years ago. But so basically when Chateappity came out and with the kind of rise of large language models, it just became a huge amount of interest in whether we can use these models to, as part of agent based models, to kind of simulate social behavior. And so that's kind of what we're doing, but we're trying to also use it to contribute some kind of social scientific theory and contribute to our understanding of social media and its dynamics.

那么你们具体做了什么实验？我粗略理解为：把一堆大语言模型放到一个虚拟社交网络上任其发挥。

And so what exactly was the experiment you did? I think of it as roughly speaking, letting loose a bunch of LLMs on a fake social network.

差不多就是这样。我们的核心构想是创建一个社交媒体平台，让它产生真实社交媒体上观察到的负面结果，然后尝试文献中提出的一系列解决方案。我们原以为需要反复调试系统才能产生问题性结果，这样才能观察这些结果的稳定性，以及哪些解决方案最有效。我们聚焦的问题——基于政治理论——是那些阻碍公共审议或公共对话的社交媒体环境条件，使得在这些平台上难以形成良性政治互动。目前已涉及三个具体方面：

That's that's pretty much it. But basically, our idea was to try to create a social media platform and make it produce the negative outcomes that have been observed on real social media, and then try out a bunch of suggestions from the literature on how we can address those problems. And so our expectation coming in was kind of that we would have to fiddle a lot with the system and try to make it produce problematic outcomes so we could then see how stable those outcomes are and how easy, what solutions are best for addressing the problems. And so the problems that we focus on, it's kind of conditions of social media that make public deliberation or public conversation, difficult, that makes it difficult to have a kind of a functioning politics playing out on these platforms, drawing on kind of political theory. And so there are three different things that we've already touched on a little bit, but so one of them are echo chambers that you do need to have, if you're gonna have a kind of constructive conversation across the political divide, you need to have both sides of the political divide present.

首先是回声室效应，要实现跨越政治分歧的建设性对话，必须让对立双方都参与进来。否则就...这是第一个必要条件。其次是注意力和平等性问题——要实现良性政治讨论，个体间必须保持相对平等。不能让两三个人垄断整个对话，那就不是公共讨论而是广播了。最后是被称作社交媒体棱镜的现象。

Otherwise, it's gonna be really So that's one condition you need. And then the second is this kind of question of attention and equality that we also touched on. If you're going to have functioning political discourse, you need to have relative equality among individuals. So you can't just have two or three individuals dominating the entire conversation because that's not a public discourse, that's just broadcasting. And then finally, what's been referred to as the kind of social media prism.

即需要构建一种建设性辩论，让人们真正试图达成解决方案。这指向社交媒体往往放大极端、极化、冲突性声音的问题，严重破坏有效对话。这三个现象就是我们试图复现的研究目标。

So this is the idea that you need to have a kind of constructive debate where people are actually trying to come to a solution. And so that speaks to this question of that social media has kind of tended to benefit loud, polarizing, conflictual voices. Right. And that is very much kind of undermining functioning conversations. And so those are the three outcomes that we were trying to kind of see if we could produce.

说实话，我们原本预计这会非常困难。

And we were expecting that to be quite hard, to be honest.

你觉得那会很难，这真贴心。

That's so sweet that you thought that would be hard.

嗯，我是说，文献已经指出或论证过，很多这类现象，特别是社交媒体棱镜效应或这种极化倾向，实际上是参与度算法的表现。它们就像是社交媒体识别出最极端的言论，然后硬塞到你面前，让你感到不安，从而增加你评论或参与帖子的概率。

Well, I mean, the literature has kind of pointed to or argued that a lot of these are, especially the kind of the social media prisms or this kind of the polarizing tendency, that those will be expression of engagement algorithms. And so that they would be like the expression of social media identifying the most outrageous things that are being said, and then shove it in your face to to kind of make you upset and and increase the probability that you will comment or engage with a post.

所以我们在这里试图验证的两种可能性是：一是当你遇到回声室效应和极化现象时，这是算法或平台的责任；二是这纯粹是人类天性使然。

So So the sort of two alternatives that we're trying to test here are, one is that when you get these echo chambers and polarization, things like that, it's the algorithm's fault or the platform's fault versus this is just human nature.

呃，我不确定是否该这样设定背景，因为在这项研究中，我们其实只是假设这至少来自算法，而不是单纯从人类行为中自然产生的。

Well, so I'm not sure if I would put that as the context really, because it's also like in this study, we don't was honestly just kind of assuming that it was from the algorithms at least, that it's not just something that would emerge from human behavior.

但是

But

这些其实是人与平台规则互动的结构性结果。但对我来说，至少这种社交媒体棱镜效应是个相当具体的现象。最极端的声音获得更多关注是个奇怪的产物。我原本预期会是——我之前也写过相关文章，称之为'触发泡泡'。它不是过滤泡泡，而是触发泡泡，是社交媒体算法试图刺激你、让你不安，从而促使你参与互动，因为这才是平台最终盈利的方式。

these are kind of structural outcomes from the interaction between people and the rules of the platform. But to me, at least this kind of social media prism, it's a rather specific thing. It's kind of odd outcome that the most extreme voices get more attention. And so I was expecting that to be, and I've written about this before arguing that it's the kind of what I've called the trigger bubble. So it's not the filter bubble, but the trigger bubble, it's the social media algorithm trying to trigger you, make you upset in order to make you engage, but because that's how the platforms ultimately make money.

让你确信。

Make you Sure.

这是我们大家都听说过的。

That's what we've all heard.

它们确实收集信息，分析你是谁并出售广告。这基本上就是我的预期。我们最初构建的是一个最基础不过的平台，只有这些智能体。需要说明的是，这些智能体的个性特征来源于美国国家选举调查（ANES），该调查包含美国公民的详细信息，包括政治倾向和是否喜欢钓鱼等，我们掌握了非常详尽的数据。

They do draw information, they figure out who you are and they sell ads. And so that was kind of my expectation. And basically what we started was just building the most bare bones platform we could imagine, which is just the agents. So I should say also that the agents, their personalities are We take the ANES, so the American National Election Survey, which has very detailed information about US citizens, including their politics and if they like to go fishing. We have a very detailed information.

本质上我们将这些数据转化为人设描述。因为大语言模型特别擅长模仿人物，比如你可以让它用莎士比亚的风格解释微波炉原理，它会表现得非常出色。

Basically we turn that into a description for persona. Because, you know, LLMs, they really love to impersonate people. You you can can ask you to, you know, explain your microwave in the voice of Shakespeare, and it will do an excellent job.

让我确认一下，所以你们投放在社交媒体上的这些大语言模型，并非都是亚当夏娃式的空白状态，而是从一开始就赋予了背景故事？

So so just to be clear, so the the LLMs that you let loose on social media, they weren't all, you know, Adam and Eve, they weren't all tabula rasa from the start, you gave them a backstory.

没错。关键不在于要让它们完全真实地再现特定个体的全部特征，而是希望保持多样性并体现一定的文化丰富性。然后它们就可以在这个极其简单的平台上互动——基本上就是查看我们模拟当天随机选取的最新新闻。

Exactly. So the point here isn't necessarily that, you know, we want them to be completely realistic encapsulations of this particular individual that they're enacting. We just want to have kind of a diversity and want to capture a little bit of that cultural richness. And so then they are allowed to interact on this platform, and the platform is very simple. So basically they look at the, they see the most recent news, random selection of news from the specific day that we're simulating.

接着它们可以选择就这些新闻发布帖子，或根据时间线转发关注对象的帖子。它们还能根据看到的帖子决定关注推荐用户。关注后就能进入对方主页，查看个人简介和最新动态。

And then they can choose to post about these news, and they can just write a post about whatever based on it. Or they see the timeline and they can choose to repost what someone else has written, someone that they follow. They can also see, based on the post that they see, they can choose to follow someone that shows up in their feed. And if they follow someone, they kind of go into their timeline, they see their little presentation about them, and then they see their most recent posts.

哇。明白了。那它们也能取消关注吗？

Wow. Okay. And can they unfollow people too?

不，我们没有取消关注功能。这个平台非常基础。

No. We don't we don't have unfollowing. It's it's really bare bones.

好的。但确实存在选择权，用户可以自行决定是否关注某人。

Okay. But there is some ability. They can choose to follow someone or not.

没错。我们正在观察的是通过这种互动形成的网络结构。是的。因为我们希望有简单可量化的指标，而不只是关注他们的对话内容之类的。

Exactly. And so what we're looking out at the at the outcomes here is the network structure that emerges through this interaction. Yeah. Okay. Because we want to have simple measurable things, and we don't want to just look at how they're talking or something.

这更像是大语言模型的直接产物。我们想要的是能体现平台结构特征的东西。网络结构的有趣之处在于它们具有涌现性，是通过结构性结果产生的。这正是我们关注的重点，因此我们试图识别这三个特征。

That's like an immediate outcome of the large language model. We want to have something that's more an expression of the structure of the platform. And network structures are interesting in the sense that they are very much kind of emergent and they are produced through their structural outcomes. And so that's what we're focusing on. And so we're trying to identify these three attributes.

令人惊讶的是，我们实际上只需要提供这个基础平台就足够了。我们观察到了这三个被广泛认为是社交媒体问题的特征。但需要说明的是，这并不意味着参与算法没有问题，它们可能仍在加剧问题，但确实表明仅移除算法并不能彻底解决问题。

And to our surprise, we didn't actually need to do anything more than provide this barebone platform. And we got these three features widely consider the problematic aspects of social media. But I should say that that doesn't mean that engagement algorithms aren't problematic. That doesn't mean that it might still be that they're making matters worse, but it does imply that removing them will not completely solve the problem.

那么你们的社交网络有多少用户？

So how many users did your social network have?

我们测试时使用了500名用户。

We ran it with 500 users.

500。好的。

500. Okay.

这种方法的一个小局限在于，与传统基于代理的模型相比，它的运行成本相当高。而传统基于代理的模型一直因计算成本过高而受到批评，所以这算是该方法的一个弱点。

Which is is a little bit of a limitation with their approach, is that it is quite expensive compared to running a conventional agent based model. Right. And conventional agent based model have always been criticized for being very expensive to run computationally, So it is kind of a weakness of of this approach.

你刚才提到出现了不良后果。能提醒我们具体是哪些不良后果吗？应该有个清单。

And you said that, the bad outcomes happened. Reminded us what the bad outcomes were? There's a list.

是的。基本上会出现回音室效应——民主党和共和党最终不再关注对方，只在自己的圈子里交流。还会产生高度不平等现象，本质上就是注意力分配的幂律分布。

Yeah. So basically, you get echo chambers. So Democrats and Republicans end up not following each other. They're just talking to themselves. You get high levels of inequality, basically power law distributions of attention.

少数用户几乎主导了整个舆论场。最后还会出现克里斯·贝尔所说的'社交媒体棱镜'现象——越是极化、极端的用户往往获得更多关注。

So a few users kind of dominate the entire discourse. And then finally, get what Chris Bail has called the the social media prism, which is that the more polarized, more extreme users tend to have more attention.

好的。我的意思是，虽然很容易用拟人化的方式谈论大语言模型，比如询问它们的动机之类。但当然这是不合理的——它们只是在模拟，并没有真正的动机。那么应该如何正确表述这个问题：为什么大语言模型会倾向于关注政治光谱上与自身立场相同的人？

Okay. And is there I mean, I it's very, very tempting to speak anthropomorphically about LLMs, right, and to ask about their motivations or something like that. But, of course, that's, you know, illegitimate that they're just faking that. They don't really have motivations. But is what is the right way of phrasing the answer to the question, why does an LLM want to follow people on its own political side of the spectrum?

这些结果的产生原因各有不同。比如我们观察到的幂律分布——注意力分配极度不均的现象，这其实源自偏好依附机制：获得新关注者的概率与你现有粉丝数成正比。某种程度上我对这个结果并不意外，因为正如之前提到的，这是社交网络的已知特征。但我们发现的另一个独特动态机制此前未被研究过——这种机制需要大语言模型与社交网络的特殊结合才能显现：我们知道转发行为（人们重新发布或分享内容）往往具有高度情绪化和应激性。我们倾向于分享那些让我们愤怒或引发强烈情绪反应的内容。

So, I mean, so so where these outcomes are kind of stemming from, I think it's a little bit different outcomes each one, but so for instance, that we get the power law distribution that we get very unequal distribution of attention, I mean, that's kind of stems from preferential attachment, right? The probability of you getting a follower is proportional to how many followers you already have. So I wasn't so surprised to a certain degree that that emerged because it is a well known feature of network, as I already mentioned earlier. But the other kind of dynamic that we identify that we haven't really seen studied before, because you do kind of need to have this kind of special large language model network combination to be able to study it, which is the fact that basically we know that retweets, people repost or sharing, is very effective, is very emotional, is very much reactive. So we see something that we're upset about or that we're really emotionally reacting to, and those are the types of things that we tend to share.

众所周知，已有观点认为这塑造了我们在社交媒体上看到的内容。但我们补充的是，它不仅影响你所见的内容，还影响着网络结构的构建与逐步形成。因此，在高度情绪化的分享内容与你的关注对象之间，存在着一种反馈效应。正是这种效应形成了某种动态——观点极端的用户往往拥有更多粉丝并获得更多关注，因为这类分享本身就是反馈循环的一部分。

That is well known and it's been kind of argued that that shapes the content that we see on social media. But what we're adding to that is that it's not only shaping what content you see, but it's also shaping the construction, the gradual formation of the network structure. And so you have a feedback effect between what is shared, which is very much emotional, and who you follow. And that is what's creating this kind of dynamic where more polarized users tend to have more followers and get more attention, because it is the kind of sharing is part of this kind of feedback process.

我理解大语言模型能如何回应查询并给出回答，但你们是否需要额外编造一套关于何时关注某人的指令？

I understand how an LLM can respond to a query and say something, but did you have to cook up an extra set of instructions for when to follow somebody?

我们只是...我们并没有提供任何...你看，这就像是设定好角色身份后问：基于这个身份，你会如何在这种情况下行动？然后就让它们根据这个基础来做出反应。

We only we we don't provide them with any you know, it's just like, this is your persona. How would you based on this persona, how would you act in this situation? And then just have them kind of respond on the basis of that.

明白了。那么这个身份设定包括他们的政治倾向，以及是否喜欢钓鱼之类的偏好吗？

Okay. And the the persona includes their political orientation as well as whether they like fishing or whatever?

没错。它包含政治立场。因此在回音室效应方面，结果或许并不那么令人惊讶。我们还可以观察大语言模型给出的关注动机——就像在回音室场景中，它会表现为：'既然我是民主党人且立场坚定，就不想与对立阵营的人互动'。

Exactly. So it contains their political affiliation. And so in terms of the echo chambers, it's also not maybe so surprising. And we can also look at the kind of motivations that the LLM is giving for why they should follow them. And it does, like in terms of the echo chambers, it's kind of like, okay, so I don't want it, like since I'm Democrat and I feel strongly about it, I don't want to engage with this person who's from the other side.

但有意思的是，这种现象超越了个人选择范畴。结构性回音室的形成动态并非仅源于个人选择，更因为他们的选择正在影响他人——这改变了被分享的信息内容，并逐步重塑着网络结构。

But I think that it is interesting that it goes beyond just that individual choice. It's not like the dynamic and the emergence of the structural echo chambers is not just because individuals are choosing it, but because their choices are impacting others, because it shapes what messages are shared and how the structure of the network is gradually built up.

那么个体智能体是否有机会被其在网络中的成功所影响？它们的受众是否被俘获？大语言模型是否学会了变得更挑衅以获取更多转发？

And is there, room for the individual agents, as it were, to, be affected by their success on the network? Is their audience captured? Did the LLMs learn to be more provocative so they get more retweets?

因此，我认为这确实非常重要，并且是理解社交媒体如何重塑政治的核心。这不仅体现在我们使用平台时变得更加两极分化，更在于平台正在塑造影响社会的激励机制——它们决定了谁能获得关注而谁不能。从这个意义上说，它正在重塑一种以出格行为为特征的政治生态。不，简而言之就是不行。不过这篇论文正是我目前正在研究的课题。

Not in so this is something that I think is is really important and I think is really central to how social media, is reshaping politics because it's not just that we get more polarized when we go on the platforms, but the the platforms are kind of shaping the incentive structures that are shaping society because they are defining who gets attention and who doesn't. And so in that sense, it's reshaping a politics built around being kind of outrageous and so on. No. Briefly put, no. But this is a paper that I'm currently working on actually.

好的。很棒。这正是我感兴趣的领域。

Okay. Good. Very much what I'm interested in.

当然会有人说：大语言模型不是人类。那么当我们使用这些大语言模型时，应该把哪些关于模仿人类行为局限性的担忧放在首位？

And, of course, someone is gonna say, look, LLMs are not people. Right? Like, what are what worries about limitations of mimicking human behavior should we, have in the forefronts of our minds when we're using these LLMs?

是的，目前这确实是个重大争议话题。虽然人们对这类生成式模拟充满热情，但我始终持更审慎的态度。我和合著者Mike Leroy即将发表的论文就专门探讨这个问题：我们能在多大程度上信任这些模型？它们作为人类行为的表征是否有效？某种程度上这是个很有趣的问题，因为这些模型确实比简单规则列表更能真实反映人类行为，但验证起来却困难得多，对吧？

Yeah. So this is it's a really kind of big debate right now because there's a lot of kind of excitement around this kind of generative simulation. I've tended to be on the more skeptical side of this. So me and my co author, Mike Leroy, we also have another paper coming out where we were specifically trying to answer this question and like, how much can we trust this and how much can we trust that they're valid as representations of human behavior? And it's a very interesting question to a certain degree because the models are more realistic as representation of human behavior than just a list of rules, But they're also much harder to validate, right?

因为你无法确切知道模型在演绎什么，也很难校准其行为以匹配人类行为。这就成了个棘手的问题——它们既不像实证方法那样具有高效度，也不像形式模型那样简洁易懂，而是处于中间地带。因此我们在这篇论文中的研究思路很明确：不关注模型生成的文本内容本身。

Because you don't know exactly what is playing out and it's very hard to calibrate their behavior to match human behavior. And so it becomes a difficult question. They're ending up somewhere in between empirical methods that have a high level of validity and formal models that are parsimonious, that are easy to understand, and they're neither. So it's a little bit unclear how we can use them. And the way that we think about that in this paper is precisely is linked to the fact that we're not looking at the text that they're producing.

我们试图研究平台架构中涌现的结构性特征。也就是说，我们刻意与其生成行为保持距离，转而观察这些行为产生的结构性结果。某种程度上我们关注的是这些结果的稳健性。什么是稳健性？简单来说就像谢林隔离模型那样——建模圈有个玩笑：只要构建包含网格结构的地理模型，就必然会出现谢林隔离效应，因为这个结果太稳定了，你必须费很大功夫才能研究其他效应。

We're trying to look at kind of structural features that are emerging as aspects of the platform. Not that you know, like we're trying to distance ourselves from their behavior and try to look at the kind of structural outcomes of that behavior. And to a certain degree, what we're interested in is the kind of the robustness of those outcomes. So in the sense, then what do I mean by robustness? Well, basically similar to the shelling segregation model, you can basically It's the kind of joke among models that whenever you build a geographical model that has any type of lattice in it, always get the Schelling Circulation effect, because it's just such a stable outcome and you have to really work hard to study any other effect.

这正是我们感兴趣的研究方向：探究平台中涌现模式的稳健程度。如果我们只是分析对话的毒性特征，我会对其实践意义存疑。但鉴于这些涌现模式展现出的稳健性，我更有信心了。不过话说回来，这方面确实还需要更多研究。

And so that's the kind of effect that we're interested in. We're interested in seeing how robust these emergent outcomes from the platform are. And so if we were just looking at the kind of toxicity of their conversation, I would feel less confident that these are something that are actually playing out in real world. But given the robustness of these emergent patterns, I have more confidence. But that being said, we definitely also need to do more studies on this.

嗯，看起来从炮击到你的LLM确实展现了某种稳健性。我是说，这些看似各异的小规模动态却导致了相似的大规模行为。听起来应该存在一个定理，类似于热力学第二定律，但关于极化或隔离这类状态具有最低自由能的原理。我不确定该怎么命名它。但在社交媒体的统计力学中是否存在这类定理呢？

Well, does seem like the robustness from shelling to your LLMs. I mean, these are these are sort of different individual small scale dynamics giving rise to similar large scale behavior. It sounds like there should be a theorem, like the second law of thermodynamics, but about polarization or segregation or something like that being the state with lowest free energy. I don't know I don't know what to call it. But are there theorems like that in the statistical mechanics of social media?

确实有研究者采用更偏向物理学家的方法研究观点动力学。我这里有位同事正在用层层嵌套的伊辛模型来理解社交媒体。而我的方法更倾向于基于实证数据并尝试成为系统本身。遗憾的是，在开放程度如此高的社会系统中，这类规律通常很难被发现。回到社会是复杂系统还是复合系统的问题，我认为它两者都不是——或者说在某种程度上两者皆是。

So there are people doing these much more physicist approaches, kind of opinion dynamics. I have a colleague here who's basically doing Ising models wrapped into Ising models in order to understand social media. My approach tends to be a little bit more based on empirical data and to trying to be the system. I think it generally is hard to find, unfortunately, this kind of loss in the social world because it is such an open system. And linking back to the question of whether society is a complex system or a complicated system, I would argue that it's neither or it's both to certain degree.

美国政治极化中让我感到奇怪的一个特点是其均衡性。纵观美国历史，基本上始终维持着两大政党各占50%左右的格局，从未出现过70:30之类的悬殊比例。这是你的模型无法验证的，对吧？

And one of the features of polarization that is a little bit weird to me in The United States is how even it is. Like, we we've essentially always in the history of The USA had two political parties with roughly 50% representation each. It never goes to, like, seventy thirty or anything like that. That's not something you can test in your model. Right?

我的意思是，你基本上已经把民主党人和共和党人的初始数量作为预设条件植入模型了。

I mean, you basically baked in into your initial conditions how many Democrats there were and how many Republicans.

是的，不，这确实不是本模型要捕捉的内容。总体而言，这些模型只是试图捕捉某种单一机制。

Yeah. No. So this is not, something I would capture in in this model. In general, I'm thinking of these models as just trying to somehow capture one mechanism. Yeah.

没错。一旦机制变得复杂，难度就会指数级上升。比如这类互动结果如何反馈到社会并推动极化，在某种程度上已经超出了本模型的边界。

Sure. And as soon as you start having a lot of mechanisms, gets very, very difficult. So in certain sense of like how the outcomes of this interaction feedback into society and how it drives polarization, that's to a certain degree outside the the kind of bounds of of this model.

你刚才提到追踪的是统计结果而非LLM发布内容中的具体词汇。但那些你赋予它们的背景故事——比如某人住在波士顿、喜欢戏剧之类的——这些因素会影响它们在社交媒体游戏中的分类、极化或成功程度吗？

So you you said that you were tracking like the statistical results, not the individual words that the LLMs were putting into their, posts or whatever. But do you know that those backstories you gave them, like, you know, this person lives in Boston, you know, they enjoy theater or whatever, Did that matter? Did did that affect how they, sorted or polarized or succeeded in the social media game?

我的意思是，在某种程度上，是的。因为如果我完全不让他们知道自己是民主党还是共和党，他们就无法以那种身份行事，也就不会触发这种反馈效应。但如果他们是否去钓鱼这件事很重要，那可能就无关紧要了。是的。

I mean, to certain degree, yes. Because if I wouldn't have given them any sense of if they were Democrats or Republicans, they wouldn't be able to act as that and you wouldn't, you know, trigger this kind of feedback effect. But if it matters that they that they know if they go fishing or not, that probably doesn't matter. Yeah.

我不知道。我是说，我...我会很好奇想知道。

I don't know. I mean, I I would be I would be curious to know.

我更多是把这看作是一种噪音或微扰，让他们不仅仅基于政治人格行事，而是还有类似'好吧，他喜欢钓鱼，他在谈论钓鱼，我也喜欢钓鱼，我要关注他'这样的行为。

I think of that more as a kind of a little bit of noise or a little bit of perturbation to to make them not only act on the basis of their political personality, but that there's also like, okay. So this you know, he likes fishing. He's talking about fishing. I like fishing. I'm gonna follow him.

所以这增加了一点那种噪音，以及我们的生活不只是政治这个事实。政治只是我们全部的一小部分，我们的身份远比这丰富得多。

So it adds a little bit of that kind of noise and the fact that, you know, our lives are not just politics. It's, that's just a small part of everything that we are. We have much more rich identities than that.

你们基本上注入了新闻？是不是就像存在一个外部扰动源，说某个新闻事件发生了之类的？

And you injected news basically? Is that like, you know, there was an external source of perturbations that said like, you know, this news event happened or whatever?

没错。基本上如果你只是给智能体提供...至少让他们在没有话题的情况下交谈，就会变成最泛泛无趣的内容。这也不会创造出那种作为噪音存在的丰富性。所以我们基本上选定某一天，获取当天的所有新闻，然后随机选择部分新闻让他们讨论。

Yes, exactly. So basically if you're just giving the agents, if you're at least letting them talk without, you know, having them something to talk giving them something to talk about, it just becomes the most generic, uninteresting And and it also doesn't create this this kind of richness that, you know, also functions as a kind of noise, you know? And so we basically focus on a certain day and we got all the news from that day and then we present them with like a random selection of those news and then have them discuss it.

明白了。所以你们不是完全编造新闻，而是从真实新闻中获得灵感的。

Okay. So you didn't totally make up the news. You actually were inspired by real news.

确实。对。我们我们我们得到了某一天的真实消息。

Exactly. Yeah. We we we got real news from from a particular day.

那么我想，你提到了注意力幂律分布。虽然这些都是大语言模型，但有些获得的关注度远高于其他。是否存在某些模型确实更擅长社交媒体运营的情况？还是说这纯粹是统计和随机性的结果？

And so I guess, you mentioned the power law distribution of attention. So some of these they're all LLMs, but some of them get a lot more followers than others. Is there any sense in which some of them are just better at social media than others, or is it is it purely statistics and randomness?

我我认为这基本是随机的。我的意思是...好吧，我不会说其中某个恰好就成为了所谓的超级网红。但确实，表现得更加政治化和极端化确实有助于扩大影响力，不过整体而言我认为这基本是随机的。而且这与多项关于幂律分布如何通过反馈效应在系统中自然形成的研究结果相符。

So I I would say that it's pretty much stochastic. I mean Okay. I I wouldn't say that it's just randomly one of them has happens to be a, you know, a great influencer as such. But there are, I mean, being more political and being more extreme does help to become more influential, but it is, I would say pretty much stochastic. And I mean, that fits also, there's been various like kind of experimental studies on on this power law distributions and how they can emerge on in systems just through the feedback effects.

被选择对象之间甚至不需要存在差异。依然会出现这种幂律分布，这基本就是随机的。

And it doesn't need to be any difference between the things that are being selected. Still get these power laws and it's just kind of random.

那么我们是否从中获得了让世界变得更好的启示？这些发现能否帮助我们改进社交媒体？

So did we learn anything about how to make the world a better place through doing this? Does this, help us suggest any ways to make social media better?

我想我们确实没怎么提及尝试过的干预措施。我们搭建了这个平台后观察到这些负面结果，这为我们提供了尝试解决问题的基准线。我们查阅文献，研究人们建议的解决方案和乐观方向，基本上收集了从简单到复杂的各种方案。其中一个是Jigsaw（谷歌子公司）发布的'桥接属性'功能。

I mean, so I guess we didn't really mention the interventions that we we tried out, because we built this platform and then we saw these negative outcomes and that gave us the baseline where we could try, okay, can we fix this problem? That became the kind of next step. And so we looked at the literature and looked at what has been suggested, what are people kind of optimistic about in terms of trying to solve these problems. And basically we had the kind of wide variety of more or less sophisticated solutions that have been presented. One of them was kind of bridging attributes, which Jigsaw has released, which is a sub company of Google.

该功能会分析消息内容，社交媒体平台可用它来排序信息流，优先展示最具建设性的评论——而不是最令人不安的内容。它能展示最具建设性、最超党派的评论，因此被称为'桥接属性'。

Which is basically they analyze the content of messages and then you can sort your newsfeed. If you're a social media platform, can use it to sort your newsfeed and you get the most kind of constructive comments. Those are the ones that you show. So instead of showing the most upsetting comments, you can show the most constructive, the most partisan. So it's called the bridging attributes.

这是一个例子。另一个例子是这些较小的解决方案，比如隐藏个人简介，当用户相互关注时只显示代理的简短描述。例如，他们不知道对方是民主党还是共和党。另一个方案是按时间顺序排序，而不是显示被分享最多的帖子。基本上，我们尝试了许多被提出的解决方案。

So that's one example. And another example was like is this smaller solutions, just like hiding the biography, little description of the agents when they follow each other. They don't know if the other person is Democrat or Republican, for instance. Another is just sorting chronologically instead of showing the most shared posts. So basically we try out a bunch of those solutions that have been suggested.

但我也要说，我们正在尝试一些相当极端的版本，这些在平台上实施可能并不现实。例如，我们展示了一个算法，它优先显示最不受欢迎的帖子，如果在现实世界中实施，这将导致一个非常糟糕的平台。因为我们想看看最极端的解决方案。但不幸的是，这些解决方案都没有真正解决我们观察到的问题，有些甚至让情况变得更糟。例如，按时间顺序排列的时间线实际上导致了更多的社交媒体棱镜效应，极端用户反而获得了更多关注。

But I should also say that we are doing fairly extreme versions of it that wouldn't necessarily be realistic to implement on a platform. So for instance, we show the one algorithm where we show least liked post first, which is how would lead to a really awful platform if you implemented it in the real world. Because we want to see the most extreme solutions kind of. But unfortunately, none of these solutions really fix the problems that we're observing, and some of them actually make matters worse. So for instance, the chronological timeline actually leads to more of a social media prism where you get more extreme users get even more attention.

因此，我们从中得出的结论是，这种涌现现象似乎对各种扰动非常顽固。基本上，它有点像谢林隔离效应，是一种非常稳健的涌现现象。

So it seems what we take from this is that this kind of emergent phenomena seems to be very kind of rigorous to to perturbations. That is it basically yeah. A little bit like the shelling segregation effect is it's it's a very robust emergent phenomenon.

是的。而且你不想拥有一个告诉每个用户该关注谁的社交媒体网络，对吧？你需要给他们一些自主权。

Yeah. And you don't wanna have a social media network that tells every user who to follow. Right? You need to give them some agency there.

是的。我的意思是，这也是人们是否会使用这个平台的问题。但基本上，对我来说，这表明了我们在社交媒体平台上看到的基本结构——你有一个网络，你关注别人，并转发内容——这与这些有问题的结果有关。

Yeah. I mean, it's also the question of if people are going to use the platform or not. But basically, I mean, to me, what this suggests is that this basic structure that we see across social media platforms where you have a network, you follow people, and you repost things, that that tends to be linked to to these problematic outcomes.

嗯，我想这正是我要问的。你已经提到这很难研究，但这些社交媒体现象真的是全新的吗？比如，我们过去满足于在三大电视网络上获取常规新闻，而现在我们有更多样化的选择。但这产生了很大的影响吗？你能看到它有影响，但我想知道我们当前的政治混乱有多少可以追溯到这一点？

Well, I guess that was where I was gonna go. You you already sort of said this is hard to study, but is it something truly new, these social media things? Like, we used to be happy just getting the conventional news on one of the three network stations on TV, and now we have a lot more variety in what we can listen in on. But has this had a big effect? Has it has it really like, you can see that it has an effect, but I guess how much of our current political mess can we be tracing to this?

我知道这是个难以回答的问题。

I know it's a hard question to answer.

是的。我是说，这最终，你知道，是无法知晓的。而且在某种程度上，将社交媒体视为某种脱离社会、偶然降临社会的事物也很棘手。因为在我看来，社交媒体现有的结构形态，很大程度上体现了向福特主义的回归——那种从工业社会向后福特主义社会的转型，在这种社会里，资本的核心诉求是广告投放和用户信息挖掘，不再服务于所有人消费同质产品的大众市场，而是不仅要识别消费细分领域，甚至要创造消费细分领域。没错。

Yeah. I mean, it's it's ultimately, you know, impossible to know. And and to a certain degree, it's also a bit tricky treating social media as something that's, like, external to society and that happened to society. Because to me, that social media is structured the way it is, is very much an expression of coming back to Fordism, the kind of transition from an industrial society to post Fordist society where the focus of capital is to is advertising and figuring out information about you, that it was not catering to a mass market where everyone, you're selling the same product to everyone, but that you're really trying to not only identify consumer niches, but even create consumer niches. Yeah.

而企业如何盈利这一基本事实——即支撑社交媒体和互联网的商业模式——极大塑造了社交媒体的现状。当然，这也会形成反哺，但很难说如果没有这种背景，社交媒体会有什么不同。

And that basic fact of how the companies make money, the business model underlying social media and the internet, that has very much shaped what social media has become. Of course, it's also feeding back, you know, and but it's very difficult to say how social media would be different if if it wasn't that context.

嗯，我想说的正是这种反哺效应——简单来说，不仅是主流媒体之外多了社交媒体，更是社交媒体在影响主流媒体。它们同样渴望点击量。

Well, I guess it's the feeding back I was gonna, mention very briefly, like, it's not just that you have social media in addition to mainstream media, but the social media affects the mainstream media. They want those clicks too.

没错。这正是人们常说的那种情况：好吧，我只要停用社交媒体就不会受到这些负面影响了。但显然事实并非如此，对吧？正如我提到的，社交媒体正在重塑我们的政治生态，也在深刻改造主流媒体。

Yeah. Exactly. I mean, I I think this is something that people often mention as like, okay, but we can just, know, I can just stop using social media and I won't be affected by these negative consequences. But of course, that's not the case, right? Because social media is, as I mentioned, it's reshaping our politics and it's very much reshaping also mainstream media.

比如去年我课程里有个学生项目，他追踪了《纽约时报》多年来的标题，并测量其'标题党'程度。基本发现是：当社交媒体在2010年兴起后，相关指标突然跃升，《纽约时报》的标题写作方式也随之改变。哇。这不过是其中一种表现，但由这些平台制造的注意力经济，确实正在重塑我们的政治、媒体乃至整个文化生态。

So a student of mine in a student project in my course last year, for instance, he looked at the New York Times headlines over time and then measured how click baity they are. And basically what he saw was that when social media entered on the scene in 2010, and so you saw it kind of jump and you saw that the New York Times also changed how they wrote their headlines. Wow. And I mean, that's just the kind of that's one expression of it, but it, of course, it's reshaping, you know, the incentives of attention produced by these platforms are reshaping our politics, our media and our culture overall.

那'标题党'具体指什么？是指故意减少信息量，用'你绝对想不到接下来发生了什么'这类表述吗？

And what does clickbaity mean? Is it a is it a function of sort of giving less information and saying, like, you won't believe what happened next?

对。其实存在一系列文本特征会增强或减弱标题党效果。但他的研究方法主要是建立点击诱饵新闻与非点击诱饵新闻的数据库，然后训练分类器进行识别。

Yeah. It's it's there is, like, actually a bunch of kind of features of text that make them more or less clickbaity. But, basically, the the way he did it was to just look at the databases of clickbait news articles and then non clickbait news articles and then train a classifier on it.

那么我想最后要讨论的是两极分化现象。你看，这些大语言模型以谢林模型的方式自我分类。我们能对信息质量说些什么吗？比如真实性与错误信息？社交媒体是否不仅让我们只与同类人交流，还通过分享错误和虚假信息让我们犯错？

So I guess a a last thing to talk about is there's polarization. So, you know, you had these LLMs. They sorted themselves in a shelling like way, etcetera. Can we say anything about the quality of the information, like the truthfulness versus misinformation? Did the or social media helping us not just only talk to people like ourselves, but to get it wrong by sharing mis and disinformation?

我想说的是，这并不是我在这个特定模型中研究的内容，部分原因是OpenAI不允许大语言模型产生错误信息，所以无法用它们来研究这个问题。我确实用真实社交媒体数据研究过。总的来说，社交媒体通过消除传统主流媒体的把关人角色，并通过制造强烈的吸引注意力动机，实际上创造了一种环境——不把言论锚定在事实上反而成为获取关注的有效策略。因为这让你可以发表惊人言论、煽动情绪，而不受现实约束。当然，这与政治也相互关联。今年早些时候我和合著者Julianne Schwery发表的论文中，我们研究了多国政客在五六年间的推特发文，分析了他们分享错误信息链接的案例。

So I I would say, I mean, this is not something I'm looking at in in this specific model, right, In part because the LLMs are under OpenAI doesn't let them produce misinformation, so you But can't really use them to study I have looked at this in the context of using actual social media data. And I mean, what I would say broadly around it is that social media is by removing the kind of gatekeepers that we used to have from mainstream conventional media, they're also and by creating kind of really strong incentives for gaining attention and shaping the kind of conditions for that, they're really producing conditions where not anchoring what you're saying to truth becomes a kind of beneficial strategy for gaining attention, right? Because you're both, it allows you to be kind of outrageous, it allows you to trigger people, and you're not really constrained by reality in the same way. And of course, that also becomes interconnected with politics. So I had a paper coming out with my co author, Julianne Schwery, earlier this year, where we look at politicians across countries, we look at their Twitter posts over a five, six year period, and we look at all the examples of when they're shared links, misinformation through that.

这样我们就能将每位政客与其分享这类信息的可能性关联起来，建立比较模型——本质上是个统计模型，用于识别政客传播错误信息的条件。这关联到社交媒体与错误信息传播之间关系的更广泛问题。近年来对此有大量争论，焦点是社交媒体是否只是整体降低了信息质量。我们的观点是：不仅如此，社交媒体还与政治运动深度交织。

And so we can link each politician to their likelihood of sharing this information, and then we can basically use that for a kind of comparative model, so basically a statistical model to identify the conditions when politicians are spreading misinformation. And so this links to this broader question of the link between social media and the spread of misinformation, which is there's been a big debate around this, especially in the last few years about whether it's just social media reducing the quality of information overall. Right. And what we argue is basically, it's not just that, but that social media becomes intertwined with politics. It becomes inter intertwined with different political movements.

结果是某些政治运动在社交媒体利益和动机的塑造下，将错误信息作为政治策略来获取竞争优势。我们的研究发现，推动错误信息增长的主要是激进右翼民粹主义政党。因此这不只是社交媒体现象，而是社交媒体与政治体系相互交织的结果。

And the result is that certain political movements are emerging shaped by the interests and incentives of social media in such a way that they use misinformation as a political strategy to gain advantages in in political competition. And what we find in that study is basically that it's specifically the radical right populist parties that are driving this rise of of misinformation. So it's not just a a social media phenomenon itself, but it's social media intertwined with politics and political systems.

那么最后请你做个宏观总结——我是否该形成这种印象：社交媒体就是有害的？它的出现是个错误？净效应是负面的？还是说我们还能保持一丝乐观？

So I'll I'll let you give, like, a last big picture kind of thought here. Like, am I getting the impression that social media are just bad, that it was a mistake, that their net effect is negative, or or can we have some shred of optimism to hold on to?

我认为某种程度上它也确实产生了积极影响，对某些群体是有益的。我成长过程中很热爱互联网，我在瑞典偏远乡村的小岛上长大，互联网为我提供了社交世界，让我接触到各种思想。某种程度上我希望能回到90年代那种纯真的网络时代。

I mean, I I think that there are also kind of, to some degree, positive outcomes from it. And for certain communities can be beneficial. And I mean, to a certain degree, growing up, I loved the internet, it was great. I grew up on this, in the countryside on this little island in Sweden in the middle of nowhere, the internet kind of provided a social world for me and allowed me to connect to ideas and everything. To a certain degree what I would hope for is also kind of going back to that innocent era of the 90s, of the internet of the 90s.

当时我用ICQ，有个按钮可以随机和世界各地的人聊天，我超爱这个功能。我整天和亚利桑那的陌生人聊天。那个时代确实更纯真，现在要恢复这种模式可能会酿成灾难。但我认为我们可以创建真正有益的平台和空间，只是需要进行比算法或界面修修补补更根本性的重构。

I was using ICQ and you had this button where you could click and talk to a random person anywhere in the world, and I just loved that. I spent my days talking to some random person in Arizona and to a certain degree, of course that was a more innocent time, and maybe if we would try to bring that back, it would lead to something horrible these days. But I do think that we could create structures, like, we could create platforms and spaces that would actually be, you know, beneficial for us, and that would actually be positive. It's just we might need to rethink it in more fundamental ways than just this cosmetic changes to algorithms or designs.

好的。这是给所有年轻人的一项功课。思考一下，我们可以做出哪些根本性的改变，因为这些问题是不会消失的。要知道，即使社交媒体总体上弊大于利，它们也有好的一面，我们必须学会与两者共存。那么，佩特·索恩伯格，非常感谢你参加《思维景观》播客。

Alright. That is something there's homework out there for all the young people. Think about, you know, fundamental changes we can make because they're not going away. You know, even if social media are are a net bad, they also are good, and we're gonna have to live with both of them. So, Petter Thornburg, thanks very much for being on the Mindscape Podcast.

非常感谢。这次对话真的太棒了。

Thank you so much. This this was really great.