The Beginning of Infinity, Part 2
《无穷的开端》，第二部分
这是我与 Brett Hall 关于《无穷的开端》访谈的第二部分。另见第一部分。
拥有好的知识理论，你就能判断其他什么是真的
David Deutsch 的理论以好的解释为核心
Naval：David Deutsch 有一个宏大的世界观，他认为一切重要的事物都可以被单个人类所理解。所谓重要，他指的是驱动大部分现实的底层基础理论。
Deutsch 聚焦于四个理论。我可以说也许还有更多，尤其是当你开始涉及 Adam Smith 和《国富论》以及其他一些更具社会学色彩的理论时。但他显然是一位物理学家，更关注现实和真理寻求，而非人类系统。
他选出的四个理论是：认识论（epistemology）；自然选择进化论；量子理论，他将相对论和其他物理学都归入其中；以及计算理论，其中包括他的量子计算理论。
这四个理论令人着迷。值得探讨一下它们各自有趣的地方。这里的突破点在哪里，哪些可能并不显而易见？
让我们从认识论开始。
我之所以喜欢《无穷的开端》，是因为 Deutsch 对认识论中什么是正确的、我们所知最好的答案，进行了非常严谨的梳理。一旦你拥有了一个好的知识理论，你就可以判断其他什么是真的。
如果你的知识理论建立在一个糟糕的基础之上，你就会在自以为真的地方得出许多实际上是假的结论。
而他的认识论以好的解释为核心。它继承了 Popper 将科学和真理寻求视为纠错机制的观点，并在此基础上加以拓展。我很想听听你对 Deutsch 所阐述的知识理论，即认识论的概述。
没有任何真理可以被证明为真
“知识”只是我们当时的最佳猜测
Brett：最初对”知识”是什么的猜测，形成了所谓的”得到辩护的真信念”（justified true belief）知识观，这也是今天最流行的观念。任何自称为 Bayesian 的人都是”得到辩护的真信念”者。
这种误解在于认为知识就是试图证明你的信念为真。如果你做到了这一点，你就可以说：“我知道那件事。“如果我能证明我的引力理论为真，那我就应该相信那个引力理论，也只有到那时我才能说它是已知的。
问题在于，没有任何方法可以证明任何一条知识是真的。Deutsch 在他的著作中所倡导的改进，是 Popper 给我们的这样一个洞见：我们拥有的只是关于现实的猜测。它们是猜想。
人们会想，“哦，这听起来有点不靠谱。不过是猜测而已。“但它不是随意的猜测。它是经受了检验、经受住了试图证明其为假的尝试的猜测。并非每个决定做出猜测的人都站在同等的立足点上。
当人们无法通过这种反驳的方法证明某个东西为假时，我们就把它作为一条知识来接受。这使我们能够接受这样一个事实：我们将能够在未来取得进步，因为我们所有的知识都是猜想性的。所有知识都是我们当时的最佳猜测。
知识内部有一种弹性，使我们能够说：“会有错误出现。我们将纠正它们，并由此能够走向无限的未来而不断进步。”
这与先前的知识观念截然不同，那种观念说：“一旦你证明了某件事为真，那它就是真的。“如果它是真的，那意味着它没有任何虚假之处，因此它不可能被反驳。那是一种非常宗教化的观念。
这种观念的现代版本就是 Bayesianism，它说：“你有一个理论，你收集更多证据，你随着时间推移对自己的理论越来越有信心。”
它比这还要糟糕一点，因为它接着说：“这种 Bayesian 推理使你能够生成新理论。“但它做不到。它最多能做到的，是向你表明你对这个理论比那个理论更有信心。
Popper 式的观点则说：“如果你能证明某个理论存在缺陷，你就可以抛弃那个理论。“
不可能有最终的引力理论
我们拥有的只是对现实越来越好的近似
Brett：在几乎所有情况下，你手头只有一个理论可供选择。
在引力的情况下，目前确实只有一个理论可供选择。那就是广义相对论。以前我们确实有两个理论。我们有 Newton 引力理论和广义相对论——但我们做了一个判决性实验。
判决性实验这个概念是科学皇冠上的明珠。你有两个相互竞争的理论。如果实验结果是其中一个方向，一个理论就被排除了，而另一个理论没有。在这种情况下，你就保留那个理论，直到出现问题为止。
这种知识观使我们能够拥有一个开放式的对进步的追求。
这与其他任何关于知识的观念完全不同。绝大多数物理学家仍然是 Bayesian。他们仍然是 Bayesian 的原因在于，大学里通常教的就是这个，而这也被视为理解世界的一种在理智上严谨的方式。
但它只不过是我称之为 scientism 的一种变体。因为它背后有一个公式，即 Bayes 定理，这是一个完全可接受的统计公式。人们一直在以完全合理的方式使用它。只是它不是一个认识论。它不能保证、甚至不能让你确信你的理论实际上是真的。
我最喜欢的一个例子：在1919年之前，所有关于引力的实验都表明它与 Newton 的引力理论一致。Bayesian 在这种情况下会怎么说？他们会说你对 Newton 的理论越来越有信心。
这怎么合理？在你最确信 Newton 理论的那一天，恰恰是它被证明为假的前一天，你怎么解释这个矛盾？
而 Popper 式的人不会有这个问题。Popper 式的人会说：“Newton 的理论在任何时刻都不是真正正确的。它包含了一些真理，但那个真理不是我们可以测量的东西。”
我说它包含了一些真理，是因为它当然比某个随机的人对引力本质的猜测更直接地与现实相连。引力确实大致遵循平方反比定律，但不完全如此。我们需要广义相对论来纠正 Newton 引力理论中的错误。
尽管广义相对论是我们目前的最佳猜测，它最终不可能成为引力的最终理论。不可能有最终的引力理论。我们拥有的只是对现实越来越好的近似。
Naval：我认为我们之所以如此容易陷入 Bayesianism，原因可能与我们之所以如此容易陷入悲观主义有关。我们在进化上就被设定为 Bayesian 的。
地球上其他每一种能够形成好的解释的动物都是 Bayesian。它们只是观察重复的事件，然后说：“太阳昨天升起了。太阳明天也会升起。“或者，“我碰到的那个东西是热的。它将来大概率也是热的。“这就是我们大部分生物系统和大部分进化遗产的运作方式。
只是现在我们有了新皮层，能够用不可见的来解释可见的，形成好的解释。这赋予了我们更高层次的推理能力，但这个更高层次的推理能力对我们来说不是本能的。它需要努力。需要深入思考。
我们默认走向 Bayesianism，因为这就是我们周围自然世界在很多层面——至少在纯生物学层面上——似乎运作的方式。
基因演化是模因演化的前奏
观点经历了与基因相同的演化过程
Brett： Popper 有一本书叫《客观知识》，副标题是”一种演化论的进路”。这并非偶然。我们所理解的认识论理论与演化论之间存在一种对称性。
在我们理解所谓的 Darwin 演化论之前，人们仅有的观念是：这些实体必定是被创造出来的。你周围看到的一切植物和动物，都一定是由某个创造者所造。没有其他的解释机制。
有些人提出了随时间逐渐变化的想法。Lamarck 就是其中之一。他的看法是，长颈鹿之所以有长脖子，是因为它们的祖先脖子稍短一些，所以它们努力伸长脖子去够那些够不着的树叶。
但同样，这里没有任何机制，除了一个个体去健身房锻炼二头肌，二头肌会随着时间稍微变大。虽然你可以在健身房锻炼并增大你的二头肌，但这并不意味着你的孩子会继承这些特征。
Darwin 提出的想法与 Popper 在知识论中的想法类似。那是一种纠错机制。其核心观念是，一个生物体在特定环境中进行试炼，如果它——如我们所说——不适应那个环境，那么它就会消亡。而如果它适应那个环境，它就能存活。
所以你看到的是生物体与环境之间与现实的碰撞。正是环境将来自现实的反馈传递给你，淘汰掉那些不够适应的生物体。
新达尔文主义的观点告诉我们，选择的单位是什么。它不是群体或兽群，甚至不是个体。它是基因。这就是 Richard Dawkins 提出的”自私的基因”的概念——如果其中任何一个基因碰巧不适应特定环境，就可能导致该生物体的死亡。物种可能存活下来，但随着环境的变化，其整个 DNA 会随着时间推移而发生极其细微的改变。
现在我们已经将此提升到了新的层次。我们人类是那个演化过程的下一步——我们能够创造解释性知识，这些知识做着同样的事情。Deutsch 喜欢说：“基因演化不过是一个序曲。接下来要到来的是模因演化。”
从此以后，宇宙的历史将是观念的历史——这些观念将经历与先前基因所经历的相同的演化过程。
人类是特殊的
人类不是一种占领了整个星球的细菌
Naval： 这四个理论中有三个都有一个有趣的模式。在认识论中的好解释方面，我们说的是猜想及其反驳和纠错是我们改进知识的方式。在基因演化中，基因突变、变异和自然选择淘汰了那些不成功的。在此之上还有模因演化——我们产生观念，然后批评淘汰掉那些行不通的观念。
与此相关的是，在发明创造中有试错。在资本主义中，创业公司不断被创立，那些想法糟糕的就失败了。我们看到这个模式一次又一次地出现。
但有趣的是，这里还有一个元模式：人类是特殊的。在认识论中，人类是仅有的非 Bayesian 推理者。在演化中，人类是我们所知唯一的模因生物。在计算理论中，人类是我们所知仅有的通用解释者——当然，除了我们自己发明的计算机之外。
科学将我们从”人类处于宇宙中心”这个观念，带到了”实际上人类没什么特别的。你不过是一个几乎无限多的 Kepler 行星中可能承载生命的一颗小小行星上的一员。“但我们所谈论的这四个理论中的三个，正在将我们引向另一个方向——人类是特殊的。人类拥有获取最大知识的能力。
对我来说一个有趣的领悟是：即使你是上帝，即使你拥有无限的知识和力量，即使你掌控了整个宇宙，你仍然不会知道自己不是在一个模拟中。你仍然永远无法证明自己不在一个模拟中。而且即使作为上帝，也不存在任何你能掌握的概念是一个人类所无法掌握的。
当然，除非物理定律不同。如果物理定律不同，那就一切皆有可能了。但在现有的物理定律框架内，人类拥有获取最大知识和最大意识的能力。这指向了一个人类是特殊的世界——而不是另一种失控并占领了这个星球的细菌。
许多这些基础理论都导向一个观点：人类是特殊的，知识是无限的，只要我们不摧毁纠错的手段，并且始终在创造新知识，那么就有充分的理由保持乐观。
我们与其他物种有质的不同
人类与黑猩猩之间不存在连续体
Brett： 你指出的是一个少数派观点。我认为文化仍然停留在你所说的第二部分。
最初我们认为自己是宇宙的中心。这是关于人类在宇宙中位置的宗教观念。地球被天球包围，万物围绕它旋转。所以我们被视为整个宇宙的继承者，上帝将这一切赐予了我们。
然后科学告诉我们，事实上，我们在宇宙中并非处于一个特别的位置。这就是宇宙学原理——即宇宙在每个地方大致相同，而我们只是其中一个毫不特别的地方。
我们不仅在宇宙学意义上不特别，在生物学上也没什么特别的。我们只是从细菌到蟑螂、再到狗和黑猩猩这个连续体上的一点。
有一位天体物理学家，Neil deGrasse Tyson——在几乎所有其他话题上我都非常喜欢他——谈到黑猩猩比我们想象的聪明得多，而我们可能也好不了多少。这几乎是所有人的想法。
而我们中许多人现在正在推动的第三种观点是：黑猩猩与我们之间的差异并非微小的定量差异。从细菌到蟑螂到狗再到黑猩猩之间确实存在一个连续体，但我们已经脱离了那条轴线。
我们在质上是不同的。我们只需睁开眼睛。你望向窗外那座美丽的城市——它就在那里——它无法用生物复杂性的逐渐增长来解释。
更多的计算能力并不能产生 AGI
即使最强大的计算机也无法回答”为什么？”
Naval： 人工通用智能那一派也完全搞错了：“只要增加更多计算能力，就能得到智能。“——然而我们并不知道是什么在底层使我们具有创造力，使我们能够提出好的解释。
人们大量谈论 GPT-3——OpenAI 发布的文本匹配引擎——它是一款令人印象深刻的软件。他们说，“嘿，我可以用 GPT-3 生成很棒的推文。“那是因为，首先，作为人类，你正在从它生成的所有垃圾中挑选出好的推文。其次，它在利用某种抄袭和同义词匹配等手段来生成听起来似乎合理的内容。
证明它生成的东西实际上没有任何意义的最简单方法，就是追问它一个问题。拿一段 GPT-3 的输出，然后追问它：“为什么是这样？“或者基于它的输出做一个预测，然后看它如何彻底崩溃——因为其中不存在任何底层的解释。
这是鹦鹉学舌。这是精妙的 Bayesian 推理。它在根据人类在网络上已生成的内容进行外推，但它没有一个底层的现实模型——无法用不可见之物来解释可见之物。我认为这一点至关重要。
这正是人类独一无二之处，我们所遇到过的任何其他生物、任何其他计算机、任何其他智能——无论生物的还是人工的——都做不到这一点。
我们不仅独一无二地做到了这一点，而且如果我们遇到一个同样拥有生成优质解释能力的外星物种，他们所能生成的任何解释，我们都能够理解。
我们拥有最大化的理解能力。在这个物理现实中，不存在任何可能的概念是一个人类在给予足够时间、资源和教育的情况下无法理解的。
我们的头脑令人惊叹地不会被震撼
人类在给予足够时间和努力后能够理解每一个思想
Brett： 有科学思维的人会说：“也许我们无法理解下一组物理定律。也许我们无法理解外星人。“这不过是诉诸超自然。它在逻辑上等同于”上帝在那里，而你不可能理解上帝是什么。上帝是这个超越我们的、无限的、全知的存在。”
你如果愿意可以相信这一点。你可以相信模拟假说。你可以相信这些东西中的任何一个。它们都是关于我们无法触及的现实的形而上学主张。无论你是否想引入那些拥有我们无法理解的想法的外星人，这一切都建立在同样的根基上。
Naval： 至少在上帝的形而上学中你可以说，“好吧，那是在另一个宇宙。那在我们的物理定律之外。“但外星人想必遵循同样的物理定律，所以我甚至看不出这种说法的依据是什么。
任何聪明到能够离开母星的物种都知道，限制性因素是想法。所以他们最想从遇到的任何其他物种那里得到的就是新的想法。他们应该做的交易就是思想的交易。
科幻小说中有一种 Malthusian 式的哲学——现在被称为黑暗森林假说——认为每一个人际物种都像细菌一样，我们会耗尽空间。不，宇宙在尺度上是无限的；它在膨胀。多重宇宙在尺度上甚至更加无限。
我们处于无穷的开端。我们不会耗尽资源。每个人都在创造想法。聪明的外星文明交换想法，成功的人类文明交换想法。因为这些想法把以前无用的东西变成了资源。
每一个外星文明都能与任何其他文明交换想法，因为他们都是通用解释者。他们都拥有最大化的理解能力。
事实上，这里最令人震撼的事情是，你的头脑不会被震撼。不存在任何你的头脑在投入时间和精力后无法吸收的想法。
所以如果我们遇到一个外星物种，我们大概应该欢欣鼓舞。他们不想要我们星球上的任何东西，除了我们的想法。而交换想法的最佳方式就是拥有一个动态的、丰裕的、繁荣的文明。
因为我从小看的是老一套的科幻作品，我过去对外星遭遇持悲观态度。“哦，如果我们遇到外星人，他们只会毁灭我们。“就像在《银河系漫游指南》中，Vogons 不假思索地拆毁地球，为超空间快速通道腾地方。
现实情况是，任何发现我们的物种都会立即把他们拥有的所有知识给我们。而且他们会渴望我们拥有的新知识，因为他们会意识到这能让他们点亮暗物质、暗能量、宇宙中未被利用的资源，让他们也能繁荣发展。
外星人在哪里？
宇宙应该充满了外星人
Naval： 既然我们谈到了外星人，让我们简要谈谈 Fermi 悖论。对于不了解的听众，Enrico Fermi 是一位著名的物理学家，曾参与曼哈顿计划。他说：“外星人在哪里？”
宇宙如此之大，可能有那么多能够以某种方式支持生命的行星。我们难道不应该已经看到他们了吗？
Brett： 几乎每颗恒星周围都有一组类似我们太阳系的行星。像银河系这样的典型星系中存在的恒星数量大约是两千亿颗，尽管估计数字最高可达约四千亿颗。我们能看到的星系数量大约在两千亿到三千亿之间。
可观测宇宙只是整个宇宙的一小部分，这意味着行星的数量绝对是天文数字。
考虑到这些数字，必然不仅存在适合生命的行星，而且宇宙应该充满了文明——有的远比我们先进，有的不如我们先进，还有一些与我们先进程度相当。
那么他们在哪里呢？
地球上只诞生过一次智能物种
创造性智能并非必然
Brett： 这是一种论证。面对这种不确定性我们必须保持谦逊，因为没有人知道答案。但我想提出一个鲜少被讨论的论证。
这个论证认为我们是孤独的。
这个论证与天文学无关；它与生物学息息相关。论证是这样的：看看地球，看看不仅现存物种的数量——数以百万计——还有曾经在地球上存在过的物种数量，数以亿计。
生命大约在三十五亿年前出现，其中大约二十五亿年里只有细菌。所以生命显然没有多大动力快速进化超越细菌阶段；它只是尽可能保持简单。
许多人有一种被 Darwin 真正消除的误解——即演化有一个方向。你会看到高中教科书上出现的演化图：一只四脚着地蹒跚而行的猴子；然后它弯腰驼背；最后它站起来了，手里拎着公文包，仿佛这就是演化预定的目标。这只是回顾往事时，看起来像是演化的目标，事后诸葛而已。
有一位美国学者 Charley Lineweaver，将此称为”人猿星球”假说——意思是，如果你把人类从一个星球上移走，类人猿自然会演化来填补智能的生态位。
他说你可以想象另一种情形：你是一头能够思考自身的大象。它们反思自己鼻子的长度，回顾生物演化的历程，看到鼻子变得越来越长。于是它们得出的结论是：“啊，演化一直致力于制造越来越长的鼻子。这就是演化的全部意义。”
当然，我们可以看出这是荒谬的。只不过是恰好有一种叫做大象的生物演化出来了，它有这么长的鼻子，但鼻子的长度并不是演化的趋同特征。
演化的趋同特征是指在生物实体中反复出现、独立产生的特征。翅膀是我最喜欢的例子。鱼类有某种翅膀，存在飞鱼。蝴蝶有翅膀，所以昆虫中也有翅膀。它们在哺乳动物中也出现了，比如狐蝠和某些种类的袋貂。当然，鸟类和恐龙也有翅膀。
在所有这些物种中，翅膀独立地反复出现。眼睛也是如此，听觉器官也是如此。
人类存在的概率微乎其微
现在让我们想想做数学或建造射电望远镜的能力——换句话说，成为一个有智慧的、有创造力的物种的能力。这种能力在地球的地质史上出现过几次？只在一个物种中，而且仅仅这一个物种。
我们能据此得出结论说，智慧物种的出现因此是必然的吗？假如你重新做一次实验，在整个宇宙中所有适宜生命存在的行星上撒下一些细菌，你能保证得到一个像我们这样的实体吗？
人类存在的概率微乎其微
从演化的角度来看，这几乎是不可能的
Brett： 还有一种思考方式，对于那些认为外星人就在那里、将来某天会造访我们的人来说，这在数学上是令人恐惧的。
我们之前谈到过已知宇宙中可能适宜生命存在的行星有数万亿颗。想象一下，从我们这些智慧人类到我们所能想象的最简单的细菌之间，只有 100 个独立的演化步骤。当然事实并非如此——可能需要一百万次或更多不同的突变，才能让生物体存活下来，从而使我们今天得以存在——但我们姑且只算 100 步。
假设这其中的每一个独立步骤发生的概率仅为十分之一。实际上，这可能更接近百万分之一，但我们大方一点，就算十分之一。
现在我们得到一条概率链：十分之一乘十分之一乘十分之一……乘 100 次。这就是十分之一的 100 次方——也就是一除以 1 后面跟 100 个零。这个数字比我之前谈论的行星的天文数字还要大得多。
换言之，按照这个论证，我们出现的概率是微乎其微的。这件事居然发生过一次，已经够让我们震惊了。
如果你无法编程实现它，你就还没有理解它
演化算法并不能产生生物
Brett： 这些都是不确定的假说，但我们也必须记住，关于自然选择驱动的演化，我们不知道的还有很多。
David Deutsch 有这样一句妙语：
如果你无法编程实现它，你就还没有理解它。
在 AGI 的情况下，这意味着我们无法编程实现它，因为我们对通用智能这个概念还不够理解。
自然选择驱动的演化也是如此。有所谓”演化算法”这样的东西，但这并不是在编程实现自然选择驱动的演化。这并不是能够在计算机内部创造出人工实体、在真实的环境压力下使它们朝着日益复杂的方向演化的那种东西。
外星人也许只是太远了
我们未必在宇宙中孤独，只是相距太远
Brett： 我们仍然面临一个问题：在大约 25 亿年的时间里——即地球上生命历史的绝大部分时间——DNA 到底在做什么？为什么在那段时间里它根本没有演化？这是怎么回事？
有一本书叫《Rare Earth》，作者是 Peter Ward 和 Donald Brownlee，这两位讨论了地球演化史上发生的种种离奇事件。我刚才只举了一个例子，即我们这些通用解释者看起来是偶然出现的，而且似乎只出现过一次；但你可以回过头去发现，从单细胞细菌演化到多细胞生物也是一件怪异而非同寻常的事，而且在实验室环境中一直无法重现。
然后从多细胞生物到类似植物的生物，再到类似动物的生物——这其中的每一个阶段，似乎都出于我们所不了解的原因才得以发生。
Naval： 可能有多重因素在共同作用。你的论证可以是统计性的而非绝对的。我们未必在宇宙中孤独，但成为通用解释者可能如此稀有，以至于当你把这个概率乘以星际之间的遥远距离时，我们可能只是彼此离得太远了。
我认为 Fermi 还有一个不合理的假设，即星际外星人能找到办法超越光速，而我们对于这如何可能连一个假说都没有。我们甚至没有任何哪怕勉强接近超越光速方向的东西。
所以，如果你受限于光速，而且跃迁到通用解释者这件事极为罕见，那么我们可能只是彼此相距太远。而且这可能需要很长很长的时间。
宇宙非常浩瀚，但也几乎是完全空旷的——至少就行星和恒星而言。鉴于此，说人类和类人的解释者相当稀有、在宇宙中的形成仍处于早期阶段、而且被极其遥远的距离所分隔以至于我们尚未相遇，这仍然是相当合理的。
如果我们确实相遇了，我想我们会知道的。
比如，当外星飞船抵达这里时，它们的无线电波早就先到了。一个文明在发明无线电并向外广播无线电波，与发明星际旅行并向宇宙各地发射火箭和文明载体之间，有一段相当长的时期。
外星人造访是为了知识，而非资源
他们唯一缺少的，是他们还没有掌握的知识
Brett： 我记得 Stephen Hawking 说过，向外太空广播无线电波是个错误，因为外星人会在那里，而且他们会像征服者一样。他们会想占领我们的星球，掠夺我们的资源以及其他各种东西。
对于”邪恶外星人来抓我们”这种想法，有几种回应方式。
第一，要在无限的未来中不断进步、拥有使你能够穿越银河系的技术，唯一的方式就是拥有 Popper 所说的那种知识观——即你能够自由地探索观念的空间，能够证伪假设，并且没有中央权威和暴力被施加于人——这些东西会压制创造力。
要拥有一个最具创造力的社会，你必须拥有自由，必须拥有自主权，因此你将拥有一个非暴力的社会。你将拥有一个将创造力本身视为目的的社会。
当我们遇到外星人时，我们不应该预期他们会是那种要来抢占我们资源的缺德家伙，恰恰相反：他们会看着我们，觉得我们是多么原始的野蛮人。他们会认为我们在道德上矮小可怜，他们会想要教导我们。他们不会想把我们关进监狱或做任何类似的事情。
知识是一个统一的整体。如果他们的物理学远胜于我们——使他们能够接近光速，或利用某种广义相对论的重力效应制造虫洞，从而能以超过光速的速度穿越太空——那么他们的所有科学家都会远远领先。他们的所有知识都会更加先进：他们的数学、他们的道德、他们的政治制度。
所以我们不必担心外星人。
顺便说一下，我们也不必担心他们窃取我们的资源。他们不会说：“啊，这里有一颗充满煤炭和水的星球。我们把它拿走吧。”
不，他们将拥有足够的技术知识，能够收集星际空间中的氢气，把它变成核聚变反应堆，并利用 3D 打印创造出他们想要的任何技术。
事实上，这也可能是 Fermi 悖论的另一个答案。他们不需要离开自己所在的区域，因为他们已经拥有了完美维持自身生存的技术。
戴森球与虚拟现实：外星文明不需要扩张
Naval：他们拥有戴森球，可以收集所需的一切能量。他们可以收集一切物质，创造他们想要的任何东西，在虚拟现实空间中拥有他们想要的任何现实。
他们唯一缺乏的，是自己尚未拥有的新知识。
思想是新石油
所有新财富都由思想创造
Naval：人类有征服的历史，因为我们争夺的是完全相同的资源，但即使在人类历史上，最早的探索者也是商人。他们远行是为了寻找香料、黄金、丝绸、可以驯化的新植物、新动物。
他们远行未必是为了征服土地。后来确实征服了，那是因为当你被困在地球上时所面临的有限资源困境。但当你拥有了离开地球的技术，有限资源的问题就不复存在了。
如果你需要某种资源，那就去找一颗中子星，或者去找一个恒星系统。你不会去争一颗小小的行星。那里有无数的 Kepler 行星，对它们来说要近得多。
人们想要的是思想和贸易。
如果你审视现代社会——虽然我知道这不是普遍信念——我们随着文明程度的提高正变得越来越好战越少。原因是，你不再为了自然资源去征服俄罗斯。当今世界上最富有的地方，是那些拥有最好思想的地方。硅谷曾一度作为财富创造引擎雄踞榜首，因为它拥有最好的思想。
新的石油是思想。一切都是数字化的。所有新财富都在思想空间中被创造出来。
事实上，如果你今天作为一个有雄心的年轻人起步，你不会去学房地产；你不会去学煤炭和石油开采；你不会进入实体资源的开采领域来创造财富。你会进入思想空间。你会进入编程、书籍、电影、博客和播客，以及制造机器人——这些东西本质上大多是知识产权。
即使作为人类文明，我们也正在远离征服实体资源，越来越多地转向思想的交易。
对人类物种来说最坏的情况是，我们太多的大国和民族国家认为自己已经达到了思想的极限，现在是时候保存资源了。他们最终会破坏掉改进、纠错和创造的手段，陷入停滞。
于是，思想生成只能来自一小批更小的城邦，而这些城邦又不得不抵御那些庞大、更具掠夺性、僵化的国家。
人们谈论中国多么了不起——“看看他们的火箭”，或者”看看他们的 GDP”，或者”看看他们建造的城市。“等他们发明出什么新东西再告诉我。等他们想出什么我们不曾拥有的惊人想法，造出什么我们不曾拥有的技术再告诉我。因为到目前为止，一切都只是模仿。他们利用从我们这里获取的技术，正在追赶。他们只是在用规模优势，因为他们人口更多。
等他们的人均 GDP 超过我们再告诉我。等他们研制出我们不知道如何制造的药物、疫苗、航天器、发电机或聚变反应堆再告诉我。等那个威权社会以自上而下的方式弄清楚如何建造全新的事物，等它更有创造力、他们的艺术更出色、他们的科学更出色、他们的技术更出色的时候再告诉我。
等那一切在一个民主、自由、资本主义的社会之上发生时再告诉我，因为我从未见过这样的先例，从未。
一个爱因斯坦抵得上一整军团博士无人机
创造力是从 0 到 1 的跨越，堆人海解决不了问题
Brett：中国每年培养的理工科学士和工程学学士比世界上任何其他地方都多。中国的大学正在输出比我们更多的理科毕业生，但他们并没有输出更多的创新者。
那些从中国大学拿到科学学位的学生并不是出去做创新的事。恰恰相反，因为他们接受的是一种特定的训练方式。他们被训练去背诵这本教科书，应对这场考试。他们无法跳出框框思考。
他们被灌输的是：“这就是正确的。这就是关于科学毋庸置疑的正确思维方式。“这可能有利于模仿——正如我们所见——但这不会是让你推进技术前沿的东西，更不用说基础物理学或其他任何领域了。
我不在乎他们的理科毕业生有多少统计数据。那毫无区别。给我 10 个有创新精神、有创造力的年轻物理学毕业生，胜过 5 万个能以 100% 正确率通过考试的物理学毕业生——任何时候都如此。
Naval：所以说，一个 Einstein 抵得上一整军团拿着物理学博士学位的无人机。这无关紧要。创造力是从 0 到 1 的跨越。无论往问题上砸多少人都不能让你到达那里。这就是模因演化的本质；这就是创造力的本质。
不要依赖可信度印章
从机构到个人的一代人转型将是混乱的
Naval：我们今天的社会中有大量依赖可信度印章的机构。
它们曾经是你在社会中获取可信度的方式。如果你是为《纽约时报》或《华盛顿邮报》撰稿的记者，你就拥有了《纽约时报》和《华盛顿邮报》的报头光环。如果你是哈佛的教授，你有可信度，因为你是哈佛的教授。
当然，这些系统被攻破了。许多根本没有资格告诉世界该怎么做的社会科学家，现在混迹其中，用荒谬的政治模型冒充经济学家或自然科学家。有些人在这些昔日伟大报纸的报头之下撰写社论文章，消耗着这些报纸长期积累起来的可信度资本。
互联网正在缓慢而持续地揭露他们，而我们正处于一个大众仍然信任机构的转型阶段。
我们被困在这个 Schelling 点——这个机构的协调点上：我怎么知道该不该雇你？你有没有哈佛的文凭？我知道它不如从前了。我知道哈佛的人文学科文凭现在八成是废话，但我没有其他可信度指标来筛选你，而我又需要一种高效的方式来做这件事。
我们看到的是权力从机构向个人的转移，但这将是混乱的，至少需要一代人的时间。
与此同时，机构正在反击。我们正处于”帝国反击战”阶段，它们试图接管 Twitter、Facebook、Patreon 等赋予个人力量的新平台。
Brett：大学和整个学术界有一根很大的大棒——他们能够自己培养下一代教师，这些教师再去教下一代中小学学生。
Naval：这是一个教士阶层。你只能说教士们批准过的内容，只有你是教士才能说，而由教士们来决定谁是教士。
创新需要分权与前沿
过去十年技术将我们推向了集中化
Naval：创新需要几个条件。其中一个似乎必不可少的条件就是分权。
我认为这并非巧合——雅典城邦、意大利城邦，乃至美国——当它更自由、联邦政府管控更少的时候——都是创新的热bed，因为你有大量的竞争。如果自己的想法不受欢迎，人们可以从一个州转到另一个州，思想之间存在着活跃的竞争。
真正重要的多样性是思想的多样性，而不是肤色的多样性。
你还需要一个前沿。你需要新的东西去探索——要么是智识前沿，要么是物理前沿。我们已经占领了加州。如果说什么的话，现在加州就是那个建制，那个既得体制。它不再是蛮荒西部的边疆了。也许我们需要一个太空中的前沿。也许我们需要像加密货币那样的智识前沿。
蛮荒西部的本质就是，那里总是充斥着骗子，总是充斥着犯罪，总是充斥着各种奇奇怪怪的东西，因为它们往往会吸引一批怪人。但与此同时，那里也是大量创新发生的地方。
我看到许多老派科学家和企业家在感叹。“新的创业者在哪能受到欢迎？“我认为 Y Combinator 创始人 Paul Graham，一个很聪明的人，发过一条大概这样的推文：”
Steve Jobs……以及像他那样的人，在自己创立的公司里活不过一天。 “他会被自己的团队取消。
但今天的 Steve Jobs 会去做加密货币。他会和所有骗子、罪犯和怪人一起待在加密货币领域，但至少在那里他有一个做怪人的空间。他有一个与众不同的地方。他有一个可以尝试新事物而不必不断向谁交代的地方。
集权与分权的钟摆
集权与分权之间存在一个钟摆。
比如，看看加密世界，中心化的金融最终变得非常僵化。政府和监管机构告诉你什么能做什么不能做。你遭遇监管俘获，转眼之间，华尔街从经济中吸走 20% 的利润——而加密货币可以取而代之。
于是你得到一股去中心化的压力，人们可以以自由、可编程的方式来做这件事。但随之而来的也有更多的骗局、欺诈和损失。
打个比方，在古代，你担心森林里有强盗和劫匪，所以你向国王求助。国王建起漂亮的城堡，国王铸造货币。但转眼之间，国王在贬损货币，国王把人扔进监狱。然后有些人跑到森林里成为强盗，因为他们想要自由。但当然，现在他们又要遭受同伴的攻击和骚扰。
所以历史上在集权与分权之间存在着一种自然的钟摆运动，我认为技术的弧线在过去十年里实际上把我们推向了集权一方。
我是 Amazon 的大粉丝，但它是一个非常中心化的实体。我认为即使在那个行业里也在发生一股去中心化的潮流。像 Shopify 这样的平台正在崛起，赋能小店去竞争。像 DoorDash 这样的本地配送服务虽然是中心化服务，但它们正在让一支去中心化的大军——餐厅和本地商店——去与中心化服务竞争。
我们会看到这根弧线来回摆动。
妥协的贫乏
妥协检验的是一开始就没人认为正确的想法
Brett： 质疑那些在某一领域内你此前认为不可撼动的东西，这个想法非常有趣。
数千年来人们一直在思考，构想民主的最佳方式是什么。
Plato 问，“什么是民主？“他提出了谁应当统治的问题。据说这就是民主的全部理念。我们需要弄清楚谁应该统治。应该由哲学王来统治吗？应该由全体公民来统治吗？
Plato 认为暴民会轻易投票剥夺少数群体的权利，那就是他所理解的民主。
但 Popper 质疑了这整套看待民主的方式。他走得更深，大致说道：“民主与谁应当统治毫无关系。民主是一种允许你以最高效率、不带暴力地移除政策和统治者的制度。而这就是你评判不同民主制度的标准。”
所以你实际上可以对法国、英国、美国、澳大利亚、加拿大做出评判。这些地方的民主制度是好是坏，取决于我们在多大程度上能够快速、高效、轻松、不带暴力地把我们不喜欢的人从民主制度中清除出去？
这才是一个好的民主制度的衡量标准，而不是试图弄清楚哪个制度能给我们最好的统治者。这和说”哪种科学方法能给我们真理论”是同样的错误。没有任何科学方法能给我们真理论。
科学是一种纠错机制。我们所能期望的只是清除坏的想法。通过这样做，我们纠正了一些错误，然后可以继续前进，找到一个比我们之前拥有的更好的理论。
这就引出了一个问题：当你与别人僵持不下时，如何做出好的决策。
有一种观念认为妥协是某种美德，但其实不是。它的好处仅仅在于，如果你有两个人完全无法达成一致，又即将陷入某种冲突，那它比暴力对抗更可取。
如果你处于这样一种局面：甲持有想法 X，乙持有想法 Y，人们对妥协的普遍理解是它处于 X 和 Y 之间的某处：甲不会得到他想要的一切，乙也不会得到他想要的一切。他们想出一个妥协方案，即理论 Z。
当理论 Z 被证明不奏效时，我们不应该感到惊讶，因为两个人从一开始就没认为它是最好的想法。甲回过头来说，“我一直告诉你 X 才是正确的想法”，乙回过头来说，“我一直告诉你想法 Y 才是最好的想法。”
他们没有取得任何实质进步。他们证明了 Z 是错的，但从来没有人一开始就认为 Z 是正确的。
这就是妥协的贫乏，这也是你在科学中某些时候所得到的东西。在政治中更是无处不在。
自由市场提供最好的反馈
替代方案是由枪杆子最多的人提供反馈
Naval： Marc Andreessen 把这一点总结得很好：“强观点，弱持有。
作为一个社会，如果你追求真理，你应该持有强观点但非常松散地持有。你想要尝试它们，看看是否奏效，如果不奏效就进行纠错。
但我们实际得到的，要么是强观点强持有——即不宽容的少数派——要么是弱观点弱持有——即那种没有人真正承担责任的妥协模式，没有人获得功劳，没有人能以自己想要的方式去尝试，然后每个人都可以退回来说，“真正的共产主义还没被尝试过。“尽管在那种情况下，真正的共产主义已经被尝试过了，只是效果不好。
顺便说一句，我常听到人们的一种常见批评是，“我们需要迈向后资本主义世界。资本主义不奏效了。“好吧，你的替代方案是什么？通常这时候人们就开始支支吾吾，因为没有太多选择。
当你试图弄清楚如何分配功劳、分配资源并奖励人们的工作时，你有两个选择：来自自由市场和现实的反馈——其最佳模型是金钱——或者来自人的反馈，也就是共产主义最终走向的地方，即由一群人来决定你做的是最好的工作。
那么，谁来决定你做了最好的工作？必须有人负责这件事，而不变的是，那个人最终会是最大的暴徒。
我认为每个共产主义国家都蜕变为独裁统治并非偶然。共产主义似乎从来不是由分布式的人民多数来管理的。它最终总是由一群攫取权力的人来管理。
这就是人性：如果我决定谁能得到黄金，它会给我的朋友、家人和我喜欢的人。而这 invariably 就是最终发生的事。
你要么需要一个目标函数来分配资源——而金钱就是已知的目标函数——要么一切就都变成主观的。而一旦变成主观的，那谁来决定是你来分配还是我来分配？我们最终只会根据谁的武力更强、谁的枪更多来做决定。
Brett： 我们在自由市场这一方的主张是，我们已经将强制从决策过程中抽离出去了。没有人被强迫购买某项服务或接受某项协议。
唯一动用强制力的时候，是政府介入的时候。那时上层的人会说：“这是最好的决定，你必须同意；否则，就会有一个人带着徽章和枪出现在你家门口。”
我们在自由市场中所说的一切，就是让个人在不被胁迫的情况下自己做决定。当然，我可能判断错误，但为什么不让他们去尝试和犯错呢？这是取得进步的唯一途径。
纠正错误的唯一方法是真正去尝试别的东西。
将事物社会化会摧毁其中的真理性
科学的最大突破来自不受欢迎的人
Naval： 将事物社会化会摧毁其中的真理性，因为社会群体需要共识才能存续——否则它们就会内斗，无法共存——而共识的本质是妥协，而非寻求真理。
科学——至少自然科学——曾经是一种独特的学科，个人可以代表社会其余成员独立寻求真理。其他个人验证他们确实拥有了关于现实如何运作的最佳当前模型，然后这些模型可以通过发明创造传播到社会的其余部分。
但社会科学是潜入学界的一种病毒，并且已经取而代之。社会科学已经彻底腐败了。
首先，它们需要向社会 appeal 以获取资金，因此带有政治动机。其次，它们自身也受到社会的影响，因为这些研究和模型被用来驱动政策。所以，它们当然也会走向腐败。现在连自然科学也受到社会科学的攻击，变得越来越社会化。
你看到的群体思维越多，你离真理实际上就越远。你可以拥有一个和谐的社会，同时仍然允许社会中的真理探寻者去发现真理，找到改变和改善现实的方法，为整个群体造福。
历史上，大多数科学突破并非来自科学机构。那些重大突破来自独立的自然哲学家，他们是非常独立的思考者，在他们的时代备受唾弃，常常遭受迫害，他们基于自己所认识的真理与社会的其余部分抗争。往往要经过数十年甚至数百年——许多时候是在他们死后——这些真理才被接受。
很多学术理论既经不起复制检验——如果你看看心理学的现状——也经不起现实的检验。
Rory Sutherland 有一句精彩的话，大意是：“营销学就是知道经济学家错在哪里的科学。” 经济学家假设完全理性行为，但人类显然是湿件生物体，所以你可以通过营销来绕过这一点。
Nassim Taleb 会更进一步说，他们假设的是虚假的理性。人类在为毁灭的风险、归零的风险定价，而学者们在遍历性推理上犯了错误。他们假设对总体有利的东西对个体也有利，但事实并非如此。
个体不想归零——不想死——所以他们不会承担毁灭的风险，不会承担破产的风险；而一个群体应该愿意承担破产的风险，因为这分散在那么多不同的人身上。
群体从不承认失败
取而代之的是分裂
Naval： 群体从不承认失败。一个群体宁愿继续活在”我们受到了打压”的神话中，也不愿承认失败。只有个体才会承认失败。即使个体也不喜欢承认失败，但最终，他们可以被迫使承认。
一个群体永远不会承认自己错了。一个群体永远不会承认：“我们犯了一个错误”，因为一个试图改变想法的群体会分崩离析。我很难在历史上找到大规模群体说”我们曾经认为 A，但答案其实是 B”的例子。
通常在这种情况下发生的是分裂，比如从天主教会到新教等等。出现的是分化，通常伴随着大量内斗。在加密货币领域也是如此，币会分叉。Bitcoin 不会突然说：“我们应该有智能合约。” ETH 不会突然说：“我们应该保持不可变。”
我曾在一个基金会的董事会任职，该基金会负责为某项事业拨款，我发现这非常令人幻灭，因为我学到的是，无论基金会做了什么，他们都会宣布胜利。每个项目都是胜利的。每个项目都是成功的。大量互相拍肩膀，大量冠冕堂皇的使命宣言和愿景声明，大量互相祝贺，大量丰盛的晚宴——但什么实际的事情也没有做成。
我意识到这是因为没有客观的反馈。因为没有亏损——全都是社交意义上的获利——他们不可能失败。而因为他们不可能失败，他们整日都在错误配置资源。当然，这样的群体最终会耗尽资金。
如果你想把世界变得更好，最好的方式是营利性机构，因为营利性机构必须接受来自现实的反馈。讽刺的是，营利性实体比非营利性实体更可持续。它们是自我维持的。你不需要一直端着乞讨碗。
当然，你会失去美好的非营利地位；你必须缴税；而且你也可能因纯粹逐利而腐败。但我认为最好的企业是那些营利的、可持续的且合乎伦理的企业，这样你才能吸引到最优秀的人才。你能维持下去，因为这是一项使命，而不仅仅关乎金钱——因为赚钱存在边际递减效应。
金钱在你生命中的边际效用是递减的。
知识使资源的存在变为无限
我们将不断创造新知识和新资源
Brett： 知识是使资源的存在变为无限的东西。知识的创造是无界的。我们将不断创造更多知识，并由此了解越来越多、越来越不同的资源。
《The Beginning of Infinity》中有一个关于铕的精彩寓言，David 谈到大约 60 年前第一批彩色电视机开始制造的时候。那时有一种阴极射线管，你向荧光屏发射一束电子流。荧光屏上有这些像素，三种不同的颜色，其中一种是红色，屏幕上那些红色荧光粉中含有元素铕。
铕的有趣之处在于，当你通电通过它、激发它时，它会发出红色光。铕更加有趣的地方在于，它是元素周期表上唯一具有这种特性的元素；它是唯一能做到这一点的化学物质。如果你向它发射电子，它会发出彩色电视所需的红色光。
据计算，地球上铕的总量有限，而铕正在被阴极射线管制造商快速消耗。于是科学家们有了一个完全站得住脚的数学理论：阴极射线管的数量是有限的；因此，我们将耗尽阴极射线管。
确实，从非常狭义的角度来说，地球上任何一种资源的总量都是有限的。当然，外太空也有铕，你也许可以去那里开采，但更深层的道理是：现在已经没有人用阴极射线管了。如今彩色电视这个概念跟提取铕已经完全没有关系了。
我们现在用的都是 LCD 屏幕、等离子屏幕，未来可能还会出现某种全新的技术，跟我们今天拥有的技术毫无关系。但我们依然会有彩色电视，会有彩色屏幕。
这一点对我们能想到的任何资源都成立。
你完全可以做一个完全站得住脚的 Malthusian 计算：如果你碰巧生活在非洲稀树草原上，我们就不能一直烧木头，因为最终所有的森林都会被烧光。显然，木头会用完。木头的量是有限的。即使你能种更多的树，最终木材的消耗速度也会超过现有的总量。而针对煤炭、石油以及我们正在消耗的其他所有东西，人们用的也是同样的论证。
Naval：即使是所谓的虚空也包含大量物质，有大量可以转化为能量的东西。那里的资源没有上限。唯一的限制在于知识。
遗憾的是，人们在这里做出了一个悲观的假设，认为人类的创造力是有限的。而我认为，恰恰是那些没有亲手建造过东西、没有从零开始创造过新事物的人，最容易有这种感觉。
对穴居人来说，极少有东西是资源
Brett：英国 ITV 有一条新闻报道，谈论 Amazon 据说产生了多少浪费，说 Amazon 经常性地销毁大量产品。
我就想，“这些人为什么要对一个他们完全不了解的生意指手画脚？“他们难道希望 Amazon 拥有完美的知识，精确知道需要生产多少产品吗？换句话说，一个在认识论上不可能达到的状态。还是说他们更希望 Amazon 生产不足的产品，让想买的人根本买不到？
当然，Amazon 实际做的只是比需要的稍微多生产一些。任何企业都是如此。它们偶尔会比需要的多生产一点。
Naval：曾经有一位风险投资家跟我争论说，鞋子的种类太多了，这是资本主义失败的例子，因为没有人需要这么多款运动鞋。
我反问他：“你是什么时候知道鞋子太多的？“在历史上哪个时间点，我们判定鞋子够了？在此之前，我们需要更多的鞋子——我们需要更有弹性的鞋子，更耐穿的鞋子，更厚鞋底的鞋子，更轻的鞋子，我们需要各种令人惊叹的鞋子创新。
然后到了某个时刻，有人拍板说：“其实我们的鞋子够了。现在得砍掉所有其他鞋款。“你是怎么产生这种想法的——恰好你出生在对的时间、对的地点，就能判定我们的鞋子够了？
这是一种人人都容易陷入的狭隘观念。
这个观念还有一个更宏观的版本，就是那种”我们正在耗尽资源”的哲学。它的起点是认为地球是有限的，资源的总量是有限的，而我们正在把它们消耗殆尽。因此如果我们不压缩消费，我们就都得死。
首先，你凭什么把界线划在地球？你怎么不认为你的城镇正在耗尽资源？为什么城镇就不是你要保护的范围，城镇之外就都是异域、不可触及？
为什么要把界线划在地球周围？我们可以去太阳系，可以去银河系，可以去宇宙，可以去多重宇宙。如果你知道如何利用它们，那里有大量的资源。
其次，你如何定义什么是资源？资源不过是某种通过知识可以从一种东西转化为另一种东西的东西。
曾经有一个时期，煤炭不是资源；铁不是资源。对穴居人来说，极少有东西是资源——不过是一些可食用的植物和一些可食用的动物，仅此而已。
驯化、收获庄稼、冶金、化学、物理、开发发动机和火箭——所有这些，都是在把那些我们曾经认为毫无价值的东西变成资源。铀从完全一文不值，变成了一种令人难以置信的资源。
这种有限资源的世界模型，隐含地假设了知识是有限的。它说的是知识创造已经终结了。我们被困在了当前的节点上，因此，基于我们目前拥有的知识，这些就是我们可得的所有资源。现在我们必须开始节约了。
但知识，是我们永远可以创造更多的东西。
术语表
| 原文 | 中文 |
|---|---|
| Adam Smith | Adam Smith |
| AGI | AGI |
| Amazon | Amazon |
| Athenian city-states | 雅典城邦 |
| Bayes’ theorem | Bayes 定理 |
| Bayesian | Bayesian（贝叶斯主义者） |
| Bayesianism | Bayesianism（贝叶斯主义） |
| Brett Hall | Brett Hall |
| Charley Lineweaver | Charley Lineweaver |
| conjecture | 猜想 |
| convergent feature | 趋同特征 |
| crucial experiment | 判决性实验 |
| Dark Forest Hypothesis | 黑暗森林假说 |
| Darwinian theory of evolution | Darwin 演化论 |
| David Deutsch | David Deutsch |
| Donald Brownlee | Donald Brownlee |
| DoorDash | DoorDash |
| Dyson spheres | 戴森球 |
| Enrico Fermi | Enrico Fermi |
| epistemology | 认识论 |
| ergodic | 遍历性 |
| europium | 铕 |
| evolutionary algorithms | 演化算法 |
| Fermi paradox | Fermi 悖论 |
| general relativity | 广义相对论 |
| GPT-3 | GPT-3 |
| Harvard | 哈佛 |
| Italian city-states | 意大利城邦 |
| justified true belief | 得到辩护的真信念 |
| Kepler | Kepler |
| Lamarck | Lamarck |
| Malthusian | Malthusian（马尔萨斯的） |
| Manhattan Project | 曼哈顿计划 |
| Marc Andreessen | Marc Andreessen |
| memetic evolution | 模因演化 |
| Milky Way | 银河系 |
| Nassim Taleb | Nassim Taleb |
| Naval | Naval |
| Neil deGrasse Tyson | Neil deGrasse Tyson |
| Neo-Darwinist | 新达尔文主义 |
| neocortex | 新皮层 |
| neutron star | 中子星 |
| Newtonian gravity | Newton 引力理论 |
| OpenAI | OpenAI |
| parroting | 鹦鹉学舌 |
| Patreon | Patreon |
| Paul Graham | Paul Graham |
| Peter Ward | Peter Ward |
| Plato | Plato |
| Popper | Popper |
| Rare Earth | 《Rare Earth》 |
| refutation | 反驳 |
| Richard Dawkins | Richard Dawkins |
| Rory Sutherland | Rory Sutherland |
| Schelling point | Schelling 点 |
| scientism | scientism（唯科学主义） |
| selfish gene | 自私的基因 |
| Shopify | Shopify |
| Silicon Valley | 硅谷 |
| simulation hypothesis | 模拟假说 |
| Stephen Hawking | Stephen Hawking |
| Steve Jobs | Steve Jobs |
| The Beginning of Infinity | 《The Beginning of Infinity》 |
| The New York Times | 《纽约时报》 |
| The Washington Post | 《华盛顿邮报》 |
| universal explainers | 通用解释者 |
| Vogons | Vogons |
| VR | 虚拟现实 |
| wetware | 湿件 |
| Y Combinator | Y Combinator |
此文章由 AI 翻译（miaoyan_chunk_translate）
Part 2 of my interview with Brett Hall about The Beginning of Infinity . Also see Part 1 .
With a Good Theory of Knowledge, You Can Decide What Else Is True
David Deutsch’s theory is centered around good explanations
Naval: David Deutsch has this great view of the world where he believes that everything important is understandable by a single human. By important he means the underlying base theories that drive most of reality.
Deutsch fixates on four theories. I could argue maybe there are a few more, especially if you start getting into Adam Smith and The Wealth of Nations and a few other more sociological ones. But he’s obviously a physicist concerned more with reality and truth-seeking, not human systems.
The four he picks are the theory of epistemology; the theory of evolution by natural selection; quantum theory, into which he subsumes relativity and other physics; and the theory of computation, which includes his theory of quantum computation.
These four are fascinating. It’s probably worth exploring what’s interesting about each of them. What is the breakthrough here that might be nonobvious?
Let’s start with epistemology .
The reason I love The Beginning of Infinity is that Deutsch does a very rigorous review of what is correct in epistemology, what we know to be the best answers. Once you have a good theory of knowledge, then you can decide what else is true.
If you’re starting with a bad basis for the theory of knowledge, then you’re going to decide on a bunch of things that are false when you think they might be true.
And his epistemology is centered around good explanations. It takes Popper ’s view of science and truth-seeking as being error-correcting mechanisms and expands on it. I’d love to hear your summary of the theory of knowledge, or epistemology, as Deutsch lays it out.
No Truth Can Be Justified
‘Knowledge’ is just our best guess at the time
Brett: Initial guesses at what ‘knowledge’ was all about amounted to what is known as the “justified true belief” vision of knowledge, and it’s still the most prevalent idea today. Anyone who calls themself a Bayesian is a justified true believer.
This is the misconception that knowledge is about trying to justify your beliefs as true. And if you’ve done so, then you can say, “I know that thing.” If I can justify as true my theory of gravity, then I should believe that theory of gravity, and only then can I say that it’s known.
The problem with this is that there is no method of showing any piece of knowledge is true. The improvement Deutsch promotes in his books is this vision that Popper gave us, that all we have are guesses about reality. They’re conjectures.
People think, “Oh, that sounds a bit wishy-washy. It’s just a guess.” Well, it’s not a random guess. It’s a guess that has stood up against trials, against attempts to show that it’s false. It’s not that everyone who decides to have a guess stands on equal footing.
When people are unable to show that something’s false—via this method of refutation—then we accept it as a piece of knowledge. This allows us to accept the fact that we’re going to be able to make progress in the future, because all of our knowledge is conjectural. All of it is our best guess at the time.
There’s elasticity within the knowledge that allows us to say, “There’s going to be errors. We’re going to correct them and, thereby, be able to make progress into the infinite future.”
This is unlike the previous conception of knowledge, which says, “Once you’ve justified something as true, well, it’s true.” If it’s true, that means there is nothing false about it and, therefore, it can’t possibly be refuted. That’s a very religious notion.
The modern incantation of this is Bayesianism, which says, “You have a theory, you collect more evidence, and you become more and more confident over time that your theory is correct.”
It gets a little bit worse than that, because then it says, “This Bayesian reasoning enables you to generate new theories.” Which it can’t. The best that it can hope to do is to show you that you are more confident in this theory than you are in that theory.
The Popperian view says, “If you can show that there’s a flaw in a particular theory, you can discard that theory.”
There Can Be No Final Theory of Gravity
All we have are better and better approximations to reality
Brett: In almost all cases, you only ever have one theory on offer.
In the case of gravity, there literally is only one theory on offer at the moment. There’s general relativity . Previously we did have two theories. We had Newtonian gravity and we had general relativity—but we did a crucial experiment.
This idea of a crucial experiment is the cherry on top of science. You’ve got these two competing theories. If the experiment goes one way, one theory is ruled out but the other theory is not. In which case, you keep that theory for so long as no problems arise.
This vision of knowledge enables us to have an open-ended quest for progress.
This is completely unlike any other idea about knowledge. The overwhelming majority of physicists are still Bayesian. And the reason they’re still Bayesian is that this is typically what’s taught in universities and this is what passes for an intellectually rigorous way of understanding the world.
But all it is is what I would call species of scientism. It’s because they have a formula behind them, the Bayes’ theorem , which is a perfectly acceptable statistical formula. People use it all the time in perfectly legitimate ways. It’s just that it’s not an epistemology. It’s not a way of guaranteeing, or even being confident, that your theory is actually true.
My favorite example of this: Prior to 1919, every single experiment that was done on gravity showed that it was consistent with Newton’s theory of gravity. What do Bayesians say in that situation? They say you’re getting more and more confident in Newton’s theory.
How does that make sense? How do you square that the day before it was shown to be false was the day when you’re most confident in it?
Now, a Popperian doesn’t have this problem. A Popperian says, “At no point was Newton’s theory actually true. It contained some truth, but that truth isn’t a thing that we can measure.”
I say it contained some truth because it’s certainly got a more direct connection to reality than some other random person’s guess about the nature of gravity. Gravity does indeed approximately vary as the inverse square law, but not exactly. We needed general relativity to correct the errors in Newton’s theory of gravity.
And even though general relativity is our best guess right now, it can’t ultimately be the final theory of gravity. There can be no final theory of gravity. All we have is better and better approximations to reality.
Naval: I think the reason we fall into Bayesianism so easily is probably related to why we fall into pessimism so easily. We’re evolutionarily hardwired for Bayesianism.
Every other animal on the planet that can form good explanations is a Bayesian. They’re just looking at repeated events and saying, “The sun rose yesterday. The sun will rise tomorrow.” Or, “That thing I touched is hot. It’s probably going to be hot in the future.” That is how most of our biological systems and most of our evolutionary heritage worked.
It’s just now we have this neocortex that can form good explanations about the seen in terms of the unseen. That gives us a higher level of reasoning, but that higher level of reasoning is not instinctual to us. It requires effort. It requires deep thinking.
We default to Bayesianism because that is how a lot of the natural world around us seems to work at least at the purely biological level.
Genetic Evolution Was a Prelude to Memetic Evolution
Ideas undergo the same evolutionary process as genes
Brett: Popper has a book called “ Objective Knowledge ,” and it’s subtitled “An Evolutionary Approach.” And that’s no accident, either. There’s a symmetry between the theory of epistemology and the theory of evolution as we understand it.
Before we understood what is known as the Darwinian theory of evolution , the only idea that people had was that these entities had to be created. All the plants and animals that you see around you had to be created by a creator. There was no other explanatory mechanism.
Some people came up with the idea of gradual change over time. Lamarck was one. His idea was the reason why giraffes have long necks is that their ancestors had slightly shorter necks, so they tried to stretch their necks to reach the leaves they couldn’t reach.
But again, there was no mechanism for this beyond the fact that an individual goes off to the gymnasium and works on their biceps and their biceps get a little bigger over time. Although you can work out in the gym and increase the size of your biceps, that doesn’t mean your children are going to inherit those characteristics.
What Darwin came up with is a similar idea to what Popper had in knowledge. It was an error correction. The idea is that an organism would trial itself out in a particular environment and if it wasn’t, as we say, fit for that environment, then it would die off. But if it was fit in that environment, then it would survive.
So you have this encounter with reality between living organisms and the environment. And it’s the environment that’s giving you feedback from reality and destroying those organisms that aren’t fit enough to survive.
The Neo-Darwinist view is to give us what the unit of selection is. It’s not the group or the herd; it’s not even the individual. It’s the gene. It’s the selfish gene idea, which comes to us from Richard Dawkins , who says if any one of those genes happens to be not fit for the particular environment, that could cause the death of that organism. The species might survive, but its entire DNA will ever so subtly change over time as the environment changes.
Now we have leveled that up. We human beings are the next step in that evolutionary process where we can create explanatory knowledge, which does the same thing. Deutsch likes to say, “Genetic evolution was merely a prelude. What’s coming next is memetic evolution.”
The history of the universe from here on out is going to be the history of ideas undergoing the same evolutionary process as the genes did previously.
Humans Are Exceptional
Humans aren’t just a form of bacteria that overran the planet
Naval: Three out of these four theories have an interesting pattern to them. With good explanations in epistemology, we’re saying conjectures and their refutations and error correction is how we improve knowledge. With genetic evolution, genetic mutations, variation and natural selection weed out the ones that didn’t work. Then there’s memetic evolution on top of that, where we have ideas and then criticism weeds out the ideas that don’t work.
Related to that, in invention there’s trial and error. In capitalism, startups get created and the ones that have bad ideas fail. We see this pattern recurring over and over.
What’s interesting, though, is there’s another metapattern here, that humans are exceptional. In epistemology, humans are the only non-Bayesian reasoners. In evolution, humans are the only memetic creatures that we know of. In the theory of computation , humans are the only universal explainers that we know of other than, of course, the computers that we’ve invented.
Science took us from this view of humans being at the center of the universe to, “Actually, humans are nothing special. You’re just one little planet out of an almost infinite number of Kepler planets that could be bearing life out there.” But the three of these four theories that we’re talking about are pointing us in this direction of humans are exceptional. Humans are capable of maximal knowledge.
One interesting realization for me was that even if you were God, even if you had infinite knowledge and power, even if you controlled the entire universe, you still wouldn’t know that you’re not in a simulation. You still could never prove that you’re not in the simulation. And even as God, there is no concept that you could hold in your head that a human being couldn’t hold.
Unless, of course, the laws of physics are different. If the laws of physics are different, then all bets are off and who knows? But working within the current laws of physics, humans are capable of maximal knowledge and maximal awareness. That points to a world where humans are exceptional and not just another form of bacteria that got out of control and overran this planet.
A lot of these fundamental theories lead to a viewpoint that humans are special, knowledge is infinite, and as long as we don’t destroy the means of error correction and we’re always creating new knowledge, then there’s good reason to be optimistic.
We Are Qualitatively Different From Other Species
There is no continuum
between humans and chimpanzees
Brett: You’re pointing out a minority opinion there. I think culture is still stuck in that second part of what you were saying.
Originally we thought that we were at the center of the universe. This was the religious conception of man’s place in the cosmos. Earth was surrounded by the celestial spheres and everything orbited around it. So we were the inheritors of the entire universe, and God had gifted us with this.
Then science showed us that, in fact, we’re not at a particularly special place in the universe. This is the cosmological principle, this idea that the universe is roughly the same at every single place and we are just one of those particularly unspecial places.
Not only are we unspecial in the cosmological sense, but biologically we’re nothing particularly special, either. We’re just on the continuum between bacteria to cockroaches through to dogs and chimpanzees.
An astrophysicist I absolutely love on almost every other topic, Neil deGrasse Tyson , was talking about how chimpanzees are a lot smarter than we think and how we might not be much better. This is what almost everyone thinks.
This third view that a lot of us are trying to promote now is that it’s not a slight quantitative difference between chimpanzees and us. There is a continuum between bacteria to cockroaches to dogs and chimpanzees, but we’re off-axis.
We are qualitatively different. All we need to do is open our eyes. You look out your window at that beautiful city that happens to be out there that cannot be explained by this gradual increase of biological complexity.
More Compute Power Doesn’t Produce AGI
Even the most powerful computers can’t answer ‘why?’
Naval: The artificial general intelligence crew gets it completely wrong, too: “Just add more compute power and you’ll get intelligence,” when we don’t know what it is underneath that makes us creative and allows us to come up with good explanations.
People talk a lot about GPT-3 , the text matching engine that OpenAI put out, which is a very impressive piece of software. They say, “Hey, I can use GPT-3 to generate great tweets.” That’s because, first, as a human you’re selecting the good tweets out of all the garbage that it generates. Second, it’s using some combination of plagiarism and synonym matching and so on to come up with plausible sounding stuff.
The easiest way to see that what it’s generating doesn’t actually make any sense is to ask it a follow-up question. Take a GPT-3 generated output and ask it, “Why is that the case?” Or make a prediction based on that and watch it completely fall apart because there’s no underlying explanation.
It’s parroting. It’s brilliant Bayesian reasoning. It’s extrapolating from what it already sees out there generated by humans on the web, but it doesn’t have an underlying model of reality that can explain the seen in terms of the unseen. And I think that’s critical.
That is what humans do uniquely that no other creature, no other computer, no other intelligence—biological or artificial—that we have ever encountered does.
And not only do we do it uniquely, but if we were to meet an alien species that also had the power to generate these good explanations, there is no explanation that they could generate that we could not understand.
We are maximally capable of understanding. There is no concept out there that is possible in this physical reality that a human being, given sufficient time and resources and education, could not understand.
It’s Mind Blowing That Our Minds Can’t Be Blown
Humans can understand every idea given enough time and effort
Brett: Scientifically minded types say, “Perhaps we won’t be able to understand the next set of laws of physics. Perhaps we won’t be able to understand the aliens.” It’s nothing but the appeal to the supernatural. It’s logically equivalent to “God is out there, and you can’t possibly understand what God is. God is this infinite, omniscient being that is beyond us.”
You can believe that if you like. You can believe the simulation hypothesis. You can believe any one of these things. They’re all metaphysical claims about a reality that we have no access to. Whether or not you want to introduce aliens who will have ideas that we can’t comprehend, that’s all standing on the same footing.
Naval: At least in God’s metaphysics you could say, “OK, that’s in a different universe. That’s outside of our laws of physics.” But the aliens presumably would be under the same laws of physics, so I don’t even see what the basis for that is.
Any species that is smart enough to get off its home planet knows that the limiting factor is ideas. So the thing that they should want the most from any other species they encounter is new ideas. And the trade that they should be making is the trade of ideas.
There’s this Malthusian philosophy in science fiction—now called the
Dark Forest Hypothesis
—that every human species is like bacteria and we’re going to run out of room. No, the universe is infinite in size; it’s expanding. The multiverse is even more infinite in size.
We are at the beginning of infinity. We’re not running out of resources. Everybody’s creating ideas. Smart alien civilizations trade ideas and successful human civilizations trade ideas. Because those ideas take things that were useless before and turn them into resources.
Every alien civilization can trade ideas with every other civilization because they’re all universal explainers. They’re capable of maximal understanding.
In fact, the mind-blowing thing here is that your mind cannot be blown. There’s no idea out there that your mind cannot absorb given the time and the effort.
So if we encounter an alien species, we should probably rejoice. They don’t want anything from our planet other than our ideas. And the best way to trade ideas is to have a dynamic, abundant, thriving civilization.
Because I grew up on rote sci-fi, I used to be pessimistic about alien encounters. “Oh yeah, if we encounter aliens, they’ll just destroy us.” Like, in Hitchhiker’s Guide to the Galaxy , the Vogons thoughtlessly demolished Earth to make room for a hyperspace bypass.
The reality of it is that any species that finds us is going to immediately give us all the knowledge that they have. And they’re going to crave new knowledge that we have, because they will realize that would allow them to light up the dark matter, the dark energy, the unused resources in the universe, to allow them to thrive as well.
Where Are the Aliens?
The universe should be teeming with aliens
Naval: Let’s talk briefly about the Fermi paradox since we’re talking about aliens. For listeners who don’t know, Enrico Fermi was a famous physicist who was part of the Manhattan Project . He said, “ Where are the aliens? ”
The universe is so large and there are probably so many planets that are capable of supporting life of some kind or another. Shouldn’t we have seen them by now?
Brett: Around almost every star there is a contingent of planets much like our own solar system. The number of stars that exist within a typical galaxy like the Milky Way is something like 200 billion, although the estimates go up to about 400 billion. The number of galaxies that we can see is around 200 to 300 billion.
The observable universe is just a small fraction of the entire universe, which means that the number of planets is absolutely astronomical.
Surely, given these numbers, it has to be the case that there are not only planets out there that are suitable for life but that the universe should be teeming with civilizations far more advanced than ours, less advanced than ours, and some that are similar in advancement to ours.
So where are they?
Intelligent Species Have Risen Only Once on Planet Earth
Creative intelligence wasn’t inevitable
Brett: Now, that’s one argument, and we have to be humble in the face of uncertainty here because no one knows. But I want to give an argument that rarely gets any air time.
The argument is that we are alone.
The argument has nothing to do with astronomy; it has everything to do with biology. The argument goes like this: Look at planet Earth and look at the number of species not only that exist right now—millions of them—but also the number of species that have ever existed on planet earth, which is hundreds of millions.
Life arose something like three and a half billion years ago, and for about two and a half billion years there was nothing but bacteria. So life apparently doesn’t have much impetus to evolve quickly beyond bacteria; it just remains as simple as possible.
A lot of people have this misconceived idea that Darwin really did away with—the idea that evolution has a direction in mind. You see these pictures of evolution that appear in high school textbooks of the monkey that’s hobbling around on all fours; then he’s hunched over; and then eventually he is standing up and holding a briefcase, as if this is what evolution had in mind. It only seems to be what evolution had in mind in retrospect, by looking backward.
There’s an American academic, Charley Lineweaver , who calls this the “Planet of the Apes” hypothesis—as in, if you remove the humans from a planet, the apes would naturally evolve to fill the intelligence niche.
He said you could imagine another situation where you’re an elephant that is able to think about themself. They reflect on the length of their trunk, and they look back through biological evolution and see that trunks get ever shorter. So what they conclude is, “Ah, evolution has been geared towards making ever-longer trunks. That’s what evolution is all about.”
Of course, we can see that that’s ridiculous. It just happens to be the case that this creature called the elephant has evolved and it’s got this long trunk, but the length of the trunk doesn’t appear to be a convergent feature of evolution.
A convergent feature of evolution is a feature that exists within biological entities that has arisen again and again, independently. Wings are my favorite example. Fish have wings of a certain kind. There are flying fish. Butterflies have wings, so we’ve got them in insects. They arose in mammals as well, with flying foxes and certain kinds of possums. And, of course, birds and dinosaurs had wings as well.
Independently, in all these species, the wings keep arising. So do eyes, and so do organs for sound.
Now let’s think about the capacity to do mathematics or to build radio telescopes—in other words, to be an intelligent, creative species. How many times has that arisen in the geological history of the Earth? In one species and one species alone.
Can we conclude on that basis that, therefore, it’s inevitable that intelligent species will arise? If you were to repeat the experiment by sprinkling a few bacteria around all the bio-friendly planets that exist throughout the universe, would you be guaranteed to get an entity like us?
The Probability of Human Existence Is Infinitesimally Small
From an evolutionary standpoint, it’s near impossible
Brett: Here’s another way to think about it that is mathematically frightening for the people who think that the aliens are out there and they’re going to visit us sometime in the future.
We were talking earlier about trillions of planets that exist throughout the known universe that might even be friendly for life to arise. Imagine that between us as intelligent human beings and the most simple form of bacteria that we can imagine, there are only 100 independent evolutionary steps. Now that’s not true—a million or more different mutations probably had to happen to allow the organisms to survive such that we exist today—but let’s make it only 100.
Imagine that each of those independent steps had a probability of just one in 10 happening. Now, in fact, it’s probably more like one in a million, but we’ll be generous and say one in 10.
Now what we have is a chain of probability: one in 10, times one in 10, times one in 10…100 times. This is one over 10, all to the power of 100—which is one over one followed by 100 zeroes. That number swamps the astronomical number I was talking about with planets earlier on.
In other words, the probability of us arising in this particular argument is infinitesimally small. The fact that it’s happened once should blow our minds.
If You Can’t Program It, You Haven’t Understood It
Evolutionary algorithms don’t produce living beings
Brett: These are all uncertain hypotheses, but we also have to keep in mind that there’s so much about evolution by natural selection that we don’t know.
David Deutsch has this little quip, “ If you can’t program it, you haven’t understood it. ”
In the case of AGI, this means we can’t program it because we don’t understand this idea of general intelligence.
The same happens to be true of evolution by natural selection. There are things called evolutionary algorithms , but this is not programming evolution by natural selection. This is not being able to create artificial entities inside of a computer that, when subject to actual environmental pressures, are able to evolve towards this increasing complexity.
Aliens Might Just Be Too Far Away
We may not be alone in the universe, just too far apart
Brett: We still have this problem of what DNA was doing for that approximately two and a half billion years—the overwhelming majority of the history of life on Earth. Why didn’t it evolve at all during that time? What’s going on?
There’s a book, Rare Earth , by Peter Ward and Donald Brownlee, and these guys talk about all the quirky things that happened in the evolutionary history of the Earth. I just picked on the fact that we universal explainers evolved seemingly fortuitously, seemingly once; but you can go back and realize that evolving from single-cell bacteria to a multicellular organism was weird and unusual and hasn’t been able to be repeated in a laboratory setting.
Then to go from the multicellular organism to something that’s like a plant and then something that’s like an animal—each of these things seems to have occurred for reasons that we don’t understand.
Naval: There could be a combination of things going on. Your argument can be statistical rather than absolute. We may not be alone in the universe, but becoming universal explainers might be so rare that when you start multiplying that by interstellar distances, which are quite vast, we might just be too far apart.
I think Fermi also had the unreasonable assumption that interstellar aliens would figure out how to get past the speed of light, when we have no hypothesis whatsoever as to how that might be possible. We have nothing even vaguely in the category of how to get past the speed of light.
So if you’re limited by the speed of light, and if the jump to universal explainers is rare, then we might just be too far apart. And it might just take a lot longer.
The universe is very big, but it’s also almost entirely empty, at least as far as planets and the stars are concerned. Given that, it’s still quite reasonable to say that humans and human-like explainers are quite rare; they’re still early in their formation across the universe; and they’re just spread out by such incredibly vast distances that we haven’t encountered each other.
If we did encounter each other, I think we’d know.
For example, by the time an alien spacecraft got here, their radio waves would have arrived long before. There’s a pretty long period in a civilization’s history when it invents the radio and starts to broadcast radio waves out, before it invents interstellar travel and it’s sending rockets and civilizations around the universe.
Aliens Would Visit for Knowledge, Not Resources
The only thing they will lack is the knowledge they don’t have
Brett: I think Stephen Hawking said that it was a mistake to broadcast radio waves out into the universe because the aliens are going to be out there and they’re going to be like conquistadors. They’re going to want to take over our planet for their resources and various other things.
There’s a couple of responses you can have to the idea of evil aliens coming to get us.
The first is that the only way to make progress into the infinite future and to have the technologies that would enable you to traverse the galaxy, is to have this vision of knowledge that Popper had—namely, that you are freely able to explore the space of ideas, able to falsify assumptions, and don’t have centralized authorities and force being used on people, which dampens down creativity.
To have a maximally creative society, you have to have freedom, you have to have liberty, and therefore you will have a non-violent society. You’ll have a society that values creativity as an end in itself.
When we encounter the aliens, we shouldn’t expect that they’re going to be immoral bastards that are going to want to take over our resources, but the opposite: They’re going to look at us and think what primitive savages that we are. They’re going to think that we’re moral midgets, and they’re going to want to teach us. They’re not going to want to put us in prison or anything like that.
Knowledge is a unified whole. If their physics are so much better than ours—which enables them to approach the speed of light or to use some weird general relativity gravity thing that creates a wormhole, so they can get through space faster than the speed of light—all of their scientists are going to be so much farther ahead. All of their knowledge is going to be farther ahead: their mathematics, their morality, their political institutions.
So we don’t have to worry about the aliens.
And, by the way, we don’t have to worry about them stealing our resources. It’s not like they’re going to go, “Ah, there’s a planet full of coal and water. We’re going to take it.”
No, they’re going to have the knowledge to be able to sweep up the hydrogen in intergalactic space and turn that into a fusion reactor and use 3D printing to create any technology that they want.
In fact, that might be another answer to the Fermi paradox . They don’t need to leave their local area because they’ve already got the technology to perfectly sustain them.
Naval: They’ve got Dyson spheres and they can gather all the energy they need. They can gather all the matter, they can create anything they want, and they can have any reality in VR space that they want.
The only thing that they would be lacking is new knowledge that they don’t have.
Ideas Are the New Oil
All the new fortunes are created with ideas
Naval: Humans have a history of conquest because we fight for the same exact resources, but even in human history the first explorers were traders. They were going out there to find spices, gold, silk, new plants to domesticate, new animals.
They weren’t going out there necessarily to conquer the land. Eventually they did because of the finite resource dilemma when you’re stuck on Earth. But the moment you have the technology to get off of the earth, finite resources go away.
If you want a resource, then you go find a neutron star or you go find a star system. You don’t go after a little planet. There are infinite Kepler planets out there that are going to be much closer to them.
It’s ideas and trade that people want.
If you look at modern society—even though I know this is not the common belief—we’re becoming less warlike as we become more civilized. The reason is that you don’t conquer Russia anymore for its natural resources. The wealthiest places in the world now are the ones that have the best ideas. Silicon Valley was on top for a while as a wealth creation engine because it had the best ideas.
The new oil is ideas. It’s all digital. All the new fortunes are being created in ideas space.
In fact, if you’re starting out today as a young, ambitious person, you don’t learn real estate; you don’t learn coal and oil mining; you don’t go into the extraction of physical resources to create wealth. You go into ideas space. You go into programming, books, movies, blogs and podcasts and building robots, which are mostly intellectual property underneath.
Even as a human civilization, we’re moving away from conquering physical resources and moving much more into trading of ideas.
The downside scenario for the human species is that too many of our larger countries and nation-states believe that they’ve achieved maximal ideas and now it’s time to save resources. They end up destroying the means of improvement, error correction and creativity, and they end up stagnating.
Then you have the idea generation coming out of a much smaller set of city-states, which then have to defend themselves against these massive, more predatory, ossified states.
People talk about China being so impressive—“look at their rocket,” or “look at their GDP,” or “look at the city that they built.” Call me when they invent something new. Call me when they come up with some incredible idea that we haven’t had and they built some technology that we haven’t had. Because so far it’s all imitative. It’s them taking advantage of technology they’ve picked up from us that they’re now catching up on. They’re just applying scale to it because they have more people.
Call me when their GDP per citizen crosses ours. Call me when they come up with pharmaceuticals or vaccines or spacecraft or energy generators or fusion reactors that we do not know how to build. Call me when the authoritarian society figures out top-down how to build something brand new, when it’s more creative, when their art is better, when their science is better, when their technology is better.
Call me when that happens over a democratic, free, capitalist society, because I’ve never seen a case of that, ever.
One Einstein Is Worth A Legion Of PhD Drones
Creativity goes from 0 to 1 and bodies aren’t going to solve the problem
Brett: China keeps graduating more bachelor of science and bachelor of engineers than anywhere else in the world. China’s universities are pumping out more science graduates than us, but they’re not pumping out more innovators.
It’s not like the students that are coming out of those universities in China with their science degrees are going off and doing innovative stuff. It’s quite the opposite, because they’ve been trained in a particular way. They’re being trained to memorize this textbook, respond to this exam. They can’t think outside of the box.
They’ve been trained that, “This is what’s true. This is the unquestioned correct way of thinking about science.” That might be good for being able to imitate, as we see, but it’s not going to be the thing that enables you to push forward the frontier in technology, let alone fundamental physics or anywhere else.
I don’t care what the statistics are on how many science graduates they’ve got. That makes no difference. Give me 10 innovative, creative, young physics graduates over 50,000 physics graduates who all are able to pass the exam with 100% efficiency any day.
Naval: So, one Einstein is worth the legion of drones with PhDs in physics. It doesn’t matter. Creativity goes zero to one. And no amount of throwing bodies at the problem will get you there. That’s just the nature of memetic evolution; it’s just the nature of creativity.
Don’t Rely on Credibility Stamps
The generation-long shift from institutions to individuals will be messy
Naval: There are a lot of institutions in our society today that rely on credibility stamps.
They used to be how you gain credibility in society. If you were a journalist writing for The New York Times or The Washington Post , then you had the masthead of The Times and The Post . If you’re a professor at Harvard, you have credibility because you’re a professor at Harvard.
Of course, those systems got hacked. A lot of social scientists who have no business telling the world what to do are now in there with nonsense political models masquerading as economists or natural scientists. You have people who are activists writing under the mastheads of these formerly great newspapers and burning up the credibility capital that these newspapers have built up over time.
The Internet is exposing them slowly but steadily, and we’re going through a transition phase where the masses still believe in the institutions.
We’re caught in this Schelling point , this coordination point for the institutions: How do I know if I should hire you? Will you have a diploma from Harvard? I know it’s not as good as it used to be. I know a Harvard humanities diploma’s probably nonsense at this point, but I don’t have any other credibility metric to filter you and I need to do it in an efficient way.
What we’re seeing is the transition of power from institutions to individuals, but it’s going to be messy and it’s going to take at least a generation.
In the meantime, the institutions are fighting back. We’re in the Empire Strikes Back phase where they’re trying to take over the new platforms like Twitter, Facebook, Patreon, which empower the individuals.
Brett: The university and all of the academia have a very big stick in terms of being able to train their own next generation of teachers who then go on to teach the next generation of primary and secondary school students.
Naval: It’s a priesthood. You’re only allowed to say what the priests have approved, and you can only say that if you are a priest, and the priests get to decide who’s a priest.
Innovation Requires Decentralization and a Frontier
Technology has swung us towards centralization in the last decade
Naval: Innovation requires a couple of things. One of the things that it seems to require is decentralization.
I don’t think it’s a coincidence that the Athenian city-states or the Italian city-states or even the United States—when it was more freeform and less federal government controlled—were hotbeds of innovation, because you had lots and lots of competition. People could switch from one state to another if their ideas weren’t welcome, and there was a robust competition of ideas.
The real diversity that matters is the diversity of ideas, not the diversity of skin color.
You also need a frontier. You need something new to explore—either an intellectual frontier or a physical frontier. We’ve occupied California. If anything, now California is the institution, the establishment. It’s no longer the front of the Wild West. Maybe we need one in space. Maybe we need intellectual ones like we have in cryptocurrencies.
It’s the nature of the Wild West that they’re always filled with scammers; they’re always filled with crimes; they’re always filled with very strange and odd things, because they tend to attract a weird crowd. But at the same time, it is where a lot of the innovation is going on.
I see a lot of lamenting from old school scientists and entrepreneurs. “Where are the new entrepreneurs welcome?” I think Paul Graham, the Y Combinator founder, a brilliant guy, tweeted something along the lines of, “ Steve Jobs… and people like him wouldn’t last a day inside the companies they created. ” He’d be canceled by his own team.
But Steve Jobs today would be in crypto. He’d be in crypto with all the scammers and all the criminals and all the weirdos, but at least there he’d have a space to be weird. He’d have a place to be different. He’d have a place to try new things without having to constantly answer to someone.
There is a pendulum between centralization and decentralization.
For example, if you look at the crypto world, centralized finance ends up very ossified. You have the government and the regulators telling you exactly what you can and can’t do. You get regulatory capture, and next thing you know, Wall Street is sucking 20% of the profits out of the economy—and crypto can replace that.
So you get decentralization pressure where people can do it in a freeform, programmatic way. But then you end up with a lot more scams, fraud and losses as well.
An analogy might be that, in olden times, you worried about brigands and robbers in the forest, so you appealed to the king. The king builds a nice keep, the king mints the money. But next thing you know, the king is debasing the currency and the king is throwing people in jail. Then some people run off into the forest and they become brigands again because they want their freedom. But now, of course, they’re subject to attacks and harassment from their peers.
So there’s a natural pendulum swing that goes on in history between centralization and decentralization, and I think the arc of technology actually swung us towards centralization in the last decade.
I’m a big fan of Amazon, but it’s a very centralized entity. I think that there’s a decentralization arc that is taking place even in that industry. Things like Shopify are coming up and enabling small stores to compete. Local delivery services like DoorDash are centralized services, but they’re allowing a decentralized army of restaurants and local shops to compete against centralized services.
We’re going to see this arc go back and forth.
The Poverty of Compromise
Compromises test ideas no one ever thought were correct in the first place
Brett: This idea of questioning things that hitherto you thought were unassailable in a particular domain is really interesting.
For millennia people have wondered about the best way to conceive of what democracy is.
Plato asked, “What is democracy?” and he had the question about who should rule. That’s the whole idea of democracy, supposedly. We’d have to figure out who should rule. Should it be the philosopher kings who should rule? Should it be the population of citizens?
Plato decided that the mob would readily vote away the rights of a minority, and that’s what he thought democracy was.
But Popper questioned this whole idea of looking at what democracy was. He went even deeper and roughly said, “Democracy has got nothing to do with who should rule. Democracy is the system which allows you to remove policies and rulers most efficiently without violence. And that’s how you judge different democratic systems.”
So you can actually make a judgment on France, England, the United States, Australia, Canada. Do these places have better or worse kinds of democracy to the extent that we’re actually able to get rid of the people that we don’t like from the democratic system quickly, efficiently, easily, without violence?
That’s the measure of a good democratic system, rather than trying to figure out which is going to give us the best rulers. That’s the same mistake as saying, “What method of science is going to give us the true theory?” No method of science is going to give us the true theory.
Science is an error-correcting mechanism. All we can hope for is to get rid of the bad ideas. And by doing that, we’ve corrected some of our errors, and then we can move forward to find something that’s a better theory than what we had before.
This raises the idea of how to make good decisions when you’re at loggerheads with someone else.
There’s this idea that compromise is supposed to be a virtue of some kind, and it’s not. It’s preferable to having a violent confrontation if you’ve got two people who otherwise can’t possibly reach an agreement and they’re going to get into a battle of some sort.
If you’re in a situation where person A has idea X and person B has idea Y, the common understanding of a compromise is that it’s somewhere between X and Y: Person A won’t get everything they want, and person B won’t get everything they want. They come up with a compromise, which is theory Z.
We shouldn’t be surprised when theory Z proves not to work, because neither person ever thought it was the best idea in the first place. Person A goes back to saying, “I always told you that X was the correct idea,” and person B goes back to saying, “I always told you that idea Y was the best idea.”
They haven’t made any progress whatsoever. They’ve shown that Z is wrong, but no one ever thought that Z was correct in the first place.
This is the poverty of compromise, and this is what you get in science at certain times. It’s everywhere in politics as well.
Free Markets Provide the Best Feedback
The alternative is feedback from whomever has the most guns
Naval: Marc Andreessen summarizes this nicely as “ strong opinions, loosely held. ”
As a society, if you’re truth-seeking, you want to have strong opinions but very loosely held. You want to try them, see if they work, and then error-correct if they don’t.
But instead what we get is either strong opinions strongly held—which is the intolerant minority—or we get weak opinions loosely held—which is this compromised model where no one really takes the blame, no one gets credit, no one gets to try the way that they want to, and everybody can then fall back on, “Real communism hasn’t been tried.” Although, in that case, real communism has been tried; it just hasn’t worked out well.
As a digression, one of the common critiques I hear people say is, “We need to move to a post-capitalist world. Capitalism isn’t working.” OK, what is your alternative? Usually this is where people start fumbling because there aren’t a lot of choices.
When you’re trying to figure out how to divvy up credit, divvy up resources and reward people for their work, you have two choices: feedback from free markets and reality—and the best model for that is money—or feedback from people, which is where communism ends up, which is a group of people who decide that you did the best work.
Now, who decides you did the best work? Someone has to be in charge of doing that, and invariably that ends up being the biggest thug.
I don’t think that it’s an accident that every communist country degenerates into a dictatorship. Communism never seems to actually be run by a distributed majority of the people. It always ends up being run by a bunch of people who are taking charge.
It’s just human nature that if I get to decide who gets the gold, it’s going to go to my friends, family and the people that I like. And that’s invariably what ends up happening.
Either you need an objective function to carve it up—and money is the known objective function—or it becomes all subjective. And if it’s subjective, then who’s to say you’re carving it up instead of me? We’re just going to decide based on who has more physical force, who has more guns.
Brett: What we say on the side of free markets is that we’ve extracted coercion out of that decision-making process. No one is forced into purchasing a service or undertaking an agreement.
The only time that force is applied is when the government gets involved. The people at the top then say, “This is the best decision and you will have to agree with it; otherwise, there’s going to be a man with a badge and a gun turning up at your door.”
All that we’re saying when it comes to the free market is that the individual gets to decide without being coerced. Now, I could be wrong, but why shouldn’t they try and make mistakes? It’s the only way to make progress.
The only way to error-correct is to actually try something else.
Making Something Social Destroys the Truth of It
Science’s biggest breakthroughs came from unpopular people
Naval: Making something social destroys the truth of it because social groups need consensus to survive—otherwise they fight and can’t get along—and consensus is all about compromise, not truth-seeking.
Science—at least the natural sciences—was this unique discipline where you could have an individual truth-seeking on behalf of the rest of society. Other individuals verify that they did, indeed, have the best current model of how reality works, and then that could be spread out through inventions to the rest of society.
But the social sciences are this virus that crept into academia and have taken over. Social sciences are completely corrupted.
First, they need to appeal to society for funding, so they are politically motivated. Then, they themselves are influenced in society because the studies and models are used to drive policy. So, of course, that ends up corrupted as well. Now even the natural sciences are under attack from the social sciences, and they’re becoming more and more socialized.
The more groupthink you see involved, the farther from the truth you actually are. You can have an harmonious society while still allowing truth seekers within the society to find truth and to find the means to alter and improve reality for the entire group.
Historically, most of the scientific breakthroughs didn’t come from scientific institutions. The big ones came from individual natural philosophers who were very independent thinkers who were reviled in their time, often persecuted, who fought against the rest of society on the basis of their truths. And it took decades or centuries—often after their deaths—before those truths were accepted.
A lot of these academic theories don’t actually stand up either to replication—if you look at what’s going on in psychology—or even to reality.
Rory Sutherland has this great quote where he said something along the lines of, “ Marketing is the science of knowing what economists are wrong about. ” Economists assume perfectly rational behavior, but humans are obviously wetware biological creatures, so you can hack around that using marketing.
Nassim Taleb would go even further and say that they assume false rationality. Humans are pricing in the risk of ruin, the risk of going to zero, and the academics are making mistakes about ergodic reasoning. They’re assuming that what’s good for the ensemble is good for the individual, and it’s not.
An individual doesn’t want to go to zero—doesn’t want to die—so they will not take risks of ruin and they will not take risks of bankruptcy; whereas a group should be willing to take a risk of bankruptcy because that’s spread out among so many different people.
Groups Never Admit Failure
You get a schism instead
Naval: Groups never admit failure. A group would rather keep living in the mythology of “we were repressed” than ever admit failure. Individuals are the only ones who admit failure. Even individuals don’t like to admit failure, but eventually, they can be forced to.
A group will never admit they were wrong. A group will never admit, “We made a mistake,” because a group that tries to change its mind falls apart. I’m hard pressed to find examples in history of large groups that said, “We thought A, but the answer’s actually B.”
Usually what happens in that case is a schism, where you go from the Catholic Church to Protestant and so on. There’s a divergence and usually a lot of infighting. This happens in crypto land, too, where the coins fork. Bitcoin doesn’t suddenly say, “We should have smart contracts.” ETH doesn’t suddenly say, “We should have been immutable.”
I was on the board of a foundation that was charged with giving out money for a cause, and I found it very disillusioning because what I learned was that no matter what the foundation did, they would declare victory. Every project was victorious. Every project was a success. There was a lot of back slapping. There were a lot of high-sounding mission statements and vision statements, a lot of congratulations, a lot of nice dinners—but nothing ever got done.
I realized this was because there was no objective feedback. Because there is no loss—it’s all social profit—they couldn’t fail. And because they couldn’t fail, they misdirected resources all day long. And eventually, of course, such groups run out of money.
If you want to change the world to a better place, the best way to do it is a for-profit because for-profits have to take feedback from reality. Ironically, for-profit entities are more sustainable than non-profit entities. They’re self-sustainable. You’re not out there with a begging bowl all the time.
Of course, you lose the beautiful non-profit status; you have to pay your taxes; and also you can get corrupted by being purely for-profit. But I would argue that the best businesses are the ones that are for-profit, sustainable and ethical so you can attract the best people. You can sustain it because it’s a mission and it’s not just about the money—because there are diminishing returns to making money.
There’s a diminishing marginal utility to the money in your life.
Knowledge Makes the Existence of Resources Infinite
We’re going to keep creating new knowledge and new resources
Brett: Knowledge is the thing that makes the existence of resources infinite. The creation of knowledge is unbounded. We’re going to keep on creating more knowledge and, thereby, learning about more and different resources.
There’s this wonderful parable of europium in The Beginning of Infinity where David talks about when the first color television started to be manufactured about 60 years ago. There was a cathode ray tube type where you’d fire a stream of electrons at a phosphorescent screen. The phosphorescent screen would have these pixels, three different colors, one of which was red, and those red phosphors on the screen were filled with the element europium.
The interesting thing about europium is, when you put electricity through it, when you excite it, it glows with this red color. The extra-interesting thing about europium is that it is the only such element on the periodic table; it’s the only chemical that will do that. If you fire electrons at it, it will glow the red that you need to have for color television.
It was calculated that there’s only a certain amount of europium on the earth, and that amount of europium was quickly being consumed by cathode ray tube manufacturers. So the scientists had a perfectly robust mathematical theory about how the number of cathode ray tubes was finite; therefore, we’re going to run out of cathode ray tubes.
It’s true, in a very narrow sense, that for any given resource you’re going to have a finite amount on planet Earth. Of course, there’s going to be europium in outer space and you could probably mine it there, but the deeper point is that no one has cathode ray tubes anymore. The whole idea of color television has nothing to do with the extraction of europium these days.
We’ve all got LCD screens, we have plasma screens, and there will probably be something else coming in the future that will have absolutely nothing to do with the kind of technology we have today. But we’re still going to have color television or color screens.
This is true for absolutely any resource that we can think of.
You might very well make a perfectly good Malthusian calculation that we can’t keep on burning wood if you happen to be living on the African savanna, because eventually all of the forests are going to be burned down. Obviously, we’re going to run out of wood. There’s a finite amount of wood. Even if you can grow more wood, eventually the consumption of wood is going to outstrip the amount that’s there. And this is the argument that’s made for coal, oil, and everything else that we happen to be consuming.
Naval: Even so-called empty space has a lot of matter and a lot of things that could be converted into energy. There is no limit to the number of resources out there. There’s purely a limit to knowledge.
Unfortunately, there’s a pessimistic assumption here that people make that human creativity is bounded, and I think it’s the people who have not built things, who have not created new things from scratch, who seem to feel this the most.
To a Caveman Very Few Things Are Resources
There was a time when coal wasn’t a resource
Brett: There was a story on ITV in the U.K. talking about how much supposed waste Amazon produces, that Amazon was routinely destroying a whole bunch of products.
I thought, “Why are these people inserting their opinion into a business that they know absolutely nothing about?” Would they prefer Amazon to have the perfect knowledge of precisely how many products need to be made? In other words, an epistemologically impossible situation to be in. Or would they prefer that Amazon made insufficient products, so the people who wanted to purchase them weren’t actually able to get ahold of them?
What Amazon does, of course, is make slightly more than what they need. That’s what happens in any business. They make slightly more than what they need now and again.
Naval: I once had a venture capitalist argue to me that there were too many kinds of shoes and it was an example of how capitalism had failed because nobody needs this many kinds of sneakers.
My question to him was, “When did you know that there were too many shoes?” What’s the point in history where we decide there are too many shoes? Before we needed more shoes because we needed more stretchy shoes, we needed more durable shoes, we needed thicker soled shoes, we needed lighter shoes, we needed all kinds of amazing shoe innovations.
And then at some point, somebody decides, “Actually we have enough shoes. Now we need to kill all the other shoe lines.” Where did you come up with this idea that you just happened to be born at the right time and the right place to identify that yes we have enough shoes?
This is a certain parochialism that everyone falls into.
There’s a more macro version of it, which is this “we’re running out of resources” philosophy. It starts with this idea that the Earth is finite, that there’s this finite set of resources and we’re consuming them all. And therefore we’re all going to die if we don’t tamp back our consumption.
First of all, how did you decide that it was the Earth? How did you decide that your town wasn’t running out of resources? Why wasn’t the town the actual area that you wanted to save and then everything outside of that was foreign and unreachable?
Why draw the boundary around the Earth? We could go to the solar system. We could go to the galaxy, we could go to the universe. We could go to the multiverse. There are a lot of resources out there if you know how to harness them.
Then, how do you define what a resource is? A resource is just something that through knowledge you can convert from one thing to another.
There was a time when coal wasn’t a resource; iron wasn’t a resource. To a caveman very few things are resources—just a few edible plants and a few edible animals and that’s it.
Domestication, harvesting crops, metallurgy, chemistry, physics, developing engines and rockets—all of these are things that are taking things that we thought were worthless and turning them into resources. Uranium has gone from being completely worthless to being an incredible resource.
This finite resource model of the world implicitly assumes finite knowledge. It says knowledge creation has come to an end. We are stuck at this current point, and, therefore, based on the knowledge that we have currently, these are all the resources available to us. Now we must start conserving.
But knowledge is a thing that we can always create more of.