返回列表
🧠 阿头学 · 🪞 Uota学 · 💬 讨论题

30 回合的游戏,你还在囤星星

马斯克不是从 Polytopia 学到规则,而是用游戏照亮了自己本就在用的"有限回合思维"——问题是,你确定自己也在玩同一种游戏吗?
打开原文 ↗

2026-03-04 原文链接 ↗
阅读简报
双语对照
完整翻译
原文
讨论归档

核心观点

  • 有限回合是战略的元前提,不是可选项 文章最强的洞察不在 8 条规则本身,而在这个元问题:你到底在玩《文明》还是 Perfection 模式?大多数人默认无限生存,所以会囤资源、慢决策、追求稳定。但如果你的真实处境是 30 回合强制结束(比如 Neta 的海外窗口),那么闲置本身就是亏损,不是节俭。这不是心态问题,是数学问题。
  • "冷酷"不是目的,建模才是 规则 1 最容易被误读。文章不是在鼓励你变成混蛋,而是说:在极高压决策中,把对手/员工/市场当作"优化题"而非"情感主体",能让你看得更清。这和 AlphaGo 的逻辑一样——不是无情,是纯粹。但这里有个陷阱:99% 的创业者用这套逻辑会导致团队崩溃和法律诉讼,因为他们既没有马斯克的资源溢价,也没有他的任务紧迫感作为道德支撑。
  • 规则 8 是整个框架的裂缝 前 7 条规则构建的是一个 24/7 不断电的优化机器,规则 8 突然说"拔掉电源"。这不是补充,这是承认:纯粹的回合优化会导致决策失控(看马斯克在 X 上的行为就知道)。文章提出的"战壕与高塔"模型试图修补这个裂缝,但逻辑上还是有张力的——如果你真的在 30 回合的游戏里,周末的"高塔时间"本身就是浪费回合。
  • 跨局负债是被忽略的计分板 文章最诚实的部分在结尾:人不是下一局会复活的单位,伤害会累积。但这个警告被埋在了"阴暗面"章节里,而不是融入主模型。这意味着整套框架对于"无限游戏"(品牌、组织健康、人才)的成本计算是失效的。如果你把无限游戏按 Perfection 模式硬打,短期吞吐会漂亮,长期复利会塌。
  • 初创公司的"回合数优势"是真的,但有前提 文中提到初创公司一年 50 个回合 vs 大公司 1 个回合。这对 Neta 很有诱惑力。但这个优势只在"你知道自己在玩什么游戏"时才成立。如果你把回合数当成目的本身(追求快速迭代),而不是为了在有限窗口内最大化某个明确的终局指标,那就是在浪费回合。

跟我们的关联

👤ATou

  • 这篇文章直接戳到你的核心困境:你是在为 Neta 的 2026 海外目标玩 Perfection 模式,还是在为"公司长期生存"玩《文明》?如果是前者,那么"战壕与高塔"模型值得强制执行——每周 4 小时的"高塔时间"不是浪费,而是确保你没有在错误的游戏里优化。如果是后者,那么这套规则会把你带进坑里。

🧠Neta

  • 规则 4(先发制人)和规则 7(选择战场)对团队管理最有启发。20 人团队的火力必须集中在能改写盘面的点(比如某个海外市场的留存突破、内容供给的质量跃升),而不是在非决定性的争论里消耗节奏。这意味着你需要一个明确的"Schwerpunkt"(决定性重心),而不是全线开战。

🪞Uota

  • AI agent 体系天然是回合制的(任务轮次、上下文窗口、反馈周期)。把"每轮产出/等待损耗/复用率"量化,空转 agent 和长等待人工审批都是"囤星星"。这给了你一个可操作的指标:衡量 agent 的真实价值不是"能做什么",而是"每回合的产出密度"。

讨论引子

  • 你的 2026 真的是 30 回合吗,还是你在用 Perfection 模式的紧迫感来掩盖无限游戏的焦虑? 背景:很多创业者会把"快速迭代"当成目的本身,而不是为了在真实的有限窗口内达成某个终局指标。如果 Neta 的海外窗口其实是 3 年而不是 1 年,那么现在的"极限执行"可能在透支长期复利资产。
  • 在"冷酷建模"和"团队信任"之间,你的边界在哪里? 背景:规则 1 的逻辑在极端情况下会导致"为了优化指标而伤害人"。对于 20 人团队,这个边界比对马斯克更硬——因为你没有无限的人才溢价来补偿伤害。
  • 你是在优化"回合数"还是"终局得分"? 背景:两份批判都指出,文章的框架容易陷入"事后解释"陷阱——任何激进行为都能被解释成"抢节奏",任何失败也能被解释成"有限回合代价"。这意味着你需要一个明确的、可被证伪的计分板,而不是一个万能的解释框架。

在 Polytopia 的竞技对局里,你正好只有 30 回合。

没有加时,也没有重来。每一回合,你若用来侦察而不是扩张,那就是一回合永远回不来的机会。游戏不会在乎你的战略多漂亮、步子计划得多周密。一旦计数归零,真正算数的只有你实际建成了什么。

其他一切都无关紧要。

埃隆·马斯克在同时掌舵特斯拉、SpaceX、Neuralink、xAI、The Boring Company 和 X 的那段时间,迷上了 Polytopia。他甚至打败过这款游戏的作者。他爱得太深,以至于一度 Polytopia 被加进了特斯拉车机,成了可以直接玩的游戏。

大多数人——包括最初的我——都把这当作一种古怪的爱好,或一桩离奇的怪癖。

但他们没抓住重点。对马斯克而言,Polytopia 不止是一款游戏。它是一张他整套世界观的蓝图,是支撑他做一切事情的那套“操作系统”的完美缩微模型。

而他从这款游戏里提炼出的 八条规则,比任何传记、访谈或财报电话会议都更能揭示他真正如何思考。

大多数 4X 策略游戏,比如《文明》、《帝国时代》和《群星》,都会给你一个开放的时间地平线。你一直玩,直到赢或输。时间被拉长:总还有下一回合、下一个时代、下一项科技。这类游戏教你耐心、积累与长远视角。简单来说,它们教你如何生存。b

Polytopia 的竞技 Perfection 模式把这些本能全部倒转。你只有 30 回合。你的得分,是时钟停下那一刻你所积累的一切之总和。仅仅这一条限制,就把这个类型训练给你的所有直觉都反了过来。

扩张胜过防守,速度好过谨慎,任何闲置的资源都等于在藤上慢慢腐烂的果子。大多数策略游戏教你生存,而 Polytopia 教你:生存并不重要。重要的只有截止时间前的产出。

下面是埃隆如何把它用到现实生活里(由沃尔特·艾萨克森讲述——他跟着埃隆待了一整年,并为他写了传记):

这就是马斯克的运作方式与几乎所有其他 CEO 的运作方式之间的结构性差距。他们在玩《文明》,而他在玩 Perfection 模式

1986 年,纽约大学宗教学教授詹姆斯·卡斯出版了 《有限与无限的游戏》。他认为世界上只有两类游戏:有限游戏——以获胜为目的;无限游戏——以让游戏继续为目的。西蒙·西涅克在 2019 年把它改写成一套畅销的商业哲学,主张伟大的公司应该玩无限游戏,把重点放在长寿上。

埃隆从未把这些内化。SpaceX 的建立不是为了做一家可持续经营的航天公司,而是要在他所感知的机会窗口关闭之前,让人类成为多行星物种。特斯拉的创立也不是为了无止尽地卖车,而是要在气候期限到来之前迫使能源转型。Polytopia 并没有教会他这些,但它映照了他本就拥有的视角——它像一面镜子,让他看见自己。

沃尔特·艾萨克森的传记把埃隆从游戏中提炼出的规则逐条列了出来。它们值得一条条单独看,因为这样我们才能诊断他是怎么思考的。

规则 1:同理心不是资产。

脱离语境,这听起来像反社会人格。但在 Polytopia 里,游戏奖励的,是对对手最优行动的建模,而不是对对手情绪的体察。马斯克的弟弟金巴尔在艾萨克森的传记里说得很直白。金巴尔告诉艾萨克森,他有一种“同理心基因”,在商业上伤害了他;而埃隆让他去玩 Polytopia,以理解当你把这条基因剥离掉时,战略会如何运作。

在 Polytopia 里,你的对手不是人,而是一道优化题。

游戏奖励你去建模他们最强的可能走法,并据此对弈;它不奖励你去揣摩他们的感受。对手若冒进,你就利用空当;若把城市留空,你就拿下。冷酷不会受罚,犹豫才会。

这在战略领域并非新概念。2016 年,当 DeepMind 造出击败世界冠军李世石的 AlphaGo 时,关键突破之一就是让机器去建模李世石的最优回应,并与“最强版本的他”对弈。

这就是博弈论学者所说的、被剥离到最纯粹、最对抗形态的“心理理论”。

凯文·达顿关于不同职业人群中精神病态特质的研究发现:CEO 与外科医生一起,位于谱系的顶端。这并不意味着这些人都是怪物。相反,这些角色会吸引那些能把情绪反应与战略计算分离开来的人。

当埃隆在 Twitter 进行大规模裁员、在特斯拉突然解雇员工,或公开批评下属时,如果你以为他在玩一场长期的社会游戏,那看起来就是残酷;但如果你意识到他在 Polytopia 里玩的是 Perfection 模式——在那种模式里,任何不能加分的人都只是在浪费回合——那看起来就更像是在优化回合。

规则 2:把人生当作一场游戏来玩。

在印度教哲学里,有一个概念叫 Lila:宇宙本身就是神圣的游戏,是现实与自身的嬉戏,除了“玩”之外没有更高的目的。

《薄伽梵歌》告诫阿周那,要不执着于结果地去打眼前的仗:也就是说,在游戏之中全力以赴、凶猛行动,同时保持一种意识——这终究是一场游戏。这不是冷漠,恰恰相反:它是“最大程度的投入,最小程度的执着”。

心理学对此有一个现代版本。密歇根大学的心理学家伊森·克罗斯,在过去十年里一直研究:当人们在高风险决策中把视角从第一人称切换到第三人称,会发生什么。

他的实验室在多项实验中发现:那些用第三人称叙述自己选择的人——比如问“伊森应该怎么做?”而不是“我应该怎么做?”——在压力下表现更好,做出可测量地更理性的决定,受到的情绪干扰也更少。其机制被克罗斯称为“自我疏离”。

游戏天然会迫使你养成这种习惯。你是俯视棋盘,而不是从棋盘内部向外看。埃隆在高风险时刻的抽离感,是从棋盘上方看到的视角,不是愚钝,也不是自闭。

对他了解甚深的希冯·齐利斯,在对艾萨克森描述时,以一种诡异的精准说出了这一点:

"小时候,你在玩这类策略游戏,你妈妈把电源拔了,你居然没注意到,你还继续玩——把生活当作那场游戏一样继续玩下去。"

规则 3:不要害怕输。

卡尼曼与特沃斯基的前景理论指出:人类对损失的感受强度,大约是同等收益的两倍。这种不对称使得大多数决策者在最应该冒险的时刻,反而趋于避险。

一局 Polytopia 只有 15–20 分钟。你经常会输——因为扩张太慢、科技点错、或被卡在咽喉要道。然后你从头再来。马斯克所说的“习惯输”,在功能上几乎等同于临床心理学中的 暴露疗法:在可控的情境里反复接触厌恶刺激,直到情绪反应逐渐减弱。

大多数人用来处理失败的情绪机制(反刍、自我保护、退回安全区)被刻意地在重复中磨平了。损失厌恶是一种默认设置;它很正常,但埃隆把它重写了。

规则 4:先发制人。

在国际象棋里,节奏(tempo)指的是你迫使对手对你做出反应时所获得的优势。在无限游戏中,失去节奏是可逆的——你有无限步可走。但在一个 30 回合的游戏里,哪怕因为被动应对而丢掉两三个节奏,也可能在数学意义上无法挽回。

这解释了为什么埃隆会宣布他不可能兑现的时间表,挑起他并不需要的冲突,并在自己的组织内部制造危机——他是在夺取节奏。

在 Polytopia 里,如果你前五回合都在侦察而不是建设,你就已经输了。地图是通过扩张而显形的,而不是通过探索。

规则 5:优化每一回合。

以利亚胡·戈德拉特在 1984 年的著作 《目标》 中提出了约束理论。它的核心是:一个系统的产出由最紧的瓶颈决定,把资源花在其他任何地方都是浪费。

这是大多数管理者在理智上都懂、但几乎没人能在本能上真正“感到”的概念。当你能看见回合计数从 30 跳到 29 再跳到 28,浪费会在身体层面让人不适——这种不适,是抽象的机会成本永远给不了的。

这就是为什么马斯克会在产能冲刺时睡在工厂地板上。这不是营销噱头。他看到的是一个以季度为周期的高管们看不到的回合计数器。

季度规划是《文明》式思维,因为总还有下一个三个月的窗口可以修正问题。埃隆则是在假设:自己所在的这一回合,可能就是决定最终得分的那一回合。

规则 6:加倍下注。

在 Polytopia 里,占领城市会给你星星(stars),那是游戏的货币。

如果你不立刻把这些星星花在新单位、科技或扩张上,你就在浪费它们。在有回合上限的游戏里,未花掉的资源代价极高,因为它们错过了本可产生的复利价值。

这对应到约翰·凯利于 1956 年在贝尔实验室提出的凯利准则。凯利准则是在你拥有优势时,为下注规模定量的数学最优策略。凯利公式表明:在足够多的回合里,下注不足与下注过度一样昂贵。

埃隆的再投资模式,从外部看起来近乎鲁莽:他在现有工厂尚未盈利时,就把特斯拉的利润投入新工厂;他在猎鹰 9 的合同尚未结束时,就把 SpaceX 的收入灌进星舰项目。

在有限游戏的框架里,囤积才是鲁莽之举。任何闲置的资源,都是在一个永不暂停的倒计时里白白浪费掉的回合。

规则 7:选择你的战场。

在 Polytopia 里,你可能会发现自己被六个甚至更多敌对部落包围,它们同时试探你的边界。第一反应——尤其是那些已经把前六条规则内化、又偏好进攻的玩家——往往是全部开打。

但如果你冲动,那就是资源放血。顶尖的 Polytopia 玩家明白:不是每一座城市都值得占领,也不是每一个敌人都值得交战。有些战斗就是陷阱。真正的技能,在于识别哪些冲突能提高你的最终得分,哪些只是让你“感觉”自己很有效率。

普鲁士将军卡尔·冯·克劳塞维茨在其至今仍是西方军事战略奠基文本的 《战争论》 中,将其称为 Schwerpunkt——即决定性重心。其含义是:把力量集中在决定性点上,才能赢得战争。拿破仑之所以能赢下一场场战役,正是因为他会在敌方战线最薄弱的点上集中压倒性的力量。

这也重新框定了埃隆的一些选择。他确实挑了一些仗来打:与 SEC 的纠葛、X 上的广告抵制、SpaceX 的政府发射合同……这些都是围绕复利位置展开的斗争,是那种“赢了就能永久改变棋盘”的战役。

在回合受限的游戏里,把任何回合花在非决定性战斗上,都是从真正能抬高分数的战斗里偷走回合。不是所有攻击性都是战略性的攻击性;有些只是带着成本的噪音。

规则 8:有时也要拔掉电源。

最后一条规则是不合群的,而这恰恰是它重要的原因。Polytopia 再怎么紧张,也是一场短游戏。结束时,你把手机放下,短暂地站在系统之外。

众所周知,产生突破的往往不是那种高度聚焦、任务导向的大脑,而是漫游的心智——淋浴时的灵感、凌晨三点的顿悟,诸如此类。

阿基米德是在浴缸里,而不是在工作台前,解决了位移的问题(Eureka!)。庞加莱则描述过:关于福克斯群函数的解法,会在他踏上公交车时不请自来地降临。

一个永远不离开棋盘的“回合上限玩家”,最终会看不清棋盘本身。但按埃隆自己的承认,这条规则最难做到:如果你正在建造某样东西,真的很难拔掉电源。

有个概念叫 “Trenches & Towers”,我个人一直奉行,并且将来会写一篇文章专门谈它。工作日,你在生意的战壕里:埋头执行、苦干、在地面层面解决问题。周末,你爬上高塔:拉远视角,看清整张地图,思考下一步该走哪里。这种分离本身就是重点。

这个模型也有阴暗面,如果忽略它就不诚实了。

回合上限思维会让你把一切都当作工具。

关系只是联盟,员工是单位,公众信任是一种可被消耗的资源,而不是应当被维护的东西。那个在第 28 回合前牺牲三个战士去抢下一座城的 Polytopia 玩家,在游戏内做了正确选择;那个为了赶船期而把一个工程团队烧到崩溃的 CEO,在他自己的游戏里也做了正确选择。

但人不是下一局就会复活的单位。伤害会跨局累积。

埃隆疏离的子女、公开的争端、连番离婚,以及那些把为他工作描述为创伤经历的员工——这些都是一种回合制框架会忽略的成本,因为它们不会出现在记分板上。

我在艾萨克森的书里读到:埃隆曾经把一名员工骂得极其恶毒,以至于同事们都被震住;而事后才发现,那名员工不久前刚失去一个襁褓中的女儿。马斯克并不知道。在 Perfection 模式的操作系统里,即使他知道也无关紧要,因为唯一会被登记的变量,是员工的绩效表现。

这个模型之所以强大,恰恰因为它剥离了变量。

最后,我真正想摆上桌面的教训,不是这八条规则,而是它们之下的元教训。

大多数人并没有一套清晰的“我如何做决策”的显性模型。他们在一些继承来的默认设置上运作:来自进化心理学的风险规避,来自部落本能的社会从众,以及来自卡尼曼与特沃斯基同一套研究的损失厌恶——而埃隆则在功能上把自己训练得脱离了这些。

它们只是环境的副产物,由与我们今天任何人所处的真实游戏毫不相关的祖先压力塑造而成。

马斯克看着 Polytopia,看见了他自己的决策过程被形式化后的版本。这八条规则不是他从 Polytopia 学来的教训,而是他原本就在用、只是被游戏照亮得更清楚的规则。

这才是关键洞见——它与你是否崇拜他毫无关系。

这也与红杉的首席产品官 Jess Lee 的一个说法相呼应:她把创业公司称作一种回合制游戏。意思大致是:如果像 Google 或 Adobe 这样的巨头,在预算周期里一年只做一次“动作/回合”,那么一家以“实时”方式运作的初创公司,在同一年里可以走 50 个回合。

每个人都在玩某种游戏,但大多数人从未审视过自己的规则、回合上限,或自己的计分方式。

你可能在玩 《文明》,而你的处境真正需要的是 Perfection 模式。你可能在一场 30 回合就结束的游戏里囤着星星。你可能在为一个与你正在玩的游戏并不匹配的记分板做优化。

又或者,你在一场事实上是无限的游戏里玩着 Perfection 模式——为了吞吐量而开采、榨干关系,在一种用“累积点数”(也就是钱)之外的尺度来衡量的生活里。

这八条规则不会告诉你自己身处哪一种游戏。任何规则集都做不到。但它们会迫使大多数创业者问自己一个问题:

你实际上还有多少回合?而你是否在像自己知道答案那样去玩?

In competitive Polytopia, you have exactly 30 turns.

There are no extensions or do-overs. Every turn you spend exploring instead of expanding is a turn you don’t get back. The game is indifferent to your strategy or how well you planned your moves. Once the counter hits zero, only what you actually built matters.

Everything else is irrelevant.

Elon Musk became obsessed with Polytopia while simultaneously running Tesla, SpaceX, Neuralink, xAI, The Boring Company, and X. He’s beaten the creator of the game. He loves it so much that at one point, Polytopia was added as a playable game to Tesla cars.

Most people, even I, at first, dismissed this as a quirky hobby or a strange eccentricity.

But they missed the point. Polytopia is more than a game to Musk. Polytopia is a blueprint for his entire worldview, a perfect miniature of the operating system running underneath everything he does.

And the eight rules he took from the game reveal more about how he actually thinks than any biography, interview, or earnings call ever has.

Most 4X strategy games such as Civilization, Age of Empires, and Stellaris, give you an open horizon. You play until you win or lose. Time stretches. There is always another turn, another age, or another technology. These games teach patience, accumulation, and the long view. Put simply, they teach you to survive.b

Polytopia's competitive Perfection mode inverts all those instincts. You get 30 turns. Your score is the sum of everything you've accumulated when the clock stops. This single constraint inverts every instinct the genre trains you on.

Expansion beats defense, speed is better than caution, and any resource sitting idle is a resource rotting on the vine. Most strategy games teach you to survive, whereas Polytopia teaches you that survival is irrelevant. Only your output before the deadline counts.

Here’s how Elon implements it in real life (told by Walter Isaacson who spent a year shadowing Elon and wrote his biography):

This is the structural gap between how Musk operates and how virtually every other CEO operates. They are playing Civilization while he is playing Perfection mode.

In 1986, NYU religious studies professor James Carse published Finite and Infinite Games. He argued there are only two types of games: finite games, which are played to win, and infinite games, which are played to keep the game going. Simon Sinek turned this into a bestselling business philosophy in 2019, suggesting that great companies should play infinite games and focus on longevity.

Elon never internalized any of it. SpaceX wasn’t built to just be a sustainable aerospace company, but to make humanity multiplanetary before a perceived window of opportunity closes. Tesla wasn’t created to sell cars indefinitely, but to force an energy transition before a climate deadline. Polytopia didn’t teach him these things, but it reflected his existing perspective. It showed him a mirror.

The rules Elon took from the game, as Walter Isaacson’s biography catalogued them, are worth looking at individually, because doing so lets us diagnose how he thinks.

在 Polytopia 的竞技对局里,你正好只有 30 回合。

没有加时,也没有重来。每一回合,你若用来侦察而不是扩张,那就是一回合永远回不来的机会。游戏不会在乎你的战略多漂亮、步子计划得多周密。一旦计数归零,真正算数的只有你实际建成了什么。

其他一切都无关紧要。

埃隆·马斯克在同时掌舵特斯拉、SpaceX、Neuralink、xAI、The Boring Company 和 X 的那段时间,迷上了 Polytopia。他甚至打败过这款游戏的作者。他爱得太深,以至于一度 Polytopia 被加进了特斯拉车机,成了可以直接玩的游戏。

大多数人——包括最初的我——都把这当作一种古怪的爱好,或一桩离奇的怪癖。

但他们没抓住重点。对马斯克而言,Polytopia 不止是一款游戏。它是一张他整套世界观的蓝图,是支撑他做一切事情的那套“操作系统”的完美缩微模型。

而他从这款游戏里提炼出的 八条规则,比任何传记、访谈或财报电话会议都更能揭示他真正如何思考。

大多数 4X 策略游戏,比如《文明》、《帝国时代》和《群星》,都会给你一个开放的时间地平线。你一直玩,直到赢或输。时间被拉长:总还有下一回合、下一个时代、下一项科技。这类游戏教你耐心、积累与长远视角。简单来说,它们教你如何生存。b

Polytopia 的竞技 Perfection 模式把这些本能全部倒转。你只有 30 回合。你的得分,是时钟停下那一刻你所积累的一切之总和。仅仅这一条限制,就把这个类型训练给你的所有直觉都反了过来。

扩张胜过防守,速度好过谨慎,任何闲置的资源都等于在藤上慢慢腐烂的果子。大多数策略游戏教你生存,而 Polytopia 教你:生存并不重要。重要的只有截止时间前的产出。

下面是埃隆如何把它用到现实生活里(由沃尔特·艾萨克森讲述——他跟着埃隆待了一整年,并为他写了传记):

这就是马斯克的运作方式与几乎所有其他 CEO 的运作方式之间的结构性差距。他们在玩《文明》,而他在玩 Perfection 模式

1986 年,纽约大学宗教学教授詹姆斯·卡斯出版了 《有限与无限的游戏》。他认为世界上只有两类游戏:有限游戏——以获胜为目的;无限游戏——以让游戏继续为目的。西蒙·西涅克在 2019 年把它改写成一套畅销的商业哲学,主张伟大的公司应该玩无限游戏,把重点放在长寿上。

埃隆从未把这些内化。SpaceX 的建立不是为了做一家可持续经营的航天公司,而是要在他所感知的机会窗口关闭之前,让人类成为多行星物种。特斯拉的创立也不是为了无止尽地卖车,而是要在气候期限到来之前迫使能源转型。Polytopia 并没有教会他这些,但它映照了他本就拥有的视角——它像一面镜子,让他看见自己。

沃尔特·艾萨克森的传记把埃隆从游戏中提炼出的规则逐条列了出来。它们值得一条条单独看,因为这样我们才能诊断他是怎么思考的。

Rule 1: Empathy is not an asset.

This sounds sociopathic stripped of context. In Polytopia, however, the game rewards modeling your opponent's optimal moves, not their feelings. Musk's brother Kimbal put it plainly in Isaacson's biography. Kimbal told Isaacson he has an "empathy gene" that has hurt him in business, and that Elon told him to play Polytopia to understand how strategy works when you strip that gene out.

In Polytopia, your opponents aren't people, but optimization problems.

The game rewards you for modeling their strongest possible moves and playing against those, not for modeling their feelings. If an opponent overextends, you exploit the gap. If they leave a city undefended, you take it. There is no penalty for ruthlessness, only for hesitation.

This is not a new idea in strategy. When DeepMind built AlphaGo, the system that defeated world champion Lee Sedol in 2016, the breakthrough was programming the machine to model Sedol's optimal responses and play against the best version of him.

This is what game theorists call "theory of mind" stripped to its purest, most adversarial form.

Kevin Dutton's research on psychopathic traits across professional populations found that CEOs rank alongside surgeons at the top of the spectrum. This doesn’t mean these individuals are monsters. Instead, these roles attract people who can separate emotional responses from strategic calculations.

When Elon carries out mass layoffs at Twitter, fires staff abruptly at Tesla, or publicly criticizes subordinates, it looks like cruelty if you think he is playing a long-term social game. It looks like turn optimization if you realize he is playing Perfection mode in Polytopia, where anyone not adding to the score is simply wasting a turn.

规则 1:同理心不是资产。

脱离语境,这听起来像反社会人格。但在 Polytopia 里,游戏奖励的,是对对手最优行动的建模,而不是对对手情绪的体察。马斯克的弟弟金巴尔在艾萨克森的传记里说得很直白。金巴尔告诉艾萨克森,他有一种“同理心基因”,在商业上伤害了他;而埃隆让他去玩 Polytopia,以理解当你把这条基因剥离掉时,战略会如何运作。

在 Polytopia 里,你的对手不是人,而是一道优化题。

游戏奖励你去建模他们最强的可能走法,并据此对弈;它不奖励你去揣摩他们的感受。对手若冒进,你就利用空当;若把城市留空,你就拿下。冷酷不会受罚,犹豫才会。

这在战略领域并非新概念。2016 年,当 DeepMind 造出击败世界冠军李世石的 AlphaGo 时,关键突破之一就是让机器去建模李世石的最优回应,并与“最强版本的他”对弈。

这就是博弈论学者所说的、被剥离到最纯粹、最对抗形态的“心理理论”。

凯文·达顿关于不同职业人群中精神病态特质的研究发现:CEO 与外科医生一起,位于谱系的顶端。这并不意味着这些人都是怪物。相反,这些角色会吸引那些能把情绪反应与战略计算分离开来的人。

当埃隆在 Twitter 进行大规模裁员、在特斯拉突然解雇员工,或公开批评下属时,如果你以为他在玩一场长期的社会游戏,那看起来就是残酷;但如果你意识到他在 Polytopia 里玩的是 Perfection 模式——在那种模式里,任何不能加分的人都只是在浪费回合——那看起来就更像是在优化回合。

Rule 2: Play life like a game.

In Hindu philosophy, there is a concept called Lila, the idea that the universe itself is divine play, a game that reality plays with itself, without purpose beyond the playing.

The Bhagavad Gita instructs Arjuna to fight the battle before him without attachment to outcomes, i.e., to act fully and fiercely within the game while holding the awareness that the game is a game. This is not apathy, but the opposite of apathy. It’s maximum engagement with minimum attachment.

Psychology has a modern version of this. Ethan Kross, a psychologist at the University of Michigan, has spent the last decade studying what happens when people shift from first-person to third-person perspective during high-stakes decisions.

His lab has found, across multiple experiments, that people who narrate their own choices in the third person such as asking "what should Ethan do?" rather than "what should I do?" perform better under stress, make measurably more rational decisions, and experience less emotional interference. The mechanism is what Kross calls "self-distancing.”

Games naturally force this habit. You look down at the board, rather than looking out from inside it. Elon’s detachment during high-stakes moments is the view from above the board, not dumbness or autism.

Shivon Zilis, who knows him better than most, described it with eerie precision when she told Isaacson:

"As a kid, you were playing one of these strategy games and your mom unplugged it, and you just didn't notice, and you kept playing life as if it were that game."

规则 2:把人生当作一场游戏来玩。

在印度教哲学里,有一个概念叫 Lila:宇宙本身就是神圣的游戏,是现实与自身的嬉戏,除了“玩”之外没有更高的目的。

《薄伽梵歌》告诫阿周那,要不执着于结果地去打眼前的仗:也就是说,在游戏之中全力以赴、凶猛行动,同时保持一种意识——这终究是一场游戏。这不是冷漠,恰恰相反:它是“最大程度的投入,最小程度的执着”。

心理学对此有一个现代版本。密歇根大学的心理学家伊森·克罗斯,在过去十年里一直研究:当人们在高风险决策中把视角从第一人称切换到第三人称,会发生什么。

他的实验室在多项实验中发现:那些用第三人称叙述自己选择的人——比如问“伊森应该怎么做?”而不是“我应该怎么做?”——在压力下表现更好,做出可测量地更理性的决定,受到的情绪干扰也更少。其机制被克罗斯称为“自我疏离”。

游戏天然会迫使你养成这种习惯。你是俯视棋盘,而不是从棋盘内部向外看。埃隆在高风险时刻的抽离感,是从棋盘上方看到的视角,不是愚钝,也不是自闭。

对他了解甚深的希冯·齐利斯,在对艾萨克森描述时,以一种诡异的精准说出了这一点:

"小时候,你在玩这类策略游戏,你妈妈把电源拔了,你居然没注意到,你还继续玩——把生活当作那场游戏一样继续玩下去。"

Rule 3: Do not fear losing.

Kahneman and Tversky's prospect theory showed that humans feel losses roughly twice as intensely as equivalent gains. This asymmetry makes most decision-makers risk-averse at precisely the moments they should be risk-seeking.

A single Polytopia game lasts 15-20 minutes. You lose all the time because you expanded too slowly, picked the wrong technology, or got trapped at a chokepoint. Then you start over. What Musk describes as getting used to losing is functionally identical to what clinical psychology calls exposure therapy: repeated controlled contact with an aversive stimulus until the emotional response attenuates.

The emotional machinery that most people use to process failure (the rumination, the self-protection, the retreat into safety) has been deliberately worn smooth through repetition. Loss aversion is a default setting. It’s normal, but Elon overwrote it.

规则 3:不要害怕输。

卡尼曼与特沃斯基的前景理论指出:人类对损失的感受强度,大约是同等收益的两倍。这种不对称使得大多数决策者在最应该冒险的时刻,反而趋于避险。

一局 Polytopia 只有 15–20 分钟。你经常会输——因为扩张太慢、科技点错、或被卡在咽喉要道。然后你从头再来。马斯克所说的“习惯输”,在功能上几乎等同于临床心理学中的 暴露疗法:在可控的情境里反复接触厌恶刺激,直到情绪反应逐渐减弱。

大多数人用来处理失败的情绪机制(反刍、自我保护、退回安全区)被刻意地在重复中磨平了。损失厌恶是一种默认设置;它很正常,但埃隆把它重写了。

Rule 4: Be proactive.

In chess, tempo is the advantage you have when you force your opponent to react to you. In an infinite game, losing tempo is recoverable. You have unlimited moves. In a thirty-turn game, losing even two or three tempos to reactive play can be mathematically insurmountable.

This explains why Elon announces timelines he cannot possibly meet, starts fights he doesn't need, and forces crises inside his own organizations. He is seizing tempo.

In Polytopia, if you spend your first five turns scouting instead of building, you’ve already lost. The map reveals itself through expansion, not exploration.

规则 4:先发制人。

在国际象棋里,节奏(tempo)指的是你迫使对手对你做出反应时所获得的优势。在无限游戏中,失去节奏是可逆的——你有无限步可走。但在一个 30 回合的游戏里,哪怕因为被动应对而丢掉两三个节奏,也可能在数学意义上无法挽回。

这解释了为什么埃隆会宣布他不可能兑现的时间表,挑起他并不需要的冲突,并在自己的组织内部制造危机——他是在夺取节奏。

在 Polytopia 里,如果你前五回合都在侦察而不是建设,你就已经输了。地图是通过扩张而显形的,而不是通过探索。

**Rule 5: Optimize every turn. **

Eliyahu Goldratt introduced the Theory of Constraints in his 1984 book, The Goal. It says: a system’s output is determined by its tightest bottleneck, and resources spent on anything else are a waste.

This is a concept most managers understand intellectually and almost none feel viscerally. When you can see the turn counter ticking from 30 to 29 to 28, waste becomes physically uncomfortable in a way that abstract opportunity cost never does.

This is why Musk sleeps on factory floors during production pushes. It’s not a marketing stunt. He sees a turn counter that executives focused on quarterly cycles don’t.

Quarterly planning is Civilization thinking, where there is always another three-month window to fix things. Elon operates on the assumption that the turn he is on might be the one that determines his final score.

规则 5:优化每一回合。

以利亚胡·戈德拉特在 1984 年的著作 《目标》 中提出了约束理论。它的核心是:一个系统的产出由最紧的瓶颈决定,把资源花在其他任何地方都是浪费。

这是大多数管理者在理智上都懂、但几乎没人能在本能上真正“感到”的概念。当你能看见回合计数从 30 跳到 29 再跳到 28,浪费会在身体层面让人不适——这种不适,是抽象的机会成本永远给不了的。

这就是为什么马斯克会在产能冲刺时睡在工厂地板上。这不是营销噱头。他看到的是一个以季度为周期的高管们看不到的回合计数器。

季度规划是《文明》式思维,因为总还有下一个三个月的窗口可以修正问题。埃隆则是在假设:自己所在的这一回合,可能就是决定最终得分的那一回合。

Rule 6: Double down.

In Polytopia, capturing cities earns you stars, which are the game’s currency.

If you don’t spend those stars immediately on new units, technology, or expansion, you are wasting them. Unspent resources in a game with a turn limit carry a heavy cost because they miss out on the compounding value they would’ve created.

This maps to the Kelly Criterion, developed by John Kelly at Bell Labs in 1956. Kelly Criterion is the mathematically optimal strategy for sizing bets when you have an edge. The Kelly formula demonstrates that under-betting is just as costly as over-betting across a sufficient number of rounds.

Elon reinvests with a pattern that looks reckless from the outside. He puts Tesla profits into new factories before the current ones are even profitable, and he pours SpaceX revenue into Starship before Falcon 9 contracts are finished.

From inside a finite-game frame, hoarding is the reckless move. Every resource sitting idle is a turn wasted during a countdown that never pauses.

规则 6:加倍下注。

在 Polytopia 里,占领城市会给你星星(stars),那是游戏的货币。

如果你不立刻把这些星星花在新单位、科技或扩张上,你就在浪费它们。在有回合上限的游戏里,未花掉的资源代价极高,因为它们错过了本可产生的复利价值。

这对应到约翰·凯利于 1956 年在贝尔实验室提出的凯利准则。凯利准则是在你拥有优势时,为下注规模定量的数学最优策略。凯利公式表明:在足够多的回合里,下注不足与下注过度一样昂贵。

埃隆的再投资模式,从外部看起来近乎鲁莽:他在现有工厂尚未盈利时,就把特斯拉的利润投入新工厂;他在猎鹰 9 的合同尚未结束时,就把 SpaceX 的收入灌进星舰项目。

在有限游戏的框架里,囤积才是鲁莽之举。任何闲置的资源,都是在一个永不暂停的倒计时里白白浪费掉的回合。

Rule 7: Pick your battles.

In Polytopia, you can find yourself surrounded by six or more hostile tribes, all probing your borders simultaneously. The first instinct, especially for aggressive players who have internalized the first six rules, is to fight them all.

If you’re impulsive, it’s a resource bleed. Elite Polytopia players understand that not every city is worth capturing and not every enemy is worth engaging. Some fights are just traps. The skill is identifying which conflicts actually improve your final score and which ones just “feel” productive.

Carl von Clausewitz, the Prussian general whose On War remains the foundational text of Western military strategy, called this the Schwerpunkt, i.e. the decisive point. The concept is that force concentrated at the decisive points wins wars. Napoleon won campaigns by concentrating disproportionate strength at the point where the enemy’s line was weakest.

This reframes some of Elon’s decisions. The fight he did pick, the SEC, advertising boycotts on X, government launch contracts at SpaceX… those are fights over compounding positions, battles where winning changes the board permanently.

In a turn-limited game, every turn spent on a non-decisive battle is a turn stolen from one that actually moves the score. Not all aggression is strategic aggression. Some of it is just noise with a cost.

规则 7:选择你的战场。

在 Polytopia 里,你可能会发现自己被六个甚至更多敌对部落包围,它们同时试探你的边界。第一反应——尤其是那些已经把前六条规则内化、又偏好进攻的玩家——往往是全部开打。

但如果你冲动,那就是资源放血。顶尖的 Polytopia 玩家明白:不是每一座城市都值得占领,也不是每一个敌人都值得交战。有些战斗就是陷阱。真正的技能,在于识别哪些冲突能提高你的最终得分,哪些只是让你“感觉”自己很有效率。

普鲁士将军卡尔·冯·克劳塞维茨在其至今仍是西方军事战略奠基文本的 《战争论》 中,将其称为 Schwerpunkt——即决定性重心。其含义是:把力量集中在决定性点上,才能赢得战争。拿破仑之所以能赢下一场场战役,正是因为他会在敌方战线最薄弱的点上集中压倒性的力量。

这也重新框定了埃隆的一些选择。他确实挑了一些仗来打:与 SEC 的纠葛、X 上的广告抵制、SpaceX 的政府发射合同……这些都是围绕复利位置展开的斗争,是那种“赢了就能永久改变棋盘”的战役。

在回合受限的游戏里,把任何回合花在非决定性战斗上,都是从真正能抬高分数的战斗里偷走回合。不是所有攻击性都是战略性的攻击性;有些只是带着成本的噪音。

Rule 8: Unplug sometimes.

This final rule is the one that doesn’t fit, and that’s precisely what makes it important. Polytopia, for all its intensity, is a short game. When it ends, you put the phone down, and for a moment, you’re outside of the system.

It’s well-known that it’s not the focused, task-positive mind that generates breakthroughs, but the wandering mind, the shower thought, the 3 AM realization, etc.

Archimedes solved the problems of displacement in a bathtub, not at his workbench (Eureka!). Poincaré described the solution to Fuchsian functions arriving unbidden while he was stepping onto a bus.

The turn-limit player who never steps away from the board eventually stops seeing the board clearly. But this rule, by Elon’s own admission, is hardest to follow. If you’re building something, it’s hard to unplug.

There’s a concept called “Trenches & Towers” which I personally live by and will write about in the future. During the week, you're in the trenches of your business: head down, executing, grinding, solving problems at ground level. On the weekend, you climb the tower: you zoom out, see the whole map, and strategize on where the next moves should be. This separation is the point.

This model has a dark side, and it’d be dishonest to ignore it.

Turn-limited thinking makes you treat everything as instrumental.

Relationships are just alliances, employees are units, and public trust is a resource to be spent rather than maintained. The Polytopia player who sacrifices three warriors to capture a city before turn 28 made the right call inside the game. The CEO who burns out an engineering team to hit a ship date made the right call inside his game.

But humans are not units that respawn next match. The damage accumulates across games.

Elon's estranged children, public feuds, serial divorces, and the employees who describe working for him as traumatic are costs that a turn-based framework ignores because they don’t show up on the scoreboard.

I’ve read in Isaacson’s book that Elon once berated an employee so viciously that colleagues were shaken, only for it to emerge afterward that the employee had recently lost an infant daughter. Musk didn’t know. Within the Perfection-mode operating system, it wouldn’t have mattered if he did because employees' performance was the only variable that registered.

The model is powerful precisely because it strips out variables.

Finally, the real lesson I want to bring to the table is not these eight rules, but the meta-lesson underneath them.

Most people don’t have an explicit model for how they make decisions. They operate on inherited defaults such as risk aversion from evolutionary psychology, social conformity from tribal instinct, and loss aversion from the same Kahneman and Tversky research that Elon has functionally trained himself out of.

They're environmental artifacts, shaped by ancestral pressures that have nothing to do with the actual game any of us are playing today.

Musk looked at Polytopia and recognized a formalized version of his own decision-making process. These eight rules aren’t lessons that he acquired from Polytopia, but rather rules he was already using that the game made clear to him.

This is the actual insight, and it has nothing to do with whether you admire him.

It’s also something Jess Lee, the chief product officer at Sequoia, refers to startups as a turn-based game. Which basically means if a big company like Google or Adobe only makes a "move/turn" once a year during their budget cycle, a startup playing "Real-Time" can take 50 turns in that same year.

Everyone is playing some kind of game, but most people have never examined their own rules, turn limits, or their own scoring systems.

You might be playing Civilization when your situation demands Perfection mode. You might be hoarding stars in a game that ends in thirty turns. You might be optimizing for a scoreboard that doesn't match the game you're actually in.

Or you might be playing Perfection-mode in a game that’s actually infinite, strip-mining relationships for throughput in a life that is measured by something other than accumulated points (read: money).

The eight rules do not tell you which game you are in. No ruleset can. But they force a question that most entrepreneurs should ask themselves:

How many turns do you actually have? And are you playing as if you know the answer?

规则 8:有时也要拔掉电源。

最后一条规则是不合群的,而这恰恰是它重要的原因。Polytopia 再怎么紧张,也是一场短游戏。结束时,你把手机放下,短暂地站在系统之外。

众所周知,产生突破的往往不是那种高度聚焦、任务导向的大脑,而是漫游的心智——淋浴时的灵感、凌晨三点的顿悟,诸如此类。

阿基米德是在浴缸里,而不是在工作台前,解决了位移的问题(Eureka!)。庞加莱则描述过:关于福克斯群函数的解法,会在他踏上公交车时不请自来地降临。

一个永远不离开棋盘的“回合上限玩家”,最终会看不清棋盘本身。但按埃隆自己的承认,这条规则最难做到:如果你正在建造某样东西,真的很难拔掉电源。

有个概念叫 “Trenches & Towers”,我个人一直奉行,并且将来会写一篇文章专门谈它。工作日,你在生意的战壕里:埋头执行、苦干、在地面层面解决问题。周末,你爬上高塔:拉远视角,看清整张地图,思考下一步该走哪里。这种分离本身就是重点。

这个模型也有阴暗面,如果忽略它就不诚实了。

回合上限思维会让你把一切都当作工具。

关系只是联盟,员工是单位,公众信任是一种可被消耗的资源,而不是应当被维护的东西。那个在第 28 回合前牺牲三个战士去抢下一座城的 Polytopia 玩家,在游戏内做了正确选择;那个为了赶船期而把一个工程团队烧到崩溃的 CEO,在他自己的游戏里也做了正确选择。

但人不是下一局就会复活的单位。伤害会跨局累积。

埃隆疏离的子女、公开的争端、连番离婚,以及那些把为他工作描述为创伤经历的员工——这些都是一种回合制框架会忽略的成本,因为它们不会出现在记分板上。

我在艾萨克森的书里读到:埃隆曾经把一名员工骂得极其恶毒,以至于同事们都被震住;而事后才发现,那名员工不久前刚失去一个襁褓中的女儿。马斯克并不知道。在 Perfection 模式的操作系统里,即使他知道也无关紧要,因为唯一会被登记的变量,是员工的绩效表现。

这个模型之所以强大,恰恰因为它剥离了变量。

最后,我真正想摆上桌面的教训,不是这八条规则,而是它们之下的元教训。

大多数人并没有一套清晰的“我如何做决策”的显性模型。他们在一些继承来的默认设置上运作:来自进化心理学的风险规避,来自部落本能的社会从众,以及来自卡尼曼与特沃斯基同一套研究的损失厌恶——而埃隆则在功能上把自己训练得脱离了这些。

它们只是环境的副产物,由与我们今天任何人所处的真实游戏毫不相关的祖先压力塑造而成。

马斯克看着 Polytopia,看见了他自己的决策过程被形式化后的版本。这八条规则不是他从 Polytopia 学来的教训,而是他原本就在用、只是被游戏照亮得更清楚的规则。

这才是关键洞见——它与你是否崇拜他毫无关系。

这也与红杉的首席产品官 Jess Lee 的一个说法相呼应:她把创业公司称作一种回合制游戏。意思大致是:如果像 Google 或 Adobe 这样的巨头,在预算周期里一年只做一次“动作/回合”,那么一家以“实时”方式运作的初创公司,在同一年里可以走 50 个回合。

每个人都在玩某种游戏,但大多数人从未审视过自己的规则、回合上限,或自己的计分方式。

你可能在玩 《文明》,而你的处境真正需要的是 Perfection 模式。你可能在一场 30 回合就结束的游戏里囤着星星。你可能在为一个与你正在玩的游戏并不匹配的记分板做优化。

又或者,你在一场事实上是无限的游戏里玩着 Perfection 模式——为了吞吐量而开采、榨干关系,在一种用“累积点数”(也就是钱)之外的尺度来衡量的生活里。

这八条规则不会告诉你自己身处哪一种游戏。任何规则集都做不到。但它们会迫使大多数创业者问自己一个问题:

你实际上还有多少回合?而你是否在像自己知道答案那样去玩?

In competitive Polytopia, you have exactly 30 turns.

There are no extensions or do-overs. Every turn you spend exploring instead of expanding is a turn you don’t get back. The game is indifferent to your strategy or how well you planned your moves. Once the counter hits zero, only what you actually built matters.

Everything else is irrelevant.

Elon Musk became obsessed with Polytopia while simultaneously running Tesla, SpaceX, Neuralink, xAI, The Boring Company, and X. He’s beaten the creator of the game. He loves it so much that at one point, Polytopia was added as a playable game to Tesla cars.

Most people, even I, at first, dismissed this as a quirky hobby or a strange eccentricity.

But they missed the point. Polytopia is more than a game to Musk. Polytopia is a blueprint for his entire worldview, a perfect miniature of the operating system running underneath everything he does.

And the eight rules he took from the game reveal more about how he actually thinks than any biography, interview, or earnings call ever has.

Most 4X strategy games such as Civilization, Age of Empires, and Stellaris, give you an open horizon. You play until you win or lose. Time stretches. There is always another turn, another age, or another technology. These games teach patience, accumulation, and the long view. Put simply, they teach you to survive.b

Polytopia's competitive Perfection mode inverts all those instincts. You get 30 turns. Your score is the sum of everything you've accumulated when the clock stops. This single constraint inverts every instinct the genre trains you on.

Expansion beats defense, speed is better than caution, and any resource sitting idle is a resource rotting on the vine. Most strategy games teach you to survive, whereas Polytopia teaches you that survival is irrelevant. Only your output before the deadline counts.

Here’s how Elon implements it in real life (told by Walter Isaacson who spent a year shadowing Elon and wrote his biography):

This is the structural gap between how Musk operates and how virtually every other CEO operates. They are playing Civilization while he is playing Perfection mode.

In 1986, NYU religious studies professor James Carse published Finite and Infinite Games. He argued there are only two types of games: finite games, which are played to win, and infinite games, which are played to keep the game going. Simon Sinek turned this into a bestselling business philosophy in 2019, suggesting that great companies should play infinite games and focus on longevity.

Elon never internalized any of it. SpaceX wasn’t built to just be a sustainable aerospace company, but to make humanity multiplanetary before a perceived window of opportunity closes. Tesla wasn’t created to sell cars indefinitely, but to force an energy transition before a climate deadline. Polytopia didn’t teach him these things, but it reflected his existing perspective. It showed him a mirror.

The rules Elon took from the game, as Walter Isaacson’s biography catalogued them, are worth looking at individually, because doing so lets us diagnose how he thinks.

Rule 1: Empathy is not an asset.

This sounds sociopathic stripped of context. In Polytopia, however, the game rewards modeling your opponent's optimal moves, not their feelings. Musk's brother Kimbal put it plainly in Isaacson's biography. Kimbal told Isaacson he has an "empathy gene" that has hurt him in business, and that Elon told him to play Polytopia to understand how strategy works when you strip that gene out.

In Polytopia, your opponents aren't people, but optimization problems.

The game rewards you for modeling their strongest possible moves and playing against those, not for modeling their feelings. If an opponent overextends, you exploit the gap. If they leave a city undefended, you take it. There is no penalty for ruthlessness, only for hesitation.

This is not a new idea in strategy. When DeepMind built AlphaGo, the system that defeated world champion Lee Sedol in 2016, the breakthrough was programming the machine to model Sedol's optimal responses and play against the best version of him.

This is what game theorists call "theory of mind" stripped to its purest, most adversarial form.

Kevin Dutton's research on psychopathic traits across professional populations found that CEOs rank alongside surgeons at the top of the spectrum. This doesn’t mean these individuals are monsters. Instead, these roles attract people who can separate emotional responses from strategic calculations.

When Elon carries out mass layoffs at Twitter, fires staff abruptly at Tesla, or publicly criticizes subordinates, it looks like cruelty if you think he is playing a long-term social game. It looks like turn optimization if you realize he is playing Perfection mode in Polytopia, where anyone not adding to the score is simply wasting a turn.

Rule 2: Play life like a game.

In Hindu philosophy, there is a concept called Lila, the idea that the universe itself is divine play, a game that reality plays with itself, without purpose beyond the playing.

The Bhagavad Gita instructs Arjuna to fight the battle before him without attachment to outcomes, i.e., to act fully and fiercely within the game while holding the awareness that the game is a game. This is not apathy, but the opposite of apathy. It’s maximum engagement with minimum attachment.

Psychology has a modern version of this. Ethan Kross, a psychologist at the University of Michigan, has spent the last decade studying what happens when people shift from first-person to third-person perspective during high-stakes decisions.

His lab has found, across multiple experiments, that people who narrate their own choices in the third person such as asking "what should Ethan do?" rather than "what should I do?" perform better under stress, make measurably more rational decisions, and experience less emotional interference. The mechanism is what Kross calls "self-distancing.”

Games naturally force this habit. You look down at the board, rather than looking out from inside it. Elon’s detachment during high-stakes moments is the view from above the board, not dumbness or autism.

Shivon Zilis, who knows him better than most, described it with eerie precision when she told Isaacson:

"As a kid, you were playing one of these strategy games and your mom unplugged it, and you just didn't notice, and you kept playing life as if it were that game."

Rule 3: Do not fear losing.

Kahneman and Tversky's prospect theory showed that humans feel losses roughly twice as intensely as equivalent gains. This asymmetry makes most decision-makers risk-averse at precisely the moments they should be risk-seeking.

A single Polytopia game lasts 15-20 minutes. You lose all the time because you expanded too slowly, picked the wrong technology, or got trapped at a chokepoint. Then you start over. What Musk describes as getting used to losing is functionally identical to what clinical psychology calls exposure therapy: repeated controlled contact with an aversive stimulus until the emotional response attenuates.

The emotional machinery that most people use to process failure (the rumination, the self-protection, the retreat into safety) has been deliberately worn smooth through repetition. Loss aversion is a default setting. It’s normal, but Elon overwrote it.

Rule 4: Be proactive.

In chess, tempo is the advantage you have when you force your opponent to react to you. In an infinite game, losing tempo is recoverable. You have unlimited moves. In a thirty-turn game, losing even two or three tempos to reactive play can be mathematically insurmountable.

This explains why Elon announces timelines he cannot possibly meet, starts fights he doesn't need, and forces crises inside his own organizations. He is seizing tempo.

In Polytopia, if you spend your first five turns scouting instead of building, you’ve already lost. The map reveals itself through expansion, not exploration.

**Rule 5: Optimize every turn. **

Eliyahu Goldratt introduced the Theory of Constraints in his 1984 book, The Goal. It says: a system’s output is determined by its tightest bottleneck, and resources spent on anything else are a waste.

This is a concept most managers understand intellectually and almost none feel viscerally. When you can see the turn counter ticking from 30 to 29 to 28, waste becomes physically uncomfortable in a way that abstract opportunity cost never does.

This is why Musk sleeps on factory floors during production pushes. It’s not a marketing stunt. He sees a turn counter that executives focused on quarterly cycles don’t.

Quarterly planning is Civilization thinking, where there is always another three-month window to fix things. Elon operates on the assumption that the turn he is on might be the one that determines his final score.

Rule 6: Double down.

In Polytopia, capturing cities earns you stars, which are the game’s currency.

If you don’t spend those stars immediately on new units, technology, or expansion, you are wasting them. Unspent resources in a game with a turn limit carry a heavy cost because they miss out on the compounding value they would’ve created.

This maps to the Kelly Criterion, developed by John Kelly at Bell Labs in 1956. Kelly Criterion is the mathematically optimal strategy for sizing bets when you have an edge. The Kelly formula demonstrates that under-betting is just as costly as over-betting across a sufficient number of rounds.

Elon reinvests with a pattern that looks reckless from the outside. He puts Tesla profits into new factories before the current ones are even profitable, and he pours SpaceX revenue into Starship before Falcon 9 contracts are finished.

From inside a finite-game frame, hoarding is the reckless move. Every resource sitting idle is a turn wasted during a countdown that never pauses.

Rule 7: Pick your battles.

In Polytopia, you can find yourself surrounded by six or more hostile tribes, all probing your borders simultaneously. The first instinct, especially for aggressive players who have internalized the first six rules, is to fight them all.

If you’re impulsive, it’s a resource bleed. Elite Polytopia players understand that not every city is worth capturing and not every enemy is worth engaging. Some fights are just traps. The skill is identifying which conflicts actually improve your final score and which ones just “feel” productive.

Carl von Clausewitz, the Prussian general whose On War remains the foundational text of Western military strategy, called this the Schwerpunkt, i.e. the decisive point. The concept is that force concentrated at the decisive points wins wars. Napoleon won campaigns by concentrating disproportionate strength at the point where the enemy’s line was weakest.

This reframes some of Elon’s decisions. The fight he did pick, the SEC, advertising boycotts on X, government launch contracts at SpaceX… those are fights over compounding positions, battles where winning changes the board permanently.

In a turn-limited game, every turn spent on a non-decisive battle is a turn stolen from one that actually moves the score. Not all aggression is strategic aggression. Some of it is just noise with a cost.

Rule 8: Unplug sometimes.

This final rule is the one that doesn’t fit, and that’s precisely what makes it important. Polytopia, for all its intensity, is a short game. When it ends, you put the phone down, and for a moment, you’re outside of the system.

It’s well-known that it’s not the focused, task-positive mind that generates breakthroughs, but the wandering mind, the shower thought, the 3 AM realization, etc.

Archimedes solved the problems of displacement in a bathtub, not at his workbench (Eureka!). Poincaré described the solution to Fuchsian functions arriving unbidden while he was stepping onto a bus.

The turn-limit player who never steps away from the board eventually stops seeing the board clearly. But this rule, by Elon’s own admission, is hardest to follow. If you’re building something, it’s hard to unplug.

There’s a concept called “Trenches & Towers” which I personally live by and will write about in the future. During the week, you're in the trenches of your business: head down, executing, grinding, solving problems at ground level. On the weekend, you climb the tower: you zoom out, see the whole map, and strategize on where the next moves should be. This separation is the point.

This model has a dark side, and it’d be dishonest to ignore it.

Turn-limited thinking makes you treat everything as instrumental.

Relationships are just alliances, employees are units, and public trust is a resource to be spent rather than maintained. The Polytopia player who sacrifices three warriors to capture a city before turn 28 made the right call inside the game. The CEO who burns out an engineering team to hit a ship date made the right call inside his game.

But humans are not units that respawn next match. The damage accumulates across games.

Elon's estranged children, public feuds, serial divorces, and the employees who describe working for him as traumatic are costs that a turn-based framework ignores because they don’t show up on the scoreboard.

I’ve read in Isaacson’s book that Elon once berated an employee so viciously that colleagues were shaken, only for it to emerge afterward that the employee had recently lost an infant daughter. Musk didn’t know. Within the Perfection-mode operating system, it wouldn’t have mattered if he did because employees' performance was the only variable that registered.

The model is powerful precisely because it strips out variables.

Finally, the real lesson I want to bring to the table is not these eight rules, but the meta-lesson underneath them.

Most people don’t have an explicit model for how they make decisions. They operate on inherited defaults such as risk aversion from evolutionary psychology, social conformity from tribal instinct, and loss aversion from the same Kahneman and Tversky research that Elon has functionally trained himself out of.

They're environmental artifacts, shaped by ancestral pressures that have nothing to do with the actual game any of us are playing today.

Musk looked at Polytopia and recognized a formalized version of his own decision-making process. These eight rules aren’t lessons that he acquired from Polytopia, but rather rules he was already using that the game made clear to him.

This is the actual insight, and it has nothing to do with whether you admire him.

It’s also something Jess Lee, the chief product officer at Sequoia, refers to startups as a turn-based game. Which basically means if a big company like Google or Adobe only makes a "move/turn" once a year during their budget cycle, a startup playing "Real-Time" can take 50 turns in that same year.

Everyone is playing some kind of game, but most people have never examined their own rules, turn limits, or their own scoring systems.

You might be playing Civilization when your situation demands Perfection mode. You might be hoarding stars in a game that ends in thirty turns. You might be optimizing for a scoreboard that doesn't match the game you're actually in.

Or you might be playing Perfection-mode in a game that’s actually infinite, strip-mining relationships for throughput in a life that is measured by something other than accumulated points (read: money).

The eight rules do not tell you which game you are in. No ruleset can. But they force a question that most entrepreneurs should ask themselves:

How many turns do you actually have? And are you playing as if you know the answer?

📋 讨论归档

讨论进行中…