特斯拉财报电话会 2016–2018
下一篇特斯拉财报电话会 2019–2021- 活动/形式: 特斯拉 2016–2018 年的季度财报电话会,共十二场(2016 年第一至第四季度、2017 年第一至第四季度、2018 年第一至第四季度),召开时间从 2016 年 5 月到 2019 年 1 月。这是 Model 3 时代:电话会从 Model X 收尾阶段的第一轮生产地狱和SolarCity 合并(2016 年)开始,经过 Model 3 上市及其“制造地狱”和 S 曲线爬坡(2017 年),一直到“资金已到位”风波、对过度自动化的纠正、第一个可持续盈利的季度,以及对成本毫不留情的收尾(2018 年)。每场电话会都是有多位发言人的投资者活动:投资者关系负责人开场,埃隆·马斯克(董事长兼首席执行官)做陈述并回答问题,首席财务官(2016 年是杰森·惠勒,之后是迪帕克·阿胡贾)、首席技术官 JB·斯特劳贝尔和其他高管轮流接话,卖方分析师提问。
- 时代脉络: 如果说 2013–2015 年是那些持久的思维模型首次成熟定型的时期,那么 2016–2018 年就是它们被实时压力测试的时期:马斯克在职业生涯最严峻的运营压力下,讲述自己的所思所想。最有代表性的一条线,是工厂就是产品的论点。它从一句话(“工厂将是比汽车本身更重要的产品”,2016 年第四季度)逐渐硬化为完整阐述的胭脂河工厂学说(“工厂……那是胭脂河工厂……这才是特斯拉真正的长期竞争优势”,2017 年第四季度),然后在 2018 年自我纠正(“我们在自动化上确实走得太远了”)。与此同时:S 曲线生产模型在 2017 年诞生并不断完善;自动驾驶时间表变得空前激进(“离 L5 只差一次软件升级”,2017 年第一季度),随后又因横穿美国的期限一再落空而一步步收回;“生产地狱”这个词被造出来,也被亲身经历;他的脾性也赤裸裸地展露出来:“AI 让我害怕极了”,“我当时真的很抑郁”,“蠢问题”发作,以及一个季度后的道歉。思维材料就在这些线索跨越十二个有日期的数据点的变化之中。
- 可信度: 已核实。每份原始材料都是 stockanalysis.com 上的完整文字稿(
verified: true)。十二份中有十份带有明确的Elon Musk -- Chairman and CEO, Tesla发言人标注;2017 年第三季度和 2018 年第二季度的文字稿没有标注(各段发言只是一段段文字),所以在这两场里,是根据明确无误的第一人称特征、标志性措辞和问答中的回答位置来认定马斯克(首席执行官做开场独白,回答战略和愿景问题;首席财务官和首席技术官接话的部分,比如“埃隆,你描述得非常好”和财务摘要,一眼就能分辨,不予引用)。只引用马斯克的话,不引用首席财务官或斯特劳贝尔的发言,也不引用分析师的提问。 - 引语出处(按季度锚定的约定): 下面每段引语都是所属季度原始文字稿去掉多余空白后的逐字子串(这是权威的核验标准);原始文字稿常在句子中间用双换行断开,所以凡是跨段落的引语,都用
...省略号表示,而不是悄悄拼接。每段引语都用#:~:text=片段锚定到该季度的 stockanalysis.com 文字稿页面。每个片段都不含撇号(,→%2C、-→%2D),解码后是引语的原文子串;stockanalysis.com 是在客户端加载文字稿正文的,所以浏览器里未必总能实时高亮。这里的保证是与原始材料的逐字子串匹配,而且每个链接都指向正确的季度(十二份文字稿之间张冠李戴,是本页要防范的首要风险)。不用原始文件路径。
概要
2016–2018 年的财报电话会,是特斯拉相关记录里心理上最赤裸的一段:马斯克几乎是实时地讲述自己如何熬过 Model 3 的“生产地狱”。每场电话会的大部分内容是财务、业绩指引和生产细节,这些要么在正文中交代,要么略去。作为思维材料留下来的,是他那些持久信念在极限压力下、跨越十二个有日期的数据点所发生的变化。
这个时代的主干是工厂就是产品的论点。2016 年,它以一句制造业信条开场(“一门心思要成为地球上最好的制造商”;“制造上的创新潜力……远比汽车设计上的大”),还给那台制造机器的机器起了个俏皮的名字:“外星无畏舰”。到 2016 年第四季度,它变成了“工厂将是比汽车本身更重要的产品”,还配上了一个跨领域的妙喻(“把火箭方程用到制造上”)。整个 2017 年,这个论点不断加深:“制造机器的机器……其他制造商很难复制”(第一季度),“速度是终极武器……如果你能看见机器人在动,那它就太慢了”(第三季度)。到 2017 年第四季度,它结晶为完整的胭脂河工厂学说:T 型车不是产品,工厂才是。随后,2018 年给出了这个时代最清楚的思想演变标志:“我们在自动化上确实走得太远了,把一些相当愚蠢的东西也自动化了”(第一季度)。这是对过度自动化的纠正,后来又被归纳为“提高制造效率的途径是速度和密度”,以及“大规模制造主要是一个软件问题”。
第二条线是 S 曲线生产模型,它完全是在这个时代诞生并完善的。它最早出现在 2017 年第一季度(难以预测的“S 曲线的起始部分”),在 2017 年第二季度变得更犀利(它会“因为有东西坏了而倒退”,还有那句坦率得让人卸下防备的“我们……希望随着时间推移少犯点蠢”),到 2017 年第三季度被提炼成一个关于认知偏差的观察:“人的直觉往往是直线外推”。
第三条是自动驾驶时间表,这是这个时代最可证伪的演变。2016 年,他说“完全自动驾驶来得会比任何人想的都快得多”,还给出了道德上的理由(“不允许自动驾驶在道德上是错的”);2017 年第一季度,他提出了最激进的说法(“只要升级软件,我们就能实现 L5”);然后是一再落空的横穿美国期限(“丢人现眼”,2017 年第二季度→“三个月,最多六个月”,2017 年第四季度→“钻系统的空子”,2018 年第二季度),反激光雷达的“拐杖/局部最优”信念(2017 年第四季度),AlphaGo 的类比,以及车队数据构成的护城河(“路上跑着几百万辆车”,2018 年第三季度;“特斯拉里程数的 5%”,2018 年第四季度)。2018 年第四季度,一切又回到了“今年年底前后”,为他在自动驾驶上的乐观画上了一个完美的首尾呼应。
这个时代的他也异常情绪外露。“生产地狱”这个词被造了出来(“头六个月我们都在生产地狱里”;“心理上的疤痕组织”;“我个人大概折了一年寿……一直守在弗里蒙特工厂”),被分出地狱的层级(“第 9 层是最糟的。我们当时就在第 9 层”),还被坦白为抑郁(“大概三四周前我真的很抑郁”)。他对 AI 的恐惧说得直截了当(“AI 让我害怕极了”),也有细腻的分辨(“洞察和监督是两回事”)。他的脾气还在那次臭名昭著的“蠢问题可不酷”打发中浮出水面(2018 年第一季度),一个季度后他又道了歉(“无礼真的没有任何借口”),这本身就是一个演变标志。使命被重申为“特斯拉存在的理由”(“加速可持续能源……关乎人类的生存风险”,2018 年第四季度),并与平价绑在一起(“造出人人都买得起的电动车”;谈到成本时说“我们在这件事上必须是绝对的狂热分子”)。整个 2018 年,一种资本纪律的信念反复出现(“你不盈利就不是一家真正的公司”;“我们在任何时候都不会进行股权融资”),而后来几年证明,这个信念并不稳固。
语气说明:本百科把这些记录为马斯克在 2016–2018 年各个数据点上表达的观点、预测和信心姿态,不作裁断。其中有几条是有日期、可证伪的预测:靠软件升级实现 L5 的说法,一个接一个的横穿美国自动驾驶期限,“今年年底”的 FSD 时间表,“不进行股权融资”的承诺。它们都作为公开的预测和信心姿态记录下来,既不背书也不反驳,恰恰因为是乐观情绪和时间表的数据点而有用(自动驾驶的时间表都没有兑现;2019 年随即进行了股权融资)。所有财务数字、SolarCity 合并和“资金已到位”的具体经过,以及产品和工程参数,都只在正文中交代,从不作为事实放进引语块。
关键引语(逐字引用,按季度锚定文字稿,仅限埃隆·马斯克本人)
工厂就是产品:从“外星无畏舰”到胭脂河工厂(2016–2017)
这是这个时代的主干。2016 年,它以一种公开宣称的身份开场,特斯拉被重新定位为一家制造公司,并坚信创新所在是制造,而不是设计:
“I think Tesla is gonna be hell-bent on becoming the best manufacturer on Earth. You know, thus far, I think we’ve done a good job on design and technology of our products.” ↗
“我觉得特斯拉会一门心思要成为地球上最好的制造商。你知道,到目前为止,我觉得我们在产品的设计和技术上做得不错。”
“We believe that manufacturing technology is itself subject to a tremendous amount of innovation. In fact, we believe that there’s more potential for innovation in manufacturing than there is in the design of the car, by a long shot.” ↗
“我们相信,制造技术本身就有巨大的创新空间。事实上,我们认为制造上的创新潜力远比汽车设计上的大,大得多。”
2016 年第二季度,那台制造机器的机器有了一个俏皮的内部代号,也得出了一条关于如何把新技术规模化的普遍教训:
“the internal name for the designing the machine that makes the machine is, we call it the Alien Dreadnought. At the point at which the factory looks like an Alien Dreadnought, then you know you’ve won.” ↗
“设计那台制造机器的机器,内部代号是,我们叫它‘外星无畏舰’。等工厂看起来像一艘外星无畏舰的时候,你就知道你赢了。”
“If it’s cutting-edge technology, it’s really hard to scale up production because you gotta design the machine that makes the machine, not just the machine itself.” ↗
“如果是尖端技术,扩大生产就真的很难,因为你得设计那台制造机器的机器,而不只是机器本身。”
到 2016 年第四季度,这个论点达到了最强的形态:工厂的地位高于汽车。支撑它的是一个跨领域的物理类比:
“I’ve refocused most of Tesla engineering, including design engineering, into designing the factory. I think in the future, the factory will be a more important product than the car itself.” ↗
“我已经把特斯拉的大部分工程力量,包括设计工程,重新集中到设计工厂上。我觉得将来,工厂会是比汽车本身更重要的产品。”
“It’s sort of just applying the rocket equation to manufacturing. You know, essentially, well, the rocket equation is taking its mass efficiency, but, like volumetric efficiency of the factory, as Jason was mentioning, and exit velocity of product from the factory.” ↗
“这差不多就是把火箭方程用到制造上。你知道,本质上,嗯,火箭方程讲的是质量效率,而这里讲的是工厂的体积效率,就像杰森刚才提到的,还有产品离开工厂的出口速度。”
2017 年第一季度,这个论点变成了一条竞争护城河的论证:制造机器的机器主要是软件,而且很难复制。
“The set of steps necessary to achieve that outcome seem pretty obvious. Heavily involve Tesla getting incredibly good at the machine that builds the machine. Which involves, by the way, a tremendous amount of software, right? It’s not just a bunch of robots that are sitting there. It’s the programming of the robots and how they interact and it’s far more complex than the software in the car. I mean, I think this is just gonna be a very difficult thing for other manufacturers to copy. I would not know what to do if I were in that position.” ↗
“要达到这个结果,需要哪些步骤似乎很明显。关键在于特斯拉要把那台制造机器的机器做到极致。顺便说一句,这涉及大量的软件,对吧?不只是一堆机器人摆在那儿,而是机器人的编程以及它们如何相互配合,这比车里的软件复杂得多。我是说,我觉得这对其他制造商来说会非常难以复制。换我处在他们的位置,我都不知道该怎么办。”
到 2017 年第三季度,这套学说有了一句口号,速度是终极武器,还被推到了生动的极致:
“It’s remarkable how much can be done by just speeding up robots, shortening the path, densifying the factory, adding additional robots to choke points and just making lines go really, really fast… Speed is the ultimate weapon.” ↗
“光是让机器人加速、缩短路径、让工厂更紧凑、在瓶颈处加机器人、让生产线跑得非常非常快,就能做成这么多事,真是了不起……速度是终极武器。”
“I think speed is the ultimate weapon when it comes to innovation or production… It’s like, if you can see the robot move, it’s too slow.” ↗
“我觉得无论是创新还是生产,速度都是终极武器……就好比,如果你能看见机器人在动,那它就太慢了。”
到 2017 年第四季度,它结晶为胭脂河工厂学说,完整地表达了工厂而非汽车才是持久优势的信念;与之并列的,还有“拄助行器的老奶奶”这个第一性原理式的速度比较:
“The competitive strength of Tesla long term is not gonna be the car. It’s gonna be the factory. We’re gonna productize the factory. Really, this is a lesson that is kind of obvious in history because the Model T wasn’t the product, it was River Rouge. The Model T was a very simple car. Anybody could have made that car… Not anyone could make Rouge River. That’s really what will be Tesla’s long-term competitive advantage.” ↗
“特斯拉长期的竞争力不会是汽车,而是工厂。我们要把工厂产品化。其实这是一个历史上挺明显的教训,因为 T 型车不是产品,胭脂河工厂才是。T 型车是一款非常简单的车,谁都造得出来……可不是谁都能造出胭脂河工厂。这才是特斯拉真正的长期竞争优势。”
“The most fundamental difference is thinking about the factory as really as a product, as a quite vertically integrated product.” ↗
“最根本的区别在于,真正把工厂当作一个产品来思考,一个相当垂直整合的产品。”
“Like grandma with a walker can exceed the speed of the fastest production line on Earth. Really not that fast. Walking speed is 1 meter per second, so 5 x faster than the fastest production line on Earth.” ↗
“就好比,拄着助行器的老奶奶都能比地球上最快的生产线还快。其实并不怎么快。步行速度是每秒 1 米,所以比地球上最快的生产线快 5 倍。”
对过度自动化的纠正:这个时代的思想演变标志(2018)
随后,这套学说自我纠正了。这是这个时代里他思想更新最清楚的一次记录:把自动化和速度推到极限之后,马斯克公开掉头。这也是他后来“人被低估了”那一观点的种子:
“However, as I mentioned in a tweet, a few months ago, we did go too far on the automation front and automated some pretty silly things.” ↗
“不过,就像我几个月前在一条推文里说的,我们在自动化上确实走得太远了,把一些相当愚蠢的东西也自动化了。”
纠正后的信念被推而广之:效率来自速度和密度,而不是资本开支,而且要靠倾听一线工人:
“However, that said, I do believe that the path to manufacturing efficiency is velocity and density. That is absolutely what we’ll be working on. Rather than just trying to, you know, spend billions of dollars on duplicating a factory, if you can make Like, if two companies are competing and one has to double its CapEx in order to double production, and the other one can, with minor CapEx, can just speed up the line by double, it’s a game over.” ↗
“不过话虽如此,我确实认为提高制造效率的途径是速度和密度。这绝对是我们要努力的方向。与其,你知道,花几十亿美元再复制一座工厂,如果你能让……比如说,两家公司在竞争,一家要把资本开支翻倍才能把产量翻倍,而另一家只花一点资本开支,就能把生产线速度提高一倍,那就胜负已分了。”
“I think in general, our understanding of production is improving dramatically, exponentially in fact. We are seeing ways to achieve improved volume with dramatically less CapEx by simplifying the production line by really engaging all of our associates, no matter how junior, in improving the way the parts are made. It’s amazing how everybody’s got good ideas, just need to solicit those ideas and implement them.” ↗
“我觉得总体上,我们对生产的理解正在大幅提升,事实上是指数级的提升。我们正在找到办法,用少得多的资本开支提高产量:简化生产线,真正让每一位员工,不管资历多浅,都参与到改进零件制造方式中来。真是了不起,每个人都有好点子,只要去征求这些点子,然后落实就行。”
到 2018 年第二季度,纠正后的模型得到了完整的重述:手动与自动之间的平衡,工厂是一个受瓶颈制约的系统,以及“生产是一个软件问题”:
“What we found as we’ve spent a lot of time debugging wide range of manufacturing issues that the potential for our existing lines to be able to produce far more cars is much greater than expected. That by simplifying production lines, by speeding them up, by in some cases, having been done manual instead of automatic and in other cases, having been done automatic instead of manual, we’ve been able to achieve dramatic improvements to the output of existing lines” ↗
“我们花了大量时间排查各种各样的制造问题,结果发现,现有生产线多产很多车的潜力比预想的大得多。通过简化生产线、给生产线提速,有些地方把自动改成手动,另一些地方把手动改成自动,我们让现有生产线的产量大幅提高了”
“It’s like basically a production system is like a giant cybernetic collector. And it’s and then it moves as fast as the slowest part.” ↗
“基本上,一个生产系统就像一个巨大的控制论集合体。然后它的速度,取决于最慢的那个部分。”
“Yes. In fact, it’s amazing how much of production is actually software. We’re really quite good at software relative to other car companies and manufacturing at volume is mostly a software problem. I think that is not well appreciated.” ↗
“是的。事实上,生产里有多大一部分其实是软件,这真让人吃惊。和其他汽车公司比,我们的软件确实做得相当好,而大规模制造主要就是一个软件问题。我觉得这一点没有得到充分的认识。”
这个教训被内化为一条面向制造的设计原则:让设计师直接感受到生产的痛苦。
“I’m really excited about Model Y manufacturing and the design for manufacturing. Like, essentially, how do we design out all the pain that we’re currently going through? We do not wanna experience it again.” ↗
“Model Y 的制造和面向制造的设计让我非常兴奋。本质上就是,我们怎样在设计阶段就把眼下经历的所有痛苦都消除掉?我们可不想再经历一遍。”
“And it’s given them tremendous insight into how they need to change the designs in the future to make it easier to produce because you feel the pain directly. Yes, once you feel the pain, like, okay, I didn’t realize I was like torturing people with my terrible design. Now I know.” ↗
“这让他们深刻认识到,将来该怎样修改设计,好让生产更容易,因为你直接感受到了痛苦。是的,一旦你感受到了痛苦,就会想,好吧,我之前没意识到我那糟糕的设计简直是在折磨人。现在我知道了。”
“It is remarkable, like, although the amount of money spent in the beginning is really quite low, at the beginning of a development program, decisions made at the beginning of development program have massive implications for future CapEx. It is better to spend a bit more time making the right design decisions and really thinking through the producibility of a product before racing ahead with CapEx decisions. There’s no question we could have made the Model 3 much easier to produce than we have.” ↗
“这很值得注意:虽然开发项目刚开始时花的钱其实很少,但项目初期做出的决定,对未来的资本开支影响巨大。与其急着做资本开支的决定,不如多花点时间做出正确的设计决策,把产品的可生产性真正想透。毫无疑问,我们本可以让 Model 3 比现在容易生产得多。”
S 曲线生产模型:诞生与完善(2017)
这是一个完全存在于这个时代内部的思维模型。它最早出现在 2017 年第一季度,描述的是新产品爬坡那条难以预测的曲线:
“The trick with the when you’ve got a whole new product and a whole new factory is, trying to predict exactly what that initial portion of the S-curve looks like is extremely difficult. Inevitably, the production starts off slowly, and then you gradually eliminate the constraints, and eventually it starts taking off exponentially. Because of that sort of initial slow ramp that then grows exponentially, a small change in where that lands in a quarter can have quite a big impact on total volume.” ↗
“难就难在,当你有一款全新的产品和一座全新的工厂时,要准确预测 S 曲线起始那段是什么样子,是极其困难的。生产不可避免地开始得很慢,然后你逐步消除各种限制,最终它开始呈指数级起飞。正因为开头是缓慢爬坡,随后呈指数增长,所以它在一个季度里落在哪个点上,哪怕只差一点,对总产量的影响也会相当大。”
2017 年第二季度,这个模型得到了完善:曲线还可能倒退。他还坦率得让人卸下防备,承认自己也不知道答案:
“You know, when we make mistakes, it’s because we’re stupid, not because we’re trying to mislead anyone. I just want to emphasize, you know, we sort of aspire to be less dumb over time. So if I knew it, I would tell you… It’s just fundamentally impossible to predict the exponential part of the manufacturing S-curve. It’s crazy hard.” ↗
“你知道,我们犯错,是因为我们蠢,而不是因为我们想误导谁。我只想强调一下,你知道,我们算是希望随着时间推移少犯点蠢。所以要是我知道,我会告诉你的……制造 S 曲线里的指数部分,从根本上就是没法预测的。难得要命。”
“The S-curve is a simplification, because it’s really running through a series of constraints. It’s just like a really jagged sort of upward growth, and it’ll plateau, and then it’ll grow rapidly, and it’ll plateau again, and then sometimes it’ll go backwards because something’s broke… Signed up for it. Not blaming hell because we bought the ticket.” ↗
“S 曲线是一种简化,因为实际上是在一个接一个地突破限制。它就像一种非常参差不齐的上升,会进入平台期,然后快速增长,又进入平台期,有时候还会因为有东西坏了而倒退……这是我们自己报名参加的。不怪地狱,因为票是我们自己买的。”
到 2017 年第三季度,它被提炼成一个关于认知偏差的观察:为什么人们(包括投资者)一再看错特斯拉。
“human intuition tends to be a straight line extrapolation, but we’re really on a very steep exponential. So it’s really an S curve.” ↗
“人的直觉往往是直线外推,但我们其实走在一条非常陡峭的指数曲线上。所以它其实是一条 S 曲线。”
同样的指数框架,在这个时代的结尾成了特斯拉的自画像(2018 年第三季度),也被他拿来回敬怀疑者(2017 年第三季度):
“when you have an exponential growth rate like we do, I mean, if you, if you look at Tesla cumulative deliveries over time, that’s like the cleanest exponential curve fit that I’ve ever seen.” ↗
“当你有我们这样的指数增长率时,我是说,如果你,如果你看特斯拉历年的累计交付量,那简直是我见过的拟合得最干净的指数曲线。”
“for the skeptics out there, I’d like to say, ask them, which one of you predicted that Tesla would go from 2,500 units delivered to 250,000 units delivered now? I suspect the answer is 0.” ↗
“对外面那些怀疑者,我想说,问问他们:你们当中谁预测到特斯拉会从交付 2,500 辆走到现在交付 25 万辆?我猜答案是 0。”
“So consider your assumptions for the future and whether they are valid or perhaps pessimistic.” ↗
“所以,想想你们对未来的假设,看看它们是否站得住,还是也许太悲观了。”
自动驾驶时间表:这个时代最可证伪的演变(2016–2018)
这是这个时代里有日期可查的乐观情绪最清楚的一条弧线。2016 年,它以对时间表的乐观开场,并第一次给采用自动驾驶加上了道德理由:
“full autonomy is gonna come a hell of a lot faster than anyone thinks it will. I think what we’ve got, under development is gonna blow people’s minds. Blows my mind.” ↗
“完全自动驾驶来得会比任何人想的都快得多。我觉得我们正在开发的东西会让大家大吃一惊。连我自己都大吃一惊。”
“I want to emphasize narrow AI. It’s, like, not gonna take over the world, but it needs to be really good at driving a car.” ↗
“我想强调的是狭义 AI。它不会接管世界,但它得非常擅长开车。”
“You really start to get quite statistically significant at that point, and can make quite a strong argument, I believe at that point, that it would be, you know, morally wrong not to allow autonomous driving.” ↗
“到那个时候,数据就真的开始具有相当的统计显著性了。我相信到那时,你可以提出相当有力的论证:你知道,不允许自动驾驶在道德上是错的。”
2017 年第一季度是整个时代最激进的说法:他断言硬件从 2016 年底起就已经装在每一辆出厂的车上,于是 L5 被重新表述为只差一次软件升级:
“It’s a matter of upgrading the software and we can achieve Level 5.” ↗
“只要升级软件,我们就能实现 L5。”
接着是那个著名的横穿美国期限,以及它的一再推迟。每一次重述,他都用“丢人现眼”“看起来像个傻瓜”这类自知之明的幽默给自己留了余地:
“Obviously over time, an autonomous vehicle is going to be far safer than a person. It’s just It’s really hard for a person to compete.” ↗
“显然,随着时间推移,自动驾驶汽车会比人安全得多。人真的很难跟它比。”
“The coast-to-coast drive, autonomous drive by the end of the year, I believe we are still on track for that. It is certainly possible that I may have egg on my face on that front. If it is not at the end of the year, it’ll be very close.” ↗
“年底前横穿美国的自动驾驶,我相信我们还在按计划推进。当然,我在这件事上完全有可能丢人现眼。就算不是在年底,也会非常接近。”
“It’s gonna kind of be like that for self-driving… I mean, timing wise, I think we could probably do a coast-to-coast drive in three months, six months at the outside.” ↗
“自动驾驶也会差不多是这样……我是说,时间上,我觉得我们大概三个月就能完成一次横穿美国的驾驶,最多六个月。”
“Yes. We can do a coast to coast drive, especially if we like if we pick a specific route, and then write code to really make that route work, we could do a coast to coast route drive, but that would be kind of gaming the system. And I think it’s really important for the water bottle team to be focused on fundamental safety of the existing features.” ↗
“是的。我们可以完成横穿美国的驾驶,尤其是如果我们挑一条特定路线,然后专门写代码让这条路线跑通,我们就能完成一次横穿美国的路线驾驶,但那有点像在钻系统的空子。我觉得真正重要的是,让团队(文字稿原文作 water bottle team)专注于现有功能的基本安全。”
2017 年第四季度的电话会加进了至关重要的反激光雷达信念,即“拐杖/局部最优”,还有 AlphaGo 的类比,用来说明他为什么预期会出现突然的指数级进展。与之配套的,是他那种先留余地、再亮出确信的招牌模式:
“what is the point in having active optical, meaning lidar, which cannot read signs?.. In my view, it is a crutch that will drive companies to a local maximum that they will find very difficult to get out of. If you take the hard path of a sophisticated neural net that’s capable of advanced image recognition, then I think you achieve the global maximum.” ↗
“用主动光学,也就是激光雷达,有什么意义?它连路标都读不了……在我看来,它是一根拐杖,会把公司带进一个局部最优,而它们会发现很难从里面走出来。如果你走那条难走的路,用一个能进行高级图像识别的复杂神经网络,我觉得你就能达到全局最优。”
“I am pretty excited about how much progress we’re making on the neural net front. It’s also one of those things that’s kind of exponential, where the progress doesn’t seem like much progress, then suddenly, wow… You look at, say, something like what Google’s DeepMind did with AlphaGo, went from not being able to beat even a pretty good Go player, to suddenly it could beat the European champion, then it could beat the world champion, then it could thrash the world champion.” ↗
“我们在神经网络方面的进展让我相当兴奋。这也是那种有点指数级的事情:进展看起来没多少,然后突然,哇……你看,比如谷歌的 DeepMind 用 AlphaGo 做到的事,一开始连一个相当不错的围棋手都赢不了,突然就能赢欧洲冠军了,然后能赢世界冠军了,再然后能把世界冠军打得落花流水。”
“Now, perhaps I am wrong, in which case I’ll look like a fool. I am quite certain that I’m not.” ↗
“当然,也许我错了,那样的话我会看起来像个傻瓜。但我相当肯定我没错。”
到 2018 年,论证从技术转向了车队数据护城河加上监管瓶颈,最后又回到了“今年年底”:
“Well, the hardest thing to predict about the timing is regulatory approval. You know, there’s the thing that’s tricky with autonomous vehicles is that autonomy doesn’t reduce the accident rate or fatality rate to zero. It improves it substantially.” ↗
“嗯,时间上最难预测的是监管审批。你知道,自动驾驶汽车棘手的地方在于,自动驾驶并不能把事故率或死亡率降到零,只是大幅改善。”
“If I could kinda say from a technical standpoint, I think we’ll probably be ready by the end of next year.” ↗
“如果只从技术角度来说,我觉得我们大概明年年底就能准备好。”
“The advantage that Tesla will have is that we’ll have millions of cars in the field with full autonomy capability, and no one else will have that. I think that will end up putting us in the strongest competitive position long term.” ↗
“特斯拉的优势在于,我们会有几百万辆具备完全自动驾驶能力的车在路上跑,而别人都没有。我觉得这最终会让我们在长期竞争中处于最有利的位置。”
“Our miles of training that we have, if you added everyone else up combined, they’re probably 5%, I’m being generous, of the miles that Tesla has. This difference is increasing.” ↗
“就我们拥有的训练里程来说,把其他所有人的加在一起,大概也只有特斯拉的 5%,这还是往多了说。而且差距还在扩大。”
“I mean, in a nutshell, when do we think that the capability will be there for when will we think it’s safe for Full Self-Driving? It’s probably towards the end of this year. It’s up to regulators to decide when they wanna approve that.” ↗
“我是说,简单讲,我们认为什么时候能具备这个能力,什么时候我们认为完全自动驾驶是安全的?大概是今年年底前后。至于什么时候批准,那要由监管机构来决定。”
生产地狱:亲历的心理(2016–2017)
这个时代给这段经历起了名字,还把它分了类。“生产地狱”这个词在 2016 年第二季度被提出,连同它的情绪代价;2016 年第三季度,他又讲了一遍个人付出的代价:
“we were in production hell for the first six months of this year. I mean, man, it was hell. Then we just managed to sort of climb out of hell, and like basically fought way through June.” ↗
“今年头六个月我们都在生产地狱里。我是说,老天,那真是地狱。后来我们总算差不多从地狱里爬了出来,基本上是一路拼杀熬过了六月。”
“I’ll get a whole lot of mental scar tissue from the first six months this year.” ↗
“今年头六个月会给我留下一大堆心理上的疤痕组织。”
“I personally probably took a year off my life or more, camping out in the Fremont factory, solving that along with a number of other members of the Tesla team, went through bloody hell in the first half of this year.” ↗
“我个人大概折了一年寿,甚至更多,一直守在弗里蒙特工厂,和特斯拉团队的许多成员一起解决那些问题。今年上半年我们经历了该死的地狱。”
2016 年第一季度那个睡袋的细节,是他身体上自我牺牲的一个反复出现的标记;还有他用“打地鼠”来比喻爬坡中的种种限制:
“I have a sleeping bag in a conference room adjacent to the production line, which I use quite frequently.” ↗
“我在生产线旁边的一间会议室里放了个睡袋,用得相当频繁。”
“There are new issues that pop up every week, and we attack them and get them to solve the schedule. Another issue will pop up in the following week. It’s schedule Whac-A-Mole.” ↗
“每周都会冒出新问题,我们就去解决,好让进度跟上。下一周又会冒出另一个问题。这就是进度上的打地鼠。”
2017 年第三季度,他在超级工厂的车间里,把生产危机说成地狱的层级,还罕见地说到了抑郁:
“I mean, this is sort of imprecise. I’m not sure what each level means really, but let’s say level 9 is the worst. We were in level 9. We’re now in level 8. I think we’re close to exiting level 8.” ↗
“我是说,这不太精确。我也说不清每一层到底是什么意思,但就当第 9 层是最糟的吧。我们当时在第 9 层,现在在第 8 层。我觉得我们快要走出第 8 层了。”
“I have to tell you, I was really depressed about 3 or 4 weeks ago when I realized that we were kind of in level 9, then we got to level 8. Now I can see sort of a clear path to sunshine. And so I feel really pretty optimistic.” ↗
“我得告诉你们,大概三四周前,当我意识到我们差不多在第 9 层的时候,我真的很抑郁。后来我们到了第 8 层。现在我算是能看到一条通往阳光的清晰道路了。所以我感觉相当乐观。”
到 2017 年第四季度,这份坦率转向了反思:他诊断出自己过度自信,重拾了对人类的信心,还把混乱的爬坡比作做香肠:
“I’m reminded of, I think it may have been Churchill’s line about sausage. If you like sausage and respect the lawyer, you should watch neither being made. And to some degree, that is true of a production ramp.” ↗
“这让我想起,我记得可能是丘吉尔关于香肠的那句话:如果你喜欢香肠,又尊重法律,那这两样东西是怎么做出来的,你都别去看。在某种程度上,生产爬坡也是这样。”
“I think in part we were probably a little overconfident, a little complacent in thinking that this is something we know and understand and put a lot of attention on other things and just got too comfortable with our ability to do battery modules since we’ve been doing that since the start of the company.” ↗
“我觉得部分原因是我们可能有点过度自信,有点自满,以为这是我们熟悉、理解的东西,就把大量注意力放到了别的事情上。我们从公司创立起就在做电池模组,所以对自己做电池模组的能力太放心了。”
“It’s actually, to some degree, renewed my faith in humanity, that the rapid evolution of progress and the ability of people to adapt rapidly has is quite remarkable.” ↗
“这其实在某种程度上让我重拾了对人类的信心:进展演变得这么快,人们适应得这么快,真是相当了不起。”
脾性与坦率:发作与道歉(2016–2018)
这个时代把他的脾性展露得异常充分。对怀疑者咄咄逼人的自信(2016 年第三季度),对炒作直言不讳的经验主义(2017 年第二季度),还有那次臭名昭著的打发分析师(2018 年第一季度);一个季度之后,他真诚地道了歉,这本身就是一个演变标志:
“Well, I suggest that they do not bet against us.” ↗
“嗯,我建议他们别跟我们对赌。”
“here’s my opinion of the, you know, battery breakthrough of the week of, you know, battery breakthrough du jour… Send us a sample. If you don’t trust us, send it to an independent lab where the parameters can be verified. Otherwise, STF.” ↗
“对于那些,你知道,本周电池突破,你知道,每日电池突破,我的看法是……给我们寄个样品来。要是不信任我们,就寄给一家独立实验室,让他们核实参数。否则,就闭嘴(STF)。”
“Yeah. Everything works on PowerPoint. You know, you could, like, I give you a PowerPoint presentation about teleportation to the Andromeda galaxy, that doesn’t mean it works.” ↗
“是啊。在 PPT 上什么都行得通。你知道,你可以,比如说,我给你做一份关于瞬移到仙女座星系的 PPT,那也不代表它行得通。”
“We have no interest, we have no interest in satisfying the desires of day traders. Like, couldn’t care less. Please sell our stock and don’t buy it.” ↗
“我们没兴趣,我们没兴趣满足日内交易者的愿望。真的,一点都不在乎。请卖掉我们的股票,别买。”
“Next. Next. Boring bonehead questions are not cool. Next.” ↗
“下一个。下一个。无聊的蠢问题可不酷。下一个。”
“First of all, I’d like to apologize for being implied on the prior call. Honestly, I think there’s really no excuse for bad manners. I was kind of violating my own role in that regard. So we have some excuse at the reasons for it and that I’ve gotten no sleep and working sort of 10 hour or 20 hour weeks, but nonetheless, there’s still no excuse.” ↗
“首先,我想为上次电话会上的无礼道歉。说实话,我觉得无礼真的没有任何借口。在这方面,我算是违反了我自己的规矩。当然,事出有因,我一直没睡觉,每周工作大概 10 小时或者 20 小时,但无论如何,这都不是借口。”
两个更安静的坦率时刻:感谢团队拦住了他自己的坏主意(2017 年第二季度),这让人看到他在更新自己;还有他难得地因为顾客的忠诚而哽咽(2018 年第三季度):
“I would like to thank my executive team for stopping me from being a fool.” ↗
“我要感谢我的高管团队拦住了我,没让我干傻事。”
“Maybe this has happened before, but I’ve never heard of a case where a company’s customers actually cared about the future of the company so much that they volunteered their time to help the company succeed. I think that’s amazing. Just don’t see that anywhere. Yeah, it’s like really makes me choked me up actually.” ↗
“也许以前有过这种事,但我从没听说过,一家公司的顾客会这么在乎公司的未来,自愿花时间去帮公司成功。我觉得这太了不起了。别的地方根本见不到。是啊,说实话,这真让我有点哽咽。”
AI 风险与“狭义 AI”那条线(2017)
这个时代里,他对 AI 恐惧的直白个人表态,以及他成熟的“洞察而非监督”监管模型。后者把他的真实立场和外界的漫画式描述区分开来:
“Well, as you know, I’m terrified of AI.” ↗
“嗯,你们知道的,AI 让我害怕极了。”
“It’s just something that I think anything that represents potential risk to the public deserves at least insight from the government… That’s Insight is different from oversight… I’m not advocating for that we stop development of AI or any of the sort of straw man hyperbole things that have been written. I do think there are great benefits to AI. We just need to make sure that they are indeed benefits and we don’t do something really dumb.” ↗
“我只是觉得,任何可能给公众带来风险的东西,至少都应该让政府有所洞察……那是……洞察和监督是两回事……我不是在主张停止开发 AI,也不是外面写的那些稻草人式的夸张说法。我确实认为 AI 有巨大的好处。我们只是需要确保它们真的是好处,别干出什么真正愚蠢的事。”
管理哲学:身先士卒,承担过错(2017)
出自没有标注发言人的 2017 年第三季度电话会,两条决定性的领导原则:哪里问题最大就去哪里,以及拒绝把责任推给别人。
“Actually, we’re doing this cold from the Gigafactory because that’s where the production constraint is for Model 3s and the most important thing for the company. And I always move my desk to wherever well, not everybody has a desk, actually. I move myself to wherever the biggest problem is in Tesla.” ↗
“其实,我们这次电话会是在超级工厂开的,因为那里是 Model 3 的生产瓶颈所在,也是公司最重要的事。我总是把办公桌搬到……嗯,其实也不是每个人都有办公桌。特斯拉哪里问题最大,我就把自己搬到哪里。”
“So I really believe that one should lead from the front lines, and that’s why I’m here.” ↗
“所以我真心认为,领导者应该在一线带头,这就是我在这里的原因。”
“everything is our fault and my fault, most of all. If we pick the wrong subcontractor, we’re the fault. So I don’t want this to be sort of an externalizing responsibility.” ↗
“一切都是我们的错,最主要是我的错。如果我们选错了分包商,那也是我们的错。所以我不希望这变成一种推卸责任。”
第一性原理与竞争世界观:护城河很逊(2016–2018)
这个时代充满了他的招牌推理。第一性原理与类比推理之争被用到了 Model 3 爬坡上(2016 年第一季度);还有那条著名的护城河很逊、创新速度才重要的信念,以及亚马逊对沃尔玛的例证(2018 年第一季度):
“It’s always tempting for people to reason by analogy instead of first principles. That would be the mistake of assuming that anything to do with the Model X production has bearing on Model 3.” ↗
“人们总是忍不住用类比推理,而不是从第一性原理出发。那样就会犯一个错误:以为 Model X 生产上的任何事都跟 Model 3 有关。”
“You can kill a fly with a thermonuclear weapon, you can with a MOAB, with a cruise missile, with a machine gun, or a fly swatter. The end result is the same, but the difficulty is considerably more significant from one to the other, and the collateral damage is considerably more significant.” ↗
“你可以用热核武器打死一只苍蝇,也可以用‘炸弹之母’、巡航导弹、机关枪,或者苍蝇拍。最终结果一样,但难度一个比一个大得多,附带损害也大得多。”
“First of all, I think moats are lame. I mean, they’re like nice in sort of a quaint, vestigial way. Like if your only defense against like invading armies is a moat, you will not last long. What matters is the pace of innovation. That is the fundamental determinant of competitiveness.” ↗
“首先,我觉得护城河很逊。我是说,它们有点古色古香、作为遗迹还挺好看的。可如果你抵御入侵大军的唯一防线就是一条护城河,你撑不了多久。重要的是创新的速度。那才是竞争力的根本决定因素。”
“Whichever company has the highest rate of innovation, unless that company is actively killed by its competitors in some way that’s nefarious or shoots itself in the foot, it will at some point exceed those competitors. This was obvious that this would occur with Amazon and Walmart because Walmart’s rate of innovation was negligible, and Amazon’s was very high. The outcome was obvious a long time ago.” ↗
“哪家公司创新速度最快,只要它没有被竞争对手用某种卑劣手段主动扼杀,或者自己搬起石头砸自己的脚,它总有一天会超过那些竞争对手。亚马逊和沃尔玛之间会发生这种事,是显而易见的,因为沃尔玛的创新速度微不足道,而亚马逊的非常高。结果老早就一目了然了。”
他所依赖的 SpaceX 与特斯拉之间的跨领域知识迁移(2017 年第二季度),以及把截止期限当作强制机制的管理理论(2016 年第一季度):
“This cross-fertilization of knowledge from the rocket and spacecraft industry to auto back and forth is I think is really been quite valuable. Certainly been very valuable for me in thinking about how do we make mass-optimized vehicles… On the space side, it’s helpful because I understand what really goes into high-volume manufacturing of something that has to be extremely reliable.” ↗
“火箭和航天器行业与汽车行业之间这种来来回回的知识交流,我觉得真的非常有价值。对我思考怎样造出质量最优化的车辆,肯定非常有价值……在航天这边,它也有帮助,因为我明白,要大批量制造一种必须极其可靠的东西,真正需要投入些什么。”
"the very same people will transition from saying it was impossible to saying it was obvious. I’m like, “Wait a second, was it obvious or impossible? It can’t be both.” ↗
“同样那批人会从说‘这不可能’变成说‘这显而易见’。我就想:‘等一下,到底是显而易见还是不可能?总不能两样都是吧。’”
还有那个“单项改进就是决定性杠杆”的洞见(2018 年第三季度),可以看出他如何推演不断累积的运营优势:
“I mean, it occurs to me that even if the only thing, like even if this was the only thing that Tesla did different was to shorten the time from factory to the end customer, any given company that did that would outcompete all other companies over time. It would not be a contest.” ↗
“我是说,我突然想到,就算特斯拉唯一与众不同的地方,只是缩短了从工厂到终端客户的时间,任何一家做到这一点的公司,随着时间推移都会胜过所有其他公司。那根本不是一场较量。”
使命、平价与资本纪律(2016–2018)
使命被重述为公司为什么存在(2016 年第三季度),被重述为交通全面电动化的信念(2017 年第一季度),而在这个时代的结尾,又被重述为与生存风险相连的“特斯拉存在的理由”(2018 年第四季度):
“Then I think it ends up being a good outcome for shareholders, because the whole purpose of any company existing is to make compelling products and services… Sometimes people lose sight of why companies should even exist.” ↗
“那样我觉得最终对股东也是好结果,因为任何公司存在的全部意义,就是做出引人入胜的产品和服务……有时候人们会忘了公司到底为什么应该存在。”
“I’m absolutely confident that electric vehicles will occupy every segment without exception… In fact, I’m highly confident that all transport will go fully electric with the ironic exception of rockets.” ↗
“我绝对相信,电动车会无一例外地占领每一个细分市场……事实上,我非常确信,所有交通工具都会完全电动化,讽刺的是,火箭除外。”
“The fundamental goodness of Tesla that, you know, like, the why of Tesla, the relevance, what’s the point of Tesla comes down to two things, acceleration of sustainable energy and autonomy. Acceleration of sustainable energy is absolutely fundamental 'cause this is an existential risk for humanity.” ↗
“特斯拉根本的好处在于,你知道,就是特斯拉存在的理由、它的意义,特斯拉的意义归结为两件事:加速可持续能源,以及自动驾驶。加速可持续能源是绝对根本的,因为这关乎人类的生存风险。”
“Also very important is autonomy. This has the potential to save millions of lives, tens of millions of serious permanent injuries, and give people their time back so that they don’t have to drive.” ↗
“同样非常重要的是自动驾驶。它有可能挽救数百万人的生命,避免数千万起严重的永久性伤害,还能把时间还给人们,让他们不用自己开车。”
使命被绑定在平价上(2018 年第三季度),也绑定在一种近乎道德义务的成本纪律上(2018 年第四季度):
“Our goal really is to make electric cars that everyone can afford, not to sort of mine high, you know, high option value cars. It’s like if we could produce a $35,000 car today, we would do it.” ↗
“我们的目标真的是造出人人都买得起的电动车,而不是去挖那些,你知道,高选装价值的车。就好比,如果我们今天就能造出一辆 3.5 万美元的车,我们就会去造。”
“And getting those costs down, variable cost and fixed cost is what allows us to lower the price and be financially sustainable and achieve our mission of environmental sustainability. We have to be absolute zealots about this.” ↗
“而把这些成本,可变成本和固定成本,降下来,才能让我们降低售价、在财务上可持续,并实现我们环境可持续的使命。我们在这件事上必须是绝对的狂热分子。”
“we have to be relentless about costs in order to make affordable cars and not go bankrupt. That’s, that’s what our headcount reduction is about. Yeah, we have to It’s, we have to be super hardcore about it. It’s the only way to make affordable cars.” ↗
“我们必须对成本毫不留情,才能造出买得起的车,同时不至于破产。这,这就是我们裁员的原因。是的,我们必须……我们在这件事上必须超级硬核。这是造出平价汽车的唯一办法。”
整个 2018 年贯穿着一种资本纪律的信念:盈利是成为“真正的公司”的门槛(第一季度),还有面对逆境的韧性说法(第四季度)。后来几年证明,这种姿态并不稳固(他还在第二季度电话会上承诺特斯拉“在任何时候都不会进行股权融资”,而 2019 年随即进行了股权融资;这一融资承诺在这里只写进正文,不作为信念放进引语块):
“We need to become a profitable company. That is a good criticism that has been leveled at Tesla, an accurate one. It’s high time we became profitable. You know, and the, and the truth is, like, you’re not a real company until you, until you are, frankly.” ↗
“我们需要成为一家盈利的公司。这是外界对特斯拉的一个中肯批评,也是准确的批评。我们早该盈利了。你知道,而且,而且事实是,老实说,你不盈利,就,就不是一家真正的公司。”
“I mean, I do think that, you know, the economy moves in cycles, and there’s clearly a significant risk of a recession over the next 12-18 months… And be all the stronger for it when the recession ends.” ↗
“我是说,我确实认为,你知道,经济是周期性的,未来 12 到 18 个月显然存在相当大的衰退风险……等衰退结束时,我们会因此变得更加强大。”
还有他向股东力推的长线投资姿态(2018 年第一季度),以及押上整家公司去搞登月式项目的意愿(2018 年第四季度):
“Really the pros, like, people get too focused on, like, what’s happening in the space of a few weeks or a few months. This is, you know, an old maxim of investing. You should not be focused on short-term things. You should be focused on long-term things.” ↗
“说真的,那些专业人士,就是,人们太关注几周或几个月之内发生的事了。你知道,这是一条古老的投资格言:你不应该关注短期的东西,你应该关注长期的东西。”
“SpaceX actually has two absolutely insane projects that would normally bankrupt a company, is Starship and Starlink. SpaceX has to be incredibly spartan with expenditures until those programs reach fruition.” ↗
“SpaceX 其实有两个绝对疯狂的项目,正常情况下足以让一家公司破产,那就是星舰和星链。在这两个项目开花结果之前,SpaceX 在开支上必须极度节俭。”
自我认同、专注与长线博弈(2016–2018)
一组自我暴露的标记:终身投入的表态(2016 年第四季度/2017 年第一季度),延续上一时代“狂妄”教训、对 Model X 过度设计的自我批评(2017 年第一季度),专注的纪律(2018 年第一季度),以及火星能激励公众的世界观(2016 年第四季度):
“I expect to remain with Tesla essentially, you know, for forever, unless somebody kicks me out. That remains my intention.” ↗
“我预计基本上会,你知道,永远留在特斯拉,除非有人把我踢出去。这仍然是我的打算。”
“I intend to be actively involved with Tesla for the rest of my life… You know, that doesn’t mean I should be CEO. It’s like, you know, I think my main, the, like, most valuable thing I can contribute is kind of product design and technology. That’s my forte. That’s what I like doing.” ↗
“我打算余生都积极参与特斯拉的事务……你知道,这不代表我应该当首席执行官。就是,你知道,我觉得我主要的,那个,我能贡献的最有价值的东西,大概是产品设计和技术。那是我的强项,也是我喜欢做的事。”
“Model X became kind of like a technology bandwagon of every cool thing you can imagine all at once. It’s like everything all at once. That is a terrible strategy.” ↗
“Model X 后来差不多成了一辆技术花车,把你能想到的所有酷东西一股脑全装了上去。就像是所有东西一下子全来。那是个糟糕透顶的策略。”
“Like, the idea is, idea generation far exceeds the ability to execute it, we just need to stay focused and not divide our attention on too many products at one time.” ↗
“就是说,想法是,产生想法的能力远远超过执行的能力,我们只需要保持专注,别同时把注意力分散到太多产品上。”
“Although I think a Mars initiative would be amazing, and really energize the public, domestically and worldwide, just as the Apollo mission to the Moon did almost half a century ago.” ↗
“不过我觉得,一个火星计划会非常了不起,会真正激励国内乃至全世界的公众,就像差不多半个世纪前阿波罗登月任务那样。”
特意不引用的内容
- 所有财务数字和量化指引:营收、毛利率、交付量、预订数量、Model 3 爬坡速度的数字,以及 SolarCity 合并和“资金已到位”的财务与法律经过。这些作为商业和财务参数,要么在正文中交代,要么略去。(在“资本纪律”下引用的,是推理姿态和信念,从不是交易数字。)
- 所有产品和工程细节:Model 3 电池和驱动单元的具体参数,Autopilot/FSD 的传感器硬件和算力,超级工厂生产线的工程,“外星无畏舰”各个版本的具体机制,特斯拉 Semi、Roadster、Powerwall 的产品参数。这些是工程,不是思维。“狭义 AI”和“火箭方程”那两句之所以保留,只是因为它们体现了思维模型。
- 其他所有发言人:投资者关系主持人,首席财务官(杰森·惠勒、迪帕克·阿胡贾),首席技术官 JB·斯特劳贝尔,其他高管,以及所有分析师的提问。只引用埃隆·马斯克。在那两份没有标注的文字稿(2017 年第三季度、2018 年第二季度)里,首席财务官和首席技术官接话的部分以及财务摘要,都被明确排除在外。
关联页面(涉及的页面)
- 工程算法——扩充了这个时代“工厂就是产品”的主干及其自我纠正:“一门心思要成为地球上最好的制造商”/“外星无畏舰”(2016 年)→“工厂将是比汽车本身更重要的产品”加“把火箭方程用到制造上”(2016 年第四季度)→“制造机器的机器……很难复制”(2017 年第一季度)→“速度是终极武器”/“如果你能看见机器人在动,那它就太慢了”(2017 年第三季度)→胭脂河工厂学说(2017 年第四季度)→对过度自动化的纠正:“我们在自动化上确实走得太远了”,加上“速度和密度”,以及“大规模制造主要是一个软件问题”(2018 年)。
- 自动驾驶——扩充了这个时代最可证伪的时间表弧线:“完全自动驾驶……比任何人想的都快”加“不允许自动驾驶在道德上是错的”(2016 年)→“只要升级软件,我们就能实现 L5”(2017 年第一季度)→一再推迟的横穿美国:“丢人现眼”→“三个月,最多六个月”→“钻系统的空子”(2017–2018 年)→反激光雷达的“拐杖/局部最优”加 AlphaGo 类比(2017 年第四季度)→车队数据护城河:“几百万辆车在路上跑”/“特斯拉里程数的 5%”,以及“今年年底”的收尾(2018 年)。
- 第一性原理——扩充了在 2017 年诞生并完善的 S 曲线生产模型(“S 曲线的起始部分”→“因为有东西坏了而倒退”→“人的直觉往往是直线外推”),护城河很逊、创新速度才重要的信念及其亚马逊对沃尔玛的例证(2018 年第一季度),类比推理与第一性原理之争以及“用热核武器打苍蝇”的说法(2016 年第一季度),SpaceX 与特斯拉之间的知识交流(2017 年第二季度),以及“缩短从工厂到终端客户的时间”这一单项杠杆的洞见(2018 年第三季度)。
- 可持续能源使命——扩充了“任何公司存在的全部意义”(2016 年第三季度),“所有交通工具都会完全电动化”(2017 年第一季度),“特斯拉存在的理由……人类的生存风险”这一收尾(2018 年第四季度),以及平价与成本纪律的绑定(“造出人人都买得起的电动车”;“绝对的狂热分子”;“对成本毫不留情”)。
- AI 生存风险——扩充了这个时代直白的“AI 让我害怕极了”和成熟的“洞察而非监督”监管模型(2017 年第二季度),以及“狭义 AI”的区分(2016 年第二季度)。
- 工作强度——扩充了生产地狱的心理:睡袋(2016 年第一季度),“生产地狱”/“心理上的疤痕组织”(2016 年第二季度),“折了一年寿……一直守在弗里蒙特工厂”(2016 年第三季度),地狱的层级和罕见的“我真的很抑郁”的自白(2017 年第三季度),以及“无礼没有任何借口……一直没睡觉”的道歉(2018 年第二季度)。
- 特斯拉——扩充了一段“2016–2018 年财报电话会”的说明,串起 Model 3 生产地狱时代、作为背景的 SolarCity 合并和“资金已到位”风波、工厂即产品与过度自动化的弧线、第一个可持续盈利的季度,以及资本纪律的收尾。
- 埃隆·马斯克——扩充了“2016–2018 年特斯拉财报电话会透露了什么”一节,串起工厂即产品的论点及其反转,先变激进、后一再推迟的自动驾驶时间表,S 曲线思维模型,赤裸的生产地狱心理(地狱的层级、抑郁、发作与道歉),直白的 AI 恐惧,身先士卒、承担过错的管理原则,以及着眼长远的资本纪律。