访谈·中文精读

2014马斯克访谈:Tesla的未来与生产价值

2014 Elon Musk Interview on the Future of Tesla and its production value

中文精读由 AI 基于原始视频与文稿改写,非官方翻译,可能存在偏差或错误;表述以原视频为准。说明

2014年,特斯拉全年交付了三万一千多辆车。就在这个量级上,马斯克对着镜头说,2020年要做到年产五十万辆。

同一年,德意志银行的分析师给出的预测是:2027年,十万辆。

两个数字差着六年和五倍。当时的行业媒体更愿意相信后者。

2020年结束,特斯拉交付了499,550辆。差550辆,卡在线上。

这段采访来自2014年,频道是Celebrity Universe,地点就在特斯拉的展厅里。主持人跟着马斯克边走边看车,从一把车钥匙聊到各州的直销禁令。马斯克在对话里讲了他判断事情的几个底层习惯,也做了几个后来被现实修正的判断。

有意思的地方全在细节里。

一把车钥匙,藏着一家公司的产品观

采访从一把钥匙开始。那把钥匙做成了Model S的等比例缩小版,形状、轮廓、颜色都对得上。按一下小车车头的位置,真车的前备箱弹开;按一下车尾,后备箱弹开。

马斯克管它叫"一辆车的超现实版本",又称它是"一辆车的印象派版本"。他一边说一边按,后备箱应声而起,里面的空间大得能塞进一个人。

这东西在2014年被当成彩蛋。往深一层看,它说明特斯拉愿意把用户每天要摸十几次的物件当成产品的一部分。传统车厂的钥匙是功能件,塑料壳加三个按钮,成本压到几毛钱。特斯拉肯在一个钥匙上复刻整车造型,理由很朴素:这是车主每天唯一会拿出来看的品牌物件。

后来的Model 3把这套逻辑推向另一个方向,钥匙变成一张卡片和一部手机。形态一直在变,判断没变:凡是高频接触的界面,都值得重做一遍。

仪表盘只剩两个物理按钮,其中一个还是双闪

主持人说了一句后来流传很广的话:坐进特斯拉,感觉像坐在一台iPhone里。

我经常说,坐进一台特斯拉就像坐在一台iPhone里。你看那块屏幕,上面能直接显示车停在48街和第六大道。(主持人)

Model S的中控是一块17英寸竖屏。马斯克说,车里严格来讲只有两个物理按钮,一个开手套箱,一个是紧急双闪。

这个设计在2014年招来的质疑最多。空调、音量、后视镜、天窗、悬挂高度,全部塞进屏幕。习惯物理旋钮的人第一反应是危险,开车时得低头找菜单。传统豪华车的定义方式正好相反:按钮越多,木纹越厚,越显得值钱。

马斯克赌的是另一件事:屏幕能升级。物理按钮装上去那天就固定了,屏幕上的功能可以一版一版长出来。2014年的Model S还没有后来的自动辅助驾驶、哨兵模式、宠物模式,这些全是靠软件更新塞进那块屏幕的。

把功能从硬件搬到软件,代价是学习成本,收益是这台车卖出三年后还能变新。今天所有新势力都在用这套逻辑,2014年敢赌的是一家年销三万辆的小公司。

主持人当时还提到一个细节,屏幕上能显示车具体停在哪条街。这个功能在手机导航里稀松平常,装进车里就变成了另一种体验:车知道自己在哪里,也知道你知道它在哪里。

7,579辆和7,800辆之间,隔着一条全球供应链

主持人报了数字:上一个季度交付7,579辆,创了纪录。当前季度,分析师预期7,800辆。他直接问,你能超过吗。

要看季末最后几天的情况。如果什么都不出岔子,我们可能会稍微高一点。但物流上的意外总是有的,车要发到全世界。欧洲或者中国卡一下海关,或者运车的火车坏在半路,这些都算不进去。(马斯克)

主持人追问,听起来你们踩过不少坑。马斯克回了一句,全都踩过。

这段对话的价值在于它把一家制造企业的黑箱拆开了。车从弗里蒙特工厂下线,上火车,进港口,装船,清关,落地,再上卡车。任何一环延误,季末的数字就掉。投资者看到的是一个整数,公司内部看到的是几十个交接点。

马斯克还补了一句更少见的表态:

我尽量把我真实的不确定程度说出来,而不是故意报低一个数再去超额完成。(马斯克)

科技公司的CEO普遍喜欢打埋伏,给一个保守指引,然后每季度超一点点,市场就奖励你。马斯克反着来,他给区间,给条件,给可能出问题的地方。这套沟通方式让他在华尔街吃过不少亏,也让特斯拉的股东名单里多了一批死忠。

那个季度的官方数字最后落在7,800附近,勉强踩线。

五十万辆这个数字,需要六条曲线同时抬头

主持人提到德意志银行分析师的预测,2027年做到十万辆级别。接着他问,等超级电池工厂(Gigafactory)投产,产线一年能推下来多少台。

我有把握的是,2020年至少能做到年产五十万辆。也许能比这更早一点,但这是我敢给的预测。(马斯克)

把两个数字放在一起看:2014年全年销量31,655辆,2020年目标500,000辆。六年,十五倍,年复合增长率大约58%。

汽车工业的正常扩张节奏是个位数。丰田、大众、通用这些公司的产能曲线都爬了几十年才到千万级。要在六年里翻十五倍,特斯拉必须同时打通五件事:电池产能、廉价车型、直营网络、充电网络、各国政府准入。

这五件事里最紧的是电池。2014年锂电池组的成本占整车成本三分之一以上,谁能把电池成本压下来,谁就能把车卖到三万美元以下。Gigafactory就是这个赌注的物理形态,2014年它刚在内华达动工,特斯拉当时宣称的目标是把电池组单位成本砍掉三成以上。

分析师们并不蠢,他们只是用了线性外推。线性模型套在指数曲线上,误差会随时间越滚越大,所以2027年十万辆的预测在第六年就被证明差了一个数量级。这个教训对今天看任何硬件公司的研报都管用:先问一句,这个模型假设的是直线还是S曲线。

2020年的答案:499,550辆。差550辆。最后那一段爬坡,靠的是2019年底投产的上海工厂,从动工到出车不到一年。

Model X把明年卖光了,他仍然宁愿晚交

主持人问Model X的预订情况。马斯克说,基本上把明年的产能卖光了。

具体多少?他给的答案是一串模糊词:大概一万五这个区间,也许两万,具体很难说。

主持人抓住这个"十五上下",开玩笑说这也是个技术术语。马斯克接了一句,我希望我的预测能力比现在更强。

我不是想故意报低。我尽量把我不确定到什么程度如实讲出来。明年Model X的爬坡怎么样,取决于很多事。一万五千是比较合理的猜测,如果一切顺利可能到两万。(马斯克)

然后他讲了一段产品哲学,这段话后来被反复引用:

对特斯拉的任何新产品,最重要的问题是:我们把这个产品做对了吗。我宁可推迟量产,推迟爬坡,直到产品足够好,也不想提前推出一个有缺陷的东西。短期看,如果我们在产品性质或者车辆质量上妥协,可能会多赚一点钱。(马斯克)
别的公司拿出那种不行的车,我实在想不通,他们为什么要把这种东西拿出来卖。我们要让所有人知道,特斯拉拿出来的东西,必须是我们自己认为惊艳的。(马斯克)

这话在2014年听着像宣传。后来的Model X证明了它的另一面。

Model X在2012年就发布了概念车,鹰翼门、六座布局、纯电SUV,把胃口吊得很高。量产一路拖到2015年9月才开始交付,2015年全年只交出去几百辆。鹰翼门的密封、铰链、传感器在量产阶段反复出问题,供应商体系也被这套复杂度拖住。

按马斯克自己定的标准,这次延期是对的。按现金流的标准,这次延期代价很高。两条标准同时成立,这才是做产品最难的地方:原则能保住口碑,保住口碑不等于保住账本。

每二十小时一座充电站,三分之一的需求在亚洲

采访中途,主持人走到一台深蓝色Model S旁边,说这是他最喜欢的颜色,然后话锋转到充电网络。

现在平均每二十个小时就有一座新的超级充电站上线。站点地图和建设排期都放在官网上,谁都能查。(马斯克)

每20小时一座,一年下来四百多座。2014年全球电动车保有量还很小,建充电站是纯烧钱,没有第二家公司在用这个速度铺。当时Model S车主充电免费,成本全由公司扛。

主持人接了一条更具体的信息:特斯拉和中国联通签了协议,要在全中国建400个充电站。

马斯克顺势讲了他对需求分布的判断:

长期看,需求大概三分之一来自亚洲,包括日本和其他地区,三分之一来自北美,三分之一来自欧洲和世界其他地方。(马斯克)

这个判断在2014年显得超前。当年特斯拉在中国的交付量很小,充电标准还没统一,和中国联通的合作最终也没能承担起打开中国市场的角色。真正改变局面的是2019年上海工厂投产。充电桩解决的是体验,本地制造解决的是价格。

但三分之一这个比例是对的。2020年之后,中国市场长期占到特斯拉全球交付的三成上下。方向和路径是两件事,马斯克在这段对话里对方向有把握,对路径明显没那么有把握,他的措辞也确实是这么摆的。

新泽西不让卖,那就去法院和议会打

主持人提了当时最棘手的一块战场:新泽西州。2014年3月,新泽西机动车委员会裁定特斯拉必须通过经销商卖车,直营店不能继续营业。得克萨斯、亚利桑那也有类似限制。与此同时,纽约和马萨诸塞放行了直营。

主持人用了"经销商卡特尔"这个词,然后问,你敢不敢跟这种游说集团正面硬碰。

我们会一直为终端客户去争。我们真正在乎的是买车的人。这件事我们不会停。(马斯克)

直营模式在2014年是行业异端。传统链条是车厂卖给经销商,经销商卖给消费者,中间那层靠信息不对称赚钱:同一辆车,不同店、不同月份、不同砍价水平,成交价能差几千美元。

特斯拉把价格挂在官网上,全国一个价,不许议价。这套做法动了整条链的奶酪,所以反对它的不只是经销商协会,还有各州的税收账本和就业叙事。

十年后回看,这场仗没打完。美国仍有多个州限制特斯拉直营店的数量,特斯拉的应对方式是把店开在印第安保留地,或者把下单流程搬到线上。中国的直营模式倒是变成了标准答案,2020年之后几乎所有新品牌都照着抄。

马斯克在这段对话里的态度很清楚:选边站,站在付钱的那个人那边,然后一直站着。这话听着像口号,后面十年它被反复执行过,包括2020年疫情期间特斯拉起诉阿拉米达县政府那一次。

最后,主持人把主持人的头衔念了一遍

采访收尾,主持人说:这就是终极创业者,SpaceX的CEO,特斯拉的CEO,别忘了还有SolarCity的董事长,PayPal的联合创始人。

2014年这个组合的分量比今天更容易被低估。SolarCity当时是美国最大的住宅光伏安装商,SpaceX刚拿到商业载人合同,特斯拉年销三万辆。三家公司共享一套方法论:把成本结构从头拆一遍,然后按物理学的下限重新设计。

马斯克在这段采访里没讲方法论,他讲的全是具体物件和具体数字:一把钥匙、两个按钮、7,579辆车、一万五千台Model X、二十小时一座充电站、三分之一的需求在亚洲。抽象的词一个没有。

这可能就是这段老视频还值得看的原因。十年过去,他给出的数字有些准得离谱,有些被现实打了折,但判断问题的方式没变:先把不确定的部分摊在桌面上,再把最难的那件事排在前面。

节目最后,主持人预告下一段采访:奥迪,还有AutoNation的CEO迈克·杰克逊,还补了一句,他们卖车卖到手软。

一个靠经销商体系卖到手软的老行业,坐在一个扬言要拆掉经销商体系的人对面。2014年的这期节目,把两条路线摆在了同一天里。

可以带走的几条

  • 汇报项目时把不确定的部分说出来。给区间、给假设、给可能出问题的环节,比给一个漂亮整数更省事。马斯克那句"我尽量把真实的不确定程度说出来",可以直接抄进你的周报模板。
  • 判断一个产品的减法是否成立,看被删掉的东西有没有搬到软件里。特斯拉删掉物理按钮,把功能塞进可升级的屏幕,这叫减法。删掉功能却不补上,那叫缩水。
  • 看到长期目标,先算复合增长率。五十万除以三万,六年,年复合58%,然后列出这家公司必须同时打通哪几条曲线,哪条最可能先卡住。用线性模型去预测S曲线,误差会随时间指数放大。
  • 评估延期决策时把两本账分开算。产品账看质量,现金账看节奏。Model X证明原则能保住口碑,也会烧掉现金,提前把这笔代价写进计划里,比事后解释划算。
  • 面对渠道或监管阻力,先确认自己服务的是终端用户还是中间商。选定了就别摇摆,摇摆一次,两边的信任都要重新买。
  • 看季度交付这类数字,先弄清口径和物流边界,再判断好坏。季末冲刺出来的数字里,往往藏着下个季度的问题。
YouTube 平台字幕(自动或人工),可能有识别错误,人名与数字请以视频为准。已按语句合并成段落,并去掉 um、uh 等口头填充词。 文稿来源:YouTube(自动/人工字幕)。

0:00go inside Elon musk's second company actually it's your third because you co-founded PayPal right SpaceX of course you run Tesla right let's talk about this one open the door so first of all it's worth starting off with the the key the key is a essentially a sort of surrealistic version of the car it's sort of an impressionist version of the car so if if you want to make something happen like for example we could open the trunk just by pressing the the trunk on the little car and when there it goes so you just press the trunk part on so if you want to open the trunk you just open the trunk and it's it's got It's got a huge huge amount of storage space I think

0:43it's more interesting yeah no sorry it's like trunks I love trunks but it's more interesting to see the dashboard of which there are no buttons right there's there's there's actually two buttons technically one is for the glove compartment and the other is for the emergency blinker I have often said that inside a Tesla is like sitting inside an iPhone and you see that dashboard you can see exactly where the car is parked on 48th and 6th it's absolutely fascinating I need some numbers on deliveries the model S delivered record numbers of deliveries what 7,579 for the most recent quarter for the current quarter people are expecting Elon around

1:237,800 will you beat that number well I I think we'll sort of see what the final day the quarter are looking like but I think we're sort of on track to to be something like that okay something like that are you going to vest it that's what I guess I'm pushing you it's really you know it really depends if if there are no if if nothing goes wrong we might be slightly higher but there are always these logistical issues that you encounter because we're shipping cars all over the world so if like if we encounter a customs issue in Europe or China or something like that or you know A Train That's carrying our cars breaks down it's difficult to account

2:00for for those extenuating circumstances it sounds like you've been through a lot of this yeah we've been through all of those so so but I mean I feel I feel fairly comfortable with you know where things are headed for for this quarter you've got some some analysts for example the Deutsche Bank analyst and he he really a lot of these guys are very bullish on it and they they predict that you would get about I guess a 100,000 in sales by 20 I mean 27 what

2:29what can we eventually expect when it comes to ramping up with the gigafactory in place the number of Teslas that you push off the assembly line well I I mean I feel comfortable that we'll be able to achieve at least a half million cars a year by 2020 and you know maybe we can do a bit sooner than that but I think that's that's what I feel comfortable predicting half a million cars by 2020 that is is pretty pretty interesting and impressive I hadn't heard that number let's get to the brand new crossover SUV the model X right what kind of pre-orders are you looking at We've basically sold out next year's production which is how many well it's

3:11sort of somewhere in the I don't know 15,000 is range maybe 20,000 is it's difficult to say exactly when it's I'm not trying to be I'm not trying to I mean I try to convey the level of uncertainty that I actually have as opposed to you know sort of sandbag or whatever but so depending upon how the ramp on the X goes next year it's you know 15ish thousand is probably a reasonable guess maybe 20 if things go especially well that's a technical term 15ish okay 15ish to I wish I had a better ability to forecast than than I do but the the thing that for for for the XO for any new product from Tesla that's actually the most important is did we get did we

4:01get the product right i' I'd much rather delay the the production delay the production ramp until we we've got the product to be super great then come out early with something that's that has deficiencies I mean I I of course and I mean we we could probably make a bit more money if if we did I think in the short term certainly if we you know if we compromised on the on the nature of the product or the quality of the vehicle but but really want to have I mean with with each product like I it drives me crazy that other comers will come out with these products these cars that are that are not that are not good like why would they even come out with that it's just

4:42so we just we want people to know that when ever Tesla comes out with a product it'll be something that we we we think is going to be amazing let's look at this one as we walk over this is by the way my favorite Tesla color this sparkly Navy deep blue a friend of mine he's I went to ele school with them Nathan Hawkman in Los Angeles bought one of these but as you look at these will China be what ramps you up because you've struck a deal with I I believe it's China Unicom to build 400 charging stations throughout China when will those be up and running so we actually there's there's a MTH of the supercharges on our website as well as the the schedule for the superchargers

5:23the the the rate at which we're establishing supercharge locations is very r right now we're putting one we're bringing one live every 20 hours on average so it's there there's a huge number of supercharged locations that are going active in terms of China demand I think you know the way we roughly think of of demand is a long-ter demand is about you know a third from from about a third from Asia which includes Japan and and other parts about a third from North America about a third from from Europe and the rest of the world so it's so China could be a huge help and here here in your own country here in the United States you had to battle against the state of New Jersey the

6:10dealers you know some are calling them the dealer cartel I find it fascinating that that some states New York Massachusetts are saying welcome Tesla you can directly sell to people without the dealer model will you ever be afraid to take on real critics and real tough people like dealer Lobby no I I I think I think we're always going to just we're going to fight for the End customer that that's who we care about and yeah we'll never stop fighting for the the at the End customer that's that's who that's who we're really trying to serve at the end of the day and here you have the ultimate entrepreneur right here Elon Musk the CEO of SpaceX the CEO of Tesla the

6:51chairman of let's not forget Solar City and the co-founder of PayPal is really one of these people who takes life and and digs your teeth in so thank you so much for coming on exclusively to talk about that yeah thanks yeah thanks for having me and guess what we we got a sort of a last Millennium but really fighting back car company Audi coming up auton Nation Auto Nation CEO Mike Jackson will be joining me live like to know your thoughts on that one later because they are selling cars hand over fist

资源与出处

相关视频

同频道、同主题或相邻年份