访谈·中文精读

2014马斯克完整访谈:私人飞机、太阳能、SpaceX与Hyperloop

Elon Musk - Jet travel, Solar power, SpaceX and the Hyperloop - Full interview 2014

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

一枚猎鹰9号(Falcon 9)火箭的造价大约6000万美元。它加满推进剂,花20万美元。

剩下的钱全花在箭体、发动机、航电和人工上。而这堆东西用完一次,就掉进海里。

马斯克用这组数字解释他为什么要死磕火箭回收。他的类比很朴素:飞机能变成大众交通工具,靠的是加完油接着飞。如果每飞一趟都得换一架新飞机,天上基本不会有飞机。

如果每次出门都得买一架新飞机,那该有多贵。那样的话,坐飞机的人会非常少。(马斯克)

访谈发生在2014年。那一年特斯拉股价坐过山车,Model S 刚烧了三辆,SpaceX 还在给国际空间站送货,Hyperloop 白皮书发布不到一年。马斯克在镜头前把这些事串起来讲了一遍,讲的是同一套逻辑:一个行业的成本结构,决定了它能长多大。

能送货的龙飞船,两年后加上人

镜头先给到龙飞船(Dragon)。马斯克说这是一号版本,能往返空间站运货,包括生物载荷,鱼之类的东西。

载人版还要多久?他的回答是大约两年,完整版的龙飞船会具备宇航员运输能力。

主持人问得实在:SpaceX 现在主要靠什么赚钱?马斯克的回答同样实在,发射卫星,这是付账单的方式。

这句话里藏着 SpaceX 的整个结构。卫星发射是现金流业务,火星是愿景业务。用前者的收入和信誉,养后者的研发。2014年的 SpaceX 已经能把货运飞船送上天,商业发射订单排着队,但真正让它区别于其他发射公司的,是那些订单背后的用途。它挣的每一笔钱,都在给一个十几年后才可能兑现的目标供血。

这种结构在商业史上不算罕见,罕见的是它的比例。多数公司的愿景业务是主业的延伸,SpaceX 的主业是愿景的燃料。理解这一点,就能理解它后来为什么敢把火箭一次次摔在海上。

他去 NASA 网站查火星计划,什么也没找到

故事要从 PayPal 卖掉之后讲起。马斯克说,他一直在想接下来该做什么。他对太空一直有兴趣,但从没觉得个人能在太空里做成什么事,那看起来是大政府的专属领域。

后来他跑去 NASA 官网,想查清楚美国什么时候去火星。按逻辑,登月之后下一步自然是火星。他什么也没找到。

他最初的判断是,美国失去了做这件事的意愿。过了很久他才推翻这个判断,认为自己错得离谱。在他看来,美国尤其是一个由探险者组成的国家。

这段自述里有意思的地方在第二次判断。第一次他归因于意志衰退,第二次他归因于没人把这件事摆上桌面。前一个判断让人绝望,后一个判断留出了行动空间。归因方式决定了下一步是抱怨还是动手,很多创业故事的分岔就在这里。

主持人追问,为什么你认定运输是最该做的事?

马斯克的说法是:如果人已经到了火星,住在那里是相对容易的部分,难的是怎么到那儿。今天去火星根本不可能。所以问题得倒着解,先解决怎么去,而且必须是能运送大量人和货物的方式。只送几个人过去,显然建不成一个自给自足的文明。

阿波罗计划是人类历史上极其鼓舞人心的事件。但人类最后一次登月是1973年,也可能是1974年。他不想让火星重演这一幕:插一面旗,留几个脚印,然后再也不去。

一次任务可以非常鼓舞人心,但它不会根本改变人类的未来。这是他对阿波罗的判断,也是他给 SpaceX 定的分野。单次任务和持续存在,中间差的是成本和规模。把成本降到某个量级以下,把规模抬到某个量级以上,事情才会从壮举变成生活。

蚱蜢火箭:一枚能自己落回来的火箭

访谈里提到了蚱蜢火箭(Grasshopper)。绝大多数火箭是一次性的,这一枚能垂直起飞、垂直降落,轻轻落回地面。马斯克说,SpaceX 官网能看到这些视频。

为什么要回收?回到火星运输系统那个目标。去火星必须便宜到一个数量级之内,如果一个人要花几十亿美元,火星基地就无从谈起。所以火箭必须能快速复用,尽可能接近完全复用。

然后就是那组成本对比。猎鹰9号造价约6000万美元,推进剂只值20万美元。

把这两个数字放在一起看,结论很直接。一枚火箭几乎全部的价值,都押在只用一次这个前提上。箭体、发动机、航电系统,这些东西的设计寿命远远超过一次飞行,但它们全部在第一次飞行后被丢弃。改变这个前提,成本曲线的形状就变了。这不是省一点钱的问题,这是换一条曲线。

这套逻辑在航空史上有现成的对照。飞机之所以成为大众交通,因为它买一次可以飞几千次,燃料成了主要边际成本。航天一直昂贵,原因在于每一次飞行都在重新购买整架飞机,燃料从来只占零头。两个行业看起来都在天上飞,成本结构完全不同。

顺带说一句,复用这件事的难点不在把火箭收回来,而在收回来之后还能快速再飞。落回来只是第一步,检查、翻修、加注、再飞,这套流程的周期和成本才决定复用到底值多少钱。马斯克在访谈里用的词是尽可能接近完全复用,这个限定词很关键。

纽约到伦敦,五十年没有变快

主持人提到他们之前聊过的一个话题:过去三四十年技术突飞猛进,唯独从纽约飞伦敦这件事几乎没变,甚至更慢。

马斯克同意,确实更慢了。他说自己最喜欢的大型客机是波音747,因为它飞得相当快,而它的第一版是六十年代设计的。从那以后,民航客机没有超过747。他觉得这很荒唐。

荒唐的事为什么能持续这么久?他给了一个组织行为学的解释。

干线客机市场基本是空客和波音的双寡头,这类项目周期极长、投入极大。坐到那两家公司的高管位置上,瞄准一点点渐进改进是更安全的选择。瞄准颠覆性改进,一旦判断错了,你就会被解雇。

这段话值得停一停。它解释了为什么很多行业在技术上明明有空间,却长期不动。工程师没问题,问题出在决策者的收益函数上。失败的惩罚远大于成功的奖励,理性的人就会选择不动。

这个框架可以往外延伸。判断一个行业会不会出现颠覆,看它现在的玩家做颠覆的期望值是不是负的。只要做对了奖励有限、做错了代价极高,这个行业里的聪明人就会集体选择渐进。技术进步的速度,很多时候被激励结构卡住,不被物理规律卡住。

主持人顺势问,那 SpaceX 为什么不去造世界上最快的喷气客机?马斯克的回答只有一句:我们得专注火箭。

一句话同时回答了另一个问题。他清楚激励结构的毛病在哪儿,也清楚自己如果分散精力会掉进同一个坑。

十二年后送第一个人上火星,前提是公司没他也行

主持人问他什么时候能实现把人送上火星的愿望。他说,大概12年后能送出第一个人。

那你本人会是那个人吗?他给了一个带条件的答案:只有在他确信自己死了 SpaceX 也不会出问题时,他才会去。如果他不在了,任务还能继续,那他就去。

这个回答把我和公司分得很清楚。先确认系统能自我维持,再谈个人意愿。听起来像工程师在设计冗余,也像创始人给自己定的一条安全边界。

把它放到当时的语境里看,分量更重一些。2014年的 SpaceX 还没完成火箭回收,载人版龙飞船还在研发,马斯克本人是公司的技术决策核心。在那样的阶段说出这个条件,等于承认一件事:一家公司的成败不该绑在一个人身上。这句话从创始人嘴里说出来,比从投资人嘴里说出来难得多。

三起没有伤亡的起火,和25万起汽油车火灾

话题转到特斯拉。主持人说,最近负面报道很多,给他一个一吐为快的机会:三辆车起火了,你怎么回应?

马斯克给了一组对比。三起没有造成任何伤亡的特斯拉起火,拿到的全国性和国际性头条,比全美所有汽油车起火加起来还多。从去年年中到访谈当天,美国大约有25万起汽油车火灾,造成约400人死亡、约1200人重伤。

数字摆在这儿,落差是几百倍的量级。三起无伤亡的起火,压过了25万起致命起火。

他接着说了一句分寸感很好的话:新技术身上应该打一束聚光灯,但不该打一束激光。

聚光灯和激光的差别在范围和强度。公众审视新技术是合理的,新技术也确实该被盯着看。当审视的强度超出比例几个数量级,它带来的效果就开始偏离本意。一个刚出现的事物,任何一点异常都会被放大成系统性问题;一个存在了一百年的技术,每年烧掉25万辆车、死掉400人,却成了背景噪音。

这段对话里还有一层没被说透的东西:风险的实际大小和风险的感知强度,几乎不相关。人们害怕陌生的、罕见的、被报道的风险,对熟悉的、常见的、没人报道的风险毫无感觉。这是人的直觉,媒体的选择顺着这个直觉走,于是偏差被放大一轮再放大一轮。

顺着这个视角看,新技术公司面对的真正难题,往往落在信任的建立上,技术本身倒在其次。第一起事故会被当作整个品类的证据,第一百起事故才被当作个例。这个不对称没法靠公关解决,只能靠时间和规模慢慢磨平。等到电动车满街跑的那天,人们对它的态度会和今天看汽油车一样平常。

220亿美元市值,和20多亿美元销售额

主持人提到产量:一年3万辆车。也提到市值:几乎和传统车企一样高。

马斯克的回应算克制。他说自己有一次讲过,当时特斯拉的估值超过了它应得的水平,公司会尽力去兑现这些期待。他相信长期看价值会显著高于当时,甚至高于几周前的高点。

但他也补了一句:如果让他断言特斯拉毫无疑问值220亿美元,而今年销售额只有20多亿美元,那就是对未来执行力的巨大预支。

这句话放在2014年看,是一个创始人在主动给自己的股价降温。放在更长的时间里看,它划出了一条评判标准:估值能站多久,取决于未来能兑现多少,当下的情绪热不热几乎不作数。十倍以上的市销率,本质上是让今天的买家提前支付了几年后的成绩。

他从没说市场错了,只说市场给的信用额度很高,需要拿执行去还。这个措辞值得记住。承认预期可能存在透支,同时表态会全力兑现,两件事可以同时成立。

除了火箭,所有交通工具都会电动

主持人回到核心质疑:混动车是一种妥协,电池不够强,也看不出改善的速度能支撑一个电池驱动的世界。

马斯克的回答很干脆,他能看到那个世界。他的判断是,除了火箭这个带有讽刺意味的例外,所有交通工具最终都会完全电动。汽车、火车、飞机、船,全部。

为什么火箭是例外?因为火箭的功率需求和工作环境太极端,能量密度和氧化剂都得自己带。这个例外本身说明他思考的落点在物理约束上,不在立场上。凡是电池的能量密度够用的场景,电气化就会赢;不够用的场景,它就一直输。

他也没有把话说满。转换不会达到100%,今天还有蒸汽机在运转,还有人骑马,但比例极小。汽油时代的结局会和蒸汽时代一样:后人回头看烧汽油的车,就像我们今天回头看蒸汽机车。

这个类比之所以成立,因为蒸汽机并没有消失,它只是退到了极小的角落里。技术更替的常态是旧技术不死,只是变得无关紧要。今天还有人在用蒸汽机,规模小到不影响任何统计。判断一项技术会不会被替代,看它还有什么不可替代的场景,而不是看它今天占多大份额。

我们本来就活在一个太阳能驱动的世界里

太阳能是不是未来?马斯克的回答是超级明显。然后他给了一个视角。

世界本来就几乎完全由太阳能驱动。没有太阳,地球是一个3到4开尔文的冰球,接近绝对零度。太阳驱动了整个降水循环,驱动了整个生态系统。除了极少数例外,所有生命都是太阳能驱动的。

这个回答的高明之处,在于它把太阳能从新能源这个分类里拎了出来,放回唯一的老能源那个位置。风能是太阳加热大气的结果,水能是太阳蒸发水汽的结果,化石燃料是几亿年前封存的太阳能。我们做的所有事,本质上都是在花太阳的账。

顺着这个视角往下看,能源问题的核心就变成了转换效率和储能成本。转换效率决定同样的阳光能拿出多少电,储能成本决定这些电能存多久、运多远。这两件事做好了,能源问题就从一个资源问题变成一个制造问题,而制造是可以靠规模降本的。

这也是他同时押注太阳能和电动车的底层原因。两头接的是同一件事:把能源从地下的存量换成天上的流量。存量有限且分布不均,流量几乎无穷且人人都有。

Hyperloop:比迪士尼太空山还温和

主持人请他用普通人能懂的方式解释 Hyperloop。

他的描述是这样的:你坐进一个舱里,把自己绑在座位上,从洛杉矶去旧金山,感觉有点像迪士尼乐园的太空山(Space Mountain)过山车。但加速度比太空山还小。能坐太空山,就能坐 Hyperloop。

悬浮方式像气垫球桌(air hockey)。区别在于气垫的喷口装在舱体上,不在管道上。所以运行起来平滑如玻璃。

这个比喻选得很准。对普通人来说,真空管道加磁悬浮是一串听不懂的词,比太空山还稳是马上能判断的体验。技术传播的门槛常常卡在这儿:原理讲得再对,找不到一个可感的参照物,听众就没法把它放进自己的经验里。

顺便说一句时间尺度。他在2013年的白皮书里给出的数字是洛杉矶到旧金山35分钟,访谈里没有重复这个数字。这个沉默可能是有意的,白皮书引发的讨论里,最多的质疑集中在造价、征地和安全上,那些问题比35分钟更难回答。

十年后的马路,和两栖动物的比喻

主持人问,十年后汽车出行会是什么样。

马斯克说,路上会有大量电动车,比现在多得多,而且会处在电动车普及曲线的陡峭段。他提到了无人驾驶,但没有展开,重点还是放在电动化上。

混动车会不会消失?他给了一个生物学比喻。混动车像两栖动物。生命从海洋走向陆地的时候,两栖动物有它的位置,它是必要的过渡形态。但到最后,剩下的两栖动物很少。

主持人说这个比喻适合收尾。确实适合。它同时承认了两件事:过渡形态有历史价值,过渡形态也不会永存。

回头看这次访谈,最能站住脚的,是他反复用的那套方法。看一个行业,先看它的成本结构;看一段停滞,先看决策者的激励;看一场争议,先看基率;看一个过渡方案,先看它替谁解决了什么问题。这些方法不需要认同他的任何一个目标,也能直接拿来用。

  • 评估一家公司时,先拆它的成本结构。找到占比最大的那一项,问一句这一项有没有可能被复用、被压缩、被绕开。猎鹰9号的故事就从这个动作开始,6000万里有5980万押在只用一次这个前提上。
  • 看到一个行业长期没有进步,别急着归因于技术不行。先看决策者的收益函数:做成颠覆性改进的奖励有多大,失败的下场有多重。奖励和惩罚不对称的地方,就会长期不动。
  • 读到关于新产品的负面新闻,先找基率。三起无伤亡起火对上25万起致命起火,这个对比能让你在一分钟内判断报道的比例是否失真。
  • 判断一个过渡方案值不值得投,用两栖动物的框架:它替谁解决了什么问题,这个问题在被替代方案彻底解决之后还会不会存在。会,它就是长期方案;不会,它就是过渡。
  • 给自己的判断加一条检查:如果我不在了,这件事还能不能继续。这条检查在马斯克那里用来决定要不要上火星,在你这里,可以用来判断一个项目、一个团队、一门生意是否真的成立。
  • 遇到一个需要向非专业听众解释的技术,别急着讲原理。先找一个对方已经体验过的参照物,把太空山放在真空管道前面。
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)。

0:00Elon Musk pleasure to have you on thanks Perry so this is the dragon that goes up into space and is going to go up on Monday correct yeah that's our dragon spacecraft version one which is capable of taking cargo to and from the space station including a black biological cargo like fish and - and that kind of thing so when do when do human beings start going up and do you think so we expect a complete version - of Dragon which will have astronaut transport capability in about two years most of what you do now is send satellites up into space right that's how you pay the bills exactly but you want to do something much more ambitious with SpaceX yeah I mean the

0:39long term aspiration is to develop the technologies necessary to transport a large number of people in cargo - Tamar's in order to create a self-sustaining civilization there and then that's really why I started the company was because it seemed as though to create the possibility for life on other planets yeah exactly I mean it sort of started off when I was thinking about what should you have to PayPal and I had always been interested in space but I didn't think there's anything that an individual could do in space I mean it seemed like the province of large governments and but I sort of

1:16started looking into it and I went to the NASA website to say well to find out when we're going to Mars it seems like obviously the head is the next thing after the moon and I couldn't find anything so like wow what's this is seems very strange initially I was under the impression that that it was a then we'd lost the will to do that and I later came to conclusion that I was quite wrong about that I think the United States in particular as a nation of explorers and do you do you look at it as why did you decide transport was the most important thing I think we would be living on Mars there's no that's that's the relatively easier easier thing yes yes absolutely so we

1:57could live on Mars just getting there is the problem yes but getting I mean right now getting to Mars is impossible so like it kind of doesn't matter what you know what you do when you got you what you do when you get there if you can't get there so first orbán's is to figure out how do you how to get there and it needs to be in a way that enables large numbers of

2:20people and cargo can't just be like a handful of people because that's obviously not going to create a self-sustaining civilization the ana paula was an amazing inspiring thing for all of humanity but the last time we went to the moon was like 1973 l4 I believe so we don't want to just have flags and footprints or and and then never go to Mars the game if we just have one mission that that that will also be a super inspiring thing but it's not going to fundamentally change the future of humanity so you have this grasshopper rocket which which unlike most rockets which which can't you know which I can be usable this one gently comes back down in a kind of vertical and yeah

3:04absolutely and you can see the videos of this on the SpaceX website the the fundamental breakthrough going back to the point of that developing a Mars transportation system it has to be affordable to go like it can't just be like billions of dollars per person to go to Mars then that you you just no way you could establish a base on laws at that cost so we have to develop rockets that are rapidly and as close to completely reusable as possible as an example the

3:41Falcon 9 cost about sixty million dollars I mean sort of like a yet you know it's like but the cost of the propellant is only about two hundred thousand dollars so it's just like you can imagine how expensive it would be if you had to buy a new plane every time you went somewhere a very few people would fly but refueling a plane is pretty easy we once talked about airplane travel and I asked you why it is that with all this technological improvement over the last 30 or 40 years the one thing that doesn't seem to have improved and it is in the domain of the private sector sort of is Jen from New York to London it takes about as long as it took 40 or

4:2150 years ago in fact it takes actually that this is it it actually just take longer because in the 70s like my my favorite sort of commercial air airliner is the 747 and because it actually goes quite fast and it's actually incredible that that that thing was I mean the first iteration of it was was designed in the 60s I mean since then I don't think we've exceeded the seven-point the 747 which is nutty well it for the commercial

4:51airliner business you essentially have a duopoly between Airbus and Boeing and these big airplane programs are really long term and they're quite expensive and I think if you're in the senior management of one of those companies it's it's it's a safer bet to take it to just aim for a little incremental improvement than to try to aim for a radical improvement because generally if you if you aim for a radical improvement and you're wrong you'll you'll get fired why doesn't SpaceX create the fast the fastest jet plane in the world we've got to focus on rockets when do you think you'll be able to achieve the aspiration of SpaceX of actually moving

5:30people to a place like Mars well I think we could probably send the first person in about 12 years will you be that person only if I'm confident that SpaceX will be fine if I die that's you know if I just maybe if I first confident that the mission would continue if I wasn't around then I would do it you've heard all the press about Tesla so let me first give you a chance to get it off your chest they're like pistol-whipped three cars caught on fire what's your response so the amount of national international

6:12news headlines are dedicated to three tesla fires that caused no injury is is greater than all of the gasoline fires that occurred in the in the united states and and with all the other cars yeah which from mid last year to today is about a quarter-million gasoline car fires which caused about 400 deaths something like 1200 serious injuries out of three non injurious fires got more national headlines been a quarter million deadly gasoline car fires that's mad what the heck is going on I mean I realized like a new technology should just should have a spotlight on it but it shouldn't have a laser on it so when you look at Tesla the big concern many people have is scale you're

7:08producing about 30,000 cars a year for two years producing oh yeah well millions so way too much attention and your muscle cap is almost as high as some of these car these cards yeah it's good I think the recent weeks but yeah do you think it's crazy or do you think it's it's an appropriate indication of future growth yeah I mean what if you know what it actually said on one on one occasion is that then I thought our valuation was more than we had any right to deserve that we would do our best to

7:41to fulfill those value expectations and that I thought I thought over time the value test would probably significantly exceed where where it is today or even where it was a relatively high point a few weeks ago but but it would be I think silly for me to assert that Tesla is unequivocally worth at one point where like worth twenty two billion dollars and we'll have you know somewhere over two billion dollars in sales this year that that is an enormous amount of credit if you for future execution the heart of what interests you about Tesla at least as I recall when we've talked is that it's not a compromise that that a hybrid ism is something of a compromise between an

8:24internal combustion engine and a battery powered engine and and here you can do it cleanly but batteries just aren't that powerful enough for and they don't seem to improve enough to imagine a world powered by batteries right III certainly can see that and in fact my opinion is that all transportation with the ironic exception of rockets will be fully electric on vez lanes yes everything when planes trains by when everything it'll never be like a 100% conversion like we still have some steam engines and some people still ride horses but it's a pretty tiny percentage so I think it'll be the same with gastly and I think in the future who will look back

9:08on the gasoline era the way we look back on the steam era today so you're using battery technology your deal you're doing Tesla which is an electric car you have solar panel you have an interest in business and solar panels yeah do you think solar energy is the future I think it's super obvious in the future absolutely in fact I mean if you think about it the

9:32world is almost entirely solar powered already in that we would be a frozen ice ball at about 3 or 4 degrees Kelvin when you're close to absolute zero without the Sun and the Sun powers our entire system of precipitation the the ecosystem yeah but from the tiny number of exceptions almost all life is solar-powered all right another great technology you've talked about that most people think of as out of a movie explain how it would work let's say it's new it's New York or Boston or LA to San Francisco I get into a achoo and it's

10:10almost like air hockey the way it works is I strap myself into a seat well it's it would actually feel maybe like like the Space Mountain ride at Disneyworld and the g-load anywhere I should be less than than what people experience at Space Mountain so if you can handle Space Mountain at Disneyland you can handle which people to handle it the Hyperloop it'll feel super smooth because as you mentioned it's it would use air skis like it like a air hockey table but with the the air jets on the on the pod side as opposed to the the tube side so it would just be it would be smooth as glass ten years from now what would car travel look like an ten years from now I think there's

10:56gonna be a lot of electric cars in the road well we're certainly faster more than there are now driverless I think I think we'll be at in the steep portion of the adoption of electric cars in in ten years you think hybrids are a transition that will fade away you leave yeah they're like the like a sort of an amphibian yeah I mean is there was a role for amphibians when life was moving from the oceans to land but what's but been in the end very few amphibians remain it's a good good boy to end on thank you

资源与出处

相关视频

同频道、同主题或相邻年份