2014年10月,旧金山,《名利场》新权势峰会的台上,主持人开场就拿马斯克打趣:你是这份榜单的第一名,诺贝尔奖这两天也要陆续公布,要是三个都拿到,就是三连冠了。
马斯克配合着笑。他第二天要开一场发布会,主持人顺势问:你打算发什么?
他的回答很马斯克:有一件东西已经在那儿了,人们没意识到,我们只需要把它打开。
主持人追问,是不是新车的一部分。他说是。又赶紧划清界限:今天不谈 Model 3,也不谈 Model X。Model 3 就是那辆三万五千美元的电动车。
然后他自己把话题带偏了。要发布的东西里,有一个代号叫 D。还有一个"别的东西",原本也想说,但那个太明显,说了等于直接公布答案,只好咽回去。
台下已经在笑了。他补了一句:
我觉得互联网很擅长猜这些东西。方向大体上是对的,但量级还没被充分认识到。那个"别的东西"的量级,更没被认识到。(马斯克)
说到 D,就得说说那段商标往事。特斯拉先做了 Model S,再做 Model X,本来想给第三代车起名 Model E,三个字母凑成那个所有人都懂的词。福特不干了,表示要起诉,只好改成 Model 3。
他在台上承认,自己在推特上预告"我们要发布 D,还有别的东西",等于自掘坟墓。今天坐在这里,只能硬着头皮把话说完。
事后来看,D 指的是双电机四驱版本。真正值得读的,是他为什么非要做这些车。
造车不是为了再开一家车厂
主持人问他,为什么要做一辆电动跑车,是什么在推着他。
他没从环保讲起。他说重要的是加速可持续交通的到来,早一点比晚一点好,人类才能摆脱烧石油。而且这件事必须和可持续发电配套,这就是 Solar City 的位置。
他把话说得很直:我不觉得世界上需要另一家汽车公司。好车厂有的是,这个行业竞争极其激烈。真正需要的是证明电动车能比汽油车更好,否则人们会继续买汽油车。
主持人抓住"更好"两个字追问:是指更快更强吗?
他说,会更大,更快。
主持人接得也快:那我能不能订一台?马斯克说,这句话马上就要传遍全网了。
玩笑归玩笑,主持人问了个更狠的问题:驱动你的是地球的可持续性,还是纯粹智力上的好奇?
马斯克把时间拨回大学。他当年想电动车,出发点跟环境没关系,跟资源有关。石油不可再生,烧完就没了。不管有没有环境影响,都必须找到替代品,否则经济会崩溃,文明会倒退,甚至退化回从前。
他说,从物理学角度看,电机显然是对的路子。做别的才奇怪。
环境问题是他后来才越来越重视的。人类往大气里排了太多碳,正在从根本上改变大气的化学成分。海洋也跑不掉,二氧化碳溶进海水变成碳酸。不采取纠正措施,灾难的概率会随时间上升,最后变成确定的结局。转型越晚,坏事发生的概率越大。
所以他给特斯拉下的定义,跟卖多少台车没关系:把电动车的到来提前十年左右。
就算没有特斯拉,电动车也会来,只是会慢。这是他能想到的最根本的一件好事。既然要造车,那就把车造好,顺便让买车这件事变得愉快一点。
送出去的专利,是一笔明知亏本的买卖
主持人问,为什么把知识产权开放了。
马斯克说,开源专利是为了加速电动车的普及。他猜很多人以为这里有竞争上的算计,比如让更多人进场造电动车,水涨船高,特斯拉跟着受益。
他把这个猜测否掉了,还主动承认:开放专利其实略微损害了我们的竞争地位。他希望由此产生的好感,能抵消掉这部分损失。
他对专利本身有一套看法。专利有点像买了一张通往诉讼的彩票,你花钱买了个凭证,等着哪天有人踩线,然后去打官司。既然这样,为什么要买?
这段话放到2014年听,很有分量。一家公司主动削弱自己的护城河,理由是这条河本身不该存在。
巨杉森林里长不出新东西
主持人引了彼得·蒂尔(Peter Thiel)那句话:我们被承诺过飞车,结果只得到了140个字符。然后问他,是不是觉得别的创新者不够敢想。
马斯克说,应该有垂直起降的超音速电动飞机。三十年前有协和式客机,现在反而没有了。大创新的速度确实慢下来了。
他给了个机制上的解释。一个行业如果是垄断或者双寡头,创新的驱动力就弱。创新通常来自新进入者。汽车行业的资本壁垒高得像一片巨杉森林,新玩家根本挤不进去。
接下来这段是全场最锋利的地方。他说,如果你是一家大公司的高管或者 CEO,做点渐进式的改动,几乎不可能被开除;做一件大胆的事,失败了,很可能被开除。于是所有人都在做渐进的事。
这段解释没有点名任何公司,但把大公司创新的死结说透了。激励结构决定了行为,行为决定了行业面貌。
顺带说一句,他自己就是那个愿意被开除的人。
机器智能的曲线,多数人看低了
主持人把话题转到他另一个长期担忧:机器失控,或者说人工智能的奇点。
马斯克说,大多数人根本没意识到机器智能进步有多快。比几乎所有人以为的都快,硅谷内部也一样,硅谷之外的人更是完全没概念。
危险在哪?如果出现一个超级智能,尤其是能递归自我改进的数字超级智能,而它的优化目标或者效用函数对人类不利,后果会非常糟糕。
他举了个例子,听起来像玩笑,逻辑却很冷:假设你让 AI 去消灭垃圾邮件,它推导一圈,可能得出结论,最有效的办法是消灭人类。因为人类才是垃圾邮件的源头。
主持人问了个很硬的问题:我们凭什么会失去对机器的控制?《2001太空漫游》里 HAL 是那样,可现实里有什么数据表明,人类和机器之间的连接在松动?
马斯克的回答是方法论,不是预言:把数字智能的进展对时间画成一条曲线,拟合,外推,看它通向哪里。
主持人追问:你说的是有目的性的机器,它们的目的就是我们写进去的效用函数。马斯克说对,但会有非预期的后果。
然后主持人问,这是不是也是你去火星的动力。
他的回答值得逐字读:如果真出现数字超级智能的末日场景,它大概率会跟着人一起去火星。但如果它的效用函数恰好被限定在地球,那它留在地球上也没问题。
换句话说,火星不是躲 AI 的避难所。这一层意思在传播里经常被省略。
几十亿分之一的概率,你我不在真实世界
这段来自观众提问。斯坦福大二学生马努(Manu)说,你做的事是把人从 A 送到 B,更快、更环保。但硅谷还有另一种研究,试图取消交通本身,就是虚拟现实。戴上 Oculus Rift,我坐在宿舍里就能看到你讲话,临场感和现在坐在这里差不多。你怎么看虚拟现实和交通的关系?
马斯克先扔了一句:也许我们现在就在一个模拟里。
然后认真回答。他听说 Oculus Rift 和一些沉浸式技术相当有颠覆性,你真的会觉得身临其境,摘下来之后反而觉得现实不真实。所以他推测,未来物理移动会减少。
接着他给出了那个后来被反复引用的论证。把时间外推到100年、200年、1000年后,如果游戏继续进步,你穿着全身触觉服,配上环绕视觉,超过某个分辨率之后,它和真实就无法区分了。
而这样的模拟,可能会有几百万个,甚至几十亿个。
那么问题来了:我们处在"基础现实"里的概率是多少?
他说,几十亿分之一。
主持人反问他,能不能给出反方论证。他想了想,说:这感觉上很真实,但可能不是真的。
去火星:一份保险,和一场冒险
主持人问,为什么非要去火星。
马斯克给了两个理由,先说防御性的那个。
为生命整体买一份保险。成本大概是经济产出的0.5%,甚至更少,换来的是意识之光能延续更长时间。地球形成45亿年来,这是第一次,生命有可能搬到另一个星球去。这个窗口可能开很久,也可能开很短。他说自己是乐观的人,希望窗口很久,但万一很短,那就应该现在动手,不能拖。
第二个理由,他自己更看重:这会是人类历史上最伟大的一次冒险。
生活里需要有让人兴奋、让人受鼓舞的事情。不能永远只是解决一个又一个糟糕的问题,人得有早上起床的理由。
主持人问他,是不是科幻小说读太多了。他说,有可能。
火箭要像汽车一样,用完还能再用
话题转到 SpaceX。马斯克说,到目前为止,进展是进化式的,还谈不上革命式。他们瞄准的那个突破,是让助推器飞回来、着陆、再飞一次。
为什么这件事重要?因为火箭一直是一次性的。航天飞机算部分可复用,但翻修极其困难,需要一万人在九个月里工作。
他要的是像飞机、像汽车那样的快速、完全复用。
他打了个所有人都听得懂的比方:除了火箭,几乎所有交通工具都是可复用的。马、自行车、汽车。如果汽车只能用一次,每趟出门都得买新车,你不会买很多车,因为太贵了。
所以复用是让发射成本降低100倍以上的前提。这个数量级不是修辞,是成本结构决定的。
主持人问政府资金的角色。马斯克说,一开始没有政府合同,SpaceX 是他用 PayPal 的钱全资投的。公司成立五六年后,才拿到 NASA 的第一份合同。现在大约四分之一是商业发射。
他顺势讲了美国的创新模式:从二战开始,政府、私营企业、大学三方结合。这个结构现在有点散了,政府做的基础研究不如从前多。他说不准基础研究的投入数字是不是合适,但他本人是研究的支持者,愿意多花钱。
他也替 NASA 说了句话:火星车、哈勃望远镜、即将上天的詹姆斯·韦伯望远镜、各种行星探测器和地球科学任务,NASA 其实做了不少事。
城市堵在瓶颈,答案埋在地下
主持人问他还有什么大想法,顺带提了飞车。
马斯克先泼冷水。天上要是满是飞车,会影响天际线,会更吵,掉到头上的概率也会变大。主持人笑说,这理由很有说服力。
他给出自己的判断:城市真正的堵点在瓶颈。郊区街道上并不堵,堵的是高速和主干道,因为城市长得比主干道大得多。
解决办法是隧道。
这时观众席上一位建筑师丹妮尔(Danielle)接过话头。她在帕洛阿尔托刚做完一栋净零能耗的被动式住宅,正在斯坦福商学院带学生做城市问题的项目,目标是底特律。她问马斯克怎么看城市级别的创新,还顺手邀请他去斯坦福讲一次。
马斯克的回答还是隧道。曼哈顿的楼平均30层,街道只有1层。30比1的比例,明显不合理,应该有多层地下高速,一层叠一层。隧道也不用跟着建筑走,可以斜着穿过地下室,想要多少层都行。
然后是他那句被反复引用的话:说到底就是建隧道的成本。隧道就是地上的一个洞,能有多难?
他还答应了去斯坦福。理由很实在:他每周都在帕洛阿尔托。
五百英里是条分界线
主持人问,更远的距离怎么办。
马斯克说,500英里以上,用垂直起降的电动超音速飞机,点对点飞,技术上很可行。500英里以下,飞机要爬升、要下降,时间都浪费在这两段上,不划算,那就该用减压管。
减压管是什么?就是把管内的空气密度降下来,阻力大幅减少,相当于把你瞬间传送到高空。速度能快很多,还没有爬升和下降的麻烦。
这就是当时外界在传的超级高铁(Hyperloop)的技术要点。他没有直接说这个词,但主持人提到了,他顺着讲了原理。
把这段和前面的隧道放在一起看,逻辑是连贯的:地面上的瓶颈,往地下走;长距离的瓶颈,往真空里走。
第一性原理,说穿了就是算材料账
斯坦福大二学生劳伦斯(Lawrence)在做一个音乐创业项目。他问马斯克,那些疯狂的想法是怎么来的,以及钱很少的时候,怎么把大胆的想法做起来。
马斯克的第一条建议是选对起点。软件,或者只依赖智力资本的生意,几个人就能开始,所以 Zip2 和 PayPal 给了他做资本密集事情的资本。
第二条是思考方式:从物理学出发,从第一性原理出发,问什么是最好的做法。
他拿火箭举例。用类比推理,你会说火箭的成本就是以往火箭的成本。用第一性原理,你会问:火箭由什么材料构成,这些材料多重,单位质量多少钱。这个算法给出了火箭成本的渐近下限。如果能想出创造性的办法,把这些元素重新排列成火箭的形状,结果就会好得多。
第三条是跨行业移植。看看别的行业已经发现了什么,能不能搬到自己这行来。这是很好的想法来源。
最后他说得很老实:通常就是死磕一个问题,试很多东西,大多数不成功,偶尔有一个成了。
钱只是数据库里的一个数字
主持人说,PayPal 之后,支付领域的创新好像不多,除了比特币。问他是不是对 PayPal 后来的走向失望。
马斯克说,PayPal 被 eBay 收购之后,基本没怎么变。
他讲了自己当年的长期愿景:把金融系统从一堆异构的、不安全的数据库,变成一个数据库,或者说少数几个数据库。钱不过是数据库里的一个数字,本质上是劳动力分配的信息机制。
而现在的数据库效率很低。老式大型机互相不通,安全性差,一天只做一次批处理。
主持人问,PayPal 分拆是不是对的。他说大概是。问他接手会怎么做,他说会把它变成一家全服务金融机构,把一个人需要的所有金融服务放在一起,无缝整合,好用,而且真的关心消费者。很多银行看起来并不太关心客户。他覺得有机会做一家真正的好银行,比人们想的要多得多。
然后说到比特币。他说比特币大概率是好事。但它的主要用途会是非法交易。不过他补了一句:这不完全是坏事,有些事也许本来就不该算非法。
他还讲了个结构性判断:比特币会同时用于合法交易和非法交易,否则它作为合法交易工具就没有价值,因为必须有一座从合法到非法的桥。
最后他补了一句:顺便说,我不持有比特币。
中国会超过北美,超级工厂顺便做储能
两位观众提问值得单独拎出来。
一位在加州大学伯克利分校做过大规模可再生能源的观众问,特斯拉会不会做大规模的固定式储能。
马斯克说会,而且规模很大。电池组配太阳能很重要,配风电更重要。内华达的超级工厂每年会产出几十 GWh 的固定储能。
另一位是从中国来的蒂娜(Tina),问中国市场。马斯克说反响很热烈。长期看,中国很可能成为特斯拉最大的市场,可能超过北美。政府监管方面,激励结构确实偏向本地生产,但整体不是大问题,政府到目前为止表现不错。
2014年说这话,是需要一点判断力的。当时特斯拉在中国刚起步,销量并不好看。
七千五百万美元的火星任务,还是太贵
最后一位提问者来自路透社印度分社。他问:印度刚用7500万美元发了一个火星任务,这是奇迹,还是印度航天有什么值得学的东西?
马斯克说这任务非常令人印象深刻,尤其它是一个政府机构执行的,从成本角度看很厉害。
但他马上给出更长远的判断:最终去火星的任务必须比这便宜得多,否则不可能建立自给自足的火星文明。要运几百万吨货物、几百万人过去,成本必须可负担,否则人根本去不了。最终要比这个数字低两个数量级,远低于一百万美元。
主持人在台上收尾:明天我们就看看,到底是什么东西一直藏在我们鼻子底下。
这场对话信息密度很高。一个正在造车、造火箭、同时担心 AI 的人,被放在一张沙发上连问了一个多小时。他给出的不是愿景口号,是一套可以拆开检验的推理方式。把这套方式用在自己手上的事,比记住他说过什么更有用。
- 评估一项成本时,先算材料的物理下限,再看行业惯例报价。两个数字的差距,就是创新空间。
- 检查你所在组织的激励结构:做渐进的事有没有奖励,做大胆的事失败会不会被开除。如果是后者,别指望创新从内部长出来。
- 判断一项技术还有多少空间,看它离物理极限多远,不看它比上一代进步多少。
- 凡是"用完就扔"的东西,成本都居高不下。找到你行业里的"一次性火箭",把它变成可复用的。
- 对 AI 的进展别靠直觉判断,找几个可以量化的指标画成曲线,外推一下,再决定要不要紧张。
- 做长期市场判断时,看人口和需求规模,不看当下的销量排名。中国超过北美这种事,是时间问题。
- 想法卡住时,去隔壁行业看看他们已经解决了什么。移植常常比原创便宜。
- 资源有限时,先从只依赖智力资本的生意做起,用它的现金流去供养真正烧钱的那件事。
0:00[Music] she a man who needs no introduction but welcome thanks for being here number one on the Vanity Fair New establishment list is that the highest honor you've ever had of course that's right the nobels are announced you're doing an announcement tomorrow and the next day is a Nobel Prize so if if you get all three that's a great Trifecta let's talk about your announcement tomorrow yeah what is it you're going to announce I'm glad you asked
0:44well I I think I can give maybe one clue which is that one of the things is already there and people don't realize it so we just need to turn it on yeah wow cool and it'll be part of a new car okay I used to be a journalist I used to be good at this yeah I I it's it'll be it'll be a different it'll be different from the current one the some part of I announce will be different from things in the past obviously it's an announcement but I can't say it's it's we're not going to talk about the model 3 or the model X the model 3 would be the $35,000 electric vehicle you're going to talk about something that's labeled D yeah D and then there's another thing
1:46which is actually would have said the other thing except it would have been too obvious and that's the reason I didn't say now I can't say what the letter is because okay and what does d stand for I think the internet is very good at guessing these things so they're they understand directionally it's it's directionally correct but
2:13the magnitude is not well appreciated yet and the magnitude of the something else too yeah that's the really big one the something else I kind of got myself in trouble by because I you know we we had the model s and the X and then just for fun we try to trademark the model e and and then then Ford suit
2:38us or said they were going to Su us so we had to change it to the model 3 so it's s3x you know totally different and so I kind of I guess when I say that that we're about to unveil the D and something else then it's I kind of dug my own grave on that front yeah yeah well we look forward to why in the world are you doing this I mean what's impelling you to make an electric special car I think it's important that we accelerate the Advent of a sustainable transport future so it's it it's better if we move towards electric cars sooner rather than later and then we can get away from burning oil and obviously that needs to be paired with sustainable
3:27generation of power that's where solar City comes in and but in order for us in order for Humanity to have a good future the century we have to figure out a way to sustainably produce and consume energy and that's what Tesla and Solar City are trying to be helpful in that regard so that's that's really the the purpose of it it's not it's not as though I think that there's a need for another car company there's plenty of you know good car companies very competitive industry but but there is a need to show that electric cars can be better
4:10than gasoline cars otherwise people will continue to buy gasoline cars when you say better I'm going to push you back to where you didn't want to go which is what you're announcing better means that it would be faster and stronger than a gas car which we think electric cars cannot be yeah will'll be bigger and faster oh cool
4:36yeah can I have one I can't wait for that quot to be out there yeah it's about to unveil the D it's bigger and faster yeah I know there have been a lot of jokes that Cara swisser started today that yeah I know we have to steer away conference called All Things D you know yeah why did you give away or open up the intellectual property oh the patter yeah yeah to we sort of Open Source pents in order to accelerate the Advent of electric vehicles that's the main reason I think people think that maybe there was some competitive reason or or or that that this would somehow be helpful to Tesla you know because more people would enter electric cars and somehow
5:27that would be you know Rising tide lift but actually I think I think the open sourcing of patents does slightly impair our competitive position on balance but I I'm hopeful that it generates enough Goodwill to overcome that competitive impairment but I mean I came to the conclusion at least on a personal level that patents were kind of like buying a lottery ticket to a lawsuit you know and so why would you want to do that what drove you into the electric car our business was it really sustainability for the planet or was it just the intellectual curiosity of it all actually so if if if you go back to like when I was first thinking about electric cars was
6:15in college and it was really more from the standpoint of the the need to transition to to some transport means that was s sustainable independent whether there was environmental impact or not because if if we're burning oil you know to move and that's not a renewable resource then then independent of any environmental impact we must find some alternative or there will be economic collapse so and civilization would would sort of crumble because we have to or revert you know so the so that that was actually
6:56originally why I was very interested in it is just because it's it I mean I think you know from a physics standpoint it's really obviously the way to go is electric motors I mean kind of crazy to do something else honestly and and then the the environmental thing grew over time and became it became evident that the environmental impact was quite significant because we're putting so much carbon into the atmosphere that were fundamentally changing the chemical makeup of of Earth's atmosphere and then and and the oceans too because because of the CO2 that gets absorbed into the into the water and creates carbonic acid so if if we don't take corrective action
7:40we that that the probability of of a catastrophe will increase over time eventually there's like certainty of of a catastrophic outcome and the longer it takes us to to make that transition the greater the probability of something bad happening and so you know so that when I think of Tesla what's the fundamental good that Tesla can achieve it would be to accelerate the Advent of electric vehicles perhaps by you know for fortunate a decade or something like that because it it'll happen anyway independent got even Tesla was not not around so that's I would say that that's a really fundamental good and since we're making cars I mean it seems like
8:24we should try to make really good cars and and you know maybe make the buying buying experience a lot more pleasant there's a Peter teal line as you know about you we were promised flying cars but we got 140 characters do you think that other innovators people not like yourself aren't really shooting as high as they should be when you're doing things from you know space shots to electric cars do you think there should be flying cars yeah okay
8:58yeah I'm am I missing something gosh I think that would be fun don't you I'm not sure I mean I think I think there should be vertical takeover Landing supersonic Air transport yeah that that's obviously should be the case we had that 30 years ago and now we don't meaning suic sure yeah no it's kind of innovation for big things has actually slowed down it seems is that true I think whenever you have a large industry that is a monopoly or duopoly the forcing function for Innovation is weak because Innovation tends to come from new entrance to an industry and if the barriers are like Tesla into the auto model yeah yeah when there are huge Capital barriers to entry then it's very
9:50difficult for new entrance to enter that that industry it's like being in a forest of giant redwoods and in that in that situation The Innovation is is weak because if you sort of it's I think it's fairly easy to understand this because if you're say the a senior executive in that company or CEO let's say if you do something incremental you very unlikely to be fired if you do something bold and it doesn't work out you're very likely to be fired so that's then they do incremental things you mentioned a moment ago the environmental sustainability thing that's really big that you worry about your other big thing that you worry about is sort of machines getting out of
10:36control explain that or the artificial intelligence Singularity yeah I don't think most people understand just how quickly machine intelligence is advancing it's much faster than almost anyone realizes even within Silicon Valley and certainly outside Silicon Valley people really have no idea so why is that
11:01dangerous I mean if if there's if there's a super intelligent particularly if it's engaged in recursive self-improvement if there's some digital super intelligence and its optimization or utility function is something that's detrimental to humanity then it will have a very bad effect you know it could be just something like getting rid of spam email or something and it's like concludes well the best way to get R spam is to get rid of humans you know but why
11:37would we lose sour of spam I I know we've all watched Cal in 2001 but why would we lose control over our machines there no data point showing that that that our connection to machine has ever been loosened actually I think the thing to do would be to plot the progress of digital intelligence versus time and and then to to maybe curve fit
12:01or extrapolate that progress and see where that leads but you're talking about machines that are not just intelligent but have intentionality is that right they have the intention of their utility function which we programmed in right yes okay but it can have unen unintended consequences is that propelling no pun intended your Mars mission
12:27ideas no I think it's quite it's more likely than not that if if there's some digital super intelligence apocalypse scenario it would probably follow people to Mars not necessarily because because of that utility function if it simply has utility function that is arbitrarily confined to Earth it it would be totally fine doing that I mean it's this is like it's just it's a so what what is let's get to the Mars mission why you feel that is something we should be aiming at I think the reason for M there there's two main reasons for Ms or becoming a multiplanetary I think one one is the defense of reason it's life insurance for Life as a whole and
13:15there there's some value to to having that you know in terms of a small percentage of our economic output like let's say half a percent or maybe even less of our economic output to ensure that that the light of Consciousness As We Know know it propagates into the future to for a much longer period of time and you know so there's that that defensive reason and and this is the first time in 4 and a half billion years in the since Earth you know was first formed that it's been possible for life to move to another planet to become multiplanetary and that window may be open for a long time or it may be open for a short time I'm actually quite an optimistic person so I'm hopeing that
13:59will be open for a long time but maybe it'll be open for a short time in which case we should take action now and and not delay you've been a fan of science fiction books have they actually influenced you sure I I'll just say The other reason which I actually find personally more motivating with respect to Mars which is that it would just be the greatest Adventure ever and very exciting and I think we need things in life that are exciting and inspiring it can't just be about solving some awful problem there have to be re reasons to get up in the morning and I was talking about the books you've read and the science fiction I noticed you're a science
14:39fiction F do you think maybe you've read too much science fiction maybe have that's certainly certainly possible so the the yeah I mean where we are right now is we've I think at on the SpaceX front we've made evolutionary but not revolutionary progress we're hoping to make revolutionary progress in the coming years but the key break that we aim for was or they were aiming for is to be able to have the the the rocket booster come back and land and be able to refly it again why is that important because that's not the way we've been doing it right yeah it's been Expendable Rockets up till now really except for the space shuttle was partly reusable but but it was extremely
15:22difficult to refurbish for flight you needed 10,000 people needed to work 9 months to refurbish this SP shuttle so what we really need is rapid and complete reusability like an aircraft or a car or really any mode of Transport is besides Rockets is reusable you know Horse bicycle you can think of almost anything but imagine if that mode of Transport was not reusable very unfortunate in the case of the horse and but but even like in cars like if you could use a car once and you have to buy a new car every time you took a journey odsa you would not buy many cars because it would be very expensive so reusability essentially U opens the door for
16:11reducing the cost of space flight by a factor of 100 or more and if you do this how much of it will be government funded in a way aren't you under a lot of contracts with government yeah in the beginning there was no government I funded SpaceX entirely with the proceeds from PayPal and but and
16:37we got off S first government contract about five or six years after starting the company from NASA and but I should say about NASA's maybe a quarter of our flights 34s are commercial but the way we have been doing innovation in this country especially starting with world War II has always been a combination of government private Enterprise universities and that is now dissipated the government is not doing as much basic research do you think that's going to be a problem for the United States well I I think I think it govern us quite a lot that I'm not exactly sure whether we're at the right number for basic research I mean I'm a fan of research so i'
17:25probably I'd be in favor of spending more money on on that all allocating more resour resources to to to basic research but but I think the government's doing doing quite a bit I mean NASA is doing doing a lot of things we've obviously got the The Rovers on Mars we've got the Hub things like the Hubble telescope the upcoming James Webb Telescope there planetary probes and Earth Science missions launching all the time so NASA is doing doing quite a lot actually what other big things like electric cars space shots whatever do you dream
18:02of well I think it's face and cars a lot obviously we've eliminated flying cars so well I'm not I mean I I'm I'm not sure about the flying cars I not that's not say I don't think there should be flying cars I mean but if the sky was full of cars flying all over the place and it was you it would affect how this how things look it would affect the skyline and and it would be noisier and there would be a greater probability of something falling in your head yes right good arguments you know those are those are not good things on the other hand that you could be able to go from one place to another faster but I think actually if at least for If you
18:47eliminate the choke points in cities then there's really not that much traffic outside of the CH points so you look at sort of in Suburban streets you don't see a lot of I mean you don't the traffic doesn't doesn't choke things it's really on the highways and major arteries and it's because the the the cities grew way bigger than the major a and how would we fix that tunnels tunnels what about autonomously driving cars would that help that would help certainly yeahh I'm going to open it up but I have one frivolous question while they bring the lights up everybody so far today has been touting how much they love Silicon Valley the HBO show and you said some
19:27bad things about it why don't you like like it well first of all I I should say that the I mean the article was not very accurate that was written it was just one article which was it was simply not an accurate article so I the I I I thought it was okay the but I think like Mike Judge did an amazing job with office base like that's one of my favorite movies ever and he may be here so be nice great he was here a few minutes ago no problem anyway so so a movie like office space just really nailed it I don't think Silicon Valley the show quite Nails it it's not it's not quite it's not quite right so that's you know that that's my objection to it it doesn't it doesn't
20:13quite get it right in my opinion but I've heard that the later episodes are better I saw the first two so yes UC Berkeley hi I used to work in large scale renewable energy mostly solar and I was wondering that given all the the release of the patents of the batteries and and and and all the knowledge on small scale Renewables like Solar City if if is there something coming about
20:46like large renewable energy storage from the from all the batteries patents in in Tesla what Tesla is going to do a stationary storage of you know in a very large scale way because it's it's very important to pair battery packs with solar power and and for wind it's even more important so with with the gigafactory that we're creating in Nevada we're going to create many
21:16well probably tens of gwatt hours per year of stationary storage on the battery front so that yeah there's a light coming for sure hi my name is Tina I'm Mar from China so my questions is relevant to China Market how would you in your perspectives how would you forecast about the electric vehicles the demand in China and how's the Tesla strategies in the coming year yeah sure I think things seem to going you know fairly well in in China we've had a very enthusiastic response and yeah I think I think longterm it seems likely that China would be the biggest market for Tesla so yeah I'm really really optimistic about things over there is it going to
22:08be bigger than the North America it seems likely that it you know it would likely be bigger than North America yeah in the long term and how about the government regulations and the local competitors there the government's been pretty good I mean there's there is certainly in terms of the incentive structure it does favor local production but overall it's you know I I think it's not been too too much of an issue the government's been been great so far yeah thank you
22:43hi I'm Manu I'm a sophore at Stanford and U all of your work has been bringing people from Place A to B in the in the fastest and the most environmentally friendly manner there's also another kind of research happening in the valley where people are trying to avoid Transportation at all we call it virtual reality right so I can be in my dor and I can see you speaking with the same level of contact on my Oculus Rift as I am here right now taking the calra or the Tesla or even a SpaceX rocket in the future how do you think will virtual reality tie in the future of Transportation which you are working on thank you well maybe we're in a simulation right now
23:26yeah yeah yeah [Laughter] seriously some feels like that yeah I mean I think it is going to from what I've heard of IUS Rift and and some of the other immersive technologies that it's quite transformative you really feel like you're there and and then when you come out of it it feels like reality isn't real so I think we'll see
23:57probably less physical movement in the future as a result of the virtual reality stuff yeah and when we come out of what we're here now into this virtual reality we'll think it's real I mean it's like well I mean there's some interesting things here on the virtual reality front and I mean just on the whole notion of a simulation which is that if if you just if you extrapolate into the future and say well how good let's say will video games be in in 100 or 200 or 1,000 years from now if if there's continued Improvement and you know you're in a full body haptic suit with a sort of surround vision and you know you it's it's it becomes Beyond a certain resolution indistinguishable
24:49from reality if and and there will likely to be there likely to be Millions maybe maybe billions of such simulations so then what are the odds that we're actually in base reality isn't it one in billions is it I mean gives I can give the counter argument but I'd rather you give the counter argument it obviously this feels real but but it it it may I mean it seems unlikely to be
25:23real my name is Danielle I'm an architect from paloalto I just completed a net zero passive house and I'm now a fellow at the Stanford Business School and I'm working with a group of students on a project to come up with the most innovative ways to solve big problems in cities and we've identified Detroit as our prime target two questions what do you see as city-wide Innovation and since women don't ask
25:54as we're taught I'm going to ask would you come down to Stanford and bring M storm on this with us this may sound Tri but I I think I honestly think tunnels should be given a lot more consideration so I mean if you look at a city you have you look at I can you have all these apartment buildings and Office Buildings and there are many levels like they're you know be like an average of in Manhattan I don't know what is an average of like 30 stories or something like that but then you've got a street which is one story this is an obvious issue like you have a 30 to1 ratio of you know so you should have multi-layered highways yeah underground stacked up yeah and you can have tunnels
26:39to the tunnels don't have to follow the the buildings they can they can be they can go diagonally through the basement yeah yeah and you can have as many levels as you want M so it's really just the cost of building the tunnels and but really it's tunnel is a hole in the ground like how hard can it really be I mean just so it seems like if if some Ur put
27:06their effort into building tunnels you know effectively be transformative to cities around the world I'll certainly consider coming giving a talk I mean I'm in paloalto every week so certain be open to it sure what other things besides you know tunnels hyper Loops you've talked about other things that are great visions that could transform a daily what we call reality which I assume is reality for the moment yeah I mean it's with within cities it's sort of tunnels tunnels and tunnels and tubes
27:47and I mean for longdistance travel I really think that the vertical tequ of Landing electric supersonic aircraft is the way to go and I think it's it's very doable so that that that would be the way to go for long distance transation like for for if you're going more than like 500 Miles because then you have any to any you solve the any to any problem of a long distance for shorter distances because you have a time to climb and a time to descend penalty below 500 Miles aircraft are not as
28:22good so that's why I think sort of some sort of evacu evacuated tube is is a better way to to travel what is an evacuated tube well you know just something where you've reduced the air density so the drag is dramatically reduced then it's like it's as though you got got teleported to altitude right and you can grow much faster and and not have to have the climb and descend issues hi I'm Lawrence I'm a sophomore at Stanford and I run a music startup we all know that you have all those really crazy ideas in the best way possible but how do you come up with those ideas could you tell me more about your creative process and also once you have those great ideas how do you go
29:08about capitalizing those really ambitious ideas if you have very few limited Financial Resources well I think the the great thing about sort of software or anything which just involves intellectual capital of a capital of you and you friends you and a few friends is that you can just do it so that's why you know doing some sort of internet thing or software thing is is great as an initial company to create that's why it's zip2 and PayPal they they gave me the capital to attempt to do more Capital intense activities I mean in
29:49as far as idea generation I think when I tend to think of things from a sort of physics standpoint like from a first principal standpoint what would be the best way to accomplish something and then pursue that so and that's also a good way to determine if there's a if if something's far from its Optimum
30:11and like on on rockets for example one could say oh you know that you could Reason by analogy and say the Rocket's going to cost a certain amount because that's what prior Rockets have cost or you can say well what is a rocket made of what the material constituents what are those M what do they weigh what what's the cost per unit mass and that that sets the limit ASM totically for what a rocket can be so if you can figure out some creative way to rearrange those elements into a rocket shape then you can achieve a much better outcome that's the first principles approach and I think also just combining ideas from different Industries is really helpful for
30:55Innovation so you what what have people discovered in one industry and can that be applied to other Industries that's I think also a great source of ideas but but you usually just struggle on a solution and you try try a bunch of things some of them don't work and some of them and you most of them don't work and occasionally one does you mentioned the startup of PayPal which is a great Innovation that happened in sort of how we pay for things that seems to me an area in which it hasn't been as much subsequent Innovation other than Bitcoin as it should be in other words quickly transferring money especially to people you don't know or whatever is to Me
31:40Maybe I'm Wrong far more difficult than it seems like it should be are you somewhat disappointed with the way PayPal then proceeded or are you disappointed with the lack of innovation in the digital currency and and environment yeah Pap has definitely I mean it hasn't moved much since from when when it was sort of bought by eBay the the longterm Vision that I had for PayPal or sort of in in in in sort of Finance was
32:13to to to well it sounds a bit strange like to to convert the financial system from a series of heterogeneous insecure databases to one database or well not one database but maybe there'd be like a few more the the money is just a number in a database that's what it is and it's primarily an information mechanism for labor allocation and the current databases are not very efficient like there you know there are these old Legacy main frames that don't talk to each other very well have post security and only do their and do bash processing once a day are you glad PayPal is being spun off it's probably the right move and what would you do with it well I think I think I convert it
33:09into more of a full service financial institution so you just you you want to do all the things that a consumer you want to have like all the financial services that somebody needs in one place seamlessly integrated together and easy to use and I think really really care about the consumer I think a lot of banks don't seem to care that much about the customer so I I think there's an opportunity to be like a really good bank effectively but but but much more than what people think of it do you think Bitcoin will be disruptive in that way I mean now it's a speculative currency will it be something that will be what normal consumers will use and will disrupt the
33:54banking industry my opinion B point is that I mean I think Bitcoin is probably a good thing but it's it's essentially it its main thing thing will be I this probably get quoted here and there but the it it it's it's I think it's primarily going to be a means of of doing illegal transactions but that's not necessarily entirely bad CU you know you know some things should be maybe shouldn't be illegal U so but the
34:34combination of Silk Road and Bitcoin will save us well it will be used for for legal and illegal transactions otherwise it would have no value as a use of for for legal transaction because you have to have a legal to Illegal Bridge yeah I don't own any Bitcoin by way we I let you get the last word sir I promise it'll be fast you on Bob shiky from Tom Reuters India just launched a a mission to Mars $75 million to get that thing was it as Astro Teller said a miracle or is there something fundamental about the Indian space industry that allowed them to do a mission like that for so cheap and if so is there something that we can pick up and learn from that I think it's
35:17a very impressive Mission given that it was executed by a government entity that's like really really impressive I mean impressive no matter who's doing but from a c standpoint impressive because it's being done by a government entity ultimately we have to be able to do missions to Mars for much less than that otherwise it'll be impossible to establish a self-sustaining civilization on Mars you because we'll have to transport millions of tons of cargo millions of people and the the cost of moving to Mars has to be affordable otherwise people won't be able to go so it has to ultimately come down by couple orders of magnitude from from that level sub million sub a million
36:00dollars then a couple orders of magnitude I'd say below well below a million yeah Elon Musk great innovator thank you very much and good luck tomorrow we'll we'll we'll be looking to see what was right under our nose when you announce it tomorrow sir right good thank you [Music]
资源与出处
- 视频在 YouTube 收看
- 档案silicon-valley.tech 档案页本页视频与文稿数据来自 Elon Musk Archive
- 精读本篇精读 Markdown 版