2019年2月,马斯克对激光雷达(lidar)的评价只有一句话:依赖它是不明智的。当时整个自动驾驶行业把激光雷达当成安全冗余的标配,Waymo 的车顶顶着它,Cruise 的车顶也顶着它。马斯克说,那条路走不到终点。
他的理由不复杂。激光雷达能把你送到某个水平,然后你就卡在那儿了。
本质上,把一堆 if then else 的规则堆在激光雷达上,解决不了这个问题。放弃吧,游戏结束了。你必须解决视觉感知。(马斯克)
这场访谈发生在2019年2月20日,全长不到一小时,信息密度很高。全自动驾驶的时间表、自研芯片的真实性能、监管机构的真实态度、以及他为什么觉得比特币的结构很精彩但特斯拉不会碰,全在里面。下面按主题拆开讲。
他先纠正了一个身份问题
访谈开场不久,主持人问:这项技术难度这么高,你凭什么有信心它能被解决,又凭什么认为特斯拉是解决它的那家公司?
马斯克的回答先绕了一下。他说有必要先澄清一件事,很多人以为他是商人,或者金融从业者。
我是个工程师,我做工程,一直都是。我写软件写了十五二十年,我在相当基础的层面理解技术和软件。我知道要让全自动驾驶真正完成,需要解决哪些问题。(马斯克)
这句话的分量在后面。它决定了整场对话里所有技术判断的底色:他谈芯片时说的是帧率、分辨率、像素裁剪,谈感知时说的是视觉输入的信息带宽,谈监管时说的是用里程数说话。这些都是工程师的语言,CEO 不会这么讲话。
他还补了一句,说自己有一个非常好的技术团队,能和他们共事是荣幸。然后给出了时间判断:这件事今年就能做完。
什么时候能安心在车上睡着
主持人追问得更具体:普通人什么时候可以上车就睡,醒来已经在目的地?
马斯克的估计是明年年底,也就是2020年底。他紧接着加了一个限定条件:他觉得从技术安全性上讲,那时候够用了,但监管机构什么时候点头,他说不准。
主持人顺着问,那会不会是中国先批?
马斯克的回答给出了一张很有意思的监管地图:
- 中国:监管者考虑周到、相当聪明,实际接触下来感觉很好。他提到中国的城市管理者懂技术,因为技术正是他们实现跨越式发展(leapfrog)的路径。
- 美国联邦层面:同样很好。
- 加州:有时候会过度热情(overzealous)。
- 欧洲:态度偏保守,节奏慢一些。
这个排序跟很多人的直觉相反。通常认为欧美监管更成熟、更开放,中国更谨慎。马斯克给出的观察是反过来的:监管效率的差异,跟技术理解力直接挂钩。懂技术的管理者知道该怎么衡量风险,不懂的只能用最保守的默认值兜底。
一百倍的数据,和一百万种你想不到的情况
整场访谈里,马斯克最用力强调的优势只有一个:数据量。
他估计特斯拉积累的真实道路里程,比所有其他公司加起来还多一百倍。主持人提到之前有说法是特斯拉占全部里程的5%,马斯克纠正说,其他所有人加起来大概只有1%。
这些数据从哪来?来自每一次人工接管。
每次有人从 Autopilot 接管车辆,系统就把这段信息保存下来上传。我们不知道是哪辆车,也没有针对个人的归因,我们只知道发生了一次接管,然后看需要做什么才能修复它。(马斯克)
他特意说明了两件事。第一,数据是脱敏的,没有个体归因。第二,人工标注正在被淘汰。司机自己开过那个路口的动作本身就是训练信号,系统看人怎么开,就学到了该怎么开。
这些数据里到底装了什么?马斯克列了一长串:
不同国家的道路标线不一样,规则不一样,其他司机的行为预期也不一样。天气有雨、雪、雨夹雪、冰雹、飓风、洪水、火灾、烟尘、沙尘。还有那些规则根本没被遵守的场面:施工人员摆错了锥桶,路面突然出现一个大坑,或者前方掉下来一个形状古怪、系统从未见过的物体。
他的原话是,这些角落案例(corner cases)又多又怪,你不会相信。自动驾驶的难点从来都堆在长尾里。前面百分之九十九的场景几个月就能做出来,最后那百分之一要用掉几年。
为什么他敢说激光雷达是死路
主持人抛出了一个行业内部的争论:Waymo 和 Cruise 的做法看起来更确定,规则明确、逻辑可验证;特斯拉用的是概率性的深度学习和神经网络。哪条路对?
马斯克给的答案借了一个优化概念:局部最优和全局最优。
用启发式(heuristics)的方法做这件事,结果会是一个局部最大值,达不到全局最大值。你必须用一个足够复杂的神经网络,才可能达到全局最大值。这也是为什么依赖激光雷达是不明智的。(马斯克)
他把规则系统比作一串 if-then-else 语句。这种系统在边界清晰的场景里非常好用,一旦进入模糊地带就崩了。
然后是这场访谈里最漂亮的一段类比:
我们人类开车,靠的是一对成像质量并不好的摄像头,装在一个转得并不快的云台上。而一个职业司机几乎从不出事故。(马斯克)
这句话把问题重新定义了。自动驾驶真正的瓶颈在"能不能理解看到的东西","感知得多精确"退到了次要位置。人类的眼睛分辨率有限、视野有盲区、反应有延迟,照样开得比任何自动驾驶系统都安全。说明视觉输入的信息量早就够了,缺的是对图像的理解。
他的结论很硬:把视觉感知解决了,你就解决了,别的感官一个都不需要。
一颗芯片,两千个百分点
访谈里被马斯克抱怨大家没怎么注意到的一件事,是特斯拉的自动驾驶 AI 芯片即将量产。所有订购了全自动驾驶(Full Self-Driving)的车主,会免费换装这颗芯片。
性能提升有多大?马斯克先说大概一个数量级,然后自己改口:
实际上更接近两千个百分点以上。(马斯克)
主持人提到自家分析师的判断。这位分析师在英伟达做了九年深度学习,看完规格后的结论是,特斯拉至少领先其他厂商三年。马斯克没有反驳,只说芯片项目三年前就启动了。
为什么要自己做芯片?他给了一个游戏行业的类比:软件和硬件是一体的。软件再聪明,也没法在老旧硬件上跑出今天的游戏画面。神经网络也一样。
具体差距在哪,他讲得很细:
- 现有英伟达硬件:每秒大约处理100帧。为了跑得动,必须做大量妥协,裁剪画面、对像素做变形、没法让所有摄像头都跑全分辨率。
- 特斯拉自研硬件:所有摄像头、全分辨率、全帧率,而且还没跑满。
"还没跑满"这句话的信息量比参数更大。它意味着后面还有余量,不需要靠软件技巧去硬挤算力。
马斯克承认,用旧的英伟达硬件理论上也能做到全自动驾驶,但那是一个难得多得多的软件问题,你得不停地在算力预算上做取舍。他给了一个量化对比:
用现有硬件加上大量努力,也许能做到比人类安全50%到100%;用硬件三代(Hardware 3),大概是比人类安全1000%。
同一个软件团队,换一颗芯片,安全倍数差一个数量级。这就是他坚持自研的理由。
监管者看的是里程数,不是发布会
主持人问,你会把这些拿去给监管机构看吗?他们看得懂吗?
马斯克的回答很务实:他们看得懂数据。你去说我们有一颗很快的电脑、一切都会好的,那只是一句声明。你拿出几十亿英里的行驶数据,展示事故率和接管率,那才是论据。
他还说了一句更强的话:不管你从哪个角度切这堆数据,结论都明确,开着 Autopilot 比不开更安全。
访谈里主持人提到一个细节:第一起致死事故之后,监管机构看完数据,态度反而转向支持特斯拉。马斯克没否认,只是补充了一个他观察到的现象:人会过度自信。
所以系统每次启用都会警告:必须关注路况,必须把手放在方向盘上。手离开太久,车会开始鸣叫提醒,然后减速。他把这套设计说得很平淡,但它揭示了一个真问题:辅助驾驶的安全边界,一半在算法,一半在人的行为。警告文案写多少遍都不嫌多,因为人一定会松懈。
至于监管是否构成障碍,他的判断是:除了少数几个更保守的辖区,基本没有被卡住。他举了个例子,自动辅助导航驾驶(Navigate on Autopilot)预计下周在欧洲获批。这种级别的延迟,放在整体进度里不算事。
先高速,后路口,最后才是别的国家
Waymo 的做法是一个城市一个城市地开,画好地理围栏,逐步放开。特斯拉没走这条路,它是在所有已售车辆上同时推功能。主持人问这个 rollout 策略怎么排。
马斯克的排序依据落在价值密度上,技术难度排在第二位:
- 高速和拥堵路段优先。全世界的城市快速路几乎都堵,走走停停最消耗人,自动驾驶在这里的收益最大。
- 路口第二。这是最难的部分,红绿灯和路口的组合有大量变体。他承认在开发层面识别停止标志和红绿灯已经没问题,但复杂路口的歧义仍然存在:好几个灯同时亮着,到底该看哪一个?他说,就算是人,有时候也看不清楚。
- 然后才是区域扩展。先让这套功能在美国跑通,再往其他地区移植。欧洲的高速公路结构不一样,角落案例也不一样。
他还透露了优先级:挪威是重点,因为那里客户多;荷兰、德国、瑞士也在推进范围里。
访谈里有一段很生动的自白。他说自己现在会为了用上 Autopilot 而选择更远的路:
哪怕有更近的路线,我还是走高速,因为能用 Autopilot。我甚至不用 Waze 了。穿过一堆弯弯绕绕的小街道,认知负担太重;坐在高速上巡航,舒服得多。(马斯克)
他还补了一个安全层面的理由:高速上的事故速度更高,致死率跟车速高度相关。低速状态下很难伤到自己,超过某个速度就危险了。高速既是体验价值最高的场景,也是安全收益最大的场景。两个维度指向同一个优先级,这个判断不难做。
开放平台:可以,但别改我的东西
主持人问,愿不愿意让其他车企在特斯拉的平台上造车。之前马斯克公开过专利,但好像没什么人用。
马斯克的回答分两层。第一层是现实:跟传统车企合作并不容易,他们也没有踏破门槛来找特斯拉。他顺口纠正了一句,专利其实有人用了。
第二层是条件。他说如果哪家车企想用和特斯拉一样的硬件加软件,特斯拉非常开放,超级充电网络(Supercharger)和 Autopilot 系统都可以谈。
但经常出现的情况是,他们说想合作,然后列出一二三四五六项要改的地方。不行,那会严重拖慢我们。想用我们这套东西,就原样用。(马斯克)
这条界线画得很清楚:开放的前提是别给特斯拉的工程团队增加负担。听上去强硬,背后是一个典型的平台经济学判断。任何为单个客户做的定制化都会分裂代码库、拖慢迭代速度,最后所有人都受损。
比特币、能源,和非洲的跨境汇款
访谈后半段换了话题。主持人提到加密货币,马斯克先开了个玩笑:认真对待,非常认真。
他讲了一件真事。当时推特上比特币和以太坊的诈骗机器人泛滥,他为了调侃,发了一句想买比特币,结果账号被自动规则封了,因为系统判定他在推销比特币。
接着他给了两个判断,方向相反。
第一个偏负面:加密货币的计算极其耗能。他理解为什么需要给加密货币的创造设定某种约束,但挖出增量比特币的能耗确实很高。他明确说,不确定特斯拉投入资源做加密货币是不是好主意,特斯拉真正想推动的是可持续能源的进展。
第二个偏正面。他谈到了比特币的结构设计,说它相当精彩,和以太坊以及一些其他项目有相似之处。然后主持人补了一组数据:当时比特币链上有1.3万亿美元的交易额,这些交易跟买披萨、买可乐毫无关系,是非洲的企业对企业(B2B)转账。因为在那些地方,把一种货币换成另一种货币的成本高得离谱,必须先过一道美元。
马斯克认同这个观察。他用了两个说法:
对他们来说,这是货币 over IP,是免费的资金传输。这件事对打开世界很重要。它绕过了货币管制。(马斯克)
他还做了一个长期判断:纸币正在消失,加密货币作为价值转移的方式,比纸片好得多。他说这话时用了毫无疑问。
最后主持人追问,特斯拉会不会开始卖比特币。马斯克说不会。
这期访谈最值得记住的三个判断
时间过去几年,回头看这场对话,有些预测过于乐观,尤其是2020年底可以在车上睡着那一条。但其中的技术判断和商业逻辑,价值反而更清楚了。
第一,视觉路线和激光雷达路线之争,本质是"理解"和"测量"之争。激光雷达测的是距离,视觉要解决的是意义。前者精确但封闭,后者模糊但开放。选哪条路,等于选了一个天花板。
第二,数据规模是自动驾驶唯一难复制的护城河。算法可以挖人,芯片可以流片,只有几百万辆车跑在路上收集的长尾场景,买不到也追不上。这个优势还会自我放大:车队越大,数据越多;数据越多,系统越好;系统越好,车卖得越多。
第三,硬件决定软件能有多简单。算力不够的时候,所有聪明都花在节省算力上;算力够了,团队才能专注解决真问题。两千个百分点的性能差距,换算过来是几年时间,外加一个数量级的安全倍数。
至于他自己,那句"我是工程师"贯穿了整场对话。他判断一件事能不能做,用的是"这个物理问题有没有解",市场规模和财务模型排在这个问题之后。这也解释了他为什么敢在2019年就断言激光雷达走不通:在他眼里,那是一道工程题,不是一道商业题。
如果你是做技术的,可以从这期访谈里拿走这些
- 先问天花板在哪。评估一条技术路线时,别只看它现在跑得多好,先问这条路最多能走到哪。规则系统在早期往往比神经网络表现更好,因为边界清晰、可调试;但它的上限是确定的。选路线等于选上限。
- 把"人类能做到"当成参照系。马斯克用职业司机的低事故率反推:视觉输入的信息量早就够了,卡住的是模型。这个方法能迁移到很多领域。如果一个受过训练的人靠有限输入就能完成这件事,那说明输入早就够用了,卡住的地方在别处。
- 把数据获取设计进产品里,别事后补。特斯拉的每一次人工接管都自动上传,用户在日常使用中无偿完成了标注。数据管道的设计优先级,应该和模型架构一样高。
- 警惕人的过度自信。系统的警告做多少层都不嫌多:每次启动提示、定时检测、离手鸣叫、自动减速。任何需要人保持注意力的产品,都要假设人会松懈。
- 按价值密度排优先级,别按技术难度排。高速和拥堵路段先做,因为收益最大、事故速度最高。同一个排序同时优化了体验和安全,这种交集值得优先寻找。
- 平台开放要给硬边界。愿意让别人用,但不为单个客户改架构。定制化带来的短期收入,往往低于分裂代码库的长期成本。
- 跟监管沟通,带数据不带愿景。"我们有一颗很快的芯片"是声明,"几十亿英里、事故率、接管率"是论据。对方听得懂哪个,取决于你给哪个。
- 对新技术保持双向判断。马斯克一边说加密货币耗能、不适合特斯拉投入,一边说它的结构精彩、是比纸币更好的价值转移方式。这两件事可以同时成立,不必强行站队。
0:00my guess as to when we would think it's safe for somebody to essentially fall asleep and wake up the destination probably towards the end of next year that's that's when when i would think it's most likely it will be safe enough for that i don't know when regulators will agree do you think they'll agree in china first could you imagine that happening china really wants to leapfrog the rest of the world and as you've said the mayors of cities and presumably the regulators are tech savvy because that's what's going to get them there we've generally found the regulators in china to be very thoughtful quite smart we have really found them to be very good we've also
0:43found the us at a federal level to be very good as well california can be a little overzealous but u.s federal china federal is good europe is a little conservative in this regard yeah we've heard sort of similar things about europe that the environment's moving a little bit slower there interesting so i mean this is an amazing technological feat what gives you the confidence that this is a solvable problem and then why should tesla be the one to solve this well first of all i think it's helpful to clarify with fuel people think sometimes that i'm like a business person or finance person or something like that i i'm an engineer i do engineering i
1:17always have so i've been i wrote software for like 15 years 20 years and i understand technology and software at quite a fundamental level i know what we need to solve to make full self-driving future complete i think we've got an extremely good technical team i think we really have the best people it's an honor to work with them i'm certain that we will get this done this year so can i ask you about autopilot so to what degree is consumer use of autopilot important in terms of furthering along the process because it seems like you know not every consumer is going to use it it might vary person to person should you sort of incentivize people to use it
1:57maybe auto initiate is that even possible here i think we're talking about quality of the data is that what you're talking about the data that you're getting because you can monitor all the data but the quality with autopilot is superior i would imagine the advantage that we have that i think is very difficult to overcome is that we have just a vast amount of data on interventions so effectively the customers are training the system on how to drive and there are millions of corner cases they're so obscure and weird you wouldn't believe it i mean there's different road markings different rules in different countries different expectations you've got rain
2:33snow sleet hail you know hurricanes floods fires smoke dust it's insane i know but we've got cars in almost all really in all those environments and so we every time somebody intervenes takes over from autopilot it saves that information on and uploads it to our system we don't know which car it was or how what happened you know there's no individual attribution for the car we just know that that intervention took place and then we see what is required to fix that intervention and we're really starting to get quite good at not even requiring human labeling basically the person would say drives the intersection we know then and is thereby training
3:16autopilot what to do right right do you think one of the questions we debate a lot when we look at what happenings at what is happening at waymo and cruz automation is that it seems as though the artificial intelligence or deep learning or deep reinforcement learning that you're using you know is probabilistic they seem to be more deterministic in their approach and because we're just looking at what's going wrong there and that's what that suggests is does that what yeah essentially and heuristics approach to this will result in a local maximum of capability so not a global maximum i think you really have to apply a sophisticated neural net
4:01to achieve a global maximum and this is why the reliance on lidar is unwise it gets you to a certain point but no further and so basically a series of if then else statements in light are is not going to solve it forget it there's game over you have to solve vision perception essentially understanding and with vision and then it's solved you don't need anything else no other senses at all i mean we drive cars with basically two cameras that aren't very good and on a gimbal that doesn't move very fast and a professional driver will almost have almost never have an accident yeah very interesting yeah you know a lot of the time we get a lot of pushback in this idea of data and
4:40sort of what data is more important than others i guess maybe highlight for us like a i mean tesla has more real real world miles than anyone else by far i think we must have more than everyone else maybe a hundred times more than everyone else yeah you said five percent i think on you you said they have all together five percent of it's probably closer to one percent yeah i mean that's amazing so there must be sort of events that are specific to tesla in there that no one else has i guess could you sort of talk about some of that uniqueness well i think it's just really the the long tail of weird events the million weird situations you know all sorts of weather conditions all
5:18kinds of road conditions situations where the road rules aren't even followed like they're not they're not always followed like you know say somebody working on the road might make a mistake and then suddenly you've you've got a situation where there's no road cones or there's a big hole in the road or something really nutty or a completely strange object that's on the road that's not recognized so it's really the the reason tells us i think is making rapid progress is because we have vastly more data and this is increasing exponentially with but as our fleet is increasing exponentially our data is increasing exponentially then i think we've got like so the best
5:54technical team and although i brought it up before people don't seem to really be taking note of the fact that the tesla autopilot ai computer is about to roll into production you know anyone who's ordered full self driving will will get that replace we'll get that for free it's really you know i've said it has like maybe it's like a order of magnitude improvement over the nvidia system that we have but it's really more like over two thousand percent yeah our analyst on ai deep learning spent nine years at nvidia and concluded that you were at least three years just looking initially at the specs at least three years ahead of anybody else out
6:33there any other auto manufacturer yeah and we started the chip program about three years ago because it just seems as though we we would want to advance making things if you want to have a complex neural network you need a combination of software and hardware and your software needs to be that much better in order to compensate for hardware if the hardware is weaker you know i feel like sort of like say have video games and how they progress it's a combination of software and hardware no amount of clever software could produce a video game on old hardware that you have today just doesn't matter you know so same thing with neural nets so right now we can
7:06process on the order of you know a hundred frames a second and we really need to do a lot of work in terms of cropping the frames and sort of bending the pixels and not going to full resolution on all cameras and that kind of thing with the current hardware we're full frames full resolution with the tesla hardware all cameras at full resolution full frames wow that's crazy still hasn't tapped out yeah so is the amount of compute processing power that you need for full autonomy fixed or would you have to upgrade this every two years how do you think about that well i mean i do think that we could we can achieve it we couldn't principal achieve full autonomy
7:45with the nvidia hardware but it's a much harder software problem and like so yeah you'll really have to try to budget your compute and do all sorts of tricks to manage how you use your compute so it's a harder software problem with hardware that's you know two thousand percent better you just you don't have to do that that constant budgeting it and so the software problem is much easier so like i think with the current hardware and a lot of effort we could we could get to full self driving with maybe maybe being like 50 to 100 safer than a person but with hardware three i think it's probably like a thousand percent safer than a person are you presenting
8:20this to the regulators do they understand this do they want to understand this i think they'll they understand data so if we show you know billions of miles with a given safety level then they will they sort of appreciate that i mean simply saying like hey we've got this really fast computer and everything's going to work it's like well you know that's just a statement like but if you've got hard data billions of miles and you can show the accident rate and the intervention rate and that essentially it's unsafe if you don't have autopilot on right which i think really is unequivocal at this point no matter how you slice the data it is unequivocal at this point that
8:55it's safer to have autopilot on the the data comment you just made we were very interested that the regulators after they examined the first fatality i think it was the the day the data around the first fatality that they just swung in your favor i i think it was shocking and i'm surprised more people haven't really taken note of that yeah it there is a phenomenon that we that we noticed that would that would happen essentially it's like if you yeah people just get get over confident with the system even though we repeatedly warn them you must pay attention to the road like literally every time you use autopilot it says you must pay attention
9:33the road you must keep your hands on the wheel every single time you use it it says this and if you take your hands off the wheel for too long it will start beeping at you and then slow down and that kind of thing so anyway so it's just but really at this point this is just just flat out no question that it's safer was you know like i would recommend it to to anyone it's just getting better so yeah so have you noticed a change in the regulatory environment are you feeling sort of more confident than previously or is this sort of you know it's still such an open question that i mean tesla's in a really unique position because you have the data to show them
10:05so it almost seems like you know you could be having sort of more advanced conversations about sort of proving that safety level i guess what are your what are your thoughts right now we're with a few exceptions we are not being held back by regulators is that a u.s comment or is it global i mean it's that there are a few jurisdictions in the world that are more conservative right now it's it's i would not say that we're held back by regulators in minor ways like for example like we expect to i think get the latest autopilot approval navigator or applied i think is going to get approved in europe next week or something like that but this is these are tiny delays and
10:42the grand scheme of things doesn't really matter yeah that's great and so as you think about sort of this development of autonomous driving and the rollout strategy you know tesla's not taking this other approach that say waymo and others are doing going city by city you're going to do a wider rollout they're still sort of i don't know if you want to call it geofencing but there's you know it's you start with highways and then you go sort of off highway and then are you going to go country by country or sort of how are you thinking of this evolving in terms of sort of the full steps to i guess to get to full autonomy you know are you then tackling intersections or you're
11:15sort of adding future future feature and then country by country yeah we started off with with highway because that tends to be the what actually matters the most if like you particularly if you're stuck in traffic and stop and go traffic that's very painful and it has the most benefit to have autonomy on freeways which are usually congested in almost every city in the world so in fact myself and i was doing is that even if there's a shorter way home you still take the highway because you can use autopilot i stopped using waze for example just like you know take the highway because then i have autopilot on you know going through a bunch of windy
11:49streets versus which is kind of like a lot of mental overhead as opposed to just sitting on the highway and cruising along is is better so we generally got gone with okay what's gonna add the most value to people also you know highway accidents tend to be higher velocity and so more potentially more dangerous like fatalities are very proportionate to speed below at certain speed it's very difficult to hurt yourself or die in in particular in tesla above some speed it's more dangerous so it's it's like okay we're where can it be most helpful from a convenience and a safety standpoint so if we focus on highways for that reason then intersections are
12:25the next thing and some of them they have a lot of variants and intersections so that's what you know we're working right now and you know it's working at a development level no problem recognizing stop signs and traffic lights but you do get ambiguity and some complex intersections with traffic lights like which one's the light the right light to focus on even if you're a person it's not always clear so that that's what we're working on there so we'll try to make that work in the us and then we'll extend that functionality elsewhere so it in like highways in europe are different particularly when you deal with corner cases so we're kind of making it work in like norway is a
13:00priority for us because we've got a lot of customers there and i think we're close to having that ready netherlands germany switzerland i guess one general question what do you think of letting other automakers build on tesla's platform i mean generally we found it's like quite it's not it's not easy to work with traditional automakers it's not like first of all they're not exactly banging down our door to work with us nobody took you up on your patents right now i think i think they have actually really yeah okay that's very different from say creating an integrated system okay like so you know if if there's a automaker out there that wanted to implement the same
13:37hardware system as as tesla and and use our software we would be very open to it but we're not going to change it what does happen is they want to work with us but then they'll say oh but we want you to change the following like six things like no because it's gonna slow us down massively yeah and and so we know it's like if you want to use exactly our thing that's fine but then they don't want to use exactly that thing we're open to other automakers using a supercharged network or open them to use it using our autopilot system they just need to make it work without a ton of overhead on tesla engineering absolutely the right thing to do i have a little
14:09bit of a different question for the it's a little bit off topic but it's in the news or it's becoming more in the news so it's about crypto crypto assets crypto you know crypto seriously chris seriously seriously very seriously i think you and jack dorsey have chatted a bit recently about this right bitcoin and ethereum scammers were so rampant on on twitter that i i said it just join in so that's ss said like one point one buy some bitcoin and then i got my account got suspended because i obviously had like some automatic rule that would if you're trying to sell bitcoin or something and i was just joking yeah yeah so my question is it's a little different from
14:56cryptocurrencies although given your history in the payments ecosystem it would be very interesting to know if you agree with jack and there is going to be one native cryptocurrency when it comes to the internet he thinks it's bitcoin it's interesting i have some friends of mine that are really involved in crypto i mean i think like the brit coin structure was quite brilliant yeah it seems like they're similar to ethereum as well and maybe some of the others but you know i'm not sure it's like i'm not sure that it would be a good use of tesla resources to get involved in crypto i mean we're really just trying to accelerate the advance of sustainable energy and i mean
15:33i think actually one of the downsides of crypto is that it computationally it's like quite energy intensive so like there had to be some kind of constraint on the creation of crypto so but it's very energy intensive to create like the incremental bitcoin at this point yeah but at the same by the same time there were 1.3 trillion dollars worth of transactions in bitcoin and we don't see it here because it's not for pizza or coke it's business to business in africa
16:01but we think it is business to business in africa where it is prohibitively expensive to convert from one nation's currency to another you have to go through the dollar i mean it really is very important to it's money over ip for them it's free transmission of money and that's really important to opening up the world and you know you know yeah it bypasses currency controls yes yeah yes yeah yeah so it's it's well and yeah paper money is going away yes and crypto is a far better way to transfer value than than pieces of paper that's for sure without a doubt that has its pros and cons just to clarify tesla's not going to start selling bitcoin okay you heard it here first even though
16:47everyone thought that so no well we we thank you so much for doing this and thank you for just doing what you're doing to change the world and make it a better place you know that's what we're doing in terms of it no seriously i always think of silicon valley that you know comedy have you seen that that show that's pretty funny i'm sure i didn't it's like literally every company often yeah oh that's what people say yeah like literally every company okay but you're really doing it and you're really doing making the world a better place by social mobile crypto this is so different this is so different we totally agree with you yeah so yeah okay you're doing the real thing
17:29cool so thank you thanks arc believes that the information presented is accurate and was obtained from sources that arc believes to be reliable however arc does not guarantee the accuracy or completeness of any information and such information may be subject to change without notice from arc historical results are not indications of future results [Music]
资源与出处
- 视频在 YouTube 收看
- 视频在 B 站收看UP 主「Circleloo」搬运
- 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