小丹尼：我是小丹尼，谈车说科技。这还是我们比较俗套的Slogan。今天有幸请到一苒还有Emma。一苒是《火星人马斯克》这书的作者，和电动Emma一起合著的，这本书马上发了。另外一苒其实是之前蔚来的PR总监，也是知乎大V。I'm
Danny, I talk about tech and cars. This is our unaltered slogan. Today
is also the lucky day to invite a Yiran and Emma. Yiran is the co-author
of "Mars man Musk" with electric Emma and the book is forthcoming.
Yiran was also the former PR Director of Nio, as well as a KOL in Zhihu. 其实最近自动驾驶在国内还挺火，包括上周文远知行和高德开始在广州测试，还有滴滴和CCTV也开始直播自动驾驶上路。从一苒的角度来讲，你觉得自动驾驶，大概现在是一个什么样的发展情况？Automated
driving has been quite fashionable in China recently. Last week, Wen
Yuan Zhixing and Gao De started testing in Guangzhou. Didi and CCTV also
started broadcasting live automated driving cars on the road. From your
perspective, what the developement stage is autonomous driving at now?01 自动驾驶现在发展到什么阶段了？
一苒：
我觉得自动驾驶其实过了第一波最狂热的阶段。首先我认为这是一个非常伟大的科技，但是也是一个特别难的科技，因为它不是腾讯读书，也不是今日头条那样的新闻APP，也不是让我们到线下店去买外卖。它不是那么轻的东西，它是一个非常重的东西。其实新科技在人类社会，从大家关注到它，发展到最后能用，可能要好几十年，但是一个新的东西刚出来的时候，大家往往会过于激动，对吧？就好像你今天刚约会一个漂亮姑娘，往往你也会过于激动，然后特别投入。
I
think automated driving has passed over the first wave of the most
feverish stage. Firstly, I think it is a very great but also
particularly difficult technology. It is not Tencent reading, a headline
news APP, or something that let us go to the offline shop to buy
takeaway. It’s not an easy thing but a very serious one.
Therefore, it may take decades for new technology to develop from
discovery to application in human society. But when a new thing just
comes out, people are often too excited, right? Just like you dated a
pretty girl today, usually, you will be very excited and devote all
yourself into it.
自动驾驶的发展状态，几年前好像是大家刚刚跟自动驾驶美女约会，都特别激动，投资人砸非常多的钱，创业者层出不穷，对吧？但是我觉得现在是处于初次约会之后，激情逐渐退却，双方都有一定的反思，然后反思之后，大家觉得对方还是挺好的。所以我觉得现在自动驾驶处于一个过了高峰掉下来，然后再往上，很有耐心地去攀爬的这么一个阶段。
For
autonomous driving, a few years ago, everyone was like dating a
beautiful woman and was very excited. Investors spent a lot of money on
it, and a lot of entrepreneurs appeared, right? But now, I think the
passion gradually recedes after the first date. Both sides have a
certain degree of reflection. After reflection, everyone feels that the
other party is still good. So I feel that the autonomous driving is at a
stage where it passes the peak and falls off, and now it climbs up
patiently.
电动Emma:
有同学问，就是大家一直都想知道的，关于自动驾驶哪家强。其实我们上周四的直播也有聊过，我不知道一苒你怎么看这个问题。你之前不是说过你觉得自动驾驶是有不同的门派吗？有六大门派，你可不可以说一下，你觉得这六大门派到底哪一家能够最先跑出来？
Some one asked
something they always wanted to know. Which company is better in
automated driving? We also talked about the question in the live show
last Thursday. What do you think about this question? You said that you
think there are different schools of autonomous driving before, didn't
you? There are six major schools, can you tell us, which one of these
six can make it first?
一苒：01出行公司对，我最近又做了一个新的梳理。6个门派中，第一个门派我觉得是出行公司，比如说美国的Uber，中国的滴滴。这类公司显然是非常重视自动驾驶技术的，因为如果他们不掌握这个技术，就会有一个掌握技术的公司，重新做一个自动驾驶出行公司，然后瞬间就会把以前的出行平台给干掉，所以这些公司可以说是第一个门派。这些公司的特点，他们最强的优势就是掌握了大量的出行车队，所以他们离应用场景是超级近的，而且他们已经有成熟的商业模式可以向客户收费。I
have done research recently. The first school of the six is the travel
companies. For example, Uber in the United States and Didi in China.
Obviously, these companies need to pay attention to automation. Because
if they do not master this technology, there will be a new company that
masters the technology to re-create an automated driving travel company,
and then instantly wipe out the previous travel platform. So these
companies can be concluded to the first school. I think their strongest
advantage is that they have a large number of travel fleets, so they are
super close to the application scenario, and they already have a mature
business model that can charge customers. 02造车公司第二个门派我觉得是离造车靠得最近的。这类公司，在美国最典型的就是特斯拉，在中国现在来看的话，蔚来也是非常典型的一家。这类公司的特点是他们已经掌握了制造汽车的能力，而且他们不断地在内部研发自动驾驶，把自动驾驶和整车制造合在一起，融合之后就能做出自动驾驶汽车。这类公司也是非常有竞争力的一个选择。I
think the second school is closest to the cars production. The most
typical one in the United States is Tesla. NIO is also a very typical
company of this kind in China. Its characteristic is that it has
mastered the ability to produce cars. It continuously develops the
automated driving internally and combines automated driving with vehicle
manufacturing to make automated driving cars. These companies are also
very competitive contenders. 03科技巨头第三类公司我觉得是类似于谷歌这样的公司。这种互联网巨头最大的优势是非常非常有钱，市值非常非常高，同时这些公司往往对人才的吸引力是超强的，所以他们非常容易吸引到大量的AI工程师、算法科学家。而且他们还有一点特别的好处，就是他们是彻底的门外汉，这就意味着他们对所有的事情都很中立，选什么样的车，在什么样的出行平台服务，他们都无所谓。所以理论上，他们可以框到的潜在市场份额可以说是最大的。The third type of companies I think are companies
similar to Googe. Internet giants' biggest advantage is that they are
very rich, and the market value is very high. These companies tend to be
very attractive to talents. So they are very easy to attract a large
number of AI engineers and algorithm scientists. And they have a special
advantage that they are thorough laymen, which means
they are neutral to everything. It doesn’t matter what kind of car they
choose, what kind of travel platform they serve. Theoretically, the
potential market share it can obtain is the largest. 04一级供应商我觉得还有第4类。第4类是一些，可能对于普通的朋友来讲没有那么熟悉，但是对于汽车从业者非常熟悉的公司，就是传统的所谓的汽车行业的一级供应商。这类供应商最典型的就是德国博世，像这样的公司他们常年掌握了为汽车提供一级的核心零部件的能力，所以说他们其实在造车这件事情上是掌握了非常多的核心技术的，他们也会非常希望加入到新的自动驾驶时代的竞争。I
think there is a fourth category, which involves companies that are
very familiar to automotive practitioners, but not to ordinary people.
These companies are the so-called traditional first-tier suppliers in
the automotive industry, and the most typical one is Bosch from Germany.
These companies have the ability to provide first-class core components
for automobiles, so they actually have mastered core technologies in
the car making process, and they also very much hope to join the
competition in the new era of autonomous driving.05电商巨头最后两类公司我觉得可能大家一般不太会想得到，第一个是电商巨头，典型的就像美国的亚马逊，在中国的话，我觉得美团包括菜鸟，其实也是这个品类。为什么电商跟自动驾驶关系很大，原因很简单，因为你要买了东西，需要有人帮你送，现在帮你送的这些物流车辆都是有人驾驶的。The
last two types of companies might be unexpected to you. The first is
the e-commerce giants, like Amazon in the United States. In China, I
think Meituan and Cainiao can be counted. Why e-commerce is closely
related to autonomous driving? The reason is very simple. If you buy
something, you need to have someone to deliver it, which is now done by
the manned logistics vehicles. 06视频传媒巨头那么以后这个车辆一旦成为了自动驾驶的车辆，最终实际上成为了一个人生活的第三空间，它肯定会比星巴克更适合做你的第三空间。所以那个时候，你坐在车里边，你不再去花精力操心车的驾驶，也不用去观察路况，你会把所有的时间用来游戏、娱乐或者学习、看新闻。在这个时候传统的视频传媒巨头就会跟这个业务有非常强的相关性，因为扩张了用户使用时间。所以你会看到，像腾讯这样的公司，它陆续投资过蔚来和特斯拉，也投资过一些地图公司，所以它其实对这个领域也是一直非常关注的。In the future, once the self-driving car is put into use, it will eventually become the third space of our life,
which will definitely be more suitable than that of Starbucks. At that
time, when sitting in the car, you no longer have to worry about the
driving stuff and keep your eyes on the road, you can spend all your
time playing games, entertaining, learning, or watching the news. At
that time, the traditional videos and social media giants will have a
very strong correlation with this business, because users stay longer on
these platforms. You can find that companies like Tencent have invested
in Nio, Tesla, and some map companies successively. They are very
concerned about this field. 总体来讲，现在的全球自动驾驶的格局，大家都看到了一个遥远的模糊的皇冠在那里，现在就像黄巾之乱的时候，所有人都带兵起义，大家都奔着皇冠去，但是确实很难讲谁、哪个属性的玩家最终有机会去称王称霸。So
generally speaking, in the global pattern of current autonomous
driving, people see a fuzzy crown afar. Now, just like the Yellow
Turbans Uprising, all of the companies are vying for the crown. However,
it is really hard to say which kind of players will eventually be the
winners.02 电动车搭载自动驾驶技术有优势吗？电动Emma：你觉得为啥现在说起自动驾驶，都会跟电动汽车联系在一起？燃油车不可以做自动驾驶吗？为什么总是电动车和自动驾驶在一起？From
your perspective, why autonomous driving is always related to electric
vehicles? Are fuel cars not applicable to autonomous driving? 一苒：我觉得这其实是一个误会，但这个误会在行业里还是很普遍的。因为有很多的电动车的相关方，他们都会努力声称电动车特别适合做自动驾驶，有巨大的经营优势。但是这个事情如果往本质里分析，其实可以说没有逻辑必然性，但又有一定合理性。为什么说没有逻辑的必然性？因为对于技术的可实现来讲，油车和电动车都可以实现自动驾驶，这其实是没有任何问题的。I
think it is actually a misunderstanding, which is very common in the
industry. Many parties involved in electric vehicles work hard to claim
that electric vehicles are particularly suitable for autonomous driving
and have huge business advantages. However, in essence, this saying has
certain rationality but no logical necessity. Why? Because from a technology perspective, both fuel and electric vehicles can realize autonomous driving, that’s for sure.你来感受一下，你就发现这个逻辑是可以被攻破的。就是一个80岁的意大利老太太，照样能够开着菲亚特500在马路上到处跑，没有问题，但是她的手不可能特别灵活，反应不可能比20岁小伙子快，四肢也不发达，对吧？这样的情况下人也能开车，而且也不撞车也不出事故，说明对于车这种速度的物体，它比蜗牛肯定是快很多了，但跟飞机火箭比它是很慢的一个东西。Let me give you an example and you can better understand. There
is no problem that an 80-year-old Italian lady drive the Fiat 500 on
the road, but her limbs and reaction cannot be as agile and fast as a
20-year-old boy, right? People with inflexible limbs can drive and there
is no crash or accident. This indicates that the speed of cars, which
is definitely much faster than snails, is very slow compared to aircraft
and rockets. 以人类非常有限的反应能力、手脚协调能力和控制油门的脚的敏感性，都能够操控一台车。这个事情是已经被世界反复证明过了。那么油车再不灵敏，它在达到人的灵敏度上，在AI的加持下是完全没有问题的。这也就是为什么你看现在的沃尔沃、奔驰，凡是搭载了L2自动驾驶辅助的油车，它开启自动驾驶辅助，在高速路上跑的时候就是比你开得稳，比你开得速度更均匀。People
can control a car with relatively poor reaction and hand-foot
coordination ability, and low sensitivity of the foot to control the
throttle. This has been proved repeatedly. With the assist of AI, fuel
cars are definitely more sensitive than people and are competent for
autonomous driving. This is why a fuel car, such as Volvo or Benz, which
equipped with an L2 autonomous driving-assisted system, can keep a more
constant speed on the highway than you can, as long as the system is
turned on. 虽然有这样一个现象，但是你会看到市场上很多搞电动车的公司，或者是一些科技创业公司，他们还是更倾向于说以后电动车跟自动驾驶是一个很好的组合。其实这不是技术问题，这是经济问题。Although
there is such a phenomenon, you will see many companies engaged in
electric vehicles or the technology startups still incline that electric
vehicles and autonomous driving will be a good combination. In fact, it is not a technical problem but an economic problem.长期来讲，在最终一定是电动车加自动驾驶是最完美的经济解决方案，因为电动车的燃料费其实比较便宜。所以长期来讲，如果按照一个最理想的模型，假设你把一个城市里所有的车都替换成自动驾驶车，然后这些车又全都是电动车，它们靠充电来补充能源，再靠自动驾驶来驱动它们，最后算下来的每公里的平均成本一定比油车少很多。因为到那一天的话，电池的成本早就降下来了。所以如果是一个终极方案，我认为确实是电车加自动驾驶，这也是这本书里面提到过的。In
the long run, electric vehicles plus autonomous driving must be the
most economical solution. After all, the cost of electricity for
electric vehicles is cheaper than fuel. So in the long run, if you
follow an ideal model that all the cars in a city are replaced with
self-driving cars, and all of them are electric cars, which rely on
charging to supplement energy, and rely on autonomous driving, the
average cost per kilometre must be much cheaper than the fuel
cars. Because by that day, the cost of the battery has already dropped.
So if there is an ultimate solution, I think it should be electric
vehicles plus autonomous driving indeed, which is mentioned in this
book.作者劳伦斯（Lawrence Burns），以前是通用全球集团的研发和规划的一把手，应该说在通用是仅次于CEO的这种职级，而且控制核心的业务。他从通用退休以后，去了一个研究型机构，然后他在那边做过一个study，在一个美国的小城市，叫安娜堡，好像这个书里面介绍过。他当时带着高等数学非常好的学术人才一起做的研究，算出了一个结果，就是最终一台纯电动的高级自动驾驶车的成本，大概是今天人力的燃油出租车成本的1/7左右，几乎是下降一个数量级。The
author, Lawrence Burns, was the leader of GE's R&D and planning in
the past, which was second only to the CEO at GE and he controlled the
core business. Then after he retired from GE, he worked in a
research institution in a small American city called Ann Arbor, as if
introduced in this book. Then all the research and testing he did there,
which he did together with very good academic talents in advanced
mathematics at that time, has come up with a result: a fully electric and advanced self-driving car will cost about 6/7 less than the human-powered fuel taxi today. It is almost an order of magnitude lower. 电动Emma：这个是在美国。That's in American. 小丹尼：对，这本书里的主要数据都是在美国测试的，在中国肯定还是会有些不一样的。而且我们也想强调，每个人，包括这本书的作者和我自己，分析问题的时候都有不同的角度。通过这本书，我觉得可以给我们多一些角度去看问题。我看到有一位同学问，10年内实现自动驾驶，这个可能吗？Yes,
most data in this book was tested in the United States. In China, there
will definitely be some differences. Besides, we are trying to
emphasize that everyone including the author of this book and myself,
has a different perspective when looking at a problem, and the book can
give you some more perspectives. I saw one of our audience asked that is
it possible to realize automatic driving within 10 years. 03 自动驾驶10年内能普及吗？一苒：我觉得这其实是一个语文题。大家每次说起自动驾驶的时候，每个人的理解不一样。有的人谈到自动驾驶就会认为是全自动，你手都不用去碰，眼睛也不用去看。其实在汽车工业界有一个美国那边的标准，大家引用得比较多，就是从L0一直到L5，把L5定义为最终极理想化的，人完全不用做任何事情，就看你的电影，看你的书，这样的理想化的自动驾驶。 I think this is actually a language problem,
because every time people talk about autonomous driving, everyone’s
understanding is different. Some people think of automatic driving as
fully automatic, and you don’t have to use your hands to touch and eyes
to watch. In fact, there is an American standard in the automotive
industry from L0 to L5, and L5 is defined as the ultimate idealized
version, that is people don’t need to do anything. 但是中间其实分很多阶段。现在市场上在售的车辆中，其实最高的规格也只不过是做到L2，特斯拉、蔚来，包括沃尔沃都是在这样一个阶段，唯一一个声称自己已经做到L3的是奥迪A8，他们在两年前就声称发布了L3。他们当时是在巴塞罗那发布的，我那次还在他们的现场，但是他们很鸡贼，因为这个L3仅仅在奥迪A8的顶配上才有，顶配意味着非常昂贵对吧？买的人就很少。 所以你可以认为当时更多的是一个公关和宣传的行为，而不是真正做了一个已经达到大规模量产的L3的车。But
in the middle, it is divided into many stages. Now, among the vehicles
on the market, the highest level of self-driving is only L2. Tesla, Nio,
and Volvo are all at this stage. The only one claims that has done the
L3 is the Audi A8. They claimed that they released the L3 technology two
years ago in Barcelona and I was on site. They were very cunning
because the L3 autonomous driving was only equipped on the top-level
configurations of the Audi A8, which means that it was very expensive,
right? Very few people can afford it. So
you can think it as more of a public relations and propaganda act
instead of releasing an L3 self-driving car that had already reached
mass production. 所以如果你从L5来讲，现在是L2，其实还有非常长的路要走，也许需要30年，也许需要40年，这是我的看法。但是今天这些公司必须出发，必须行动，只有这样才有机会让30年后的人类真的可以不用自己开车了。这个事情一定是非常缓慢的，渐进地去推进，不要指望第一步就把司机完全替代掉。可能第一步先是让司机开车比以前轻松一些，比以前安全一些，每年路上少很多的车祸，少牺牲很多的人命，少很多的事故，改变一些司机的工作方式和生活方式。有可能司机的定位会逐渐从Driver走向Operator，从驾驶员变成运营管理者，只是临时偶尔地去驾驶一下。所以这个事情我觉得还有非常漫长的道路要一步一步去走。 So
if you are speaking about L5 autonomous driving, now we are at L2 and
there is still a very long way to go, maybe 30 years, maybe 40 years, but
today these companies must start, must take action, only then they will
have a chance. That is, 30 years later, human need not to drive by
themselves, this must grow slowly, and it progressively advances. Do not
expect the first step is to completely replace the driver. It is
possible that your first step is to let the driver drive easier than
before, a little less than before. Reducing many accidents on the road
each year, and fewer victims. And then it will change the way drivers
live. It's possible that the driver's role will gradually move from
driver to operator, from driver to operations manager. They temporarily
or occasionally drive, so I think there is a very long way to go.04 自动驾驶包会比整车贵吗？电动Emma：有人问，FSD（Full Self-Driving）一直在涨价，车一直在降价，以后会不会FSD比车还贵？的确特斯拉就是一个很明显的例子。Someone
asked if FSD (Full Self-Driving) had been rising prices, while cars
were cutting prices, would FSDs be more expensive than cars? Indeed,
Tesla is an obvious example. 一苒：这个很有意思。我以前也一直都很关心这个问题。我们这次推荐的这本书里，他就做了一定的测算。根据他的估算，按照美国那边的情况，最后车辆的造价和自动驾驶的整个软硬件套装的价格，成本价差不多是1:1的，都是几千美金这样一个数量级。我觉得在中国应该也差不多，最终一台小型的自动驾驶车，可能车几万块钱，整个的自动驾驶套件几万块钱，然后合在一起可能是一个十几万的东西，可能最后就成了公共交通的主力了。This
is very interesting. I have been concerned about this issue for a while
I think the book that we recommend here, the author, he made a certain
calculation, in fact, according to his estimate, based on the
situation in the United States, the final cost of the vehicle and the
price of the entire hardware and software package of self-driving, the
final cost price is almost 1:1. They are all about thousands of
dollars magnitude. I think in China, it should be similar. In the end,
if a miniaturized self-driving car. The car maybe tens of thousands of
dollars, and then the whole system can be tens of thousands of dollars,
and then together they may be of a hundred thousand. They may end up as
the main force of public transport.05 《自动时代》一书颠覆了你对自动驾驶的哪些认知？电动Emma：你们俩看了这个书，你们觉得对自动驾驶有什么不同的理解？我觉得我最没想到的一个地方就是，原来自动驾驶最开始技术的起源是美国军方想用这个来做战争前线的技术，所以就搞了三次相关的比赛，就推动了自动驾驶。我后来就想其实很多真的很高精尖的技术发展，幕后的推手都是战争。你看以前美国和苏联冷战，推动了航天技术，然后登月计划也是那个时候孵化出来的。在某种意义上，中东局势不稳定，也让大家想到要发展电动车去提升能源安全。那你们俩看了这个书之后，对于自动驾驶有什么新的看法或者知识点吗？Then
you two read this book, you have different understandings of autonomous
driving. Like me, I think the biggest surprise to me is that the origin
of this technology is from the U.S. military. It wanted to use this as a
front-line technology. So it did three competitions to promote the
development of autonomous driving. And then I thought a lot of high-tech
development, the pusher behind is the war. As you can
see, the cold war between the United States and the Soviet Union
promoted space technology. The moon landing program also hatched at that
time, in a sense, the unstable situation in the Middle East also pushes
people to develop electric vehicles to enhance energy security. After
you two read this book, do you have any new ideas or knowledge about
autonomous driving? 小丹尼：我觉得最大的收获就是，这本书让我们认识到了自动驾驶发展过程的从0到1。其实大众知道自动驾驶，基本上已经是从1到n的阶段了。我们可以看出来硅谷的科技公司和底特律的一些汽车公司，对这件事的态度是截然不同的。比如说Google就是通过DARPA比赛，挖掘出了一个非常优秀的团队，也就是现在我们比较熟悉的谷歌的自动驾驶团队Waymo。当时我记得拉里佩奇拉拢团队，有一句话让我印象非常深刻，就是说如果自动驾驶这件事成功了，它在未来可能会做得比谷歌本身还要大，这件事如果有10%的可能性，那也是非常值得做的。现在来看大家的冲击感可能不是那么强，但是大家要想一想，这个时间点是在2008年，12年之前。我觉得还真是非常了不起，在12年前谷歌就已经重点投入做这件事情了。 I think my biggest takeaway is that I feel that the
book makes me realize that the 0 to 1 development of autonomous
driving. In fact, when the public knows that the words of autonomous
driving, basically at this time is the 1 to n period. We can
see that these technology companies in Silicon Valley and some of
Detroit car companies, they took a very different attitude towards this
matter. For example, through the DARPA competition, Google found out a
very good team. That is the name that we are now more familiar, Waymo. I
remember the way that Larry Page built up the team. There is a sentence
that I was
very impressed. He said that if the autonomous driving gets success, it
may be larger than Google itself. If there is a 10% chance of success,
it's worthwhile to do so. I understand that we may not feel such strong
impact right now, but it was 12 years ago, in 2008, in fact, I think it is very remarkable. Twelve years ago, Google started to focus on doing this.
一苒：这本书让我最印象深刻的一个地方，是他们在安娜堡做的实验。这本书的作者当时应该已经预感到了这个行业会变化，但我觉得特别有意思的是，他们真正把演算的数学模型最后算出结果以后，连他们自己都觉得很惊讶。当然我今天读到这个模型我也超级惊讶，因为这个模型非常违背人的常识。One
thing that I am most impressed in this book is about the experiment
they did in Ann Arbor. The author should have foreseen that the industry
would change at the time, but I find it particularly interesting that
after they finally got the results by calculating the mathematical
model, even they themselves were surprised.I was also super surprised when I read this model today because it is contrary to people's common sense. 我来跟大家叙述一下是怎样一个模型。他们做了这样一个测试，如果你在美国找一个小城市，这个城市可能一共就20万台汽车左右，如果在这个城市里，把所有的现在人类手动驾驶的车，替换成全自动驾驶的车。他们设置了各种各样的数学模型，比如说虽然当时Uber还没有被发明，但他们已经预估到了以后会有Uber这样的东西，如果你以后拿着手机或者电脑给自己叫一部车，从A点到B点，他预估车大概会在两分钟左右来接你。I
will tell you what kind of model it is. They did such a test: if you
were in a small city in the United States, there may be a total of about
200,000 cars in the city. Assuming that they replaced all the currently
human-driven cars in this city with fully-autonomous driving cars. They
set up a variety of mathematical models. Although there was no Uber at
the time, they have foreseen that there would be something like Uber in
the future. They estimated that in the future, if you hold your mobile
phone or computer to get a car from point A to point B, the car would
pick you up in about two minutes. 这个用户场景跟今天我们用的Uber、滴滴打车是非常类似的。它在这个基础上去预估，最后发现如果用全部的自动驾驶车来替代掉今天的手动驾驶车，车辆要减少90%。这意味着绝大多数人的车都可以送进回收站，送进垃圾场，不需要再被持有。那么这个模型在第一步演算的时候，是没有考虑早晚高峰的，这也是一个非常重大的变数。The
user's scene is actually very similar to today's Uber and Didi taxis.
If it is estimated on this basis, it is finally found that if all the self-driving cars are used to replace today's manual driving cars, The answer is to reduce the vehicle by 90%,
which means that most people’s cars can be sent to the recycling bin,
and they don’t need to be held when they are sent to the dump. Then when
the first step of this model is calculated, he is I did not consider
the morning and evening peaks, which is also a very important variable. 我在蔚来工作的时候，也曾经跟我以前的老板斌哥交流过自动驾驶的问题，我也听过他谈论这个事情。他就特别提到，早高峰晚高峰，包括伴随的潮汐现象，其实是对自动驾驶非常大的一个挑战。很多人都希望在早晚高峰有出行的独占权，所有人都希望移动。在这个时候如果靠出租车，你会发现在北京上海，真正早高峰的时候，出租车是很难打的。如果你不想耽误时间，你还是得自己独立拥有一台车。这台车早高峰晚高峰度过了以后，就成了一个闲置资产，因为你去办公或者去休息了，这个车就不得不趴在一个地方静置，造成浪费。When I was working in Weilai, I had also communicated with my former boss Bin about the problem of autonomous driving. I
have also heard him talk about this. In fact, he specifically mentioned
the morning and evening peaks, including the accompanying tidal
phenomenon, which is actually a very big challenge to autonomous
driving. Many people hope that there will be morning and evening
peaks because at this time everyone wants to move. At this time, if you
rely on a taxi, you will find that it is very difficult to get a taxi in
Beijing and Shanghai during the morning rush hour, but if you don’t
delay your time, you still have to own your own. A car, this car becomes
an idle asset after the morning peak and evening peak, because you go
to the office or go to rest, this car will have to lie in a place to
stand still, causing waste. 但这种浪费似乎是一种必然，因为如果没有这种浪费的话，早晚高峰大家坐什么车？所以对高峰期的处理，会是这个模型非常重要的一步。作者他们的团队也做了这个模型，做完就发现，如果要解决早晚高峰问题，只需要把10%的自动驾驶汽车变成15%就可以解决。But
this waste seems to be inevitable, because if there is no such waste,
what car will everyone take in the morning and evening rush hours, so
the processing of the peak period will be a very important step in this
model. The authors and their team They also made this model. They found
that if they want to solve the morning and evening peak problem,
they only need to change 10% of the self-driving cars to 15% to solve
it. 这个事情对我的冲击力非常大，因为我以前从来没有想象过，我们人类社会现在的车完全被替换，对吧？我们多少都有点感觉，如果车辆变成自动驾驶，车的数量会减少，但是到底减少到多少呢？特别厉害的一点是在大概快10年前，他们就已经做过了研究，而且做的所有的定量分析显示15%就够了。这会是一个非常惊人的变化，对于我们今天人类生活的出行社会来讲。这是我觉得这个书最让我震惊的地方。This
matter has a great impact on me because I used to Never imagined that
if we humans were completely replaced from the current car to the
future, right? We all feel a little bit if the number of vehicles
that become autonomous vehicles will decrease, but how much? A
particularly powerful point is that they have already done research
about 10 years ago, and all the quantitative analysis done shows that
15% is enough. This will be a very amazing change for our travel in
human life today. In terms of society, this is what shocked me the most
about this book.
字数限制，完整中英文稿在公众号：DannyData小丹尼
