无人驾驶 | 经济学人商业
1
思维导图作者:
Qianna,对语言有点敏感,对逻辑十分执拗,对摇滚太过着迷
9.15晚八点,
小熊Lucian老师给大家带来一节免费的二笔实务课!
有兴趣的小伙伴记得来哈!
老师简介:
Catti二笔二口
雅思7.5,口笔译实践、竞赛经验丰富
笔译实践量达100万字,
含医疗机构官网、重大会议论坛材料、
公共卫生政策、城市形象宣传片等类型
口译项目涉及教育、农业、医疗领域
直播平台:
小鹅通
链接:https://jdyth.h5.xeknow.com/sl/1QaPHV
有兴趣的同学可扫码进群
2
Business | Schumpeter
商业 | 熊彼特专栏
Business | Schumpeter
商业 | 熊彼特专栏
No one behind the wheel
无人驾驶
Cherish your Uber drivers. Soon they will be robots
珍惜Uber司机吧,不久就是无人驾驶了
This column was not written by Chatgpt. But it did have the benefit of another breathtaking technology that could change life as we know it. It was partly typed in the back of a Waymo self-driving taxi that, with ghostly serenity, climbed the hills of San Francisco, through the fog of Twin Peaks (it was still “Fogust”), around the spaced-out hippies of Haight-Ashbury to the Golden Gate Bridge. Rather than driving into wet concrete, as one of its hapless rivals did recently, it politely gave way to a cement mixer that swerved across its path. It was a scenic—and entirely trustworthy—office-on-wheels.
本期专栏虽然不是用Chatgpt写的,但成文确实受益于另一项能够改变我们现有生活方式的的硬核科技。这篇文章有一部分就是作者坐在 Waymo 自动驾驶出租车后座上写下来的。车内非常安静,一路翻上旧金山的山丘,穿过双峰镇的浓雾(当时还是 “浓雾八月”),绕过海特·阿什伯里嬉皮士区,开往金门大桥。不久前,竞争对手的自动驾驶汽车很不幸地一头扎进了未干的水泥里,而Waymo自动驾驶出租车却能给一辆占道的水泥搅拌车让路。这辆车就是完全是“车轮上的办公室”,窗外有风景,安全有保障。
Like Chatgpt, self-driving vehicles are one of those marvels of artificial intelligence (ai) that make you pinch yourself when you encounter them because they seem so strange, and then pinch yourself afterwards because they become so familiar. The strangeness is quirky rather than scary. The robotaxi arrives at the tap of an app, with your initials quaintly lit up on the laser cone on the roof. You wave at it to stop, but there’s no driver with whom to make eye contact, so you run up the hill in hot pursuit until it finds a safe spot. Get in and a disembodied voice advises you that though the experience may be “futuristic”, you still have to buckle up. Then the steering wheel gently turns itself, and at a speed steady enough that you can use your laptop without feeling sick, you set off on a journey up the foothills of the ai revolution.
和 Chatgpt 一样,自动驾驶汽车也是人工智能技术助力下一大奇迹。见到自动驾驶汽车这么不可思议的东西,你不禁会掐自己一把,而当你习惯了自动驾驶汽车,你会再掐自己一把。这种陌生感虽然有些古怪但并不可怕。点击应用程序,自动驾驶出租车如约而至,激光圆锥投影仪用古雅的字体显把的你名字的首字母缩写投影在车顶上。你挥手示意停车,但车上没有司机和你眼神交流,于是你加紧往坡上赶了上去,直到车子找到安全的停车位置停下来。上车后,会有模拟人声提醒你,虽然这段旅程颇具“未来感”,但你仍需系好安全带。然后,方向盘缓缓转动,车子行驶平稳,使用笔记本电脑工作也不会感到晕车。人工智能革命的这段山麓之旅由此开始。
This part of the revolution is not yet on the breakneck scale of Chatgpt. Self-driving cars are ai in the physical rather than digital realm, and though it is annoying when chatbots “hallucinate”, any mischief-making by a robotaxi could be fatal. That is why safety, not speed, is paramount. Yet once you become accustomed to the experience, it is easy to imagine a future where more of life is spent in self-driving taxis; where commuters can work, watch videos or snooze while stuck in traffic; where tourists can sight-see without having to speak a foreign language; where lovers can escape the prying eyes of a driver. The questions are: how far away is that future, and what is the cost?
Chatgpt的革命性势头发展迅猛,排山倒海,而自动驾驶汽车领域的革命还没有如此规模。自动驾驶汽车是实体领域的AI应用,而非数字领域。聊天机器人的“跳脱”固然叫人不爽,可自动驾驶出租车稍有差池便涉及人身安全。因此,自动驾驶安全最为重要,而非速度。一旦人们习惯了无人驾驶,便可预见未来我们会有更多时间在自动驾驶出租车中度过;上班族可以在堵车时工作、看视频或打盹;游客出国观光无需外语沟通;情侣不必再忍受司机八卦的目光。但问题是:这样的未来还有多远,代价又是什么?
An immediate answer to the first question is that there are no signs of it yet in the drug-ravaged dystopia of San Francisco’s city centre. Schumpeterarrived there after taking a train from the airport. He tried to call a Waymo, but it is an area where robotaxis still fear to tread. He had to hail an Uber to reach Waymo territory. That is telling. After years of testing, Waymo (owned by Alphabet) and its rival Cruise (majority owned by General Motors) got the go-ahead from California’s regulators in August to sell driverless taxi rides across the city 24 hours a day. But they still lack permits to serve the airport, indicating how cautious supervisors remain. Both firms operate in parts of Arizona, and are expanding into more American cities. But the staggered San Francisco roll-out suggests self-driving taxis will not become ubiquitous with anything like the speed that Ubers or electric scooters did in the 2010s.
第一个问题(实现自动驾驶的时间)的答案可以直接给出:脏乱差的旧金山市中心毒品泛滥成灾,这里还没有无人驾驶的迹象。我同事从机场乘火车到旧金山市中心,想叫一辆Waymo出租车,但机器人出租车还不敢在此地域运营。他只好叫了Uber到达Waymo的运营范围。这很能说明问题。经过多年试验,Waymo(Alphabet公司旗下)和竞争对手Cruise(通用汽车控股)在八月份获得加州监管部门批准,可以在州内提供24小时无人驾驶出租车服务。但是Waymo和Cruise仍未获准在机场区域提供服务,这表明有些监管者对自动驾驶仍然持有谨慎的态度。目前,两家公司在亚利桑那州部分地区运营,并且也正在向美国更多城市进军。但是,自动驾驶出租车在旧金山推广进度缓慢,推广速度和普及程度都不及2010年代的Uber或电动滑板车。
How quickly they proliferate will depend a lot on technology. Waymo and Cruise use a combination of detailed maps, extensive sensors and ai to achieve full autonomy, but only in the geofenced areas where they are trained. That means they can only advance step by step. Tesla hopes to muscle into their territory with full-self-driving (fsd) technology that learns more the more its cars travel, using not maps but cameras and computer power to mimic a driver’s eyes and brain. Yet Elon Musk’s claims about Tesla’s fsd capabilities create more confusion than clarity. So far the technology requires too many human interventions to compete with Waymo and Cruise. No driver can safely remove themselves from behind the wheel of a Tesla. When they eventually can, it could accelerate the robotaxi revolution considerably.
自动驾驶出租车的大规模普及很大程度上取决于技术发展。Waymo和Cruise利用详细的地图、大量传感器、人工智能等手段实现完全自动化,但仅限于车辆经过训练的地理围栏区域。这意味着这两家公司要发展自动驾驶汽车只能循序渐进。特斯拉试图利用“全自动驾驶技术”进军自动驾驶领域。全自动驾驶技术不使用地图,而是利用摄像头和计算机模拟司机的眼睛和大脑,因而车辆走过的路越多,学得就越多。不过,埃隆·马斯克的说法反而让人对特斯拉全自动驾驶能力更加疑惑。特斯拉的全自动驾驶技术需要过多的人工干预,难以和Waymo与Cruise竞争。没有司机,特斯拉就无法保证安全行驶。如果他们最终能攻克这一难关,这将对无人驾驶出租车变革起到巨大的推动作用。
As for the costs, there are business and societal ones. Currently Waymo and Cruise have around 500 cars on San Francisco’s streets, a fraction of the number of ride-hailing vehicles. For every robotaxi, there may be at least one highly paid Silicon Valley engineer tinkering with the technology. The service might also need to be cheaper than a human-driven taxi to attract the masses. So the firms will probably run at a steep loss for a while. Later this year Waymo and Uber intend to introduce driverless technology to users of the Uber platform in Phoenix, Arizona. That sounds promising. As Kersten Heineke of McKinsey, a consultancy, says, Waymo can take advantage of access to Uber’s large user base while Uber can start tackling the limits to its growth caused by a shortage of drivers. But self-driving technology will need to be deployed at grand scale to become cost effective.
至于成本,要考虑社会和经济两个层面。目前,Waymo和Cruise在旧金山街面上有约500辆车,只出租车数量的一小部分。每辆自动驾驶出租车的背后可能至少都有一名硅谷高薪工程师进行技术维护。自动驾驶出租车服务也许要比普通出租车更便宜,才能吸引大众。因此,这两家公司可能在一段时间内大幅亏损。今年晚些时候,Waymo和Uber打算向亚利桑那州凤凰城的Uber平台用户推行无人驾驶技术。此举听起来很有前景。麦肯锡咨询公司的克尔斯滕·喜力(Kersten Heineke)表示,Waymo可以借助Uber的庞大用户基础优势,而Uber也可以借此突破因司机短缺造成的增长瓶颈。但是,自动驾驶技术将需要大规模铺开才能实现低成本高效益。
The societal costs are harder to gauge. Waymo says it has 100,000 would-be passengers on a waiting list in San Francisco, out of a population of about 800,000. That suggests pent-up demand. But there is also scepticism. The media are mostly focused on mishaps, such as when robotaxis disrupt emergency services or freeze en masse. An anti-car outfit called Safe Street Rebel has found a low-tech way to disable the vehicles: it puts a cone on the car bonnet to confuse their sensors.
社会成本更难估量。Waymo 表示,旧金山人口约80万,其中10万人在排队报名乘车,他们都是潜在的客户。这表明需求是被抑制的。但也有人对此表示怀疑。媒体主要关注事故,例如自动驾驶出租车扰乱应急服务或在人群里开不动掉链子。一个名为“安全街道反叛者”(Safe Street Rebel)的反汽车组织已经找到了一种技术含量很低但却能让自动驾驶汽车失灵的方法:在汽车引擎盖上放置一个松果儿来干扰传感器。
The wisdom of cabbies
出租车司机的智慧
There are subtler concerns, too. Driving creates an unspoken chemistry between humans from all walks of life, based on a combination of self-preservation and trust. Robots may make driving safer. But that may come at a cost to human interaction. Then there are jobs. Driverless technology could eventually extinguish the taxi-driver profession, destroying a valuable source of income and a font of local knowledge, not to mention a priceless resource for journalists. In that vein, Schumpeter asked Aleksandr, a Ukrainian Uber driver in San Francisco, what he made of a driverless future, and surprisingly his eyes lit up. He hopes, he says, to buy an autonomous vehicle and use it to offer rides 24 hours a day, three times longer than he is able to drive. “I will do almost nothing while the car makes money for me,” he says. “Just like Uber.”
也有一些比较微妙的担忧。出于自我保护和信任,驾驶能在各行各业的人们之间产生难以言喻的化学反应。机器人可能会让驾驶变得更安全,但这种安全是有代价的——人类失去互动机会,以及工作岗位减少。无人驾驶技术最终可能会让出租车司机这一职业就此消失,让人们失去宝贵的收入来源,当地少了很多百事通,就连记者也将失去宝贵的资料库。为此,笔者询问了旧金山的乌克兰籍Uber司机亚历山大对无人驾驶趋势的看法,出人意料的是,亚历山大两眼放光,表示希望入手一辆自动驾驶汽车,每天24 小时都能载客,是他每天驾驶时长的四倍。“我几乎什么都不用做,汽车就能给我赚钱,”他说。“就像Uber 一样。”
3
观点|评论|思考
本周感想
Vicky,少儿英语老师+笔译新人
方向盘在谁手上
我是一个有信任危机的人。作为一个特斯拉车主,开车近半年的时间里,我从来没有使用过特斯拉标志性的自动驾驶,一次都没有。
于我而言,把方向盘交到别人手上,尤其是一台机器手上,简直是不可想象的。当我全身心的投入到我的旅程中,在舒心的办公、休闲,突然这台全自动驾驶的车,由于一颗松果儿的原因,开始不受控了起来。想到那个画面,是我半夜都会惊醒的情景。
我们可以相信科技到什么程度?当我们把所有的事物都交给人类设计的程式,那个手握命运方向盘的人,还是人类吗?作为一个科幻小说迷我可以告诉你历届大神给出的101种答案,没有一种是人类喜闻乐见的。可是我们仍然在不断地探索科技的边界,技术能给生活带来的“便利”。就像是我们一直在努力的联系三体人,告知对方我们的存在。而对方给出的回应是友好还是毁灭,我们却没有仔细思考过。
自动驾驶也许是未来不可避免的趋势,就像是火车提速,就像是飞机升空,可是在这个过程中一定会遇到提速后出轨,升空后坠毁的阵痛,人类在这种阵痛中不断完善技术,而阵痛中的牺牲品呢?他们作为人类进步被选中的“献祭”全程却是没有预期的。其实在科技的进步中,我们知道这些“献祭”一定会存在,我们只是不知道被“献祭”的那个人会不会是自己。因为在我们把生活交给科技的时候,方向盘也许已经不在我们自己手里。
4
愿景
01 第十七期翻译打卡营
3位一笔,1位二笔
授课内容全部为CATTI实务考试相关
小班直播课,随机批改,有针对性讲解
每位同学拥有一次批改机会
03 早起打卡营
微信扫码关注该文公众号作者