2012年8月,特斯拉累计造了不到3000辆车。2500辆Roadster,再加上几百辆Model S。
十年之后,2022年股东大会,马斯克站在德州超级工厂(Giga Texas)的舞台上,说特斯拉刚刚下线第300万辆车。
屏幕上那条增长曲线干净得不像真的。马斯克自己都盯着看了几秒,说这看起来就是风险投资人商业计划书里那种画出来哄人、但从来不会成真的曲线。
你会在风投的商业计划书里看到这种曲线,通常它不会成真。这次它真的成真了,这才是最神奇的地方。(马斯克)
台下欢呼。他没多解释这条线是怎么来的,只说了句感谢团队。
他把这一年的成绩单拆成几块讲:产能、现金流、FSD、制造、能源、机器人。每一块都有一组数字,也都有他没解决完的问题。
从“精神损伤”到全行业最高营业利润率
产能是这场演讲最先摆出来的数字。特斯拉已经把年化产能推到150万辆,年底的目标是摸到200万辆的年化运行率。
这个数字放在十年前近乎荒谬。
更让马斯克得意的是财务。屏幕上一条红绿相间的累计盈利曲线,他指着说,这玩意儿大致对应的是“精神上的痛苦”。
这条红绿曲线大致对应的是精神痛苦,或者叫精神损伤。(马斯克)
玩笑归玩笑,他要说的是另一件事:特斯拉成立以来产出的现金,已经超过花掉的现金,留存收益转正。
我们产出的现金比花掉的更多,留存收益为正,算是配得上自己吃的这碗饭。这非常难。(马斯克)
2017年是个坎。Model 3产能地狱,公司差点断气。从那之后自由现金流一路向上,这条曲线是他站在台上最有底气的东西。
他特意强调,这还是在自动驾驶没有真正兑现之前的数据。一旦自动驾驶解决,自由现金流会被放大。
你把数字算一遍会想,真的能夸张到那个程度吗?它可能真能。(马斯克)
整场演讲里,他几次回到同一个判断:特斯拉的盈利能力还在早期,自动驾驶和机器人是两个还没打开的放大器。
FSD:顺手把现实世界的 AI 问题解了一半
他先问台下:有多少人在FSD beta项目里?一片手举起来。
原本要发的是10.12到10.13,做着做着架构改动太大,他说这版可能够格叫10.69。台下懂这个梗的人笑成一片。
复杂左转是这次的重点。他提到一个叫Chuck的用户反复测试的那个左转路口,说成功率大概90%,正在往100%走。
我们非常尊重有价值的批评。Chuck那个左转,现在大概90%的成功率,快接近100%了。(马斯克)
“两周后”是马斯克的老梗,这次他也没忍住,说可能下周,也可能两周。团队现在是每周七天在跑。
接下来是他这场演讲里信息密度最高的一段:为什么特斯拉用摄像头加神经网络做自动驾驶。
他的逻辑很朴素。道路系统是为人眼和生物神经网络设计的。红绿灯、车道线、路牌,全都长成人类能看懂的样子。那要让机器在这套系统里开车,最自然的方案就是模仿人:摄像头当眼睛,神经网络当大脑。
道路系统是为生物神经网络和眼睛设计的,那么能奏效的硅基版本,自然就是摄像头加硅基神经网络。(马斯克)
所以他给这件事的定性是:特斯拉顺带在解决现实世界AI里很重要的一块。
他借这个机会招人。做AI的人应该考虑来特斯拉,因为这里解决的问题能救命:自动驾驶比人类驾驶安全一个数量级,意味着能拯救数百万人的生命、避免数千万人重伤。
然后他打了个电梯的比方。
以前每部电梯都有操作员,正常配置,还得配个大继电器。但操作员偶尔会犯错,把人夹成两半。后来电梯自动化了,你按个按钮,到楼层,门开,就这么简单。未来的汽车也是这样。(马斯克)
这个类比把自动驾驶的终局讲清楚了:它不会是一个需要你时刻盯着的新功能,它会变成电梯按钮,没人再讨论它。
AI Day 放在下个月底,他说会有很多有意思的东西,这次先不讲。
电动车的对手不是彼此,是汽油车
马斯克回顾了特斯拉出发时听到的那些话:电动车不可能造出来;就算造出来能跑几百公里,也没人买,因为人们爱汽油车。
他的原话是,创业造车公司是“蠢”,造电动车公司是“蠢的平方”。
当年我们决定做这件事,办一家汽车公司已经很蠢,办一家电动车公司是蠢的平方。(马斯克)
后面那句更狠:很长时间里他们确实不赚钱。但现在,特斯拉的营业利润率是全行业最高的。
台下鼓掌。他顺势纠正一个常见误解:特斯拉不只是汽车公司。
特斯拉有很多重身份,它既是硬件公司,也是软件公司。软件在车里,在神经网络训练里,也在工厂里。(马斯克)
他把工厂叫作“巨大的控制论集体”:人、机器、软件混在一起协同工作,软件越好,这个集体运转得越好。
你可以把工厂看成一个巨大的控制论集体,人和机器和软件在一起。软件写得越好,这个集体就越强。我不觉得其他车企是这么想的,但事实就是这样。(马斯克)
谈到竞争,他的框架很清楚:电动车之间互相抢不是主线,电动车从汽油车手里抢份额才是。
所以他欢迎传统车企转向电动。五年前的新闻稿可不是这个调子,那时候没几家大厂说电动车好话。现在几乎每一家都认了电动化这条路。
这正是特斯拉当初想做的事。不是为了抢最大市场份额,是想向整个行业证明电动化可行,证明只要车做得足够好,人们就会买。(马斯克)
他还讲了个有意思的现象:竞争对手每投一次电动车广告,特斯拉的销量就往上走一点。
专利方面,特斯拉早就全部开源,竞争对手可以免费使用。
我们把专利全开源了,他们可以免费用。这算是一种共同繁荣,挺好的。(马斯克)
这套逻辑和他在其他场合说的一致:特斯拉的对手从来不是某家车企,是汽油车那一整套基础设施。
Model Y 和正在铺开的超级工厂
产品层面他只挑了一个重点:Model Y。
他的说法是,Model Y今年会成为按营收计全球销量最高的车,明年会成为按销量计最高的车。
他没把话说满,“鸡蛋没孵出来之前不要数”。因为两家新工厂还在爬坡。
这些工厂不会自己神奇地运转起来。柏林和奥斯汀都还有大量工作要做,供应链不一样,问题一大堆。(马斯克)
他给产能爬坡下了个定义:所谓产能,就是解决一万个问题的速度。单个问题都不难,难的是有一万个。
没有哪个问题单独看特别难,但有一万个。产能爬坡的速度,本质上就是你解决这一万个问题的速度。(马斯克)
新工厂方面,他说今年晚些时候可能再宣布一个选址。台下有人喊加拿大,他回应说自己有一半加拿大血统。
长期规划是至少10到12个超级工厂,平均每个工厂的目标产能是150万到200万辆。
他还提了一句,加州的Fremont工厂已经是北美产量最高的整车工厂。上海团队则被他形容为“很难被打败”。
上海团队真的很难打败,他们太厉害了。(马斯克)
一笔被忽略的太阳能账
能源部分他先辟了个谣。总有人说电动车全生命周期排放比汽油车还差,这不是事实。
电动车的全生命周期排放远低于汽油车。我知道这些话你们已经听过了,但每隔一阵这套胡说八道又会冒出来。(马斯克)
然后他讲了一个自己很得意的数字。
过去十年,特斯拉太阳能板发出的电,加总起来,超过了制造特斯拉全部车辆、以及在超充和家里给这些车充电所消耗的能量。
把过去十年特斯拉太阳能板发的电全加起来,比造我们所有车、给所有车充电用掉的能量还多。这是我觉得特别酷的一件事。(马斯克)
这个账的意义在于,特斯拉卖出去的太阳能,已经抵消了它卖出去的车在能源上的欠账。一家车企可以这么算账,说明能源业务不是配角。
他把可持续能源的未来拆成三根柱子:第一,用太阳能和风能发电;第二,用固定式电池储能,因为风光是间歇的;第三,电动交通。
可持续的能源生产,主要靠太阳能和风能;固定式电池包把间歇的绿色电力存起来;再加上电动交通。这三根柱子立起来,就是一个完全可持续的未来。(马斯克)
电池回收他专门说了一段。有人担心电池报废后怎么办,特斯拉在内华达已经在做回收,现在每周处理超过50个电池包。
量之所以小,是因为车太耐用了,电池通常要用12到15年才退役,绝大多数还装在车上跑。
你可以把电池包想成高品位矿石。要去碎一堆石头,还是碎一个电池包?答案显而易见。(马斯克)
短期回收量小,长期会变得很重要。他的判断是,回收这件事会随着第一批电动车集中退役而突然放大。
第三篇章:算一笔全球的账
他把第三篇章(Master Plan Part 3)定义为“规模化”。
这事要从全球宏观的角度算:要实现一个完全可持续的经济,到底需要什么。
需要多少吨锂、多少正极材料、多少负极材料、多少隔膜、多少电解液、多少电子器件。所有这些物料加起来是什么量级,是第三篇章要回答的问题。
把所有需要的东西列清楚:锂、正极、负极、隔膜、电解液、电子器件,多少吨,什么规模。让所有人看清这件事到底怎么做。(马斯克)
他讲这段的时候语气变了。他说自己见过太多失去希望的人,觉得气候已经没救,地球要完了。
很多人已经失去希望,觉得太晚了,没机会了,地球要完了。不是这样的。地球不会被毁灭。(马斯克)
然后他说了那句被反复引用的话。
地球能被救,也一定会被救。(马斯克)
第三篇章的目标,是把“可行、正在发生、要尽快”这三件事讲清楚,让更多人从绝望里出来。这段话是整个演讲里少见的非商业表达。
真正的护城河是制造
制造部分他讲了两件事:一体化压铸和工厂布局。
特斯拉做出了业内最大的压铸件,复杂度很高。后底板原来由171个零件组成,现在变成2个。
我们把171个金属零件变成2个。同时更轻、更刚、操控更好、噪音振动更好、密封防水更好。每个维度都更好。(马斯克)
车身更刚、更轻,操控和NVH一起改善,防水密封也更好。他说这是材料团队和铸造技术的成果,本质上是在重新想一辆车该怎么造。
产线机器人数量也降下来了。现在一条产线用的机器人,大约是当年Model 3产线的30%。
工厂布局也在改。新工厂接近一个单一的整体式厂房,物流路径尽量走直线。Fremont不是这样,那里的流程绕来绕去。
Fremont我们做了很多事,但流程很复杂,不是一条顺畅的流。所以我们是在重新设计工厂本身。(马斯克)
他把制造称作特斯拉长期的可持续优势。
电动车大家都会有,所有厂商都会做电动车。自动驾驶最终所有车都会有,只是比他们想的要晚。最难复制的是特斯拉的制造技术。(马斯克)
怎么判断一家科技公司能不能成?他的方法是看最聪明的工程师想去哪。
看最聪明的工程师想去哪里工作。那家科技公司大概率会赢。就像职业体育队,看顶尖球员都去哪支队伍。(马斯克)
他说特斯拉和SpaceX经常轮流当工程师最想去的公司第一名。两家公司之间还允许员工流动,想从造车换到造火箭,可以。
工厂安全他给了一个具体做法:鼓励一线员工提安全改进建议,今年的目标是人均3条,已经超过了。
我们今年超过了人均3条建议的目标。这对提升安全来说是游戏规则级别的改变。(马斯克)
他还报了个数:去年特斯拉收到300万份求职申请。这个数字放在整个制造业里都很夸张。
超充和安全:两件靠“联网”做好的事
超充网络方面,他承认偶尔会有坏桩,但整体可用性极高。原因在于车永远在线,桩坏了反馈闭环很快。
现在主力是V3,V4可能明年开始铺。超充数量每年翻一倍,哪里缺,大概率很快就会有。
我们的超充数量基本每年翻一倍。你觉得重要的地方现在还没有,大概率很快就有了。(马斯克)
安全部分他讲得比较细。特斯拉车型在美国政府测过的所有车里,乘员受伤概率最低。
这个数据是实时更新的,因为车联网。别的车厂绝大多数不联网,撞了之后厂商拿不到数据。特斯拉能看到每一次碰撞,然后改进车身结构、安全带预紧、气囊展开策略。
其他车厂绝大多数不联网,出了事故他们其实不知道发生了什么。特斯拉能看到碰撞,然后问自己怎么把安全做得更好。(马斯克)
更往前一步的是预判碰撞。Autopilot的AI如果能以99%以上的概率判断碰撞即将发生,就会提前收紧安全带、打开气囊。
绝大多数车只能在碰撞发生的那一刻触发气囊,那一下很暴力。提前预判,气囊的展开就可以更温和、更有效。
绝大多数车只能在撞上的那一刻弹气囊,那一下很暴力。如果能提前预判碰撞,气囊的展开会好得多。(马斯克)
这段是整场演讲里最能说明“软件定义汽车”的地方:安全不再只靠钢材和结构,还靠一套能提前半秒做决定的模型。
FSD beta 的里程曲线
FSD beta的累计里程已经超过4000万英里,他预计今年会远超1亿英里。
部署节奏上,他重申今年内FSD beta会在北美大范围推开,任何申请的人都能用。
基本上到今年年底,北美任何申请FSD beta的人都能拿到。(马斯克)
AI Day第二部分放在下个月底,他说上次已经让人吃惊,这次还有不少东西。他没展开,只留下一个悬念。
被低估的 Optimus:经济学的底层变量
整场演讲里,马斯克花时间最多、语气最认真的一段,是机器人。
他说自己有点意外,分析师们似乎没理解Optimus的重要性。
我有点惊讶,至少那些分析师好像没真正理解Optimus机器人有多重要。(马斯克)
他的判断是,长期看Optimus会比汽车更值钱。
我的猜测是,长期看Optimus会比汽车更有价值。它会把“经济”这个概念整个翻过来。(马斯克)
为什么这么说?他给了一个经济学解释。
经济的体量,本质上是人均GDP乘以人口。人口受劳动力供给约束,劳动力一直是稀缺的。资本也是约束,想扩产得先有钱。
如果劳动力不再稀缺,资本也不再是约束,经济的上限就被拿掉了。
经济就是人均GDP。如果没有资本约束,经济可以任意大。(马斯克)
翻译成直白的话:过去经济增长被两件事卡着,一是能干活的人不够,二是投进去的钱不够。如果机器人能干活,而且造机器人的成本足够低,这两个卡点同时消失,经济体量的天花板就被拆掉了。
这也是他为什么说Optimus会颠覆“经济”这个概念。过去所有经济学讨论的前提是稀缺,劳动力稀缺、资本稀缺。这个前提一旦松动,很多结论要重写。
十年后:1亿辆车
有人问他十年后特斯拉会有多少辆车。现场有人喊3亿。
马斯克说1亿比较靠谱。
10年前我们造了不到3000辆车,现在累计300万辆。10年后1亿辆是相当可行的。如果10年后不到1亿辆,我会很意外。(马斯克)
他还说,除了车,还有“很多其他产品”,有些外界听过,有些没听过。
然后是股东投票的感谢,接着进入网络投票的问题环节。
整场演讲听下来,马斯克讲的其实是一条主线:特斯拉已经跨过了“能不能活”这道坎,现在要回答的是“能长多大”。产能、制造、能源、FSD、机器人,都是这道题的变量。
这场演讲里,真正值得跟进的东西
股东大会的PPT和口号每年都差不多,真正有信息量的是那些可以被后续数据验证的承诺。下面几条是可以拿去做判断依据的。
- 看产能爬坡,别看发布会PPT。核心指标是每个工厂的周产量曲线,以及新工厂从投产到满产用了多少周。200万辆年化是不是真的,看的是这个数字,不是台上的口号。
- FSD beta盯两个数:累计里程和人工接管率。里程涨得快不代表能力强,接管率才反映模型在真实路况下的水平。4000万英里到1亿英里这段时间的数据最有参考价值。
- Optimus看硬件参数,不看演示视频。自由度、负载、续航、单台成本,这四个数字决定它能不能真替代劳动力。马斯克说的“经济可以被任意大”,前提是单台成本足够低。
- 等第三篇章的原材料测算表。它会给出锂、正极、负极、隔膜、电解液的具体吨数。这张表是判断哪些上游环节会紧张、哪些会过剩的直接依据。
- 关注一体化压铸的扩散速度。171个零件变2个,改的是车身成本结构,也决定了后续车型的定价空间。哪家车企跟进得快,谁的成本曲线就更好看。
- 关注超充V4的铺开节奏。超充是特斯拉除了车之外最硬的资产之一,数量每年翻倍这个节奏能不能维持,值得按季度核对。
- 安全数据要找第三方复核。厂商自报的受伤概率需要独立验证,尤其是“提前预判碰撞”这类新功能,实际减伤效果要用真实事故数据说话。
- 求职者记一条:看最聪明的工程师往哪走。马斯克自己给的判断方法就是这个。哪家公司能持续吸引顶级工程师,哪家公司就更可能赢下下一个十年。
0:11thank you Oh Come on well I feel welcome so well it's great to see everyone here well [Laughter] [Applause] so yeah welcome to the Giga Texas [Applause] all right so let's see I'll go over kind of recount the year it's been an amazing year the Tesla team has done amazing work and I have to say like one of the things that professor's like what do you enjoy most in life I really being able to work with a super talented group of people and and manufacture those products and deliver them to people and make people happy from those products is is that that's one of the best things in life so [Applause]
1:44so yeah I just like to say what an honor it is to work with such a talented team at Tesla and that's the reason we've been able to do all these things so yeah we're aiming to achieve a two million vehicle run rate by the end of the year [Applause] [Laughter] this is the best crowd I mean [Applause]
2:27I mean [Applause] laughs [Applause] I love you guys too so anyway thanks to the the hard work of the Tesla team we've already been able to achieve one and a half million unit annualized run rates and depending on how the rest of this year goes we I think we might get close to or we'll get approximately at the one and a half million Mark and and be exiting the year at a two million unit run rate and and they're also worth noting just recently in the last few weeks we made our three millionth car so yeah
3:25you know the it's pretty wild to think like 10 years ago where things were you know 10 years August 2012. we'd we'd made 2500 roadsters and I think maybe a couple hundred Model S's ish but less than 3 000 cars so 10 years ago it made less than than 3 000 cars and here we stand 10 years later having made over three million [Applause]
4:11and actually this is one of the cleanest exponentials I've you know it it looks like one of those sort of you know business plan presentation things that that doesn't actually come true but you know you see it in the Venture Capital business plan situation but it's actually true that's the amazing part and yeah [Applause]
4:46[Laughter] we ever did okay on demand so thank you so yeah it's great here I love California too to be frank I'm going to be clear you know oh come on we got great teams in California and and we've got a great team in in China and Shanghai and a great great team in Berlin and Germany and Europe it's so
5:26yeah and I thought it's interesting to look at the the cumulative profitability since Inception and at this approximately track tracks to sort of mental pain actually his psychic damage red green curve is approximately correct but it is I'm just very proud of the fact that we've actually been able to to to produce more cash than we have spent and have positive retained earnings and be worth our salt essentially and that that's a really big deal it's very hard so I mean this is this is a big deal and I think it's gonna go up from here [Applause]
6:37so you can see we obviously had some challenges there in 2017. and then but since then our free cash flow generation has been been very good and trending trending upwards and and this is really before autonomy really kicks in we've had we have a autonomy to some degree but solving autonomy is a
7:06will really be an amplification of free cash flow to agree that is you know you run the numbers and it's like wow can it really be that crazy but it could be that crazy and you know this this year I swear [Applause] thank you [Music] we got anyone here in the FSD beta program [Applause] [Music] sweet yeah no totally totally totally okay well I I thought I thought [Applause] all right [Applause] I thought I thought you might ask that so because it's 10 to 13 we've been working on for a while and and actually what sort of happened is we've we've We've made some pretty significant architectural improvements so it's really going to be more than a 10.12 to 10.13 release it might
8:16speak too soon it might qualify for 10.69 so [Applause] it's got it obviously I can't just throw that out you know but it's it's it's that there's a lot of a lot of improvements and especially in complex left turns and yeah we're going to solve Chuck's turn yes yes 100. absolutely you know we we have a lot of respect for valid criticism so and yeah so I I I think we're at roughly 90 success rate with with your turn so yeah we're almost at 100 so so it's looking good and yeah so I mean I'm hopeful this might might it might be a week next week [Laughter] yeah so yeah two weeks so
9:27yeah it's working well for me [Applause] so yeah so anyway teams like working literally seven days a week and and making major improvements and it's really interesting because we're effectively solving an important aspect of artificial intelligence real world AI for self-driving which
10:00when you think about it kind of is what's needed to solve the to solve self-driving because how is the road system designed it's designed for a biological neural net and eyes and so naturally the thing that would therefore work the Silicon analog is cameras and silicon neural Nets and and so sort of sort of by accident we're actually solving an important I think very useful element of artificial intelligence and I definitely want people out there you know talented people who are working on AI to consider working at Tesla because I think we're solving just a very important part of AI and one that can ultimately save millions of lives and prevent tens of millions of
10:48serious injuries by by driving just an order of magnitude safer than than people and you know there used to be a time back in the day where it would have yeah this is super important so I mean there used to be a time when we'd have elevator operators it was normal to have elevator operators and have like a big relay and stuff but you know every kind every now and again you make a mistake and show somebody in half so I'd be like okay then we went to automated elevators and you press a button and you go to your floor and it just works and and that's kind of how it's going to be in the future with cars and yeah so [Applause]
11:36yeah and then we'll have lots of interesting things to say on AI day at the end of next month so I'll leave that to next month yeah so you know when we started out here obviously we were told we were told electric cars were impossible and even if really if you could make a you know an electric car with a couple hundred miles range then nobody would buy it anyway because people just love gasoline cars so you know when we sought it out it was it was dumped dumped to start a car company and then dumb squared to do an electric car company and we're told you know you're never gonna make money Etc and we didn't for a while but now we have the highest
12:19operating margin in the whole industry okay [Applause] exactly exactly Tesla is not just a car company who has those many companies in one and we're as much as a software company as we are a hardware company so and that's that's really going to be obviously essential for the future and software both in the car and obviously with neural net training but also software in the factory as well so you can think of the factory as a
12:57you know giant cybernetic Collective so Factory's just an enormous cybernetic Collective of humans and machines and software and the better the software is the better that cybernetic Collective works I don't think other oems think like that but it's that's what it is so you know there's a lot of talk of competition among like electric vehicles but really it's the the EVS are electric vehicles taking a market share from gasoline cars so yeah so the and protozo standpoint we obviously welcome this and we're very excited to see that the big car companies are embracing electric vehicles if you were to Red rewind the press releases to five years ago that
13:54was not the case they were you know percent not saying nice things about electric vehicles five five years ago but now I believe almost every major car company in the world has embraced electrification and agrees that it is the right path and this is really what we set out to do with Tesla you know it was not to like you know get maximum market share or anything it was really try to get the to show the order industry that it was possible to go electric and that if you made compelling electric cars people would buy them and and and that's that's what's happened and I think that's really profound so [Applause] yeah
14:44also when when our competitors advertise for electric vehicles every time they do that our sales go up so it's it's pretty funny so yeah but but we do we have open source all our patents so they can use our patterns for free and so we're you know helping out and I think it's sort of a mutual Prosperity thing and it's it's good so
15:12yeah so model y you don't you definitely don't want to count chickens until they're hatched but I think we're tracking to have model y be the highest selling vehicle by Revenue this year and the highest by unit volume next year so [Applause] still a lot of work to do like these these factories don't just magically work so I still a lot of work to do in Berlin
15:52and and here in Austin to spool up these two gigafactories and we've got different Supply chains and so it's always there's a host of problems none of the problems are individually all that difficult but there's like 10 000 of them so it's the rate of production is like how fast can you solve the 10 000 problems essentially and we're solving pretty fast but a lot of work to do and yeah we might be able to announce another Factory location later this year [Applause]
16:37[Music] [Applause] wait wait wait wait wait where should we okay where should we build it [Applause] [Laughter] [Music] okay we're gonna get a lot of Canada's [Applause] I'm half I'm half Canadian so maybe I should you know but yeah well I think you know ultimately we'll end up building I don't know probably at least 10 or 12 10 or 12 gigafactories and they will there will be really gigafactories like output you know aiming for output average output of like one and a half to two million units per Factory which is enormous so
17:34and the our Fremont Factory in California is is already the highest output Factory in North America so sorry well I think we'll have a friendly competition between Texas California Berlin Shanghai is hard to beat I have to say Shanghai is pretty a Shanghai team is is just awesome so that they're
18:10yeah anyway great great team there so with respect to you know sometimes you get the sort of okay you know these bogus attacks of like EVS are somehow worse than gasoline cars which is not true the average life cycle emissions of an EV are dramatically lower than that of a gasoline car am I kind of telling you things you already know but you know every now and again you think surely this nonsense has been put to bed but it it it pops up again but it's obviously electric cars just dramatically lower CO2 or lifetime than than any than a than a gasoline car so anyway that's kind of obvious now this this is the thing that I think is super cool if you add up all the energy produced
18:53over the last 10 years by Tesla solar panels it is more energy than was used to manufacture all of our cars and charge all of our cars at superchargers and at home he says I mean this is like and I mean the the mission of Tesla is to accelerate sustainability and the I think this this is really a amazing you know more energy produced than was used in making the car or and charging them all over billions of miles so and we're going to keep obviously increasing our solar activity and and vehicles and and try to keep these on
19:46par because the three elements of a sustainable energy future sustainable energy production primarily with solar and wind and then a stationary battery packs to store the the the sustainable energy because of its wind and solar intermittent intermittent and then electric transport and if you have those three pillars you have a fully sustainable future and yeah so that's yeah
20:19and obviously battery packs are various recyclable they are there you can think of it like high grade or do you wanna you know crunch up a bunch of rocks or crunch up a battery pack which is like super high grade ore so it's a no-brainer to recycle battery packs and we are already recycling I should point this out so sometimes people wonder what's happening we're already recycling at over 50 a week in Nevada so Tesla battery recycling is already underway has been underway and is scaling up Now cars are lasting quite a long time so there's not that many batteries to actually recycle because you kind of have to typically wait like 12 years or so like quite a long time
21:05before the battery actually is no longer useful sometimes 15 years so recycling starts off small but then it becomes very significant long term so to the bathroom Factory footage obviously just 50 packs a week yeah so yeah that's what I was saying like this actually just not that many packs to recycle because there's not but the vast majority of them are still in use
21:52sorry oh yes exactly where is that thing so so investment three is going to be fundamentally about scaling just looking at the overall problem from a global macro economic standpoint and saying what what are all the things that are needed to achieve a fully sustainable economy because I think a lot of people don't don't know and really like what tonnage of lithium of cathode
22:36anode separator electrolyte Electronics what are all the things that are needed that what all the things that need to be done in order to transition to a fully sustainable global economy which I think you know the sooner we do that the better for the planet and I think just articulating that and and just making it clear that this is absolutely doable and it it is being done and we just want it to be done as fast as as possible because I think a lot of times I mean I meet a lot of people out there who who have lost hope you know they think they think it's too late and they think there's there's no chance and the Earth is doomed it's like it's not doomed okay and it's yeah
23:35yeah Earth can and will be saved [Music] [Applause] so in addition to battery advancements and Electronics advancements in AI we've also done a lot to simplify the structure of cars to make it easier to manufacture and one of the things we've done is create the the largest castings that have ever been done and they're very complex castings and so we're able to take a 171 pieces of metal and go from 171 pieces to two and in the process make it lighter stiffer with better right handling better noise vibration harshness better sealing against water so it's really better in every way and yeah so
24:36[Applause] so and we're going to keep enhancing the the casting so this is a testament to our materials team and to have a lot of casting technology so really rethinking the whole way in which a car is made and it's yeah it's a gigantic Improvement we have at this point from going from say model s or even even model three we're at about 30 percent of the robots used for model three from for a current model line so yeah [Applause]
25:23we've also improved the layout of the factory so the factory is sort of a it's sort of a single well close to a single monolithic Factory with a very straightforward flow Fremont we do a lot in Fremont but the flow is complex and and and it's it's not not an easy flow so we're really rethinking the factory and I think if the the like the really long-term sustainable advantage of Tesla will be manufacturing I've mentioned this before but obviously everyone will have electric cars all manufacturers have electric cars and eventually it'll probably take longer than they think but eventually all cars will be self-driving and the thing that will be hardest to
26:12replicate is Tesla's manufacturing technology so this is actually very important for a long-term standpoint [Applause] you know when thinking about the competitiveness of companies especially if the companies are technology companies I recommend looking at where the smartest Engineers want to work this is wherever the the smartest Engineers want to work that's going to be that that that technology company is going to be the one that is likely to succeed just like if it's a pro sports team where are the ace players going okay probably that team will win so we put a lot of effort into ensuring that the best engineers in the world want to work at at Tesla and
27:10Frank sometimes Tesla's number one sometimes SpaceX number one but so those are just like like last year but and we do actually for those curious and maybe want to work at either SpaceX or Tesla we do allow people to move from one company to the other if they would like so if you want to spend a bit of time working on electric vehicles better time spending working on Rockets you can that that's cool we support that so yeah
27:43and we've also made a lot of improvement with Factory safety so we I think we believe we now have the the best Factory safety in the industry and so we're yeah we're excited to have you know very good safety and getting better so and a bunch of this is driven by just encouraging people within the factory to submit ideas for safety Improvement and We've passed our goal of of three suggestions per employee in this year and this this is really is a game changer for improving safety so and yeah
28:34don't you don't have to clap after every slide unless you really want to but so we we there's a lot of interest in working at Tesla with three million job applications last year so sorry yeah it's true so a lot of interest in working at Tesla now some people may occasionally have encountered a supercharger that didn't work but in general the uptime of our superchargers is extremely good and we just try to make it super smooth and the because our superchargers are always connected like the car is always connected our feedback loop for fixing a supercharger is very quick and obviously we keep upgrading the supercharger capability we're now at version three we'll start rolling out
29:27version four I don't know maybe next year so that just the supercharges get better and better and you know who knows that yeah actually I don't give away this is [Laughter] let's just say that there's there's some there's some cool stuff happening on the supercharging front so Alaska yes
29:58all right we're we're basically doubling our supercharger account every year so if there isn't a supercharger in some place that you think is important it probably will be that soon so we've also made a lot of progress on safety so our cars are already the the safest in the industry so we have the lowest probability of Entry of any cars ever tested by the US government and we update this in real time so we because our cars are connected this is a big difference all the other cars with rare exception are not connected so that so the manufacturers don't really know what happened in a crash but in the case of Tesla we are able to look at crashes and see how can
30:45we improve the safety of you know look at the crashes and prove the the the the crash structure also we improve how the seat belts tension and how the airbags deploy so in these updates if we figure out oh there's a slightly better way to deploy the airbags to improve the safety in a crash we'll actually do an over there update to improve the way that the airbags deploy or the way that the seat belt pre-tensions and we're now starting to include our Tesla autopilot AI to be able to see if
31:23a crash is about to occur and if it if it sees that a crash is about to occur with 99 plus percent probability then it will activate the seat belt pre-tensioners and deploy the airbags as opposed to what the vast majority of cars do is the the only the vast majority of cars are only able to deploy airbags when the crash is happening and so that's that makes the airbag deployment a lot more violent but if you can anticipate the crash the airbag deployment can be much better than than just impact so
32:01okay [Applause] we're also seeing FST beta Grove very rapidly and this this is definitely going to go very exponential in miles driven so we're now at over 40 million miles and I suspect by this year we'll be I don't know well over 100 million miles so and we're still tracking very much to have a
32:34white widespread deployment of FST beta this year in North America so I'm yeah I should say if if basically FSD beta will be available to anyone who requests it by the end of this year so [Applause] so AI Day part two I think people were blown away but that's at the end of next month so we'll leave that if you accept it's there's going to be a lot of cool stuff so
33:18the I'm sort of surprised that you know people or at least like analysts out there are not really understanding the importance of the Optimus robot my guess is Optimus will be more valuable than the car than long-term and in fact it will I think turn the whole notion of what's an economy on its head you know the point at which you have no shortage of Labor you know economy is GDP per capita if you do not have a capital constraint then the economy just is can be arbitrarily huge so yeah it's sort of crazy
34:09anyway thank you for voting so your vote matters and and it's great to have direct contact with individual shareholders and thank you for for voting to support our proposals foreign
34:41I wonder how many cars we'll have in 10 years 10 years ago we had less than 3 000 cars now we've made three million 300 million foreign I say 100 million is pretty doable so I'd be surprised if I'd be surprised if it's not over 100 million in 10 years and then many other products some of what you've heard about some of which you have not so [Applause]
35:31you upload it out all right so we've got some questions from the internet that were voted to the top so I'll answer these questions and then we'll take just questions from the audience
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