一家季度交付下滑、股价承压的汽车公司,CEO 在财报电话会上说,这家公司未来的市值可能等于排在它后面五家公司的总和。
他还加了个前提:前提是执行得好。
特斯拉 2025 年第一季度财报电话会,马斯克开口先讲的不是车,是华盛顿。整场会下来,信息密度其实很高:Robotaxi 的启动时间、Optimus 的产能节奏、一款还没量产的新车型、一套要把生产周期压到五秒的制造系统,还有他对关税、对中国供应链、对人形机器人全球排行榜的判断。下面按顺序拆开讲。
先说 DOGE:抗议是有人付钱的,五月起我要回特斯拉
马斯克在政府效率部(Department of Government Efficiency,简称 DOGE)的任职,是这场电话会绕不开的背景。他自己先挑明了:因为这段时间在政府工作,他承受了不少反弹。
他把这件事的动机说得很直白,美国正在以不可持续的速度积累赤字,这条路通向破产。DOGE 团队在做的,是削减浪费和欺诈。
他给出的选择只有两个。
“我们要么让浪费和欺诈继续下去,它的增长速度正在让这个国家破产;要么就跟浪费和欺诈对着干,把国家拉回正轨。我认为正确的事是后者。”(马斯克)
他还补了一句更重的:如果美国这艘船沉了,所有人都会跟着沉,包括特斯拉,包括其他所有公司。
关于外面那些抗议,马斯克的说法是:这些抗议组织严密,是有人出钱支持的。他表示抗议者当然不会承认自己是因为拿了不该拿的钱才上街,他们会找出别的理由,但真实原因就是那些收着浪费资金的人希望继续收下去。
这部分内容对投资者最有用的一句在后面:从五月开始,他投入 DOGE 的时间会大幅下降。
他的安排是,建立 DOGE、把政府的财务秩序理顺,这块大工作量基本完成。接下来的总统任期内,他会继续每周花一两天在政府事务上,目的是防止已经砍掉的浪费和欺诈卷土重来。但五月起,他的主要时间会回到特斯拉。
这句话对特斯拉的意义,不用多解释。
“我们至少十几次在死亡边缘,这次不是”
马斯克主动提到了特斯拉的历史。
他说这家公司这些年经历过很多次危机,至少十几次处在死亡的边缘。然后他给这轮下了一个判断:
“这次不是那种时候。我们不在死亡边缘,差得远。但确实有一些挑战,我预计今年会有一些意料之外的颠簸。”(马斯克)
颠簸会来自哪里,他没细说,但他把时间尺度拉长了。特斯拉的长期价值建立在两件事上:大规模的自动驾驶汽车,以及大规模、高产量的通用人形机器人。
他给出的估值判断相当夸张:一家能以低成本大规模制造真正有用的自动驾驶人形机器人和自动驾驶汽车的公司,价值是惊人的。如果执行到位,特斯拉会成为全世界市值最高的公司,而且遥遥领先,可能相当于后面五家公司加起来。
这句话要拆成两层来看。第一层是前提,“执行得好”这个 if 是他自己反复强调的,不是客套。第二层是时间,他没说这事今年发生,也没说明年。他在会上多次把“真正开始影响财务”的时间点指向 2026 年年中。
换句话说,接下来几个季度的财报大概率不会好看,这是他提前打的预防针。他甚至在会上劝投资者把目光抬起来,越过眼前这段路,看向前方那个“光亮的未来”。
自动驾驶的钱什么时候到账:2026 年年中
这是整场电话会最实在的一段,因为马斯克难得给了一个财务上的时间表。
6 月,奥斯汀。特斯拉会在那里开始提供完全无人的付费载客服务,车里不坐人。这件事他之前讲过几个月了,这次确认还在轨道上。
但他把“技术跑通”和“财务见效”分得很清楚。从财务角度看,真正的问题是这件事什么时候能对公司的利润表产生实质影响。他的回答是 2026 年年中,或者 2026 年下半年。一旦开始实质性地推动财务,后面就会变成指数级增长。
起步的规模小得有点出人意料。奥斯汀上线第一天,车队规模大概是 10 到 20 辆车。
马斯克的解释是,第一批要非常仔细地观察,然后再快速放大。他还给了个具体场景:6 月底或 7 月,你可以直接去奥斯汀,叫一辆自动驾驶的特斯拉。
为什么起步这么谨慎,他讲了一个测试逻辑。要发现那些极少见的边缘情况,比如两万英里才出现一次的问题,光靠少量车跑是不够的。普通人一年开一万英里,要在一两个月里压缩完这个测试周期,就需要大量车辆同时在路上跑。
所以现在奥斯汀街上会看到奇怪的一幕:一支接一支的特斯拉车队在城里绕圈。马斯克自己说,这看起来会相当诡异。
增长速度他用了 S 曲线来描述。S 曲线中间那段的斜率很难预测,但终点是清楚的:绝大多数特斯拉车辆实现自动驾驶。他的预测是,2026 年下半年会有数百万辆特斯拉在全自动驾驶状态下运行。
年底之前,美国多座城市会开放个人使用。他给的验收标准很朴素:你能不能在车里睡着,醒来时已经到了目的地。他认为今年年底这个场景在美国很多城市就能实现。
还有一个技术细节值得记下来。马斯克提到,针对不同地区可能需要一套本地化的参数。比如美国东北部下暴雪的天气,一个在加州开得很好的“司机”,换成曼哈顿的暴雪未必开得好。人会这样,用数字神经网络和摄像头工作的车也会这样。他明确说这不是必需的配置,但有价值。
为什么特斯拉敢只靠摄像头
会上有人问 Waymo。马斯克的回答直指成本结构。
Waymo 的车非常贵,产量很低。特斯拉的车成本大概是 Waymo 的四分之一,甚至只有 20%,而且是高产量制造。他用了“讽刺”这个词:谷歌在 AI 上非常强,却选择了昂贵的传感器套件路线;特斯拉赌的是纯 AI 加摄像头的方案,而且赌对了。
这套方案的核心优势是通用性。马斯克强调,特斯拉在解决的是自动驾驶的通用问题,不是某个城市的具体问题。
“一旦我们能在几个城市跑通,就能在这个法律辖区内的所有城市跑通。在美国几个城市跑通,就能在全美跑通。在中国几个城市跑通,就能在全中国跑通。欧洲同理,唯一的限制是监管批准。”(马斯克)
对比一下另一条路线就明白了。靠高精度地图的方案,把某个街区扫得非常精细,可这个街区经常在变,一变车就不工作了。特斯拉靠的是 AI 和一颗为这个目的专门设计的自研芯片。
这里有个容易被忽略的事实:特斯拉的 AI 软件团队和 AI 芯片设计团队都是自己建的,没有收购任何公司。
至于市场份额,马斯克的表态是:“我看不到现在有谁能跟特斯拉竞争。以后肯定会变,但至少据我所知,特斯拉会拿到 99% 这种荒谬的份额,至少九十几。”
他补了一句时间尺度:如果明年有数百万辆车部署出去,而别人没有,那用不了几年,路上就会有 1000 万辆自动驾驶汽车,而且还在增加。
还有个技术细节值得单独拎出来。有人问摄像头会不会被阳光晃瞎。马斯克的回答涉及一个叫直接光子计数(direct photon count)的方案。常规做法是光子信号先经过图像信号处理器处理,一旦对着太阳,后处理会把画面洗白,反而增加噪声。特斯拉绕过了这一步,直接读光子计数。结果是车可以基本正对太阳行驶,也能在看起来最黑的夜里看见东西。雾天里,它的表现比普通人眼还要好一点。
顺带纠正一个容易混淆的概念。Robotaxi(自动驾驶出租车)是一类服务的统称,任何一辆实现自动驾驶的特斯拉都能当 Robotaxi 用。Cybercab 是特斯拉一款具体产品的名字。马斯克自己都说这套命名“非常混乱”。已经生产出来的绝大多数特斯拉车辆,硬件上都具备成为 Robotaxi 的能力。
Optimus:一万个零件,没人能准确预测爬坡
人形机器人这块,马斯克先泼了一盆冷水,再给了时间表。
冷水是:Optimus 现在仍然是一个研发项目,不是量产项目。今年的产量是几千台,而且大部分集中在下半年。
他解释为什么不能快。Optimus 上几乎所有东西都是新的,电机、齿轮箱、电子器件、执行器,几乎找不到现成的供应链,唯一例外是那颗特斯拉 AI 计算机,和车上用的是同一个。
然后他说了一段关于制造业的实话:
“一个复杂的新产品,它的推进速度取决于整个系统里最慢、最不幸运的那个零件。粗略估计,Optimus 有一万个独特的零件。所以这里任何人如果告诉你他能精确预测一个全新产品的产能爬坡,他不知道自己在说什么。这根本不可能。”(马斯克)
他的描述是,你会经历一连串的瓶颈,这个零件卡住了,解决了,那个零件又卡住了,如此乘以一千次。产能爬坡的速度,取决于你解决每个问题的速度。
Optimus 今年就撞上了一个具体的麻烦:稀土磁铁。机器人手臂上的执行器为了把电机做小,用了永磁体。中国要求这类磁铁出口需要许可证。马斯克说正在跟中方沟通,中国希望得到保证这些磁铁不会被用于军事目的,他的回应是这些磁铁只进人形机器人,不是武器系统。
尽管有这些障碍,他的产量预测依然激进:
今年年底,特斯拉工厂里有几千台 Optimus 在干活。五年内,年产量做到 100 万台,可能四年就够。2030 年他有信心实现年产百万台,也可能是 2029 年。
他还提到一个规模判断:Optimus 会是史上放量最快的产品。
Megapack:把一座电网的年输出翻一倍
能源业务这段,马斯克讲了一个容易被忽略的物理常识。
一座电网的实际能力,远大于它现在每年的总输出。原因是发电厂不能一直满功率运行,夜里用电少,机组只能降到半功率甚至四分之一功率。如果能让电厂 24 小时满功率跑,现有电厂的年发电量可以翻倍。
缺的那一环是缓冲。夜里把电存进电池,白天再放出来。Megapack 干的就是这件事。
这也是为什么公用事业公司开始大规模采购。马斯克说,现在吉瓦级(GW)的电池订单已经很常见,特斯拉手里有不少吉瓦级甚至更大的订单。他预计固定式储能业务最终会做到每年太瓦级(TW)。
关于汽车业务,他提了一个季节性因素。第一季度通常是最差的季度,没人愿意在暴风雪里去买车。特斯拉偏偏选了这个季度做新版本 Model Y 的产线切换。
这里要澄清一个数字。文字稿里出现了“11 亿台”的说法,明显是语音识别错误。Model Y 是全球最畅销的车型,单车年产量大约一百多万辆。真正的难点在于,特斯拉是在全球所有工厂同时做这次切换的。马斯克在会上专门感谢了团队,说这是一次非常困难的转换。
五秒一台车,而这只是步行速度
制造部分的信息量可能是整场会里最大的,尽管它听起来最不性感。
马斯克提到了那套被称作“Unboxing”的新生产系统。他特意澄清,这个名字容易让人联想到拆手机包装盒的愉悦体验,但那不是它的意思。
“这是对造车方式的一次彻底重构。全世界没有任何一辆车是这么造出来的。工厂本身就是产品,和车一样是产品。”(马斯克)
目标指标是节拍时间(cycle time)五秒或更短,也就是一条产线每五秒下线一台车。
这个数字听起来像子弹发射。马斯克的换算很有意思:五秒一台,换算成线速度大约是每秒一米,也就是每小时三英里左右的步行速度。你甚至能跑得过它。
但放到汽车行业里,这个速度是碾压性的。特斯拉上海工厂目前的节拍是 33 秒,已经是他们认知里全球最快。新架构的设计能力是它的六到七倍。
他给了一个很坦率的预期管理。上海工厂在传统汽车生产体系里是 A+ 的水平,想把节拍压到 30 秒以下极其困难,因为再快就要超过人能做到的极限。新的无人化产线靠的是跑得极快的机器人。但新架构刚起步时,成绩大概是 C。潜力在于,这套架构本身的上限是 A+ 里的 A+。
工厂内部已经在跑真实场景了。Model Y 从产线末端自己开到物流场,自己停到等卡车来接的准确位置。物流场就在露天,没有专用车道,人来人往,卡车在卸货,施工在进行。马斯克特意说,从马路上就能看见,没有遮挡,任何人都可以开车经过弗里蒙特工厂看这些车自己跑。
他还给了一个更远的承诺:今年晚些时候,第一辆 Model Y 会从工厂直接开到客户家里。弗里蒙特和另一座工厂都在计划内。
顺便一个尺寸对比:奥斯汀超级工厂的面积是五角大楼的三倍。
从锂矿到电芯:福特之后最垂直整合的车企
关税话题上,马斯克把能做的和不能做的分得很清。关税决定权完全在美国总统手里,他是顾问之一,会提建议,总统会听,但决定权不在他。他的立场一贯明确:低关税总体上有利于繁荣,他会继续主张低关税。
特斯拉真正能做的,是把供应链本地化。这件事他们干了好几年,逻辑上既降成本又降物流风险:至少让供应链和整车在同一块大陆上。北美、欧洲、中国三地都有本地化供应链,特斯拉因此成了受关税影响最小的车企之一。他也承认,在利润率本来就低的时候,关税依然难受。
具体做到什么程度,会上给了不少细节。每一种电池部件,至少有两个不同国家的来源。这件事供应链团队和工程团队一起做了好几年,不是几个月能搭起来的。
更狠的是纵向整合。南德州有锂精炼厂,是中国以外最大的,还有扩产空间。奥斯汀超级工厂旁边有正极材料精炼厂。负极材料还在讨论。
马斯克在这里说了一句很有他风格的话:
“最好的零件是没有零件。这是锂电池的终极梦想,不要负极。”(马斯克)
现实当然还得有负极、正极、锂、电解液、隔膜,然后把这些东西在电芯工厂里组装起来。市面上有整家公司只做电芯,其他什么都不做。特斯拉是从锂精炼一路做到电芯下线。马斯克的判断是,他们目前是全行业每千瓦时成本最低的,至少在北美采购的电芯里是最低的。
用他的话说,特斯拉是自亨利·福特时代以来垂直整合程度最高的车企。当年福特自己挖铁矿、种橡胶树,特斯拉还没到那一步,但方向一致。
关税、无人机,和那个“第一名一定是特斯拉”的排行榜
会上有人问无人机。马斯克的回答引用了一条他在 X 上转发过的推文,作者是他的朋友纳瓦尔(Naval Ravikant):
“任何不能自己制造无人机的国家,注定成为能制造无人机的国家的附庸。”(纳瓦尔,马斯克转发)
他接着补了一句:美国目前造不出自己的无人机。中国生产了全球大约 70% 的无人机,如果看整条供应链,几乎所有无人机都对中国有依赖。他的态度是,美国需要把更多人和资源转移到制造业上,这种程度的依赖不该存在。
他对中国的评价是“了不起”,但结论很硬:美国不该在无人机上依赖到这种地步,依赖到只有中国给零件才能造出来。
人形机器人这块,他的判断分两层。第一层:目前全世界没有任何国家的任何公司能跟特斯拉比。第二层,他有点担心:
“排行榜上,第二到第十名可能都是中国公司。但我相信第一名会是特斯拉。”(马斯克)
汽车需求方面,他的说法比较克制。特斯拉不能免于宏观环境的影响,经济不确定的时候,人们会推迟买车这种大额支出。但排除宏观因素,他们没看到需求下降。
“买一辆不能自动驾驶的汽油车,就像骑着马用翻盖手机”
马斯克在这里讲了一个他讲了很多年的类比,这次讲得最完整。
翻盖手机时代,市面上一百多种设计。厂商犯的错误是继续做各种变体:换个造型,换个颜色,找下一个细分市场。诺基亚一度是全球市值最高或接近最高的公司,它继续做翻盖手机。然后 iPhone 出来了。
马斯克的判断是,用户要的只有一台什么都能做的超级智能手机,就一台。
“在并不遥远的未来,买一辆不能自动驾驶的汽油车,就像骑着马用翻盖手机。还有人这么做,但很少了。”(马斯克)
延伸出来的结论更激进:未来大多数人不会买车。
关于 FSD 的体验限制,他给了一个很具体的判断标准。真正的分水岭是:你能不能在车上读短信、回短信。人们现在手动开车时就在干这些事,80 英里时速下发短信、化妆、梳头、吃午饭、刮胡子,都是常态。
现在的问题是车太较真,坚持要你盯着路。马斯克说会逐步放宽,每几周或每个月松一点,直到你能做任何想做的事,包括睡觉。到那时候,每月 99 美元的订阅会是你这辈子花得最值的 99 美元。
给读者的行动清单
这场电话会信息量大,但真正能用来做判断的锚点并不多。下面几条是可以直接拿来跟踪的。
- 盯 6 月奥斯汀的实际车队规模,不是发布会。首日 10 到 20 辆起步,看它多久爬到百辆、千辆,以及有没有公开的接管率数据。这比任何演示视频都硬。
- 把 2026 年年中当成一个可验证的时间承诺记下来。马斯克说财务影响在那个时候显现,如果到 2026 年 Q2 财报还看不到自动驾驶收入的实质贡献,这个时间表就应该重新打折。
- Optimus 的观察重点是供应链,不是演示。稀土磁铁的出口许可是当下最具体的卡点,许可证批不批、什么时候批,比机器人走路的视频更能说明量产节奏。
- 看能源业务的收入占比变化。Megapack 的订单从吉瓦级走向太瓦级是个长期趋势,这条线的确定性比自动驾驶高,被讨论得却少得多。
- 关税的判断要盯总统的决定,不是马斯克的表态。他自己在会上说得很清楚,决定权不在他手里,他只能建议。
- 人形机器人的竞争格局,值得把中国公司单独列一张表。马斯克自己说第二到第十名可能都是中国公司,这句话里的信息量比“特斯拉第一”更大。
- 制造业那条线容易被忽略,但五秒节拍是个可以长期跟踪的指标。上海工厂 33 秒是目前的最快纪录,新架构能不能从 C 爬到 A+,会决定特斯拉未来十年的成本结构。
Hello, everyone. Well, there's never a dull moment these days. Thanks for sure. Every day is going to be exciting. As some people know, there's been some blowback for the time that I've been spending in government with the Department of Government Efficiency or DOGE. I think the work that we're doing there is actually very important for trying to sprain in the insane deficit that is leading our country, the United States, to destruction. And the DOGE team has made a lot of progress in addressing waste and fraud. The natural blowback from that is those who were receiving the wasteful dollars and the fortunate dollars will try to attack me and DOGE team and anything associated with me.
But then I'm really left with two choices. Should we just let the waste and fraud continue? And it was continuing to grow at a really unsustainable pace that was bankrupting the country. Or to fight the waste and fraud and try to get the country back on the right track? And I believe the right thing to do is to just fight the waste and fraud and get the country back on the right track and working together with President Trump and his administration. Because if the ship of America goes down, we all go down with it, including Tesla and everyone else. So, I think this is critical work. Now, the protests that you'll see out there, they're very organized, they're paid for.
They're obviously not going to say, admit that the reason that they're protesting is because they're receiving fraudulent money or that they're the recipients of wasteful largesse, but they're going to come up with some other reason. But the real reason for the protests, the actual reason is that those receiving the waste and fraud wish to continue receiving it. That is the real thing that's going on here, obviously. Now that said, I do think there's the large slug of work necessary to get the DOGE team in place and working in the government to get the financial house in order is mostly done. And I think starting probably next month, May, my time allocation to DOGE will drop significantly.
I'll have to continue doing it for, I think, probably the remainder of the President's term, just to make sure that the waste and fraud that we stop does not come roaring back, which will do if it has the chance. So, I think I'll continue to spend a day or two per week on government matters for as long as the President would like me to do so and as long as it is useful. But starting next month, I'll be allocating probably more of my time to Tesla and now that the major work of establishing the Department of Government Efficiency is done. So, at Tesla, we've gone through many, many crises over the years and actually been through many near-death experiences.
We were probably on the ragged edge of death at least on maybe a dozen times. It's been so many times. This is not one of those times. We're not on the ragged edge of death, not even close. But there are some challenges, and I expect that this year will be, there will probably be some unexpected bumps this year. But I remain extremely optimistic about the future of the company. The future of the company is fundamentally based on large-scale autonomous cars and large-scale, large volume, vast numbers of autonomous humanoid robots.
So the value of a company that makes truly useful autonomous humanoid robots and autonomous useful vehicles at scale at low cost, which is what Tesla is going to do, is staggering. I continue to believe that Tesla, with excellent execution, will be the most valuable company in the world by far. But that's an important if, we must execute well. But if we do execute well, I think Tesla will be the most valuable company in the world by far. It may be as valuable as the next five companies combined. But there'll be a few bumps along the road before that happens.
I said I think on the last earnings call that we'll start to see the prosperity of autonomy take effect in a material way around the middle of next year. We expect to have these … be selling fully autonomous rides in June in Austin, as we've been saying for now several months. So that's continued, but the real question from financial standpoint is when does it really become material and affect bottom line of the company and start to be a fundamental part of the … When does it move the financial needle in a significant way? That's probably around the middle of next year, second half of next year.
And then once it does start to move the financial needle in a significant way, it will really go exponential from there. So that's, I'd encourage people to look beyond some sort of bumps and bundles of each road immediately ahead of us. Lift your gaze to the bright shining sort of down the hill, I don't know, some Reagan-esque imagery. And that's where we're headed and not too distant future like I said. Kind of next year or two. So, let's see. With respect to supply chain risk, something that Tesla has been working on for several years is to localize supply chains.
This actually makes sense from a cost standpoint and from a logistics risk standpoint, is to have the supply chains be at least located on the continent in which the car is built. So we are, I think the least company, the least affected car company with respect to tariffs, at least in most respects. I mean, it remains to be seen. Now, tariffs are still tough on a company when margins are still low. But we do have localized supply chains in North America, Europe, and China. So that puts us in a stronger position than any of our competitors.
And undoubtedly, I'm going to get a lot of questions about tariffs, and I just want to emphasize that the tariff decision is entirely up to the President of the United States. I will weigh in with my advice with the President, which he will listen to my advice, but then it's up to him, of course, to make his decision. I've been on the record many times saying that I believe lower tariffs are generally a good idea for prosperity, but this decision is fundamentally up to the elected representative of the people being the President of the United States. So I'll continue to advocate for lower tariffs rather than higher tariffs, but that's all I can do.
Now let me walk you through why I'm so excited about the future of Tesla. So, first of all, autonomy. The team and I are laser focused on bringing robotaxi to Austin in June. Unsupervised autonomy will first be solved for the Model Y in Austin. And then … Actually you should parse out the terms robotic taxi or robotaxi and just generally like what's the Cybercab because we've got a product called the Cybercab and then any Tesla which could be an S3 extra wide that is autonomous is a robotic taxi or robotaxi. It's very confusing. So the vast majority of the Tesla fleet that we've made is capable of being a robotaxi or robotic taxi.
Once we can make the whole system work where you can have paid rides fully autonomously with no one in the car in one city, that is a very scalable thing for us to go broadly within whatever jurisdiction allows us to operate. So, because what we're solving for is a general solution to autonomy, not a city specific solution for autonomy, once we make it work in a few cities, we can basically make it work in all cities in that legal jurisdiction. Once we can make it based to work in a few cities in America, we can make it work anywhere in America. Once we can make it work in a few cities in China, we can make it work anywhere in China, likewise in Europe, limited only by regulatory approvals.
So, this is the advantage of having a generalized solution using artificial intelligence and an AI chip that Tesla designed specifically for this purpose as opposed to very expensive sensors and high precision maps of a particular neighborhood where that neighborhood may change or often changes and then the car stops working. So, we have a general solution instead of a specific solution. Then, with regards to Optimus, we're making good progress in Optimus. We expect to have thousands of Optimus robots working in Tesla factories by the end of this year, 10 years forward.
And we expect to scale Optimus up faster than any product, I think, in history, to get to millions of units per year as soon as possible. I think I feel confident in getting to a million units per year in less than five years, maybe four years. So, by 2030, I feel confident in predicting a million Optimus units per year. It might be 2029. Let's see with respect to energy, energy business is doing very well. The Megapack enables utility companies to output far more total energy than would otherwise be the case. When you think of the energy capability of a grid, it's much more than, say, total energy output per year.
If the powerplants could operate at peak power for all 24 hours, as opposed to being at half power, sometimes a quarter power at night, then you could double the energy output of existing power plants. But in order to do that, you need to buffer the energy, so that you can charge up something like a battery pack at night and then discharge into the grid during the day. So, this is a massive unlock on total energy output of any given grid over the course of a year. And utility companies are beginning to realize this and are buying in our Megapacks at scale. So, at this point, a gigawatt class battery is quite a common thing. We have many orders and offer for gigawatt and beyond batteries.
And we expect the stationary energy storage business to scale ultimately to terawatts per year. So very, very good numbers. Now, Q1, first quarters of a year are usually pretty tricky. Because it's usually the worst quarter of the year because people don't want to go buy a car in the middle of winter during a blizzard. So we picked Q1 as a good quarter to do a cutover to the new version of the Model Y and we changed production of the world's bestselling cars with … Remember the Model Y is the bestselling car of any kind on earth with a 1.1 billion unit per year output of a single model. And we did this changeover at the same time in factories all across the world.
So congratulations to the Tesla team on an amazing job in pulling off what is a very difficult transition. So yeah, it's really very impressive work. So, yeah. In conclusion, while there are many near-term headwinds for us and the border industry, the future for Tesla is brighter than ever. The value of the company is delivering sustainable abundance with our affordable AI powered robots. I like this phrase, sustainable abundance for all. If you say, like, what's the ideal future that you can imagine? That's what you'd want. You'd want abundance for all in a way that's sustainable. It's good for the environment. Basically, this is the happy future.
If you say what's the happiest future you can imagine. One which is that would be a future where there's sustainable abundance for all. Closest thing to heaven we can get on Earth, basically. So, thank you again to the Tesla team for all their efforts of the challenging time. I look forward to continuing to lead the team to great success in the future.
The expense is bringing it in from China right now.
Sure. Well, just … we must just [inaudible 00:28:23] by the … disambiguate the Cybercab from Robotaxi once again. Because the Teslas that will be fully autonomous in June, in Austin, probably Model Ys, so that's currently on track to be able to do paid rides, fully autonomously, in Austin in June, and then to be in any other cities in the US by the end of this year. It's difficult to predict the exact ramp week by week or month by month, except that it will ramp up very quickly.
So it's going to be like basically an S-curve where it's very difficult to predict the intermediate slope of the S-curve, but you kind of know where the S-curve is going end up, which is the vast majority of the Tesla fleet being autonomous. So that's why I feel confident in predicting large-scale autonomy around the middle of next year, certainly the second half of next year, meaning I predict that there'll be millions of Teslas operating fully autonomously in the second half of next year. It does seem increasingly likely that there will be a localized parameter set perhaps sort of … especially for places that have, say, very snowy weather.
Like say if you're in the Northeast or something, like there's … you can think of it, it's kind of a human. You could be a very good driver in California, but are you going to be also a good driver in a blizzard in Manhattan? You're not going to be as good. So there is actually some value in … you can still drive, but your probability of an accident is higher. So it is increasingly obvious that there's some value to having a localized set of parameters for different regions and localities. But the output [inaudible 00:31:23] the [inaudible 00:31:23] category, it's not the required category. Again, really the car is just very much like the human.
It's digital neural nets and cameras, and humans operate with biological neural nets and eyes. And so the same strengths and weaknesses will be present. Or a digital neural net and cameras versus a biological neural net and eyes. Ashok, if you'd like to elaborate on that.
I mean very basic terms. If we're sitting at an accident every 10,000 miles, well then you have to drive 10,000 miles, on average, before you get an accident or an intervention. So it's like, "Okay." We must be really … people must be very wigged out by the sheer number of Teslas doing circuits in Austin right now. We're like … it's going to look pretty bizarre.
Yeah, there's just always a convoy of Teslas going, just going all over to Austin in circles. But I just can't emphasize this enough. In order to figure out long tail things, if it's one in 10,000, let's say it's one in 20,000 miles, or one in 30, the average person drives 10,000 miles in a year. So now try to compress that test cycle into a matter of a few months, that means you need a lot of cars during a lot of driving in order to compress that, or just do in a matter of a month what would normally take someone a year.
Oh, yeah, those videos are amazing.
Frankly, I think the Chinese consumer might be the most demanding consumer, and actually the customers in China are awesome. They have a lot of fun with the cars. I saw one guy take a Tesla, autonomous, on a narrow dirt road across a mountain. And I'm like, "[inaudible 00:35:54] a very brave person." And the Tesla's driving along on a road with no barriers where, if it makes a mistake, he's going to plunge to his doom. But it worked.
Yeah, yeah. We should ask that question too. Speaker 1 : Yeah, we have our first big builds coming at the end of this quarter within Q2, and then, in the coming months, we start to large scale installation of all the equipment in Giga Texas with still on schedule for production next year.
It's in the Giga [inaudible 00:36:43].
Yeah. Yeah. Speaker 1 : It's happening and people don't know. It's just happening upstairs and along the long lines while we're still building the Model Ys and Cybertrucks every day.
Yeah. It's worth noting that the Tesla Gigafactory in Austin is three times the size of the Pentagon. Speaker 1 : Including the garden.
Yeah, including the ground zero garden. So because of the Pentagon, like this building used to look big. Not anymore.
But before the end of this year. Not necessarily … let's say, within the US … we do want to test … at Tesla, we're absolutely hardcore about safety. We go to great lengths to make the safest car in the world, and have the lowest accidents per mile, and fewest lives lost, so we want to be very careful. And we want autonomy to be definitively safer than manual driving. So it's not enough that it'd just be as safe, it needs to be meaningfully safer than if it's cars manually driven. And we want to confirm that there's not something … we just want to be cautious with the roll-out. We don't want to jump in at the deep end with an army.
But that said, I think we should be able to have it work in several studies, later this year, for personal use. The acid test being you should be able to … can you go to sleep in your car and wake [inaudible 00:38:46] your destination? And I am confident that will be available in many cities in the US by the end of this year.
Well, okay. The issue with Waymo's cars is it costs way more money. [inaudible 00:40:38] But that is the issue. The cars are very expensive, made in low volume. Teslas are, I don't know, probably cost a quarter, 20% [inaudible 00:40:54] a quarter what a Waymo costs, and made in very high volume. So, ironically, we are the ones that made the bet that a pure AI solution with cameras, and what do you have, the car actually will listen for sirens and that kind thing, is the right move. And Waymo decided that an expensive sensor suite is the way to go even though Google's very good at AI. It's ironic.
And it is worth noting that Tesla's built an incredible AI software team and AI hardware chip design team. Prospect nothing. We didn't acquire anyone. We just built it. So yeah, it's really … I mean, I don't see anyone being able to compete with Tesla at present. I'm sure that'll change eventually, but at least as far as I'm aware, Tesla will have, I don't know, 99% market share or something ridiculous. That ninety-something percent, at least, I don't know, something might change, but if we have millions of cars deployed next year, and unless others have millions of cars deployed … unless we're blocked by regulatory situations, it won't be long.
I mean, in a few years we'll have 10 million autonomous cars on the roads and counting.
Yeah.
Way more money.
With the software update, it will become autonomous. To be clear, the Model Ys that we're talking about being autonomous in Austin in June are the Model Ys we make currently. There's no change to it.
The car that they have.
Yeah. Speaker 1 : [inaudible 00:43:39] through the tunnel, the Model Ys. Everything.
Right. Exactly. It is being put to … it's doing useful work, fully autonomously, at the factories. As Ashok was mentioning, the cars driving themselves from end of line to where they're supposed to be picked up by a truck to be taken to a customer.
… Customer, and I'm confident also that, later this year, the first Model Y will drive itself all the way to the customer. From our factory in Boston and our one in… Here in Fremont, California, I'm confident that, from both factories, we'll be able to drive directly to a customer from a factory.
Yeah. Literally goes from the end of line and drives itself to your house. Speaker 2 : It's important to note, in the factories, we don't have dedicated lanes or anything. People are coming in and out every day, trucks delivering supplies, parts, construction. For the-
And people can film it. By the way, you can see this from the road. It's not covered. Speaker 2 : Exactly.
In these videos, people will take videos online and anyone who wants to go see it can just drive past our Fremont factory and see the autonomous cars- Speaker 2 : [inaudible 00:44:55].
… driving themselves. Yeah, and they drive themselves and they put themselves in the exact right spot to be picked up. Speaker 2 : Yeah, the logistics yard is right there in the open. We don't move it again to another lane.
But they go to a specific spot. Find a spot. Speaker 2 : Yeah.
Yeah, that's just a routine everyday thing now.
Yeah. Although the line won't be obviously, at this rate-
… initially, this is a revolutionary production system. I'm not sure what the right word is. Unboxing sounds like something… Like when you get your phone.
Yeah, you have a pleasant experience when you take your phone out of the box, which, of course, is nice, but this is much more revolutionary than that. This is a profound re-imagining of how to make cars in the first place. No car is made like this anywhere in the world. The factory is the product as much as the car is the product. It's really just the first principles approach to manufacturing that will ultimately allow us, I think, to… I'm trying to think. I'm confident, ultimately, it'll allow us to achieve a cycle time, meaning a unit every five seconds or less, off a single line.
Speaker 2 : And we want to incorporate some of these for testing into our existing production lines as well to [inaudible 00:47:49] Cybertruck already.
This is something I've been thinking about for a long time and I've been thinking about this for a long time. It's not a crazy thing. A car every five seconds may sound like it's coming out like bullets, but actually it's coming out at walking speed. Speaker 2 : It's like a meter a second.
A meter a second. We're still far away from caring about the aerodynamic drag of the manufacturing line because you're still at three miles an hour type of thing. Every five seconds sounds crazy, but it's three miles an hour we're talking about. Yeah, you can run away from it basically. But that's still by far the fastest line on earth. It's half… I don't know. What's the [inaudible 00:48:51] line? I don't know if it's about- Speaker 2 : Shanghai phase two.
And that's us? Speaker 2 : 33 seconds. Yes.
We're the fastest, right? Speaker 2 : I would think so.
We think we're the fastest at 33 seconds in our Shanghai factory, but this would be six times faster or seven times faster, more advanced. It'll be slower than that at first. But the point is that, when you fully optimize the design, an operation of the next generation factory that we're building right now, a five-second cycle time or less is… The design is capable of it. When you go through a radical new architecture, you go from being in A… I'd say Shanghai, in particular, is an A+ on a moderately… An advanced, but still… Speaker 2 : Traditional?
Traditional car production system. They're really doing about as good as possible to do within a conventional scenario. Trying to get much below 30 seconds is extremely difficult, but if not… And you start getting into impossible… Where you have to be faster than a human could possibly move. Then the autonomous line, it really just needs to be robots moving really fast and that's where it gets us up five seconds. But we'll start off with getting a C… Instead of an A getting a C in a new architecture, but then the potential is there over time to move that up to an A+ within an A+ architecture.
Yeah. In fact, there's no more vertically integrated car company than Tesla. We're the most vertically integrated car company since Henry Ford back in the day, when they were mining iron and stuff and growing rubber trees. We're not growing rubber trees and mining iron yet, but we have built a lithium refinery in South Texas and it's, I think, the biggest lithium refinery outside of China, I think. Is that right? Speaker 3 : Yeah. Speaker 4 : I think so. Speaker 3 : [inaudible 00:52:38].
But it's output potential would be the biggest lithium refinery outside… And we've got space to expand it if we need to build more, right? And then we've got the cathode refinery in Austin next to the Gigafactory. We've got to figure out what to do about the anode. That's an ongoing subject of discussion. The best of all possible worlds would be figuring out how to have no anode. Best part being no part. That's the dream of the lithium batteries to the anode. Not have an anode. But either way, we better have the anode, the cathode, and the lithium and the electrolytes and the separator to make a cell. But there's no other car company that is built in lithium refineries and cathode refineries.
We're ridiculously vertically integrated and that's our best position to protect against supply chain disruptions. [inaudible 00:53:45], do you want to talk about progress in the front? Speaker 4 : Yeah. Certainly for our in-house cells, we've multi-sourced every component and we have every part coming from at least two different countries of origin, which is… We started this… The supply chain team and the engineering team worked together on this for the last couple of years to put that together. It's not something we did in a couple of months. This is years of work. We're in a good position to take advantage of that.
And the insourcing of lithium and cathode, they're the two most critical parts of the battery that's run in that backyard. And we've total insulated from…
Oil. It needs to be an operation. Speaker 4 : It needs to be an operation. That's the right way to-
We also make our own cells, by the way. Cell production… If you make the anode, the cathode, the lithium, the electrolyte separator, the can, and then you got to put all that together in the cell factory. And there are entire companies that all they do is produce cells, but they don't do the other stuff. They don't refine lithium or the cathode. Our cell production is going quite well and I think we're… We're currently the lowest cost per kilowatt-hour Speaker 4 : Yeah. For all cells we purchased in North America.
Yeah. Speaker 4 : The lowest cost to us, I think.
We have the lowest cost per kilowatt-hour, all things considered. The Tesla cell is the most competitive sell for a kilowatt-hour put into a car. If it's a Tesla cell, it's lower cost than if it's a supplier sell. Speaker 4 : And the plan this year is to really build off that base. Getting to lowest cost is… It's the hardest challenge for so many batteries. It's relatively easy to build a flashy product that does one thing well, to build something at high volume and low cost is super difficult and we're using that as base to then build off and add performance in different areas for new products coming out.
Yeah.
Yeah, Tesla is not immune to the macro demand for cars. When there is economic uncertainty, people generally want to pause on doing a major capital purchase like a car. But as far as absent macro issues, we don't see any reduction in demand. Speaker 2 : Correct. That's where we continue to focus on affordability and it's a fundamental focus there.
I just want to emphasize, Optimus is still very much a development program. It's not a large volume production. That's why this year we will make a few. We do expect to make thousands of Optimus robots, but most of that production is going be at the end of the year. Almost everything in Optimus is new. There's not an existing supply chain for the motors, gearboxes, electronics, actuators, really almost anything in the Optimus, apart from the Tesla AI computer, which is the same as one in the car. When you have a new complex manufactured product, it'll move as fast as the slowest and least lucky component in the entire thing. And as a first-order approximation, there's 10,000 unique things.
That's why anyone here who tells you they can predict, with precision, the production ramp of a truly new product doesn't know what they're talking about. It is literally impossible. You go through this series of constraints where this part's a limiting factor, now that part's a limiting factor, now this part's a limiting factor, and multiply that by 1000, basically. And then the rate of the production ramp is decided by how quickly you can solve each of those problems. Now, Optimus was affected by the magnet issue from China because the Optimus actuators in the arm do use permanent magnets.
Now, Tesla as a whole does not need to use permanent magnets, but when something is volume constrained, like an arm of the robot, then you want to try to make the motors as small as possible, so we just did the design in permanent magnets for those motors, and those were affected by the supply chain, by basically China requiring an export license to send out anywhere with magnets. We're working through that with China. Hopefully we'll get a license to use the rare earth magnets. China wants some assurances that these are not used for military purposes, which obviously they're not, they're just going into a humanoid robot.
It's not a weapon system, but that is an example of a real challenge there. I'm confident we'll overcome these issues and we'll, by the end of this year, have thousands of Optimus robots.
Yeah.
Yeah. The reality is that, in the future, most people are not going to buy cars. One could say, look, if you want to continue with your phone metaphor, you can remember the days of the flip phones, when there was
A hundred different flip phone designs. And the mistake that manufacturers made was to try to make many different variants of a flip phone, which was a mistake. They should have made the iPhone. So because obviously everyone's going to want a smartphone. But in the beginning when the iPhone came out, I was like, "Wow, I can't believe these guys are not reacting as though this is death." But they didn't, they kept making different variants of flip phones. Nokia, I think at one point was the most valuable company in the world, or close to it, but they kept making flip phones, trying to find another market niche.
Maybe somebody wants a phone of a different style, maybe this different color, or whatever it is. Nope, they just want a super intelligent phone that can do everything. Just one. I said this many years ago. In the future, in the not too distant future, buying a gasoline car that is not autonomous will be like riding a horse while using a flip phone. Some people still do it, but it's rare. Speaker 5 : Great. Thank you. The next question comes from Emmanuel Rosner at Wolfe. Emmanuel, please unmute yourself.
Yeah, go ahead Joshua.
Yeah, really we just work through a long tail of unusual interventions. And these are really very rare. As I was saying earlier, you can have an intervention every 10,000 miles. I mean, that's a lot of driving you've got to do to even find one case within Athens.
Anyway, it's only a couple months away, so you can just see for yourself in a couple months in Austin. Speaker 5 : Great. Our next question comes from Edison at Deutsche Bank. Edison, please unmute yourself.
We have to see how things settle out. I don't know yet. Some things we're doing as we already talked about, which is that we've already taken tremendous steps to localize our supply chain. We're more localized than any other manufacturer, and we have a lot of things underway to increase the localization, to reduce supply chain risk associated with geopolitical uncertainty. Did you have followup, Edison?
Yeah. We're still debating the exact number to start off on day one, but it's like, I don't know, maybe 10 or 20 vehicles of day one. And watch it carefully, then scale it up rapidly after that. So we want to make sure that we're paying very close attention the first time this happens. But yeah, you'll be able to, end of June or July, just go to Austin and order a Tesla Truck, autonomous drive. Speaker 5 : Great. The next question comes from George at Canaccord.
Yeah, I mean we do need to give people more time to… A key break point is can you read your text messages or not?
Can you write a text message or not? Because obviously people are doing this, by the way, with unautonomous drives all the time. And if you just go for a drive down the highway and you'll see people texting while driving, doing 80 mph.
Yeah. Putting on makeup, doing their hair, with the mirror down, and texting, and driving, at 80 mph. This is a common occurrence. So people eating lunch, you name it, shaving. So anyway, but right now the car is very insistent that you pay attention to the road, which reduces the value somewhat, because it's very rigorous about you paying attention to the road. And we will gradually lighten up on that, with every few weeks or every month, we'll relax that a little bit, so you can be more and more able to do things you want to do and not have the car demand your [inaudible 01:15:05] attention.
So that value, it'll really be profound when you can basically do whatever you want, including sleep. And then that $99 is going to seem like the best $99 you've ever spent in your life. Speaker 5 : Great. And George, did you have a follow up?
Actually it does not blind the camera. We use an approach which is direct photon count. So when you see a processed image, so the image that goes from the silicon photon counter, that then goes through a digital signal processor or image signal processor. That's normally what happens. And then the image that you see looks all washed out because if you point the camera at the sun, the post-processing of the photon counting washes things out, it actually adds noise. So quite a big breakthrough that we made some time ago was to go with direct photon counting and bypass the image signal processor.
And then you can drive pretty much straight at the sun, and you can also see in what appears to be the blackest of night. And then here in fog, we can see as well as people can, probably better. But in fact, I'd say probably slightly better than people, than the average person, anyway. And, yeah.
Yeah.
Yeah.
Well, at the risk of stating the obvious, I'm one of many advisors to the president, I'm not the president. But I've made my opinion clear to the president, and other people made their opinion clear to the president. He talks to many people and he makes his decision. And I'm hopeful that the President will observe whether my predictions are more accurate than the predictions of others, and perhaps weigh my advice differently in the future, which we'll see. But I'm an advocate of predictable tariff structures, and generally I'm an advocate for pre-trade and lower tariffs.
Now one does need to take a look at if some country is doing something predatory with tariffs or if a government is providing extreme financial support for a particular industry, then you have to do something to counteract that. But I think that that's on a case by case basis, strategically. But the president is the elected representative of the people and is fully within his rights to do what he would like to do.
Well, I think you know the answer on the drones. I mean, a friend of mine, Naval, posted on X, I reposted it, I think a prophetic statement, which is, "Any country that cannot manufacture its own drones is doomed to be the vassal state of any country that can." And America cannot currently manufacture its own drones. But that's [inaudible 01:24:10] fortunately. So China I believe manufactures about 70% of all drones. And if you look at the total supply chain, almost 100% of drones have a supply chain dependency on China.
So China is in a very strong position, and here in America, and we need to shift more of our people and resources to manufacturing, because this is… And I have a lot of respect for China, because I think China is amazing actually, but the United States should not have such a severe dependency on China for drones and be unable to make them, unless China gives us the parts, which is currently the situation. With respect to humanoid robots, I don't think there's any company in any country that can match Tesla. Tesla and SpaceX are number one. Now I am a little concerned that on the leaderboard, ranks two through 10 will be Chinese companies, but I'm confident that rank one will be Tesla.
Speaker 5 : Great. Well, I think that's unfortunately all the time we have for today. We appreciate all your questions and look forward to talking to you next quarter. Thank you very much and goodbye.
我仍然相信,只要执行出色,Tesla将远远成为世界上最有价值的公司。
I continue to believe that Tesla, with excellent execution, will be the most valuable company in the world by far.
我预测,明年下半年会有数百万辆Tesla实现完全自动驾驶运营。
I predict that there'll be millions of Teslas operating fully autonomously in the second half of next year.
我们希望自动驾驶明确比人工驾驶更安全。所以仅仅一样安全还不够,它必须比人工驾驶的汽车显著更安全。
we want autonomy to be definitively safer than manual driving. So it's not enough that it'd just be as safe, it needs to be meaningfully safer than if it's cars manually driven.
目前我看不到有谁能与Tesla竞争。
I don't see anyone being able to compete with Tesla at present.
工厂本身就是产品,和汽车一样。
The factory is the product as much as the car is the product.
最理想的情况是想办法不用负极。最好的零件就是没有零件。
The best of all possible worlds would be figuring out how to have no anode. Best part being no part.
这里任何告诉你他们能精确预测一款全新产品量产爬坡的人,都不知道自己在说什么。这根本不可能。
anyone here who tells you they can predict, with precision, the production ramp of a truly new product doesn't know what they're talking about. It is literally impossible.
现实是,未来大多数人不会买车。
The reality is that, in the future, most people are not going to buy cars.
未来,在不太遥远的未来,买一辆非自动驾驶的汽油车,就像骑着马用翻盖手机。
In the future, in the not too distant future, buying a gasoline car that is not autonomous will be like riding a horse while using a flip phone.
任何无法自己制造无人机的国家,都注定会成为能制造无人机的国家的附庸国。
"Any country that cannot manufacture its own drones is doomed to be the vassal state of any country that can."
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