十年后,全球经济规模会是现在的 10 倍。
这句话是马斯克在 Abundance Summit 上说的。主持人彼得·戴曼迪斯(Peter Diamandis)愣了一下,反问一遍确认,马斯克点头,还补了一句,说这个预测让他觉得挺舒服。
把这个数字摊开:10 年 10 倍,年化复合增长大约 25.9%。人类历史上还没有哪个大型经济体在十年里跑出过这个速度。美国战后黄金年代的年均增速在 3% 上下,中国最猛的十年也就 10% 出头。马斯克说的是全球,而且是连续十年。
他给了一个前提:只要不爆发世界大战。这个条件他讲得很轻,但它是整段预测的地基。
这场对话在 X 上直播,两个人从长寿研究一路聊到戴森云、Optimus 机器人、通用高收入,最后收在迷你猛犸象宠物上。中间信息密度很高,值得一段一段拆开看。
硬起飞已经开始了,只是还没完全去掉人
戴曼迪斯问的第一个硬问题是:递归自我改进(recursive self-improvement)到底到了没到。
马斯克先确认了定义。如果指的是完全没有人类介入、AI 自己改进自己,那还没到。但人类参与的比例正在快速下降,每一代新模型都由上一代模型参与构建。这件事已经发生很久,只是没走完最后一步。
可能今年年底就到了。我不会等到明年。(马斯克)
戴曼迪斯追问,那到了那个点,是不是硬起飞(hard takeoff)。马斯克的回答没有停顿:我们现在就在硬起飞里面。
他描述了自己的日常感受:晚上睡觉,中间出现一个重大的 AI 突破;早上醒来,又出来一个。戴曼迪斯接了一句,说这一堆让人头晕目眩的进展里,有不少是马斯克自己搞出来的。
这段话的分量在于它换了一套时间单位。过去人们讨论 AGI,习惯用"还有几年"这种刻度。马斯克现在用的是"睡一觉"这个刻度。一个模型的能力在八小时内被另一个模型超越,工程上意味着迭代周期已经短于人的注意力周期。你还没消化完上一个版本,下一个版本已经在跑分榜上了。
这也解释了他后面为什么把预测窗口压缩到十年。当迭代周期缩短到以天为单位,任何超过十年的线性外推都会失真。
Grok 押的赛道是预测未来
谈到自家模型,马斯克说新版 Grok 4.20 表现很好,在某些指标上是最好的。他挑出来的那个指标是预测能力。
它最擅长预测事情。这大概可以算是衡量智能最好的指标。(马斯克)
这个选择值得停一下。预测能力作为智能的度量,逻辑上很扎实:一个系统如果能把接下来会发生什么说得准,说明它内部真的建立了世界的模型,而不是把训练数据里的模式背了下来。压缩即理解,预测即压缩,这条线索从香农那代人就开始了。
他也坦白了短板:编程能力落后。他迟到的原因就是刚开完一个关于编程的全员大会,把需要做的事一件件过了一遍。
我觉得我们大概今年年中能追上来,并且超过竞争对手。(马斯克)
编程这件事对模型公司来说不只是个功能点。代码能力直接决定了模型能不能参与下一代模型的构建,也决定了它在企业市场能不能站住。这块落后,递归自我改进的闭环就少一环。
用太阳的能量给智能标价
整场对话里最能让人"原来如此"的一段,是马斯克用能量给智能定价。
他的推法是这样:智能最终落在算力上,算力最终落在电力上。一个文明能调动多少能量,大致框定了它能产生多少智能。
然后他开始换算。地球文明现在的全部用电量,相对太阳的总输出,落在万亿分之一的量级。就算把地球经济的用电量放大一百万倍,也只到太阳输出的百万分之一。而地球本身能截获的太阳能,大约是太阳总输出的二十亿分之一。
结论很直接:真正的天花板不在行星表面。
如果一个实体用掉相当于全人类文明一百万倍的电,它在想什么、长什么样、会做什么?那会相当壮丽。(马斯克)
这段话解释了 SpaceX 和 xAI 合并的战略意图。算力要靠太空太阳能和太空散热来解决,数据中心的物理位置最终会离开地表。戴曼迪斯问发射时间表和第一年能拿到多少带宽,马斯克笑了笑,说 SpaceX 处在静默期,不能讲可能引发问题的内容。他只说了一句:他等不及那个速度。
顺着这个逻辑,他给了一个不太好接受的推论:人类会很快在这个星球上变成智能的少数,然后变成极少数,最后变成微观意义上的少数。这个进程还不限于地球。
太阳系里的智能,会比地球上的智能大若干个数量级。(马斯克)
把这段话当作科幻修辞会读错。它是马斯克给 SpaceX 定价的方式。如果智能的最终形态在太空,那么发射能力和在轨制造能力就是终极基础设施,火箭公司的估值锚点就不再是卫星发射市场,而是整个文明的算力上限。
为什么十年能涨十倍
戴曼迪斯问了一个方法论问题:你现在能看多远,能做出多少年还算靠谱的预测。
马斯克的回答是 S 曲线。技术进展很少走成一条平滑的指数线,它更像一串互相重叠的 S 曲线。起步慢,进入指数段,撞到瓶颈变成线性,再收敛成对数曲线,直到下一个突破把它重新拉起来。AI 这几年的表现就是这样,每隔一段时间有人宣布撞墙了,然后下一次突破把墙推倒。
他说,以前他能比较有信心地往外看一二十年。现在不行了,所以他把预测压缩到十年这个尺度,然后给出了那个数字。
这听起来可能有点疯狂。十年后经济规模是现在的 10 倍,我对这个预测感觉挺舒服。(马斯克)
戴曼迪斯提到马斯克此前说过,五年之后 GDP 增速会到三位数。两个说法放在一起,指向同一件事:增长曲线正在换挡。
这里有个容易看漏的地方。马斯克说的 10 倍是实际产出,配合他后面讲的那套通缩逻辑,名义数字可能远小于 10 倍。东西造得多了十倍,价格同时往下跌,账面上的 GDP 涨幅会被吃掉一大块。对普通人来说,"经济体量翻十倍"和"你的账户数字翻十倍"是两件完全不同的事。
十年内还能期待什么,他也列了一份清单:月球上会有基地,火星上会有人,月球上会有质量驱动器(mass driver)。质量驱动器是电磁弹射装置,用来把月球上开采的物质低成本地甩进太空。这个设想来自物理学家杰拉德·奥尼尔(Gerard K. O'Neill)的太空城市方案,马斯克说十年内能看到它变成现实。
Optimus 3 今年夏天投产,问题是初期会非常慢
机器人是这场对话里最具体的部分。
马斯克说 Optimus 3 已经进入最终完成阶段,会是世界上最先进的机器人,其他产品和它不在一个水平线上。
说实话,我还没见过任何一个机器人的演示能比得上 Optimus 3。也许有,也许藏在某个地方,但我没见过。(马斯克)
时间表他给得很明确:今年夏天开始生产,初期产量非常低,走的是制造业经典的 S 曲线爬坡。明年夏天达到高产量。明年设计 Optimus 4。他的目标是每年发布一个新的机器人设计,每一代都比上一代更好。
产能布局上有个细节。戴曼迪斯提到特斯拉的超级工厂是 1150 万平方英尺,Optimus 还要再建 950 万平方英尺。马斯克插了一句玩笑,说 9 平方英尺,取整。然后他认真了一点,说新工厂的设计会和现有工厂不一样,因为造的东西不一样。
关于机器人造机器人,马斯克的回答比较克制。特斯拉现在直接雇员大约 15 万人,其中三分之二在工厂里,要么在造东西,要么在管理造东西的人。往上还有供应商体系,大概一两百万人。他说短期内没有裁员计划,反而会增加人数。
但特斯拉的人均产出会变得非常离谱。你会不太敢相信那个数字。(马斯克)
这句话是理解他整套经济预测的钥匙。他的模型里,失业算不上主要矛盾,人均产出才是变量。一个人配合一堆机器人,产出可以是今天的几十倍,那么工资、价格、税收、福利的整个算式都要重写。这也是他为什么对"机器人抢工作"这个叙事不太焦虑,他关注的是单位人力的产出曲线,而不是岗位数量表。
逻辑上有漏洞吗?有。人均产出暴涨不自动意味着分配均匀。产出归谁、怎么分,是政治问题,不是工程问题。马斯克在这一点上承认了不确定性,后面那段"80% 概率是好结局"就是给这个漏洞留的注脚。
通缩会先到,然后是通用高收入
戴曼迪斯把话题引到分配上。两人之前在播客里聊过"通用高收入"(universal high income,UHI),这个概念比全民基本收入(UBI)更进一步:不只是发一笔够活的钱,而是发一笔让生活相当宽裕的钱。
马斯克先把风险讲清楚,这段值得逐字看,因为他很少这么讲。
我不认为我们应该自满。未来是一个可能性的区间,并不是所有分支都是好的。(马斯克)
然后他给出概率:大概 80%,可能更高,会是好结局。
机制部分他讲得很干脆。通用高收入的资金来源就是直接发钱,因为商品和服务的产出会远远超过货币供应量。通缩在他看来就是一个比值:产出除以货币供应。分子涨得比分母快,价格就跌。
如果商品和服务的增长速度远远超过货币供应量的增长速度,你就会得到通缩。我预测这会发生。(马斯克)
他还加了一个加速器:会有大量的人创办新公司,互相竞争,把价格一路压下去。AI 把创业的门槛降到极低,供给端的竞争强度会远超今天。戴曼迪斯补了一句,说这种竞争会让价格下跌发生得越来越快。
这个链条里有个隐含前提值得点出来:通缩要成为好消息,得是技术驱动的成本下降,而不是需求崩塌带来的价格下跌。前者意味着同样的钱能买更多东西,后者意味着没人买东西。马斯克押的是前者,但两者在统计口径上看起来一样,这也是政策制定者未来几年最头疼的技术难题。
人类的欲望会在 1000 倍处饱和
这段推演往下走,会到一个有点荒诞的地方。
马斯克说,AI 和机器人会制造太多东西、提供太多服务,最后会没什么事留给人类做。原因在于人类能想到自己想要什么,这件事本身有上限,跟懒不懒关系不大。
就算你把经济规模做到今天的一千倍,人类能想到的所有想要的东西,大概早就被满足了。(马斯克)
他这个判断有争议空间。人类历史上的欲望一直在被重新定义,有了马车想要汽车,有了汽车想要飞机。但他赌的是欲望扩张的速度会慢于供给扩张的速度,最终收敛。这是整场对话里最不确定的一环,也是他所有乐观结论的落点。
换个角度想,他说的"饱和"可能指的是物质需求。人对地位、认同、创作、探索的渴求很难饱和,而这些恰恰是零和或者接近零和的。所以更可能的未来是:物质极度充裕,稀缺性转移到别的地方去。这一点马斯克没有展开,访谈也没往这个方向追问。
钱会失效,AI 只认瓦特和吨位
戴曼迪斯顺着问:那货币的价值会大幅下降吗,我们会走向后资本主义吗。
我认为货币在未来的某个时点会变得不相关。(马斯克)
他猜那样的社会可能像伊恩·班克斯(Iain M. Banks)《文化》系列里描述的后稀缺文明。再往后,AI 根本不会用人类货币做计价单位。
它会关心功率和质量,瓦特和吨位。(马斯克)
戴曼迪斯开了个玩笑:就在你快要成为万亿富翁的时候,钱开始不值钱了。马斯克说差不多是这样。
然后他解释了自己那个数字的构成。所谓万亿身家,只是他在几家公司的持股比例,那些钱并没有躺在银行账户里。公司做了有用的事,估值上涨,他持有的百分比跟着变成一个看起来很高的数字。他还顺口说了一句,如果传言说他是世界首富,其实主权国家比他有米。
这段话听起来像谦虚,其实是个财务常识的提醒:股权财富是账面估值,能不能兑现取决于有没有人愿意在那个价格接盘。当整个经济体的产出增速远超货币供应,估值体系本身会重构,今天的数字到了那天未必还能对齐。
奇点这个词不是修辞
戴曼迪斯抛出了全场最重的问题:民主和现代制度能不能跟上这场超音速海啸,还是会直接被冲垮。
它叫奇点是有原因的。进入奇点之后会发生什么,很难知道。(马斯克)
他顺手解释了一下 Grok 的标志:黑洞周围的引力透镜光环,物质和光正在往里掉。掉进去之后是什么,物理学也给不出答案。
很难知道奇点里面会发生什么。但会非常有意思。未来会很有娱乐性,这一点我有信心。(马斯克)
戴曼迪斯问的是制度能不能撑住,马斯克的回答绕开了这个问题本身。他没讲监管、没讲选举周期、没讲官僚体系的反应速度,而是把话题带到了"未来很有趣"。这种绕开本身也算一种回答:在他的框架里,制度属于奇点之前的变量,奇点之后无所谓跟不跟得上。
关于制度,他给了一个具体的乐观理由:AI 和机器人会解决财政赤字,国家不至于破产。
这个判断的分量不小。大部分关于 AI 和就业的讨论都停在"多少人会失业",马斯克直接跳到了财政层面。生产率暴涨会做大税基,同时压低单位产出的成本,债务占 GDP 的比例会自己缩下去。这个推演成立的条件是税制能跟上新的产出结构,而这一步在现实中通常滞后很多年。
每个人都能获得比今天首富更好的医疗
戴曼迪斯问的是希望,马斯克给的答案落在了医疗上。
他的推理是:如果人形机器人有极高的灵巧度,同时又有极高的智能,那么地球上每个人都能获得比今天最富有的人更好的医疗。顶尖医生的知识和手感可以被复制到千万台机器上,边际成本趋近于零。
讲到这里他插了一段私事。他做过三次颈部手术,前两次做错了,背部到现在还疼。
我就在想,能不能请 AI 解决一下背痛。那会是一个巨大的胜利。(马斯克)
他还给了一个观察:人老了脾气变差,很大一部分原因是背痛。背一直疼就睡不好,睡不好就会暴躁。这个说法听着像玩笑,疼痛和情绪的关系在临床上是有据可查的,慢性疼痛患者里抑郁和焦虑的发生率远高于一般人群。
戴曼迪斯提到当天上午 David Sinclair 在台上讲他的部分表观遗传重编程药物 ER-100 要进人体试验,最近发表的论文显示它能修复关节。如果这条路走通,背痛可能在解决清单上。马斯克回了一句,那会很棒。
这段对话里有个反差值得留意。马斯克整场都在谈百万倍、十亿分之一、太阳系级别的智能,落到自己身上,最想解决的是一节疼了多年的颈腰椎。技术叙事的宏大程度和个人痛感的琐碎程度,在这几分钟里撞到了一起。
他承认自己变乐观了
对话快结束时,马斯克说了一段不太像他平时风格的话。
你对我有影响。我决定变得更乐观一点。以前我不算乐观主义者,可能太多时间花在负面的事情上了。做乐观主义者全是好处。(马斯克)
他马上又补了平衡:不能自满,不能假设一切都会自动变好,但要去努力让它变好。他把自己定位成乐观的现实主义者。
这段话放在整场对话的末尾,作用是把前面那些宏大的数字拉回到人身上。十年十倍、戴森云、太阳系智能,这些话题容易让人产生距离感。他最后落在背痛、颈部手术、老人脾气这些具体的痛点上,反而让那套推演有了落脚处。
访谈的收尾是迷你猛犸象。戴曼迪斯提到 Colossal 的 Ben Lamm 正在做去灭绝,马斯克说养一只迷你猛犸当宠物会很酷,还顺便请求有人把侏罗纪公园做成真的,说就算有生命危险他也愿意去。
整场对话从"十年十倍"开始,最后收在一只虚构的小象身上。这个落点不算突兀,它其实是马斯克那套乐观主义的最简版本:技术会把不可能的事变成可以养在家里的东西。
把这场对话翻译成可以做的事
马斯克这套预测对不对,十年后才知道。但里面有几个人物判断,现在就能拿来指导决策。
- 用产出口径看公司,别用岗位口径。他说特斯拉不裁员、还要加人,同时人均产出会高到离谱。未来评估一家公司,员工人数会越来越不重要,每人的实际产出才是关键。你所在的公司如果人均产出常年不动,风险比裁员消息更大。
- 把自己挪到产出曲线陡峭的那一侧。能被 AI 放大十倍产出的岗位,和会被 AI 直接替代的岗位,区别往往在于工作里有多少判断、多少重复。现在就可以盘点自己一周的工作,把重复部分交给工具,把判断部分做深。
- 用通缩视角理解价格战。如果你所在的行业正在打价格战,别只当成坏事。马斯克的框架里,供给端竞争加剧加上产能暴涨,通缩是结构性的。要么把成本做到行业最低,要么往价格不敏感的地方走。
- 把身体当作最重要的长期资产,尤其是脊柱。他三次颈部手术、至今背痛的细节背后是个具体问题:腰椎颈椎的毛病在四五十岁集中爆发,而那正是决策密度最高的阶段。现在就处理体态、久坐和睡眠,回报率比多数投资都高。
- 评估 AI 模型时,看它预测得准不准。马斯克把预测能力当作智能的核心指标,这个方法普通人也能用。让模型预测一件你熟悉的、下周会发生的事,连续记录一个月,比看任何跑分榜单都实在。
- 对十年做一次 10 倍思考实验。假设你所在行业的规模十年后是今天的 10 倍,什么会变成瓶颈,什么会变得稀缺,你现在的位置在瓶颈的哪一侧。这个练习不需要预测准确,它的价值在于逼你重新排优先级。
- 保持乐观,但留冗余。他自己给的概率是 80% 好结局,剩下 20% 不是零。乐观地做长期规划,同时在现金、健康、技能上留出缓冲,这两件事并不冲突。
- 补一点能量和算力的常识。马斯克整场论证的底座是瓦特和吨位。搞懂一个数据中心要多少电、一块 GPU 的热设计功耗是多少,你对 AI 行业新闻的判断力会上一个台阶,也不容易被"算力无限"这类说法带偏。
- 盯住他的时间点,用来校准自己的预期。今年夏天 Optimus 3 投产,明年夏天高产量,年中编程能力追平,年底递归自我改进可能完成。这些是可验证的承诺,到期回头看兑现率,比听任何宏大叙事都有用。
0:00audience and as you can see still trying to monetize hope. Yeah, yeah, you look like you're in great shape. I'm doing great. Last time I saw you
0:10sort of youth serum things going on or what? It's it's our longevity X prize. We're we're getting there, buddy. We're getting there. I think I think in our last conversation you're getting on board with the idea of extended longevity, yes? Yeah. Okay. [laughter] I'll leave it at that. Some degree. I mean like I don't know if we want everyone to live forever or whatever, but I think health span and and not you know, having an extended period [clears throat] of senescence where where you're just drooling on yourself sounds like a good idea. We we want to avoid that. Yeah. So, first off, congratulations on the merger of SpaceX and xAI. Baller move. going to power
0:52humanity's first Dyson swarm? So, I'm curious it truly truly is. What's your timeline for launching these data centers and how much bandwidth do you think you can get in the first year? Give us a sense of the speed at which you're going to be making this happen. yeah, so SpaceX is has filed SpaceX is in a quiet period. I I can't actually tell you things that would cause a problems. yeah. [laughter]
1:29I'll I'll leave it I'll leave it at that. I appreciate that. But I can't wait for for the speed. You know, we had a conversation here on Monday with with Eric Schmidt and with one of the leads from one of the other hyper scalers. I won't mention who, but I'm curious where you feel we are in recursive self-improvement. Are we there? Do you see Grok doing recursive self-improvement at this point? And how and what's the timeline for AGI and ASI?
1:59Give us a sense of that. Yeah, I I I think we're we've been on recursive improvement for a while here. you say if if it's really if you mean by like recursive self-improvement without a human in the loop. Is that what you mean? I do. I I on the on the AI software side. I mean humans are gradually getting less and less in the loop on the recursive self-improvement. So, you know, every successive model is is built by the one before it.
2:39So, that that that that is happening to a large degree, but it's it's not yet fully automated. it may be there end of this year, but I'm not going to wait until the next year. Do you see a hard takeoff at that point? We're in the hard takeoff. Okay. Right now. Yes. I mean, look at I mean at this point I go to sleep, there's some massive AI breakthrough, and when I wake up there's another one.
3:17Yes. Yeah, it's hard to keep track, honestly. It's a It's a bit of a head spinner. Yeah, well, I think a lot of that head spinning is happening from you, too. Yeah. Well, you know, Grok is doing pretty well, and in some metrics by some metrics it's the best. For example, it's the best at predicting things, which I you know, it's arguably the the best metric for intelligence.
3:44the new Grok 4.20 is it's really really good. we're we're currently behind on coding. The reason I was put put late for this was that I was just in a giant sort of all hands sort of on coding, just going through all of the things that need to happen to essentially catch up and exceed our competitors on coding.
4:12which I think I think we'll do. I feel Yeah, we'll we should probably get there by the middle of this year. And And then I I think people don't don't quite realize just how much intelligence there will be or you know, just how far it will exceed human intelligence. To a degree that is impossible to fully understand.
4:35but you can certainly imagine a situation where we Let's say if let's say a million times more energy is harnessed than all of Earth's current electricity usage. That would still only be roughly a millionth of the sun's energy output. So, essentially if you increase Earth's economy by a factor of a million, it's still roughly a trillion So, since we're a trillionth of the sun's energy, if you increase Earth's economy in in in terms of electricity usage by roughly a million, you'll be roughly 1 millionth only of the sun's energy harnessed.
5:17But what is it What is What is an economy or an intelligence using a million times more electricity than all of Earth's civilization think about or look like or do? It's going to be something pretty magnificent. The challenge will be even vaguely appreciating that level of intelligence.
5:37but it's it's safe to say it will it will solve everything you can possibly think of. Yes.
5:42longevity being, you know, You're certainly one of them. And I I I do enjoy your unrelenting optimism. Thank you, pal. I see you I see you if you Hope. Hope. Yeah, exactly. You've taken to heart monetizing hope, which is pretty funny. Yeah, it was it was It was Grok's It was Grok's marketing advice to me when you roasted me on [laughter]
6:08Right. Okay. Grok was seeing you and saying you monetize hope. But hey, if you you're going to manage to further monetizing misery
6:15Yes, but yes. For sure. Yeah. But but yeah, just the the when you when you have AI AI and robots are going to increase the the like economic output or or by by so many orders of magnitude we we we cannot possibly comprehend it. We're likely in the very short time to become a minority, then a vast minority, then a microscopic minority of intelligence on this planet.
6:51Yes, not even on this planet, but in the solar system.
6:54Yes, for sure. Because Like like if if you know, your best case outcome for Earth for intelligence is roughly 1 billionth of the sun's energy. that's your best case outcome. if you if you if you generate intelligence only on Earth. Intercept it, right? Yes. Yes, cuz roughly one roughly half a billionth of the sun's energy hits Earth, and that's the vast majority of energy that that's out there that that we can access.
7:26so really the intelligence in the solar system will be many orders of magnitude greater than the intelligence on Earth itself. How May I ask you a question, Elon? how far out can you see? How many years out can you make reasonable predictions right now? It's hard to predict the the path exactly, especially if it cuz often things are kind of an S curve or a series of S curves where it starts off slow grows exponentially hits linear zone and then goes logarithmic.
8:10that generally has been what what I've seen with the breakthroughs in in AI AI for example is you'll you'll there'll be some breakthrough it'll do have an S curve but and then looks like it's just going to go to infinity but then you hit logarithmic returns until there's another breakthrough. Yeah.
8:26so progress in AI is just a sort of series of you know sort of overlapping S curves or connected S curves. I mean there was a point where you could probably predict out a decade or two decades. What are your thoughts now?
8:44Yeah. Okay, this is going to sound pretty crazy. That's okay. We we've been we've been talking about crazy all night.
9:02there's a separate acceptable audience to wild prognostications. Yes. [snorts]
9:14I'd say the economy is 10 times this its current size in 10 years. Greater than. Okay. Yeah, you
9:25really saying something. Yeah, you said triple digit growth in in 5 plus years from now on on GDP. And 10x the economy. But in terms of your ability to
9:37Like I I I feel like that's that's that's a 10x in roughly 10 years I feel is a I I actually feel like comfortable prediction with There's obviously if there's like World War or something that that could put a kink in those plans or those expectations. Yeah. but in the absence of World War current trends continue I would say the the economy 10x in 10 years.
10:09Love it. Can you give us an We had a bunch of robots
10:12have a base on the moon. Yes. And we'll And we'll have
10:16have people on Mars. And we'll have mass drivers on the moon. Yes. I think so. In 10 years, I think I think we'll have a mass driver on the moon in 10 years. I love it. Gerard K. O'Neill's vision being fulfilled. we had four robots on stage here this year on the at the Abundance Summit. I I look forward to Optimus. I'm curious, Optimus 3 timeline and in [clears throat] particular when can I buy one or two?
10:46When's when do you expect Optimus to go into commercial for commercial sale? Or will you be leasing it? Well, we're in the final stages of completion of Optimus 3, which is really going to be by far the most advanced robot in the world. Nothing's even close. Yeah. In fact, I haven't even seen any seen any demos of robots that are as good as Optimus 3, frankly. Maybe they're out there or they're secret or something, I don't know, but
11:23You know, I I I have to make sure I'm saying things that are reasonably public as well, of course. Of course. We're streaming this on X. Yeah. Okay, this is pretty public knowledge, I guess. Yes. [laughter]
11:36Yeah, I I I think we'll start production on Optimus 3 this summer. but but very slow at first. Yeah. like it you know the sort of classic S curve ramp of manufacturing units versus time. and then probably reach high volume production around summer next year. and then you know, we'll we'll have Optimus 4 you know, the design likely next year. I got to try to release a new robot design every year and I'm an improved robot design every year. When when Dave Blunden and I were at the Gigafactory, it was extraordinary experience, 11 and a half million square feet for the Tesla. And then I think you said you're building out 9 and a half million square feet for
12:27Optimus there as well, which is which is extraordinary. let's let's
12:349 square feet [clears throat] round numbers. Yeah. Yeah. Yeah. I had to That'll be that'll be that'll be quite That that's going to be a a a a a new factory design, too. Like it's still not different from other factories. How far before we have robots building robots? I mean, you automated so much of the Gigafactory already where there are humans playing a small role. Do the robots just play the role that humans are playing in that regard?
13:04we still have a lot of humans building things. Mhm. you know, Tesla direct employees are building things or like basically people in the factory are either building or managing people who are building. roughly 100,000. so we have a lot of people. Is it To- Tesla total head count is around 150k.
13:25Of of which 2/3 are, you know, in the factory in one form or another. And then our suppliers there's probably maybe a a million or two million people in our suppliers type of thing. So, it's it's a lot of people. What what we do expect that is that is that the output per per person at Tesla becomes very, very high. Yeah.
13:50So, we're not planning any like layoffs or reductions in personnel. In fact, we will increase our head count. But, the output per human at Tesla is going to get nutty high. When we were to Like you can't really believe it, you know. When we were together we discussed the idea of sustainable abundance on our podcast and you reinforced the idea that we have a coming age of universal high income which has become a point of discussion beyond UBI. But, I'm just wondering if you have any thoughts on how we get there. Have you reflected on that any further?
14:28And and more so, you know, we talked about a time frame of civil unrest, you know, two, three, four, five years probably a lot of COVID-like checks in the interim until we get to a demonetization and a you know, deflation that leads us to UHI. Any more reflections on that? Cuz that's a really People need that hope and that vision. Yeah.
14:55I mean, to be clear, I don't I don't I don't think we should be sort of complacent. We need to We do need to be careful because the future is a range of possible outcomes and they're not all great. But, I I at this point I I I I do agree with you that it's it's likely to be great.
15:11you know, it's probably 80% likely, maybe more likely to be great. And and I and I do think we'll have universal high income. We're we're We're just issuing money to people, you know, and and they are really just because the output of goods and services will far so far exceed the money supply that you know, that that that effectively have deflation cuz just deflation is just the ratio of the outputs of goods and services to the money supply.
15:42so, that that's so if you if you if the rate of growth of of goods and services far exceeds the rate of growth of the money supply, which I predict will happen, then you'll have deflation. Yes. And a lot of a lot of people spinning up new companies, competing against each other, driving the price down, and increasing the variability and deflation.
16:05Faster and faster. Yeah, it's basically AI and robots are going to make so much stuff and provide so many services that they'll actually run out of things to do for the humans. They'll just run out of things to do for the humans. And then they they'll you know, there's there's only so much that humans can even express that they want.
16:27So, you go back to my example of like if you go a million times greater than the US economy, you you've long since saturated all human desire. you know, like maybe like even if you go a thousand times more than our current economy, thousand times, you you you've probably already saturated saturated the human anything people can think of that they want.
16:53So, do do you think the the value of money is going to significantly decrease? Will it Will we go post-capitalist? Yeah, I think money will stop being relevant at some point in the future. So, just as you're becoming a
17:10it's it's it's it's probably something like an in-banks culture sort of future. And I I think the AI down the road will really not use human currency. It will just care about power and mass, wattage and tonnage. Yeah. It's kind of ironic then, right? Just as you're becoming a multi-trillionaire, money starts to have less value.
17:41Yeah. Pretty much. yeah. You know, this this all this stuff is just really just trillionaire sort of represents like some percentage ownership in companies that I have
17:53Yeah. you know, built and it's not like sitting in a bank account, you know, it's just it's just literally I own a percentage of the companies. The companies are doing lots of useful things. The value of the company grows. I own a percentage of the companies and that's the sums up to that number which seems high.
18:09Yeah. You know, it's I I I was interviewed by somebody who was asking me about your your drive, what drives you. And I said Elon's driven to solve problems. He's driven to make life in the world better by just solving the biggest problems over and over and over again. And if someone else were solving them, he wouldn't need to. But no one else is solving them. So, I just want to say,
18:36Yeah. thank you for that, pal. Thank you for that. you're welcome. [applause] I [applause]
18:46I I am curious. Do you think that democracy and our modern institutions can keep up with a supersonic tsunami coming our way? Are they just going to fall in its way? They're just going to break down. How do we deal? I mean, it's called the singularity for a reason, you know, which [laughter] is that it's hard to predict what happens in the in the singularity. I mean, Grok's logo is the singularity.
19:19I love it. It's a beautiful logo behind you, by the way. It's gorgeous. Yeah, thank you. yeah. It It's sort of the light The light The The The halo around a a black hole is that the mass and light are falling in type of thing. It's hard It's hard to know what happens inside the singularity.
19:39but it's going to be very interesting. I We're going to live in The future will be very entertaining. Of that I'm confident.
19:45Yes. And I I think also like AI and robotics also I mean we're going to solve our budget deficit, frankly, and and not just go bankrupt as a country. So I I'm I'm You know, you've had an influence on me in that I'm like just I've just decided to be more optimistic. It's like we just should be more optimistic. Thank you, pal.
20:11I wasn't an optimist, but I was like maybe dwelling a little too much on the negative stuff. It's all It's all upside being an optimist. And a realist a little bit. Yeah. Yeah, yeah, exactly. You don't want to be complacent or just assume everything's going to go well, but try to make it go well.
20:31But I mean, there will be some pretty amazing things that happen. So, if you've got humanoid robots that are that have very high dexterity and and and and are incredibly smart, it means that everyone on Earth will have access to better medical care than than than the richest person on Earth, which by the way, I would say like, you know, if I'm allegedly the richest person in the world I think actually sovereigns are richer than me, by the way. But But it's like I like you know, I like I had to have like an like a neck surgery three times cuz the first two ones were done wrong.
21:06You know, like this is I'm like what the you know. [laughter]
21:11So And and I'm like my back still hurts a little bit. I'm like can AI please solve back pain? That would be a huge win. And I think it will. Yep. So you know, back pain sucks. I think that's maybe like a you know, someone asked why do people get grumpy when they get older? It's cuz back pain. You know, it's like if your back hurts all the time you can't sleep well, you're going to be grumpy. We had but You you We had David Sinclair on stage this morning and he's going into human trials with ERW 100 his partial epigenetic reprogramming. And one of the one of the papers recently published shows it enables joint repair.
21:50And so back pain may be one of the things that it eliminates. So That'll be amazing. Yeah, for sure. For sure.
21:58Every honestly average happiness level for humans would just go up so tremendously if you just solve back pain because it's not a question of of if you will get back pain, it's when you know, you'll get back pain. I I I keep on invite I keep on inviting you to come down to Fountain Life in in in Dallas. We'll we'll help you out, but sometime when you have time.
22:18you have like what do you I understand like you can get like MRI and CAT scans and everything, but like what do you do with that? You know, it's like I'm happy to happy to send you the list. I'll I'll DM you the list of therapeutics.
22:30some or something. Yeah, exactly. [laughter]
22:35you know, listen, you've been so generous. Next up on stage with me is one of another great Moonshot entrepreneur Ben Lam who runs Colossal the de-extinction company. you know, the woolly mammoth and 15 other species. I heard I heard you say you might want a mini woolly mammoth. Is that Is that true?
22:54Yeah, I think it'd be really cool to have a pet miniature woolly mammoth. That'd be pretty epic. Okay. I'll I'll put a word in with you for you with with with Ben. [laughter]
23:04They're very adorable. Little things just running around trumpeting away and it's like look at this this [laughter]
23:10would be a great little pet. Amazing. And also I can somebody please do Jurassic Park in real life? I'd definitely go out even if there was some risk of death. That'd be super cool. I think if anybody's going to do that it's it's Ben Lam in Colossal. He's re- he's engineering living life products.
23:29Someone asked him recently if he can make a Pikachu and he said probably. I think Yeah. Well, Jurassic Jurassic World or whatever that would be great. All right, I'll ask him. Elon, so grateful for you coming and joining us and sharing. Thank you my Thank you my friend. Let's give it up for Elon Musk.
23:54I have the power. Nothing's going to stop us. [music] You have the power.
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