凯蒂·米勒问马斯克,如果重来一次,你还会去搞那个叫 DOGE 的部门吗。
马斯克的回答是:不会。他说他会把时间放回自己的公司,那样“车就不会被烧了”。
这话说得挺重。2025 年那阵子,特斯拉门店和车辆被纵火、被砸的新闻不少,他把这笔账直接算在了自己推动政府瘦身的头上。
一个网友起的名字,最后成了联邦部门
DOGE 这个名字本身就带着互联网的随意。马斯克原本想叫它“政府效率委员会”(Government Efficiency Commission),后来有人在网上提议,不如叫“政府效率部”(Department of Government Efficiency),缩写正好是 Doge,那只柴犬表情包。
这名字听起来不错。于是我们就凭空造了一个部门。(马斯克)
他给自己打的分数是“小有成效”。
他说 DOGE 停掉了很多拨款,那些钱连“有争议”都算不上。项目结束十年了,承包商还在收钱;办公室早就关了,租金还在付。你看了会愣一下。
僵尸付款,和让它停下的代价
他举的例子是一类叫“僵尸付款”(zombie payments)的东西。每年大约一千亿到两千亿美元的钱从财政系统里流出去,没有对应的项目,没有验收,也没有人追问。
解决办法简单到有点荒谬:强制每一笔付款必须有一个付款代码和一句用途说明。填不出来,钱就发不出去。
马斯克说,这类钱大概占政府支出的 2% 到 3%。比例不高,但极其难停。
很少有人会主动要求政府别再给自己打钱。(马斯克)
这句话解释了为什么“削减浪费”在竞选时是全民共识,在执行时会变成战争。
钱一旦流动起来,它就长出了受益人、中介、游说团体和地方经济。你把闸门拧上,反扑的力度跟省下来的金额完全不成比例。
你如果停掉流向政治腐败的钱,他们会猛烈反扑。他们真的希望钱继续流。所以你一停,反应就会非常强。(马斯克)
他说自己一开始就没抱多少幻想。结论也很简单:政府做的事情越少越好,尽可能少。
他反复绕回的那个担心,落在人口结构上
访谈里绕不开移民。马斯克的判断很直接:美国在做一件他称为“选民进口”的事。
他的逻辑链条是这样的。如果政府提供一张巨大的支票,说任何人从任何地方来,都能拿到钱和一堆免费的东西,那就会有大量的人接受这个报价。人来了,走快速通道送进公民身份,靠政府转移支付生活,然后投票。
你不需要建边境墙。如果你是用飞机把人运进来的,那墙就没意义了。(马斯克)
他举了两个例子,明尼苏达州的伊尔汗·奥马尔(Ilhan Omar),和纽约市长马姆达尼(Mamdani)。他的说法是,这些人能当选,靠的主要不是在美国出生的人。
这套判断属于他的政治立场,访谈里他没给出数据。但这段话值得记下来,因为它是理解他后来一系列政治动作的底层动机:财政浪费只是表层,让他真正紧张的是人口构成和投票结构的变化。
用他自己的说法,他不想看着美国变成“共产主义地狱”。
预测和愿望,是两件必须分开的事
话题转到 AI,马斯克先做了一个区分,这个区分贯穿了他后面所有的表述。
我要把“我希望发生什么”和“我预测会发生什么”分开。人们经常搞混,以为我预测的就是我想要的。(马斯克)
他的预测是:如果 AI 和机器人的发展曲线延续下去,它们最终能做人类想让它们做的一切。到那时候,工作变成可选项,人可以用自己的时间做任何事。
他的愿望是:如果他能,他会让 AI 和机器人慢下来。
但我做不到。它前进的速度非常快,不管我喜不喜欢。(马斯克)
这个位置挺尴尬。他一边是这场竞赛里跑得最快的几个人之一,一边公开说自己希望它慢一点。你可以说这是虚伪,也可以说他只是把自己看到的东西讲出来了。他反复强调的一点是,他预测的那个未来,他并没有投过票。
六小时睡眠,和真正的“心灵杀手”
米勒问他睡几个小时。答案是六个。
这个数字不是他估的。他用手机 App 测过自己的自然睡眠时长:5 小时 56 分钟。
他试过睡得更少。结果是在清醒时间里多出了两小时,但白天的认知能力下降,得不偿失。
这一点值得所有熬夜的人抄下来。睡眠不足不会让你多出两小时,它会把剩下的十几个小时整体降级。
然后是噩梦。
你要问我为什么会在噩梦里醒来?AI。我连续很多天做 AI 噩梦。你让我怎么办呢。(马斯克)
他说自己不是个容易害怕的人,但他承认自己“强烈地感受恐惧”。处理办法是,一旦发现某个恐惧是非理性的,就把它掐掉。
“恐惧是心灵杀手”(Fear is the mind killer)是《沙丘》里的台词,他显然很熟。
紧接着他自己改了口。作为同时管着好几家公司的人,他每天要处理的信息量是巨量的。他的做法叫“信息分流”(information triage),并且尽量把一天切成若干段,减少在不同任务之间跳来跳去。
人们说恐惧是心灵杀手。要我说,上下文切换才是。(马斯克)
这个类比可以量化。每三秒切换一次任务,认知惩罚极高;每三分钟切换一次,仍然很高;把同一类事情攒到一段完整时间里处理,开销就降下来了。
他这里的麻烦在于,特斯拉和 SpaceX 内部的议题本来就杂,再加一个 X,随时有突发新闻、有人被绑架、某个国家出乱子,都要他去发声。他自己承认,这种日子很难完全不切换。
特朗普是个天生的喜剧演员
提到特朗普,马斯克先笑了。他说总统“非常有趣”,是那种天生的、不太费劲的幽默感。
他讲了一个场景。纽约市长马姆达尼到白宫,有记者问总统,你还认为他是法西斯吗。总统的回答是:
你就说“是”,这样说省事。别纠结,就说“是”。(马斯克转述特朗普)
马斯克讲这段的时候心情很好。你可以从中看到他对政治的一种态度:很多冲突本身是表演,识破表演的人反而轻松。
Starship 被严重低估了
问到他人生里最好的时刻,他先报了孩子出生、SpaceX 第一次入轨、特斯拉把电动车做成。然后话题停在了 Starship 上。
他的判断是,Starship 的革命性被外界严重低估了。
这是人类第一次有一个火箭设计,让“完全可复用”成为可能结果集合里的一个。(马斯克)
这句话要拆开看。以前的火箭,不管设计得多好,复用根本不在可能结果里。你可以让一级火箭回收,但整个系统不是按“复用”这个目标长出来的。Starship 是第一套从图纸阶段就把“完整回收再飞”当成正常工况的设计。
被问到为什么不直接让 V2 复用,他说 V3 的性能提升太多,V2 到 V3 之间有超过一万处改动,值得跳过去。
一万处。这个数字比任何形容词都有说服力。
他给了一个很有意思的评价框架。他说,如果未来还有历史学家,他们会把 Starship 放进“生命演化名人堂”的前十名。
这个榜单大致是这样:
- 单细胞生命出现
- 多细胞生命出现
- 线粒体被吞进细胞,成了内置的发电厂
- 生命分化成植物和动物
- 生命从海洋爬上陆地
- 生命成为多行星物种
他说,能放进这个前十名的事情本来就不多。你几乎可以用这个尺度去评价任何一个文明、任何一种生命形式。成为多行星物种,是这十件事之一。
他补了一个限定条件:必须是可持续的多行星,不是去一趟就回来。行星冗余的意思是,一颗星球上出事了,另一颗还能活。
他还顺口说了一句:Starship 没有用 AI 来设计,AI 应该会对此表示欣赏。
火星的推销词:你会不舒服,可能会死
很多人有一个误解,觉得去火星是逃离地球,是亿万富翁的避难所。马斯克说完全不是。
火星会非常危险,月球基地也一样。比地球危险得多,也难受得多。(马斯克)
早期去火星建基地的人,死亡率比留在地球高得多。空间狭小,不舒服,食物不如地球,工作强度极大,而且有可能整体失败。
这就是我的推销词:你会不舒服,吃的不好,可能会死,工作量大到离谱,而且可能不成功。你要去吗?(马斯克)
这个推销词的作用是筛选。愿意在这种条件下上车的人,才是真想去的人。
他拿詹姆斯敦(Jamestown)做类比。17 世纪初那批英国殖民者到北美,死亡率高得惊人,有一段时间被叫作“饥荒时期”。马斯克说,如果当年有社交媒体,他们大概会天天发“我们正在死”的视频,后面的船就不会来了。
他顺带提了一句,早期北美殖民地里还有一支队伍整个消失了,至今不知道发生了什么。
那个殖民地的结局无关紧要,有用的是这个类比:任何开拓性的事,早期都是烂摊子。
米勒问他,那要不要二选一:保证地球好好的,还是保证火星殖民地成功。
马斯克说这是个假二选一。
地球比火星好得多,这一点要讲清楚。但如果我们想成为多行星物种,火星是唯一的选择。(马斯克)
他引了齐奥尔科夫斯基(Tsiolkovsky)那句话:地球是人类的摇篮,但人不能永远待在摇篮里。
这句话被引用得太多了,但放在这个语境里,它其实是在回答“为什么要花这么多钱去做一件大概率失败的事”。
时尚停在了 2015 年
访谈里有一段完全跑题的内容,但特别有意思。
马斯克的儿子 Saxon 问过他一个问题:为什么所有东西看起来都像 2015 年。
他愣了一下,然后承认这话说得对。
你拿一张 2015 年的照片,说这是 2025 年,没人能分辨出来。我们在风格上十年没有动过。(马斯克)
他的观察是,60 年代有 60 年代的样子,70 年代有 70 年代的样子,80 年代、90 年代各有各的鲜明特征。到了 2000 年代和 2010 年代,风格的变化越来越少。
他甚至说,严格讲,从 2000 年到现在 25 年,男装基本没变过。你把 2000 年的一群人和 2025 年的一群人放在一起,问别人哪张是哪年,很难答对。
他喜欢看老画里的内阁部长,说那些人的外套比现在好看,高领子、领巾之类的东西,看着就有劲。现在大家穿的都是一套很普通的西装。
我觉得我们该在风格上演化一下。稍微加点料。(马斯克)
外星人、月球高尔夫与一座新建的城市
米勒问他,还有哪些阴谋论没成真。
他的回答是:这个清单已经快空了。“据我所知,它们最后都成真了。”
然后他主动说起外星人,理由是被问得太多了。
我没有看到任何外星人的证据。SpaceX 高管团队里也没有一个人看到过。因为我问过他们:伙计们,我是不是漏掉了什么?(马斯克)
他说,如果有人真看到什么,多半也只是不明飞行物。这个词本来就是字面意思:天上有个东西,你不知道是什么。
它可能是某个国家的新武器项目,可能是高超音速导弹的原型。技术上确实算“未识别飞行物”,但离外星人还差得远。
然后他抛出一个冷笑话。宇航员尼尔·阿姆斯特朗(Neil Armstrong)的名字,把中间名缩写 A 放在姓前面,拼出来正好是 ALIEN。他说,纯属巧合。
聊到登月,他补了一个细节。阿波罗计划的宇航员不只在月面上走了走,他们还打了高尔夫。马斯克说这话时的语气是:我们不但去了,还去了好几次,然后待得有点无聊,就开始在月球上挥杆了。
有人问他月球上的物理。他的回答干脆:有重力,是地球的六分之一。有重力你才站得住,没有重力人就飘走了。月球没有大气,但有这六分之一的重力,所以球被击出后的轨迹跟地球上完全不同。
名字这件事,马斯克是认真考虑过的。他手上有一串以 Star 开头的资产:Starlink、Starship。那 Starship 从哪里出发呢。答案定成了 Starbase,星舰基地。
这片地方原来是得克萨斯州南部、格兰德河边上一块沙洲,海拔只有三英尺左右,涨潮时几乎要淹。他们在那儿建了一座巨型火箭工厂,竖了两座发射塔,就在河的视线范围内。
马斯克说自己当时想,这地方大概是地球上最酷的一块地。那它就得有个配得上它的名字。
更少见的是,Starbase 最后成了一个法律意义上的城市,完成了注册。
你不会经常听说有新城市被建起来。(马斯克)
他拿历史上的公司城做对比。福特当年搞过类似的东西,迪士尼更彻底,那地方的名字直接叫迪士尼世界(Disney World),从“乐园”变成了“世界”,还成功注册成城市并且拿到了税收豁免,当年是个大事。
马斯克说他去过迪士尼世界大概十次,可能更多。推着他去的理由跟过山车无关,肯尼迪航天中心就在旁边。带孩子去看发射,孩子看完火箭就想去看哈利波特园区,行程就这么定了。
问到最喜欢哪个项目,他犹豫了一下,说了太空山(Space Mountain)。紧接着补了一句批评:太空山该升级了,坐起来有点颠,视觉上也没那么科幻了,看起来像是“昨天的明天”。
孩子最好玩的年纪是五到十岁
这句话他说得很随口,但信息量不小。马斯克有十几个孩子,这个判断大概是被反复验证过的。
五岁之前,孩子基本还在解决自己身体的问题,交流有限;十岁之后,开始有自己的世界和脾气;中间那五年,好奇心最旺,还愿意跟你待在一起。
“好”这个概念,本身依赖人类存在
访谈里有一段偏哲学的对话,马斯克回答得比平时更慢。
他的观点是,如果没有人类,“好”这个概念就不会存在。价值判断需要一个主体,主体没了,好坏也就没了落脚点。
沿着这条线,他的结论是:在宇宙中增加意识的总量,是件好事。因为只有通过更多的、更强的意识,宇宙才能被理解。理解宇宙这件事,本身需要有一个能提问的东西存在。
他还算了一笔账,用来解释人在宇宙里的位置。
构成你身体的原子,来自恒星。恒星在核心把氢聚变成重元素,爆炸时把这些元素撒出去,再凝聚成新的恒星,再来一轮。所以可以问一个问题:你身上的原子,曾经在恒星核心待过几次。他的估计是平均三到四次。
接着是第二个问题:你身上的原子,今后还会在恒星核心待几次。各种估算不一样,但大致结论是,我们大概正处在中间点。往后可能还有四次左右。
这个算法的意义在于,它给出一个尺度。如果把“存在”定义为原子待在恒星核心的次数,人类现在正好走到一半。前半程已经把单细胞变成了会发推特的东西,后半程还没开始。
他说,如果你想看大图景,这就是最大的那张图。
短视频正在腐蚀大脑
米勒问他,哪一项进步让人类变糟了。
他答得毫不犹豫:短视频。“它似乎在腐蚀人们的大脑。”
再问他不希望什么被发明出来,答案分两层。浅层是他“希望短视频从来没被发明”。深层是希望永远不要有人造出能杀死全人类的病毒。任何会摧毁意识的发明,都不该出现。
最有趣的结果,最有可能发生
马斯克在这场访谈里讲得最完整的理论出现在这一段,值得慢慢拆。
他的假设是:最有趣的那个结果,往往就是最可能发生的那个。
为什么?他先从自己的工程经验讲起。SpaceX 和特斯拉在电脑里跑仿真,会跑成千上万种情况。团队真正会盯着看的,是那些有戏剧性的、出人意料的、差点成功又失败的案例。一切都按预期走的那次,没人看,因为没有信息。
所以仿真里被反复研究的,是边界情况:火箭在什么参数下能入轨,在什么参数下会炸,那条临界线在哪里。他们不模拟“立刻爆炸”,因为那也没意思。
他把这套逻辑推广到现实本身。如果宇宙是个模拟,那么模拟它的人,也就是那些“观众”,会在剧情无聊的时候关掉它。达尔文式的筛选就发生了:只有最有趣的模拟会被继续运行。
所以从达尔文的角度套用到模拟理论上,只有最有趣的模拟能活下来。那么最有趣的结果就是最可能的,因为另一个选项是被删除。(马斯克)
这个推论的落点是一句半开玩笑的话:我们只有一个目标,把剧情保持得足够有意思。他还对着镜头补了一句请求,请不要删除我们,我们会保持有趣的,我发誓。
他顺手用这个框架解释了几件事。
第一,为什么“世界和平、一切完美”不会是最可能的未来。因为那不是故事。没人会去看一部从头到尾都顺利的电影,观众会中途离场。
第二,为什么历史总是一波三折。所有故事都是三幕结构,第一幕上升,第二幕跌落,第三幕反弹,喜剧收尾是圆满,悲剧收尾是破碎。历史不会是一条一直向右上的直线。
第三,特朗普的政治轨迹。先赢,中期选举输了,然后又赢回来。按马斯克的理论,这个剧本的“有趣度”远高于一路平推,所以它更容易发生。他甚至说,这样看的话,那个结果几乎是必然的。
传送和时间旅行,只在模拟里才成立
米勒问了一个物理问题:如果可以选,你要瞬间移动还是时间旅行。
马斯克说这两件事几乎是同一件事,因为都触碰了同一个底线。
不打破现实,你就没法超越光速。所以如果你真的能瞬间传送到别处,那意味着我们的现实被打破了。(马斯克)
时间旅行同理。回到过去改变事件,会产生逻辑矛盾,而逻辑矛盾在真实宇宙里不该存在。
他给了一个例外条件:除非我们是模拟。在模拟里,时间旅行完全不违反任何东西。
他的解释方式很工程师。想象一个电子游戏,你存了很多个档,可以回退到之前某个存档点。原来那条时间线并没有消失,只是变成了另一个存档。多个存档并行存在,彼此不需要逻辑自洽。
很多人讨论时间旅行会把自己绕进去,因为他们同时坚持两件事:事件必须逻辑一致,而时间旅行必然导致逻辑不一致。马斯克说,这个组合本身就不可能成立,毛病出在前提设定上。
X 的推荐算法,十二月见真章
谈到 X,马斯克把时间点说得很具体:十二月,推荐算法会变得“勉强像样”。
他解释这套系统的信息量级。每天大概有一亿条帖子进入系统,其中很多是垃圾和诈骗,可以快速丢掉。剩下的内容要匹配给每天三四亿用户。这个匹配问题的规模大到离谱,需要海量算力。
Grok 在这里的角色是读完所有内容。马斯克说这话时带点心疼:“可怜的家伙要读一亿条帖子,希望它不会把自己的脑子搞坏。”
他给了一个非常清晰的验收标准,所有做推荐系统的人都该记下来:
你在信息流里,有没有看到来自你从未关注过的账号、但你觉得非常有意思的内容?如果有,算法就在工作。(马斯克)
第二个标准更狠:一个新账号,零粉丝,如果内容足够好,它能被很多人看到吗。
能做到这两点,推荐系统才叫成功。做不到,就还是在做“关注流”的搬运。
标明大洲,不算泄露隐私
米勒问的是 X 显示账号所在地这件事。马斯克的回应区分了两个层级。
他指出,技术上你本来就可以只填区域,比如“亚洲”,范围很大。这套机制真正的作用是增加造假成本。如果一个人假装自己是美国公众、欧洲人或非洲某国人,而账号的一切痕迹都指向另一个大洲,那伪装就变难了。
我们不想人肉任何人。但如果只是说你在哪个洲,我觉得这不算人肉。(马斯克)
这套逻辑的潜台词是,网络上大量“民意”是有组织制造的,降低伪造成本等于放大噪音。加一个粗粒度的位置标签,成本几乎为零,收益是让操纵变贵。
社交媒体让人更表演,也让人更真实
他没有一味批评社交平台。他的观察是双向的。
一个方向是表演性变强了。人们为了多几个播放、多几个赞,会做出各种夸张的行为。
另一个方向是,真实世界发生的视频变多了。有些画面会直接挑战主流叙事,但它就是发生了,而且会迅速传播开。
你同时得到了两样东西:更强的表演,和更多挑战叙事的真实录像。(马斯克)
这两件事同时成立,并不冲突。同一个工具既放大了表演,也放大了证据。判断一个视频有没有价值,可以看它有没有提供一个此前不存在的信息,传播量本身说明不了什么。
钢铁侠是按他建模的
米勒问他愿意当漫威英雄还是邦德反派。他选了前者,然后说了一句挺得意的话:电影里的钢铁侠,原型就是他。
这事有据可查。小罗伯特·唐尼和导演乔恩·费儒(Jon Favreau)在开拍前专门来找过他,参观了 SpaceX。《钢铁侠 2》有相当一部分直接在 SpaceX 工厂里拍摄,你如果看那部片,背景里的厂房就是真的。
他还记得一个细节,剧组在大堂拍斯嘉丽·约翰逊的打斗戏,他当时站在那儿看,心想这也太不真实了。
所以他半开玩笑地说,如果这一切都是模拟,那这个设定也太离谱了。你让我相信这些是真的?
被问到“如果你是玩家,你会觉得这是现实还是模拟”,他的回答是:我大概会觉得是模拟,而且我现在正在把所有支线任务都做一遍。
两千五百加元,一只书箱,一只衣箱
访谈里有个假设性问题:如果你现在从零开始,兜里只有一千美元,你会怎么办。
马斯克先否定了这个前提。他说自己当年是 17 岁一个人到北美,带了两千五百加元,大约两千美元,一箱书,一箱衣服,落地蒙特利尔。这确实是他起步的全部家当。
但他认为这种“归零”在今天对他不成立。
我现在掌握的知识量太大了。要让我落到那种境地,得发生真正的灾难,文明熔化那种级别。(马斯克)
他的推理是这样的。如果他还活着、还清醒、还有信用,那他就可以去说服别人给他钱,承诺一个高回报。这件事他做过,而且现在能做。
反过来说,如果真到了要给一个流浪汉一千美元的场景,那说明社会已经崩了。你在辐射坑里走来走去,像《辐射》游戏那样,一千美元解决不了任何问题。
这个回答听起来像炫耀,但它其实点出了一个真实的经济学原理:可迁移的能力和信誉,比现金更抗风险。现金会归零,会赚钱的能力不会。
如果真要做点事,他说他会去写电子游戏。他年轻时就干过这个,十二岁卖过一款自己写的游戏。
他给自己的驱动力总结是:喜欢解决问题,喜欢造东西。造了很多。
早餐是牛排鸡蛋,最爱的食物是芝士汉堡
马斯克说他现在每天从牛排、鸡蛋和咖啡开始。这是稳定的一餐。
午饭他通常不吃,吃也很少。晚饭看场合,如果是社交局,菜系会很杂。他说自己什么菜都吃。
但被问到“如果余生只能吃一样东西”,答案是芝士汉堡。他的理由是这东西是天才发明,虽然天天吃会单调。
接着他讲了一个故事,这段是全场最像生活片段的。
他住洛杉矶时,带大儿子们去一家叫 Sugarfish 的寿司店。这家店很讲究,菜单上写着不要点酱油,因为主厨已经放好了正确的量,你不能加。主厨认为你不该吃酱油,你就吃不到酱油。
服务员挨个问大家想吃什么。问到 Saxon,Saxon 说,我要一个芝士汉堡。
服务员愣了大概三十秒。在这家连酱油都不能加的店里,有人点了芝士汉堡,这件事需要时间处理。
然后服务员说,我们没有芝士汉堡。
Saxon 用全场都能听见的音量说:什么?什么样的餐厅会没有芝士汉堡?那给我来个汉堡吧,我不知道你们为什么对奶制品有意见。
结果汉堡也没有。
马斯克讲这段的时候很高兴。他说这孩子当时完全是懵的,反复念着不可思议,这地方居然没有芝士汉堡。
这个故事不只是好笑。它顺带把马斯克的教育观露了一点:他的孩子不太害怕在正式场合说错话,也不觉得“别人都点寿司”这件事能约束自己。
小份的芝士汉堡,也能做得很好吃
话题延续到高端餐厅。马斯克说他喜欢烧烤,因为在奥斯汀;如果吃高级料理,他喜欢法餐,但不是天天吃。
米勒问他,如果小份的芝士汉堡被做成一道菜,能好吃吗。
他的回答很笃定:当然能。做小份汉堡不需要诺贝尔奖,物理上没有障碍,只是很少有人认真去做。
他顺便解释了为什么大多数小汉堡难吃:面包太多,肉太干,比例失衡。
然后他提了米勒说的一个场景:设一场晚宴,请三位已故的人。他给的名单是莎士比亚、本杰明·富兰克林、尼古拉·特斯拉。他说其实想聊的人很多,但这三位最有意思。
既然是千载难逢的一餐,就该上十二道菜。也许其中一道,可以是一枚小小的芝士汉堡。他说,如果他们想吃,那当然可以。
用最多的表情,是笑哭那个
被问到最常用哪个表情,马斯克说是那个笑哭的。
这个答案和他前面说的一切是自洽的。他把大量时间花在看人类出洋相上,而且他显然觉得其中很多很好笑。
你该拿这段访谈怎么办
马斯克这场对话信息量很散,但有几条线是可以直接用的。
- 把“我预测的”和“我想要的”分开标注。你判断一个技术趋势时,先写一句“我认为它会发生”,再写一句“我希望它发生”。两句话都写下来,你会发现自己很多判断其实是情绪。
- 砍掉自己的僵尸流程。任何一笔支出、任何一个项目,先要求它填一个“用途代码”。填不出来的,停掉。这一步能省下的比例大概是个位数,但清理成本极低。
- 保护你的睡眠下限。用手机 App 测一次自然睡眠时长,把它当成硬约束。少于这个数字,你多出的清醒时间是借来的,要还利息。
- 把上下文切换当成本来管。把同类任务攒成一段时间块,中间不打断。真正伤认知的地方在切换频率,工作量反而没那么要紧。
- 用“边界”来验证判断,别只盯着“最优”。像 SpaceX 跑仿真那样,去找那个临界点:什么条件下能成,什么条件下会崩。只跑“一切顺利”的场景,等于没跑。
- 给新账号一个机会。无论看内容还是招人,问一句:如果这个人零粉丝、零背景,只看他拿出来的东西,够不够好。
- 把有趣当成一个严肃指标。一件事如果连讲出来都没人想听,它大概也不会发生。
Nice to see you again, Katie.
Right, right.
Well, I guess I couldn’t believe I was there for the most part. It all seemed extremely surreal at the time. Doge was a made up name that had been made up, I don’t know, two or three months before and based on Internet suggestions. I was going to call it the Government Efficiency Commission and then someone on the Internet said, “No, it should be the Department of Government Efficiency.” That sounds great. So we just made up a department.
We were a little bit successful. We were somewhat successful. I mean, we stopped a lot of funding that really just made no sense. That was just entirely wasteful. For example, there was probably $100 million, maybe $200 billion worth of zombie payments per year, which simply by enforcing that there be a payment code and an explanation for the payment, that the payment would not go out. So we’ve made that change to the main treasury computer and a bunch of other computers. It seems insanely obvious, but there are just, call it 2 or 3% of government payments that go out that really should not be going out. And it’s actually quite hard to stop.
So it’s a pretty rare individual that would ask the government to stop sending them money.
Do you mean, would I repeat history or would I…
I mean, no, I don’t think so. I probably, I don’t know.
I mean, the thing is, I think instead of doing Doge, I would have basically worked on my companies, essentially. And the cars, they wouldn’t have been burning the cars.
Yeah. If you stop money going for political corruption, they will lash out big time. So they really want the money to keep flowing. So if you stop it from flowing, there’s a very strong reaction to stopping the money flowing.
Well, I wouldn’t say I was super illusioned to begin with. I guess it’s just you really want the least amount done by government possible. The least amount. I guess maybe the biggest thing is that there are massive transfer payments going to illegal immigrants. Massive. Essentially we’re paying people to come here from somewhere else in vast numbers, including flying them in. So it’s not like you need a border wall. If you’re flying them in, then fast tracking them to citizenship and making them beholden to government payments and voting hard left, that’s essentially voter importation.
If you create a gigantic money magnet where you say, “If anyone comes here from anywhere else, we’re going to pay you tons of money, give you lots of free stuff, come to America and get paid to do so,” you’re going to get a lot of people taking up on that offer. People say this is fake. I’m like, actually, well, let’s look at Ilhan Omar, who literally was voted into power, voted into Congress by a large group of people from Somalia who are in Minnesota, which is really far from Somalia, or Mamdani who was voted to be mayor by a majority of people who are not born in America. That’s my understanding at least. And then California, big time situation.
So I don’t know, we just don’t want to turn into a communist hellhole. Basically.
Assuming the current trend of artificial intelligence and robotics continues, which seems likely, the AI and robots will be able to do anything that humans want them to do, essentially.
People will be able to do whatever they want with their free time. Work will be optional. I just want to separate out what I wish would happen versus what I predict will happen because people get confused about that. They think that what I predict will happen is what I want to happen. What I predict to happen is not the same as what I want to happen. If I could, I would certainly slow down AI and robotics, but I can’t. It seems to be advancing at a very rapid pace. Whether I like it or not.
It used to be. At this point, I don’t know. I wouldn’t say there’s anything particularly keeping me up at night right now, except that if you say, why do I wake up in nightmares? AI. Yeah. Actually I’ve had a lot of AI nightmares. I had AI nightmares many days in a row. What am I supposed to do about it?
I try not to have irrational fears.
If I find an irrational fear, I squelch it. I don’t believe fear is… Fear is the mind killer. So I’m somebody who feels fear strongly.
Six. You can tell based on my posts.
People have actually mapped them so it’s very clear when I’m sleeping and when I’m not. I tried having less than six hours sleep, but although I’m awake more hours per day, my cognitive function is reduced. So my natural sleep, and I actually timed it with the phone. You can get a phone app, time it. It’s 5 hours 56 minutes. That’s what the phone said.
Well, I have a lot of inbound communication, so it’s information triage. I try to segment the days so that there’s not too much context switching, because arguably fear is not the mind killer, context switching is. It’s hard not to context switch if you’ve got an inbox full of stuff, but you can think of, if you had to context switch every three seconds or every 30 seconds or every three minutes, the context switching cognitive penalty would be very high every three seconds.
But even within Tesla and SpaceX, there are many different things. Getting stuff on X, random news things, people being brought alive and stuff like that. They’re like, what the hell’s going on in this country?
You know, President Trump is very funny. He’s got a great sense of humor.
He’s very funny. He’s naturally funny, somewhat effortless. I mean, when he had Mamdani in the office and they asked him if he still thought the President was a fascist, and the President said, “Just say yes. It’s easier that way. Don’t worry about it. Just say yes.” He’s like, “Yes.” Whatever. How silly.
The creator.
God is the creator.
Well, I believe this universe came from something. People have different labels.
I can’t go to things where there’s a general public because there’s an immediate “Can I have a selfie” line that forms. And these days, particularly in light of Charlie Kirk’s murder, there are serious security issues. It’s not that I don’t want to. I simply can’t.
It certainly reinforced the severity of the situation where life is on hardcore mode. You make one mistake and you’re dead. It only takes one. One mistake.
Well, I mean, obviously when my kids were born, or the first time SpaceX got to orbit, or Tesla made an electric car work. There’s a lot of things. There’s a lot coming down the pike.
Starship. The degree to which Starship is a revolutionary technology is not well understood in the world. It’s the first time that there’s been any rocket design where full and rapid reusability is possible or full reusability at all is possible. This is the first design where a reusable rocket is one of the possible outcomes. Where success is in the set of possible outcomes.
Well, we could have made V2 reusable, but there were a lot of performance improvements for V3, so it made sense to go to V3. There’s 10,000 different changes between V2 and V3. Maybe more than 10,000. Really. If there are historians in the future, they’ll look back at Starship and say it was one of the most profound things that ever happened. Now you can think of historic events as where would they fit in the evolutionary hall of fame? So you’ve got things like single-celled life, then you’ve got multicellular life, capturing mitochondria so that you have a power plant in the cell. You’ve got differentiation into plants and animals, life going from oceans to land.
And then also on that scale, probably in the top 10, is life becoming multi-planetary. There just aren’t very many things that are in the top 10 of the evolution of life where you could basically evaluate any given civilization or any given life form on that scale. Life becoming multi-planetary is in the top 10. It needs to be sustainably multi-planetary. So not just visiting, but actually multi-planetary in the sense that if you have planetary redundancy, so if there were to be a catastrophe on one of the planets, the other planet would survive.
Starship is capable of doing that for the first time in history and no AI was used to create it, so the AI will appreciate that.
You know, Tesla is mostly about making sure life on Earth is good. And xAI is about that too. Because multi-planetary means Earth’s got to be good and you need another planet. Sometimes people think because they have legacy templates, mental templates, they think that going to Mars is an escape from Earth or that it would be some place where billionaires would go or something like that. But actually Mars will be very dangerous and the moon base would be also dangerous. Much more dangerous and much less comfortable than Earth. So the people that would go in the early days to make life multi-planetary on Mars or the moon, they would have a much higher risk of death than if they stayed on Earth.
And things would be cramped and uncomfortable. So that’s the sales pitch for Mars. It’s going to be uncomfortable. The food won’t be as good as Earth. You might die. It’s going to be a mass amount of hard work and it may not succeed. That’s the sales pitch. Do you want to go?
Yeah. Didn’t want to be in Jamestown.
Yeah. Maybe if there had been social media back then, they would have been saying, “We’re all dying. Here’s videos of us dying.” Would have probably put a damper on future voyages. But yeah, they just, poof, on fuel just disappeared. We don’t know what happened to them.
I just think like from a fashion standpoint we should evolve. Like my son Saxon said at one point, “Why does everything look like it’s 2015?” I was like, damn, things do. Everything does look like it’s 2015. It’s like if you took a picture from 2015 and said in 2025 it looks exactly the same. There were still, stylistically things are the same as 2015. We have not moved the needle in a decade.
Something new, you know, like the 60s had a definitive style, the 70s had a definitive style, the 80s had a definitive style. And then the 90s also had a different style. But then you start looking at the 2000s and 2010s and it’s like less and less every year. I think we should evolve our style. And if you look at some of the older paintings, you know, past cabinet secretaries, some of them like they look cool. Like their jackets are cooler than what we have right now. You know, they have sort of like a high collar and like a sort of, I don’t know what some sort of, what do you call those things, ascot or something like that. I mean it just looks cool, but we don’t.
Everything’s like a very normal looking suit at this point. But like literally the same as 2015. I’m being generous, because arguably the same as 2010. So in 15 years. And I’m like, from a fashion standpoint, I don’t think we’ve moved since 2000. In 25 years, if you showed someone a picture of, “This is a bunch of dudes in 2000. This is a bunch of dudes in 2025. Which year is which?” So I think we should, I don’t know, spice it up a little.
I mean, which conspiracy theories haven’t come true at this point? We’ve run out of conspiracy theories that, because it will come true as far as I can tell. I mean, I don’t know of any aliens. People always ask me if there are aliens. I have seen no evidence of aliens. No one on the SpaceX senior team has any evidence of aliens. Because I’ve asked him, like, “Guys, am I missing something? Has anyone on the team, has anyone seen any evidence of aliens?”
That’s just an unidentified flying object. So UFOs, it could be like some new weapons program or whatever. That’s, you know, some hypersonic missile or something like that. That would be technically a UFO, but it’s just basically some weapons prototype. It’s not like aliens. So although Neil Armstrong, Neil A. spelled backwards is alien. Coincidence.
Yes. We went to the moon a few times, actually. And played golf on the moon. We didn’t just go to the moon. We actually got a little bored and started playing golf on the moon.
That was the jump the shark moment.
No, the playing golf on the moon.
You know, I looked at it.
Yeah, yeah. Whacked golf ball on the moon.
So, like, there is gravity. One-sixth. If it wasn’t gravity, you just float away. There’s no atmosphere.
But there is one-sixth gravity.
I don’t know. How would I know? What do you think?
Me?
Oh.
Thanks.
Yeah, I don’t yell.
Right. Well, why would they think anyone would work at the companies? Yeah, I mean, talented people can go work anywhere they want. So they’re only going to work at one of my companies if they want to. And if they’re mistreated in some way, they would leave and go work somewhere else.
Well, I think we needed something inspirational. We kind of have a lot of star things, you know, so we got Starlink, Starship. Well, where would Starship depart from? Starbase. I mean, Starbase is, as you mentioned, it’s like, I think it’s probably the coolest place on Earth.
And it used to be a sandbar down by the Rio Grande. It’s only like 3 feet above sea level. So we built a gigantic rocket factory and two giant launch towers down by the river, literally within sight of the Rio Grande. And on an actual sandbar. Kind of had to have like an inspirational name. And then we made it a city. So it’s an incorporated city, like legally a city. You don’t hear about new cities being formed that often.
Yeah, I think Ford had some kind of like company town situations. But yeah, Disney World is, it’s literally its name. I’m Walt Disney. This is my world. I’ve gone from land to world. They got like incorporated as a city and got tax exemption, which was like a whole, was a big deal. Yeah, I’ve been to Disney World probably 10 times.
Yeah, maybe more than, maybe more than 10, but at least 10 times because Cape Canaveral is right by Disney World.
So when I’d have the kids, then I would, my older kids and I was, we’re trying to get the rocket launch from Cape Canaveral. Then, you know, the thing they’d want to do is go to Disney World or Harry Potter Land.
I’m sort of tempted to say Space Mountain, I suppose. Yeah, probably Space Mountain. I mean, I do think Space Mountain needs an upgrade. A little herky jerky. It doesn’t look quite as sci-fi as it used to. You know, it’s like the day before yesterday is tomorrow or just like yesterday.
Generally, kids are the most fun between 5 and 10.
The concept of good wouldn’t exist without humanity. I do think humanity is, on balance, good. I generally think increasing the amount of consciousness in the universe is a good thing. Trying to try to understand the nature of the universe, which you can only do by increasing conscious awareness. I’ve thought about like, how did we get here? Because if we just start out as a hydrogen gas cloud that sort of condensed and then formed stars, and then these stars exploded and then they recondensed, formed stars again and then exploded, exploded again, and then eventually you get to us 13.8 billion years later.
One of the interesting questions to think about is how many times have your atoms been at the center of a star? I think it’s like on average three or four times, something like that. Then how many times will your atoms be at the center of a star? Estimates vary, but it seems like we’re roughly halfway. So your atom’s likely to be at the center of the star, maybe another four times or something like that. It depends on what your predictions are for the future. But in terms of existence as measured by the number of times your atoms will be at the center of a star, we seem to be roughly halfway that early. You know, if you want to look at the big picture, that’s the really big picture.
What’s one adventure that’s made us worse?
Maybe short form video. It seems to be rotting people’s brains.
I hope never gets invented.
Well, I mean, obviously I hope like that people don’t invent a virus that can kill all humans. Like that’s an obvious thing. I mean, yeah, generally I hope inventions that destroy consciousness are not invented. I think the future is going to look very interesting. So I do have this theory about predicting the future, which is that the most interesting outcome is the most likely. Which, if simulation theory is accurate, makes sense. Because if anyone is simulating a wide range of futures, they’re going to stop the simulation when it gets boring. Because this is what we do in our reality.
So if SpaceX is doing, or Tesla doing simulations to understand how a car would work or robot or spaceship or something like that, we run all these simulations in the computer and the simulations that we pay attention to are the ones that are the most interesting. Like the simulation where everything goes right on the rocket we actually don’t pay attention to because that’s not a, everything goes right simulation is fine. So we actually test the, when we simulate the rocket flight, we’ll actually test all sorts of oddball situations. But we don’t test it. We don’t have the simulation be totally wrong because I mean like if the rocket just explodes immediately, that’s also not interesting.
So it’s like you’d need to find the envelope of possible flight paths where the rocket can make it to orbit and without exploding. And then you find those boundaries and then when you launch the actual rocket, you try to make sure it stays within those boundaries. Or another way to think of it is like we could be an alien Netflix series and that that series is only going to get continued if our ratings are good.
Yeah, but you can think of it like from a Darwinian standpoint applied. If you apply Darwin to simulation theory, then only the most interesting simulations will continue. Therefore the most interesting outcome is most likely because it’s either that or annihilation. So really we have one goal, keep it interesting.
Well, social media makes people more performative. By the same token, you get more real life video of things that are actually happening, and anything that is very interesting will go viral on the Internet. So you have both. You’ve got more performative where people are doing anything they can to get a few more views on their TikTok video or whatever, or their reels, or maybe on their X post or something. And so that’s very performative. But then you also see real life videos that challenge the narrative but are nonetheless real.
I don’t really think about it that much. I mean, there’s in a country of origin, we have to be a little careful about this. You can actually technically just specify your region. Like you can say I’m in Asia or something like that, which is quite big. But it does make it a little harder that if somebody is trying to pretend that they’re, say, a member of the American public or in Europe or Africa, wherever. If everything about their account is from a different continent than they are pretending to be from, it gets a little harder to pretend. We don’t want to dox people, but we kind of think you’re not really doxing someone if you say which continent they’re from.
Yeah.
It’s a false dichotomy. I think I’d say guarantee Earth. Earth’s much better than Mars, to be clear. But Mars is just our best option if we want to become a multi-planet species. It’s really our only option if you want to become a multi-planet species. We’ve got Mars, which is very difficult, but not impossible. Earth is much better than Mars, but we can’t—I think it was Tsiolkovsky, or I think he said, “Earth is a cradle of civilization, but we can’t stay in the cradle forever.”
I think it would depend on which Marvel superhero or which Bond villain. I suppose I’d rather be a Marvel superhero. They did model Iron Man in the movies after me.
Yes. That’s pretty cool. Yeah. Robert Downey Jr. and Favreau met with me, toured SpaceX and stuff. In fact, Iron Man 2, a large part of the movie is filmed in SpaceX.
Yes. If you watch Iron Man 2, you’ll see it’s the SpaceX factory is the actual background.
Yeah, it was cool. We had Scarlett Johansson doing martial arts in the lobby, actually. Yeah. And you expect me to believe this is all real, the simulation?
Exactly. What are the odds?
I mean, if you were me—
Would you think this is real or a simulation?
Yeah.
Yeah. And I’m doing all the side quests and everything.
Doge, probably.
Who came up with these questions?
What do you mean? So that with an algorithm means you only see the people you follow.
Yeah, yeah. There’s the sort of people you follow. And then there’s a recommendation algorithm. I think probably in December we’ll finally have a half decent recommendation algorithm.
Yeah. So it really is just trying to show people stuff they’d be interested in. But there’s an enormous amount of AI horsepower being applied to this. Where Grok, poor thing, is reading—is going to read all 100 million posts per day, which is—
Hopefully it doesn’t destroy its mind or something. Yeah, it does take a lot of compute. Most posts are—there’s a lot of spam, scam stuff. So that can be easily discarded, I suppose. But then you’ve got to take 100 million pieces of content, match that to sometimes 3 or 400 million people per day. So that’s a lot of matching.
Well, we really are kind of—this is just the beginning kind of thing. What I mentioned, like Grok reading everything and recommending any given thing to anyone should go live in December. So the acid test for this is: are you seeing content that you find really interesting from accounts you’ve never seen before? If that’s happening, then the algorithm is working. It should be possible for somebody to post content as a new user with no followers. And if that content is excellent, it gets seen by a lot of people. So can an account with small number of followers or a new account, if the content is intrinsically excellent, can that content be seen by a lot of people? That’s our goal.
Actually those things are almost the same thing in that you can’t break the speed of light without breaking reality. And so if you could teleport somewhere instantly, if you were talking about teleportation faster than the speed of light, presumably it would be—then that would break our reality, as would time travel. Unless there’s a very important conditional here. Unless we are a simulation. Time travel does not break a simulation.
I think—well, people do tend to get wrapped up in knots with the time travel thing because they try to simultaneously say something must be logically consistent but logically inconsistent. That’s impossible. But if you think of it like a video game and say, okay, you’ve got various saved games and you can go back and restore a saved game from a prior start point. You still have your other saved games, and there are many games going on in parallel. They don’t have to be consistent with each other. That is a false assumption. If we’re a simulation, we might be somebody’s video game or TV show or something like that.
Like I said, we’re just going to keep it interesting so they don’t turn the computer off. I’m just saying, if that’s true, keep it interesting. Or they can turn off the computer and they might. Please don’t delete us. Please don’t delete us. We’ll keep it interesting. I swear.
Yeah. So if the most interesting outcome is most likely, what do you think are the most interesting things that can occur now? Most interesting is not what you want. It’s just as viewed by a third party. Let’s say this was, for argument’s sake, an alien Netflix series, and you’re trying to maximize your viewership, maximize your ratings. It’s actually an interesting thought experiment. It’s actually not that interesting if everything just blows up. It’s now over. That’s not that interesting. It’s not that interesting if there’s a calamity that wipes out all the humans. The show just ended.
But I mean, fortunately and unfortunately, if there is drama that, like war or something like that, that is interesting. People will go to movies and watch, say, a World War I movie where people are getting blown up from cannon shells and they’re in the movie theater eating popcorn, drinking a soda. You wouldn’t go to a movie where everything was just perfect and stayed that way. You’d leave the theater.
There’s always a story arc. There’s always an arc, and it’s generally not a linear arc. So it’s not going to be like things start here and just go straight up and to the right and end up in a good place or something like that. There’s usually ups and downs. Classic sort of story arcs, essentially. Act 1, Act 2, Act 3. You have an initial rise in Act 1, fall down in Act 2, surge back in Act 3 with a happy ending if it’s a comedy, or a sad ending if it’s a drama. If you look at President Trump’s story, it’s more interesting that he lost the intermediate term and then won his second term after that. Just like the story arc, initially up, then down, then resurgent again.
If you went with my theory that the most interesting outcome is the most likely, then that was the most likely outcome. It was inevitable.
I am—Irony Man. Something like that. I’m paraphrasing. What am I watching? Actually, right now? I’m watching Teenage Mutant Ninja Turtles, the TV series. Turtles in a Half Shell. Turtle power. Yeah. Because Little X wants to watch that. I’m watching things that the kids want to watch. Re-watched Dodgeball last night.
Yeah. “If you can dodge a wrench, you can dodge a ball.” “If you can dodge a wrench, you can dodge a ball.” What? Yeah. High motivation to dodge if somebody’s throwing a wrench at you.
“The Final Countdown” by Europe.
What’s the goal?
If it’s a simple thing, I’ll wing it. If it’s a complex thing, I’ll look at the instructions.
Well, I did originally come to North America with, I don’t know, $2,500 Canadian. So I don’t know, maybe $2,000 U.S., one bag of books and one bag of clothes in Montreal at age 17. So that is how I started out. At this point, I have a lot of knowledge. A lot of things have to go wrong for that to be the case. It’s like, am I just emerging from prison, perhaps with a stipend? All my company’s been confiscated. I mean, it would take Armageddon, which hopefully that doesn’t happen, like Ragnarok next level. And I lost.
What the hell.
I mean, it’s impossible for someone to have that amount of all the knowledge that I have and then be dropped down to a low resource amount. Because the reality is that either something truly catastrophic has happened, like civilization has melted, or I will be able to ask people to just give me money. And with the promise that I will have a high return, which is what I’m able to do right now. If you give me a dollar, you will get back much more than a dollar. So this is—it’s somewhat of an impossible dichotomy because civilization would have had to have been destroyed or something. In which case $1,000 is not going to solve your problems. You can’t do much with this.
If you’re wandering around radioactive craters and you’re in, like, Fallout or whatever, then $1,000 is not going to solve anything. And if civilization has melted, then probably just talk people into giving me money, which I’ve done before.
I don’t know if that’s all that random, but I’d probably write video games or something like that. I did that at one point. I like solving problems, so I like building things. I built a lot of things. Like, a lot.
Well, these days, I start off with a breakfast of steak and eggs and coffee. And then dinner tends to vary. I usually don’t have lunch or if I do something very small, and then dinner, depending on whether it’s social or not, will vary in cuisine. I like a wide range of cuisine.
American food is my favorite food.
Yeah, the cheeseburger is probably the—if I had to say, like, there’s only one thing you can ever have for the rest of time, which admittedly would be a bit monotonous, but it would probably be a cheeseburger, because cheeseburgers are amazing. It’s a genius invention. I’ll tell you a funny story about when I was living in LA and I took my older boys out for lunch to Sugarfish, which is a very kind of uptight sushi restaurant. In fact, on the menu of the restaurant, it says, “Do not ask for soy sauce because the chef has put the right amount of soy sauce, and you can’t have any more.” And if the chef doesn’t think you should have soy sauce, you can’t have soy sauce.
That’s what it says on the menu, basically. So, like, extremely strict sushi restaurant. And so the waiter is going around asking everyone what they want, and then it comes to Saxon, and Saxon says, “I’ll have a cheeseburger.” And the waiter takes a moment for the waiter to recover because no one’s ever asked for a cheeseburger at this very strict sushi restaurant. Took him, like, 30 seconds to realize he’d just been asked for a cheeseburger because you’re not even allowed to ask for soy sauce. So then when he finally recovered, he said, “We don’t have cheeseburgers.” And Saxon goes at the top of his voice, “What?
Like, what kind of restaurant doesn’t have cheeseburgers?” And says, “Fine, I’ll have a hamburger. I don’t know what you got against dairy,” but, yeah, they don’t have hamburgers either.
Yeah, but he was nonplussed. It’s like, “I can’t believe this place doesn’t have cheeseburger.” So, yeah, I mean, I guess I like barbecue, which is good because I’m here in Austin. I mean, if it’s haute cuisine, I like French food as well. But not every day. Every once in a while.
I guess the emoji I use the most, which is the laughing emoji.
Or maybe Shakespeare, Ben Franklin, Nikola Tesla. I mean, there’s actually a lot of people I’d like to—I would have liked to talk to. And we’ll eat, I guess, whatever they’d like. I think if you’re going to—if this is a once in a lifetime thing, I think you’d want to have some epic, you know, 12-course meal or something like that.
Yeah. Yeah. Do you want to go all out for that dinner? I think you’re probably not going to serve cheeseburgers unless they want it. Yeah. Maybe one of the courses could be like a tiny cheeseburger.
No, but they could. It’s just they don’t try. There’s nothing—you could make a tiny cheeseburger taste just as good as a big cheeseburger if you tried.
Rare, but yes.
1% of the time.
But usually it’s too much bread and it’s dry.
Yeah.
Yeah, but could you make a tiny cheeseburger that’s good? Of course. Like, you’re not breaking—you’re not—like, you don’t need a Nobel Prize for this. You can definitely make a tiny cheeseburger. It’s not physically possible. I’m saying it’s just rare.
You’re welcome.
假设人工智能和机器人技术的当前趋势延续下去——这看起来很有可能——那么AI和机器人将基本能够做人类想让它们做的任何事。
Assuming the current trend of artificial intelligence and robotics continues, which seems likely, the AI and robots will be able to do anything that humans want them to do, essentially.
人们将能够用空闲时间做任何自己想做的事。工作将变成可选项。
People will be able to do whatever they want with their free time. Work will be optional.
我试过睡不到六小时,但虽然每天醒着的时间更多,我的认知功能却下降了。
I tried having less than six hours sleep, but although I’m awake more hours per day, my cognitive function is reduced.
我尽量把每天分段,避免太多任务切换,因为可以说,恐惧不是心智杀手,任务切换才是。
I try to segment the days so that there’s not too much context switching, because arguably fear is not the mind killer, context switching is.
如果未来有历史学家,他们会回顾Starship,说它是有史以来最深刻的事件之一。
If there are historians in the future, they’ll look back at Starship and say it was one of the most profound things that ever happened.
我没有看到任何外星人存在的证据。SpaceX高级团队里也没有任何人掌握外星人存在的证据。
I have seen no evidence of aliens. No one on the SpaceX senior team has any evidence of aliens.
我确实有一个预测未来的理论:最有趣的结果最有可能发生。
I do have this theory about predicting the future, which is that the most interesting outcome is the most likely.
应该有可能让某人以新用户身份、在没有粉丝的情况下发布内容。如果那内容很出色,它就会被很多人看到。
It should be possible for somebody to post content as a new user with no followers. And if that content is excellent, it gets seen by a lot of people.
我最初来北美时,17岁,在蒙特利尔,带着——我也不知道——2500加元。所以,我不知道,也许2000美元,一包书和一包衣服。这就是我的起点。
I did originally come to North America with, I don’t know, $2,500 Canadian. So I don’t know, maybe $2,000 U.S., one bag of books and one bag of clothes in Montreal at age 17. So that is how I started out.
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