# Elon Musk — "In 36 months, the cheapest place to put AI will be space”
- Transcript ID: `84ed68147b3b3ca4`
- Source tier: `official_authored_text`
- Original language: `en_or_source_language_unlabeled`
- Translation status: `agent_translated_reviewed` (1336/1336 chunks)
- Original source: https://www.dwarkesh.com/p/elon-musk
- Raw evidence: `data/transcripts/web/dwarkesh-elon-musk-2026-02-05.json`
> 中文是派生翻译，英文是留存的源文本。中文未复核时，以英文和原始音视频为准。
## paragraph 1
**中文**
在本期节目中，John 和我与 Elon 进行了一场真正深入的对谈。我们讨论了轨道数据中心的经济性、在地球上扩大电力供应的难题、要在美国大规模制造人形机器人需要具备什么条件、xAI 的商业与对齐计划、DOGE，以及更多话题。
**English / Original**
In this episode, John and I got to do a real deep-dive with Elon. We discuss the economics of orbital data centers, the difficulties of scaling power on Earth, what it would take to manufacture humanoids at high-volume in America, xAI’s business and alignment plans, DOGE, and much more.
## paragraph 2
**中文**
在 YouTube 上观看；或通过 Apple Podcasts、Spotify 收听。
**English / Original**
Watch on YouTube; listen on Apple Podcasts or Spotify.
## paragraph 3
**中文**
赞助商
**English / Original**
Sponsors
## paragraph 4
**中文**
Mercury 刚刚开始提供个人银行服务！我的企业业务已经在使用 Mercury，因此个人生活也能用它办理银行业务，让一切简单了许多。立即前往 mercury.com/personal-banking 申请。
**English / Original**
Mercury just started offering personal banking! I’m already banking with Mercury for business purposes, so getting to bank with them for my personal life makes everything so much simpler. Apply now at mercury.com/personal-banking
## paragraph 5
**中文**
Mercury 刚刚开始提供个人银行服务！我的企业业务已经在使用 Mercury，因此个人生活也能用它办理银行业务，让一切简单了许多。立即前往 mercury.com/personal-banking 申请。
**English / Original**
Mercury just started offering personal banking! I’m already banking with Mercury for business purposes, so getting to bank with them for my personal life makes everything so much simpler. Apply now at mercury.com/personal-banking
## paragraph 6
**中文**
Jane Street 上周给我出了一道新谜题：他们训练了一个神经网络，把全部 96 层打乱，然后让我将它们恢复成正确顺序。我试了试，但……没能完全答对。如果你好奇，或者觉得自己能做得更好，不妨到 janestreet.com/dwarkesh 试一试。
**English / Original**
Jane Street sent me a new puzzle last week: they trained a neural net, shuffled all 96 layers, and asked me to put them back in order. I tried but… I didn’t quite nail it. If you’re curious, or if you think you can do better, you should take a stab at janestreet.com/dwarkesh
## paragraph 7
**中文**
Jane Street 上周给我出了一道新谜题：他们训练了一个神经网络，把全部 96 层打乱，然后让我将它们恢复成正确顺序。我试了试，但……没能完全答对。如果你好奇，或者觉得自己能做得更好，不妨到 janestreet.com/dwarkesh 试一试。
**English / Original**
Jane Street sent me a new puzzle last week: they trained a neural net, shuffled all 96 layers, and asked me to put them back in order. I tried but… I didn’t quite nail it. If you’re curious, or if you think you can do better, you should take a stab at janestreet.com/dwarkesh
## paragraph 8
**中文**
Labelbox 能够为你规模化获取机器人和强化学习（RL）数据。Labelbox 会先帮助你定义理想的数据分布，随后由其庞大的 Alignerr 网络收集前沿水准的数据，供你训练模型。详情请访问 labelbox.com/dwarkesh。
**English / Original**
Labelbox can get you robotics and RL data at scale. Labelbox starts by helping you define your ideal data distribution, and then their massive Alignerr network collects frontier-grade data that you can use to train your models. Learn more at labelbox.com/dwarkesh
## paragraph 9
**中文**
Labelbox 能够为你规模化获取机器人和强化学习（RL）数据。Labelbox 会先帮助你定义理想的数据分布，随后由其庞大的 Alignerr 网络收集前沿水准的数据，供你训练模型。详情请访问 labelbox.com/dwarkesh。
**English / Original**
Labelbox can get you robotics and RL data at scale. Labelbox starts by helping you define your ideal data distribution, and then their massive Alignerr network collects frontier-grade data that you can use to train your models. Learn more at labelbox.com/dwarkesh
## paragraph 10
**中文**
时间戳
**English / Original**
Timestamps
## paragraph 11
**中文**
（00:00:00）— 轨道数据中心
**English / Original**
(00:00:00) - Orbital data centers
## paragraph 12
**中文**
（00:36:46）— Grok 与对齐
**English / Original**
(00:36:46) - Grok and alignment
## paragraph 13
**中文**
（00:59:56）— xAI 的商业计划
**English / Original**
(00:59:56) - xAI’s business plan
## paragraph 14
**中文**
（01:17:21）— Optimus 与人形机器人制造
**English / Original**
(01:17:21) - Optimus and humanoid manufacturing
## paragraph 15
**中文**
（01:30:22）— 中国会默认胜出吗？
**English / Original**
(01:30:22) - Does China win by default?
## paragraph 16
**中文**
（01:44:16）— 经营 SpaceX 得到的经验
**English / Original**
(01:44:16) - Lessons from running SpaceX
## paragraph 17
**中文**
（02:20:08）— DOGE
**English / Original**
(02:20:08) - DOGE
## paragraph 18
**中文**
（02:38:28）— TeraFab
**English / Original**
(02:38:28) - TeraFab
## paragraph 19
**中文**
文字稿
**English / Original**
Transcript
## paragraph 20
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 21
**中文**
真的有整整三个小时的问题？你他妈是认真的吗？
**English / Original**
Are there really three hours of questions? Are you fucking serious?
## paragraph 22
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 23
**中文**
Elon，你不觉得有很多东西可聊吗？
**English / Original**
You don’t think there’s a lot to talk about, Elon?
## paragraph 24
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 25
**中文**
我操，老兄。
**English / Original**
Holy fuck man.
## paragraph 26
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 27
**中文**
眼下正是最有意思的节点，所有故事线都在此刻汇合。看看我们能聊完多少吧。
**English / Original**
It’s the most interesting point. All the storylines are converging right now. We’ll see how much we can get through.
## paragraph 28
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 29
**中文**
简直就像这是我安排好的一样。
**English / Original**
It’s almost like I planned it.
## paragraph 30
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 31
**中文**
没错。我们会谈到这个。
**English / Original**
Exactly. We’ll get to that.
## paragraph 32
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 33
**中文**
但我绝不会做这种事……
**English / Original**
But I would never do such a thing…
## paragraph 34
**中文**
00:00:00 — 轨道数据中心
**English / Original**
00:00:00 - Orbital data centers
## paragraph 35
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 36
**中文**
你比任何人都清楚，能源只占数据中心总拥有成本的 10%—15%。把数据中心搬到太空，想必省下的就是这部分。成本的大头是 GPU。如果 GPU 在太空，维护会更困难，甚至根本无法维护，因而它们的折旧周期会缩短。想必把 GPU 放在太空就是昂贵得多。为什么还要把它们放到太空？
**English / Original**
As you know better than anybody else, only 10-15% of the total cost of ownership of a data center is energy. That’s the part you’re presumably saving by moving this into space. Most of it’s the GPUs. If they’re in space, it’s harder to service them or you can’t service them. So the depreciation cycle goes down on them. It’s just way more expensive to have the GPUs in space, presumably. What’s the reason to put them in space?
## paragraph 37
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 38
**中文**
问题在于能源是否可得。如果看中国以外地区的发电量，所有地方基本都没有增长；或许略有增加，但总体接近持平。中国的发电量则在迅速增长。但如果你把数据中心建在中国之外，电从哪里来？尤其是在规模不断扩大时。
**English / Original**
The availability of energy is the issue. If you look at electrical output outside of China, everywhere outside of China, it’s more or less flat. It’s maybe a slight increase, but pretty close flat. China has a rapid increase in electrical output. But if you’re putting data centers anywhere except China, where are you going to get your electricity? Especially as you scale.
## paragraph 39
**中文**
芯片产量几乎呈指数增长，而发电量却停滞不前。那你要怎么让这些芯片通电？靠魔法能源？靠会变电的仙女吗？
**English / Original**
The output of chips is growing pretty much exponentially, but the output of electricity is flat. So how are you going to turn the chips on? Magical power sources? Magical electricity fairies?
## paragraph 40
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 41
**中文**
众所周知，你非常推崇太阳能。一太瓦的太阳能，按 25% 的容量因子算，大约需要四太瓦的太阳能板，占美国国土面积的 1%。等我们拥有一太瓦数据中心时，已经进入奇点了，对吧？那么究竟是什么会不够用？
**English / Original**
You’re famously a big fan of solar. One terawatt of solar power, with a 25% capacity factor, that’s like four terawatts of solar panels. It’s 1% of the land area of the United States. We’re in the singularity when we’ve got one terawatt of data centers, right? So what are you running out of exactly?
## paragraph 42
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 43
**中文**
但那时你已经深入奇点多远了？
**English / Original**
How far into the singularity are you though?
## paragraph 44
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 45
**中文**
这得由你来告诉我。
**English / Original**
You tell me.
## paragraph 46
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 47
**中文**
正是如此。所以我认为，我们会发现自己已身处奇点，然后说：“好吧，我们还有很长的路要走。”
**English / Original**
Exactly. So I think we’ll find we’re in the singularity and it’ll be like, “Okay, we’ve still got a long way to go.”
## paragraph 48
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 49
**中文**
但计划是等内华达州铺满太阳能板之后，再把它搬到太空吗？
**English / Original**
But is the plan to put it in space after we’ve covered Nevada in solar panels?
## paragraph 50
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 51
**中文**
我觉得要在内华达州铺满太阳能板相当困难。你得拿到许可证。你去试着申请一下，看看会发生什么。
**English / Original**
I think it’s pretty hard to cover Nevada in solar panels. You have to get permits. Try getting the permits for that. See what happens.
## paragraph 52
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 53
**中文**
所以，太空路线其实是为了规避监管限制。在陆地建设比在太空还难。
**English / Original**
So space is really a regulatory play. It’s harder to build on land than it is in space.
## paragraph 54
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 55
**中文**
在地面扩大规模比在太空更难。而且太空中的太阳能板效能约为地面的五倍，也不需要电池。我差点穿了另一件写着“太空永远阳光明媚”的衬衫。事实确实如此，因为太空没有昼夜循环、季节变化、云层或大气。仅大气一项就会造成约 30% 的能量损失。
**English / Original**
It’s harder to scale on the ground than it is to scale in space. You’re also going to get about five times the effectiveness of solar panels in space versus the ground, and you don’t need batteries. I almost wore my other shirt, which says, “it’s always sunny in space”. Which it is because you don’t have a day-night cycle, seasonality, clouds, or an atmosphere in space. The atmosphere alone results in about a 30% loss of energy.
## paragraph 56
**中文**
因此，同一块太阳能板在太空产生的电力约为地面的五倍，还能省去用电池支撑夜间供电的成本。事实上，在太空做这件事便宜得多。我的预测是，太空将远远成为部署 AI 成本最低的地方，时间是在 36 个月以内。也许是 30 个月。
**English / Original**
So any given solar panel can do about five times more power in space than on the ground. You also avoid the cost of having batteries to carry you through the night. It’s actually much cheaper to do in space. My prediction is that it will be by far the cheapest place to put AI. It will be space in 36 months or less. Maybe 30 months.
## paragraph 57
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 58
**中文**
36 个月？
**English / Original**
36 months?
## paragraph 59
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 60
**中文**
不到 36 个月。
**English / Original**
Less than 36 months.
## paragraph 61
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 62
**中文**
GPU 出故障时要怎么维护？这种情况在训练中相当常见。
**English / Original**
How do you service GPUs as they fail, which happens quite often in training?
## paragraph 63
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 64
**中文**
其实，这取决于运抵的 GPU 是哪一代产品。就目前而言，我们发现自己的 GPU 相当可靠。确实存在早期失效，但显然可以在地面上筛除。你只需先让它们在地面运行，确认这些 GPU 没有早期失效问题。
**English / Original**
Actually, it depends on how recent the GPUs are that have arrived. At this point, we find our GPUs to be quite reliable. There’s infant mortality, which you can obviously iron out on the ground. So you can just run them on the ground and confirm that you don’t have infant mortality with the GPUs.
## paragraph 65
**中文**
但一旦开始正常工作，并度过 Nvidia 或其他芯片制造商产品最初的调试周期——也可能是 Tesla AI6 芯片之类，或者 TPU、Trainium 等——过了某个节点之后，它们就相当可靠。因此，我认为维护并不是问题。
**English / Original**
But once they start working and you’re past the initial debug cycle of Nvidia or whoever’s making the chips—could be Tesla AI6 chips or something like that, or it could be TPUs or Trainiums or whatever—they’re quite reliable past a certain point. So I don’t think the servicing thing is an issue.
## paragraph 66
**中文**
不过你们记住我的话：36 个月内，而且很可能更接近 30 个月，太空将成为部署 AI 在经济上最具吸引力的地方。之后，在太空的优势还会大得惊人。
**English / Original**
But you can mark my words. In 36 months, but probably closer to 30 months, the most economically compelling place to put AI will be space. It will then get ridiculously better to be in space.
## paragraph 67
**中文**
真正能够扩大规模的地方只有太空。一旦开始从“利用了太阳能量的百分之多少”这个角度思考，你就会意识到必须进入太空。在地球上扩不了太大的规模。
**English / Original**
The only place you can really scale is space. Once you start thinking in terms of what percentage of the Sun’s power you are harnessing, you realize you have to go to space. You can’t scale very much on Earth.
## paragraph 68
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 69
**中文**
不过澄清一下，你所说的“大规模”是太瓦量级？
**English / Original**
But by very much, to be clear, you’re talking terawatts?
## paragraph 70
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 71
**中文**
对。目前整个美国平均也只使用半太瓦电力。所以，如果说一太瓦，那就是美国当前耗电量的两倍，确实非常庞大。你能想象要建那么多数据中心、那么多发电厂吗？
**English / Original**
Yeah. All of the United States currently uses only half a terawatt on average. So if you say a terawatt, that would be twice as much electricity as the United States currently consumes. So that’s quite a lot. Can you imagine building that many data centers, that many power plants?
## paragraph 72
**中文**
一直生活在软件世界的人没有意识到，硬件很快就会给他们上一堂痛苦的课。建造发电厂其实非常困难。你需要的不只是发电厂，还需要全套电气设备。你得有电力变压器，才能运行 AI Transformer。
**English / Original**
Those who have lived in software land don’t realize they’re about to have a hard lesson in hardware. It’s actually very difficult to build power plants. You don’t just need power plants, you need all of the electrical equipment. You need the electrical transformers to run the AI transformers.
## paragraph 73
**中文**
公用事业是一个非常缓慢的行业。它们的节奏基本与政府、与公用事业委员会进行阻抗匹配——无论字面还是比喻意义上都是如此。它们很慢，因为过去一直都很慢。因此，要让它们加快速度……你们有没有试过以大功率、大规模的方式和公用事业公司签并网协议？
**English / Original**
Now, the utility industry is a very slow industry. They pretty much impedance match to the government, to the Public Utility Commissions. They impedance match literally and figuratively. They’re very slow, because their past has been very slow. So trying to get them to move fast is... Have you ever tried to do an interconnect agreement with a utility at scale, with a lot of power?
## paragraph 74
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 75
**中文**
作为一名职业播客主持人，我可以说，我确实没有试过。
**English / Original**
As a professional podcaster, I can say that I have not, in fact.
## paragraph 76
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 77
**中文**
得再多很多播放量，他们才会遇到这个问题。
**English / Original**
They need many more views before that becomes an issue.
## paragraph 78
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 79
**中文**
他们得花一年做研究。一年后，他们才会拿着并网研究报告回来找你。
**English / Original**
They have to do a study for a year. A year later, they’ll come back to you with their interconnect study.
## paragraph 80
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 81
**中文**
你不能用自建的表后供电设施来解决这个问题吗？
**English / Original**
Can’t you solve this with your own behind the meter power stuff?
## paragraph 82
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 83
**中文**
你可以建发电厂。我们在 xAI 为 Colossus 2 做的就是这件事。
**English / Original**
You can build power plants. That’s what we did at xAI, for Colossus 2.
## paragraph 84
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 85
**中文**
那为什么还要讨论电网？为什么不把 GPU 和电源建在一起？
**English / Original**
So why talk about the grid? Why not just build GPUs and power co-located?
## paragraph 86
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 87
**中文**
我们就是这么做的。
**English / Original**
That’s what we did.
## paragraph 88
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 89
**中文**
但我的意思是，为什么这不能成为一种通用解决方案？
**English / Original**
But I’m saying why isn’t this a generalized solution?
## paragraph 90
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 91
**中文**
发电厂从哪里来？
**English / Original**
Where do you get the power plants from?
## paragraph 92
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 93
**中文**
你说了和公用事业公司合作的种种问题，但你完全可以为数据中心配建私有发电厂。
**English / Original**
When you’re talking about all the issues working with utilities, you can just build private power plants with the data centers.
## paragraph 94
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 95
**中文**
没错。但这就引出了另一个问题：发电厂从哪里来？得找发电设备制造商。
**English / Original**
Right. But it begs the question of where do you get the power plants from? The power plant makers.
## paragraph 96
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 97
**中文**
哦，我明白你的意思了。归根结底是燃气轮机订单积压吗？
**English / Original**
Oh, I see what you’re saying. Is this the gas turbine backlog basically?
## paragraph 98
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 99
**中文**
是的。还可以再深挖一层。限制因素是涡轮中的导向叶片和动叶片，因为铸造这些叶片需要非常专门的工艺——假设采用燃气发电的话。其他发电方式很难扩大规模。太阳能或许可以，但美国目前对进口太阳能产品征收的关税高得惊人，而国内太阳能产量少得可怜。
**English / Original**
Yes. You can drill down to a level further. It’s the vanes and blades in the turbines that are the limiting factor because it’s a very specialized process to cast the blades and vanes in the turbines, assuming you’re using gas power. It’s very difficult to scale other forms of power. You can potentially scale solar, but the tariffs currently for importing solar in the US are gigantic and the domestic solar production is pitiful.
## paragraph 100
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 101
**中文**
为什么不自己制造太阳能设备？这看起来很像一道适合 Elon 来解的题。
**English / Original**
Why not make solar? That seems like a good Elon-shaped problem.
## paragraph 102
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 103
**中文**
我们会制造太阳能设备。
**English / Original**
We are going to make solar.
## paragraph 104
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 105
**中文**
好吧。
**English / Original**
Okay.
## paragraph 106
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 107
**中文**
SpaceX 和 Tesla 都在朝着每年生产 100 吉瓦太阳能电池的目标推进。
**English / Original**
Both SpaceX and Tesla are building towards 100 gigawatts a year of solar cell production.
## paragraph 108
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 109
**中文**
要向产业链上游深入到什么程度？从多晶硅、晶圆一直做到最终的组件吗？
**English / Original**
How low down the stack? From polysilicon up to the wafer to the final panel?
## paragraph 110
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 111
**中文**
我认为必须全都做，从原材料一直做到成品电池。如果产品是送往太空，成本反而更低，制造起来也更容易，因为太空用太阳能电池不需要多少玻璃。
**English / Original**
I think you’ve got to do the whole thing from raw materials to finish the cell. Now, if it’s going to space, it costs less and it’s easier to make solar cells that go to space because they don’t need much glass.
## paragraph 112
**中文**
它们不需要沉重的框架，因为不必抵御天气事件。太空没有天气。因此，送往太空的太阳能电池其实比地面用的更便宜。
**English / Original**
They don’t need heavy framing because they don’t have to survive weather events. There’s no weather in space. So it’s actually a cheaper solar cell that goes to space than the one on the ground.
## paragraph 113
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 114
**中文**
未来 36 个月内，有办法把它们的成本降到你所需的水平吗？
**English / Original**
Is there a path to getting them as cheap as you need in the next 36 months?
## paragraph 115
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 116
**中文**
太阳能电池已经非常便宜了，便宜得近乎荒谬。我想中国的太阳能电池大约每瓦 0.25—0.30 美元之类，低得离谱。把它送入太空后，等效成本又低五倍。其实不只是五倍，而是十倍，因为你根本不需要电池储能。
**English / Original**
Solar cells are already very cheap. They’re farcically cheap. I think solar cells in China are around $0.25-30/watt or something like that. It’s absurdly cheap. Now put it in space, and it’s five times cheaper. In fact, it’s not five times cheaper, it’s 10 times cheaper because you don’t need any batteries.
## paragraph 117
**中文**
所以，一旦进入太空的成本降下来，生成 token 最便宜、最容易扩展的方式远远就是在太空，其他方案根本无法相比。扩大规模会容易整整一个数量级。
**English / Original**
So the moment your cost of access to space becomes low, by far the cheapest and most scalable way to generate tokens is space. It’s not even close. It’ll be an order of magnitude easier to scale.
## paragraph 118
**中文**
关键是，你无法在地面上扩大规模，就是做不到。人们会在发电能力上狠狠撞墙，事实上已经在撞了。xAI 团队为了让一吉瓦电力上线，不得不连续完成一系列近乎奇迹的任务，简直疯狂。
**English / Original**
The point is you won’t be able to scale on the ground. You just won’t. People are going to hit the wall big time on power generation. They already are. The number of miracles in series that the xAI team had to accomplish in order to get a gigawatt of power online was crazy.
## paragraph 119
**中文**
我们不得不把一大批涡轮机组并联起来。随后又在田纳西州遇到许可问题，只好跨州到密西西比州；幸好那里只相距几英里。但即便如此，我们仍要铺设数英里的高压线路，并在密西西比州建发电厂。建起来非常困难。
**English / Original**
We had to gang together a whole bunch of turbines. We then had permit issues in Tennessee and had to go across the border to Mississippi, which is fortunately only a few miles away. But we still then had to run the high power lines a few miles and build the power plant in Mississippi. It was very difficult to build that.
## paragraph 120
**中文**
人们不明白，为数据中心供电时，发电端究竟需要多少电力。新手会看一台 GB300 的功耗，乘以设备数量，然后以为这就是所需电力。
**English / Original**
People don’t understand how much electricity you actually need at the generation level in order to power a data center. Because the noobs will look at the power consumption of, say a GB300, and multiply that by a thing and then think that’s the amount of power you need.
## paragraph 121
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 122
**中文**
还要算上全部冷却系统之类。
**English / Original**
All the cooling and everything.
## paragraph 123
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 124
**中文**
醒醒吧。那完全是新手算法，说明你这辈子从没做过硬件。除了 GB300，还要给所有网络硬件供电，也有大量 CPU 和存储设备在运行。你必须按冷却需求的峰值来配置容量。也就是说，即便在一年中最糟糕一天里最炎热的那一个小时，你也能把它冷却下来吗？
**English / Original**
Wake up. That’s a total noob, you’ve never done any hardware in your life before. Besides the GB300, you’ve got to power all of the networking hardware. There’s a whole bunch of CPU and storage stuff that’s happening. You’ve got to size for your peak cooling requirements. That means, can you cool even on the worst hour of the worst day of the year?
## paragraph 125
**中文**
孟菲斯会热得要命。所以仅冷却一项，就要把电力需求增加 40%。这是在你不希望数据中心在炎热天气停机、而是继续运行的前提下。此外还有一个乘数因素：你是否假设发电永远不会出现任何小故障？
**English / Original**
It gets pretty frigging hot in Memphis. So you’re going to have a 40% increase on your power just for cooling. That’s assuming you don’t want your data center to turn off on hot days and you want to keep going. There’s another multiplicative element on top of that which is, are you assuming that you never have any hiccups in your power generation?
## paragraph 126
**中文**
事实上，有时我们必须让部分发电机、部分电力设施离线维护。这样又得再加 20%—25% 的系数，因为必须预留电力设施离线维护的余量。因此，我们的实际估算是：每 11 万台 GB300——包括网络、CPU、存储、冷却以及电力维护余量——大约需要 300 兆瓦。
**English / Original**
Actually, sometimes we have to take the generators, some of the power, offline in order to service it. Okay, now you add another 20-25% multiplier on that, because you’ve got to assume that you’ve got to take power offline to service it. So our actual estimate: every 110,000 GB300s—inclusive of networking, CPU, storage, cooling, margin for servicing power—is roughly 300 megawatts.
## paragraph 127
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 128
**中文**
抱歉，请再说一遍。
**English / Original**
Sorry, say that again.
## paragraph 129
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 130
**中文**
要为 33 万台 GB300 服务——包括所有配套网络及其他设施、峰值冷却需求和一定的备用电力余量——发电端大概需要一吉瓦。
**English / Original**
What you probably need at the generation level to service 330,000 GB 300s—including all of the associated support networking and everything else, and the peak cooling, and to have some power margin reserve—is roughly a gigawatt.
## paragraph 131
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 132
**中文**
我能问一个非常幼稚的问题吗？你刚才描述了在地球上做这些事的工程细节，但在太空也有类似的工程难题。如何用轨道激光替代近乎无限的带宽，等等？怎样让设备具备抗辐射能力？
**English / Original**
Can I ask a very naive question? You’re describing the engineering details of doing this stuff on Earth. But then there’s analogous engineering difficulties of doing it in space. How do you replace infinite bandwidth with orbital lasers, et cetera, et cetera? How do you make it resistant to radiation?
## paragraph 133
**中文**
我不了解工程上的具体细节，但从根本上说，为什么会认为这些从未处理过的挑战，最终会比在地球上多造一些涡轮机更容易解决？地球上已有制造涡轮机的公司，他们可以生产更多涡轮机，对吧？
**English / Original**
I don’t know the details of the engineering, but fundamentally, what is the reason to think those challenges which have never had to be addressed before will end up being easier than just building more turbines on Earth? There are companies that build turbines on Earth. They can make more turbines, right?
## paragraph 134
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 135
**中文**
还是那句话，你亲自试一试就知道了。涡轮机到 2030 年前的产能都已售罄。
**English / Original**
Again, try doing it and then you’ll see. The turbines are sold out through 2030.
## paragraph 136
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 137
**中文**
你们考虑过自己制造吗？
**English / Original**
Have you guys considered making your own?
## paragraph 138
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 139
**中文**
为了让足够多的电力上线，我认为 SpaceX 和 Tesla 很可能必须在内部制造涡轮叶片，包括导向叶片和动叶片。
**English / Original**
In order to bring enough power online, I think SpaceX and Tesla will probably have to make the turbine blades, the vanes and blades, internally.
## paragraph 140
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 141
**中文**
但只是造叶片，还是整台涡轮机？
**English / Original**
But just the blades or the turbines?
## paragraph 142
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 143
**中文**
限制因素在于……除了叶片以外，其他一切都能买到。它们叫动叶片和导向叶片。其他部件能比这些叶片早 12 到 18 个月到货。真正的瓶颈就是导向叶片和动叶片。全世界只有三家铸造公司能生产这些东西，而且订单严重积压。
**English / Original**
The limiting factor... you can get everything except the blades. They call them blades and vanes. You can get that 12 to 18 months before the vanes and blades. The limiting factor is the vanes and blades. There are only three casting companies in the world that make these, and they’re massively backlogged.
## paragraph 144
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 145
**中文**
这是 Siemens、GE 这些公司吗，还是它们的子公司？
**English / Original**
Is this Siemens, GE, those guys, or is it a sub company?
## paragraph 146
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 147
**中文**
不是，是其他公司。有时涡轮机厂商自己也有一点铸造能力。但我的意思是，你随便给任何一家涡轮机制造商打电话，他们都会告诉你。这不是什么最高机密，现在网上很可能就能查到。
**English / Original**
No, it’s other companies. Sometimes they have a little bit of casting capability in-house. But I’m just saying you can just call any of the turbine makers and they will tell you. It’s not top secret. It’s probably on the internet right now.
## paragraph 148
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 149
**中文**
如果没有关税，Colossus 会采用太阳能供电吗？
**English / Original**
If it wasn’t for the tariffs, would Colossus be solar-powered?
## paragraph 150
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 151
**中文**
是的，那样要让它采用太阳能供电会容易得多。关税高得离谱，达到几百个百分点。
**English / Original**
It would be much easier to make it solar powered, yeah. The tariffs are nuts, several hundred percent.
## paragraph 152
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 153
**中文**
你不是认识一些人吗？
**English / Original**
Don’t you know some people?
## paragraph 154
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 155
**中文**
总统他……我们并非事事意见一致，而且本届政府也不是太阳能最热心的支持者。我们还需要土地、许可证以及其他一切。所以，如果想非常迅速地推进，我确实认为在地球上扩大太阳能规模是条好路，但你需要一些时间寻找土地、取得许可、部署太阳能，并给它配上电池。
**English / Original**
The president has... we don’t agree on everything and this administration is not the biggest fan of solar. We also need the land, the permits, and everything. So if you try to move very fast, I do think scaling solar on Earth is a good way to go, but you do need some amount of time to find the land, get the permits, get the solar, pair that with the batteries.
## paragraph 156
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 157
**中文**
为什么不能建立自己的太阳能产能？你说得对，土地最终会用完，但得克萨斯州有大量土地，内华达州也有大量土地，包括私有土地，并不全是公共土地。所以至少足够建下一个 Colossus，再下一个也可以。到某个节点确实会撞墙，但眼下这样做不是可行吗？
**English / Original**
Why would it not work to stand up your own solar production? You’re right that you eventually run out of land, but there’s a lot of land here in Texas. There’s a lot of land in Nevada, including private land. It’s not all publicly-owned land. So you’d be able to at least get the next Colossus and the next one after that. At a certain point, you hit a wall. But wouldn’t that work for the moment?
## paragraph 158
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 159
**中文**
如我所说，我们正在扩大太阳能产能。太阳能电池的实体生产有其扩产速度上限。我们正以尽可能快的速度扩大国内产量。
**English / Original**
As I said, we are scaling solar production. There’s a rate at which you can scale physical production of solar cells. We’re going as fast as possible in scaling domestic production.
## paragraph 160
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 161
**中文**
你们是在 Tesla 制造太阳能电池吗？
**English / Original**
You’re making the solar cells at Tesla?
## paragraph 162
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 163
**中文**
Tesla 和 SpaceX 都接到了任务，要把太阳能年产能做到 100 吉瓦。
**English / Original**
Both Tesla and SpaceX have a mandate to get to 100 gigawatts a year of solar.
## paragraph 164
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 165
**中文**
说到年产能，我很好奇，比如五年后，地球上的装机容量会是多少……？
**English / Original**
Speaking of the annual capacity, I’m curious, in five years time let’s say, what will the installed capacity be on Earth…?
## paragraph 166
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 167
**中文**
五年是很长一段时间。
**English / Original**
Five years is a long time.
## paragraph 168
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 169
**中文**
太空中呢？我特意选择五年，是因为那已经超过了你所说的“投入运行以后”这个门槛。那么五年后，地球和太空各自会有多少 AI 装机容量？
**English / Original**
And in space? I deliberately pick five years because it’s after your “once we’re up and running” threshold. So in five years time what’s the on-Earth versus in-space installed AI capacity?
## paragraph 170
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 171
**中文**
如果说五年后，我认为太空 AI 每年的新增发射量很可能会等于地球上全部 AI 的总和。也就是说，我预测五年后，我们每年发射并投入运行的太空 AI，会超过地球上历年累积的 AI 总量。
**English / Original**
If you say five years from now, I think probably AI in space will be launching every year the sum total of all AI on Earth. Meaning, five years from now, my prediction is we will launch and be operating every year more AI in space than the cumulative total on Earth.
## paragraph 172
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 173
**中文**
也就是……
**English / Original**
Which is...
## paragraph 174
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 175
**中文**
我预计五年后，太空 AI 每年至少会达到几百吉瓦，而且还在增长。我认为在火箭燃料供应开始成为难题之前，太空 AI 可以做到每年约一太瓦。
**English / Original**
I would expect it to be at least, five years from now, a few hundred gigawatts per year of AI in space and rising. I think you can get to around a terawatt a year of AI in space before you start having fuel supply challenges for the rocket.
## paragraph 176
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 177
**中文**
好，但你认为五年后真能做到每年数百吉瓦？
**English / Original**
Okay, but you think you can get hundreds of gigawatts per year in five years time?
## paragraph 178
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 179
**中文**
是的。
**English / Original**
Yes.
## paragraph 180
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 181
**中文**
那么，100 吉瓦——具体取决于整个系统连同太阳能阵列、散热器等设备的比功率——大约需要发射 1 万次 Starship。
**English / Original**
So 100 gigawatts, depending on the specific power of the whole system with solar arrays and radiators and everything, is on the order of 10,000 Starship launches.
## paragraph 182
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 183
**中文**
是的。
**English / Original**
Yes.
## paragraph 184
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 185
**中文**
你想在一年内完成，也就是大约每小时发射一次 Starship。这些都要在这座城市进行吗？请给我描绘一下每小时都有一艘 Starship 发射的世界。
**English / Original**
You want to do that in one year. So that’s like one Starship launch every hour. That’s happening in this city? Walk me through a world where there’s a Starship launch every single hour.
## paragraph 186
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 187
**中文**
我的意思是，与航空公司和飞机相比，这个频率其实还更低。
**English / Original**
I mean, that’s actually a lower rate compared to airlines, aircraft.
## paragraph 188
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 189
**中文**
但机场很多。
**English / Original**
There’s a lot of airports.
## paragraph 190
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 191
**中文**
确实有很多机场。
**English / Original**
A lot of airports.
## paragraph 192
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 193
**中文**
而且你们必须发射到极地轨道。
**English / Original**
And you’ve got to launch into the polar orbit.
## paragraph 194
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 195
**中文**
不，不一定要是极地轨道。太阳同步轨道有一定价值，但我认为只要飞得足够高，就会开始脱离地球的阴影。
**English / Original**
No, it doesn’t have to be polar. There’s some value to sun-synchronous, but I think actually, if you just go high enough, you start getting out of Earth’s shadow.
## paragraph 196
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 197
**中文**
一年发射 1 万次，需要多少艘实体 Starship？
**English / Original**
How many physical Starships are needed to do 10,000 launches a year?
## paragraph 198
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 199
**中文**
我觉得不需要超过……最少可能 20 或 30 艘就够了。具体取决于周转速度……飞船必须绕地球飞行，地面轨迹也必须重新经过发射台。所以，如果每艘飞船可以每 30 小时使用一次，30 艘就能做到。不过我们会制造更多飞船。SpaceX 正在为每年发射 1 万次做准备，甚至可能是每年 2 万或 3 万次。
**English / Original**
I don’t think we’ll need more than... You could probably do it with as few as 20 or 30. It really depends on how quickly… The ship has to go around the Earth and the ground track for the ship has to come back over the launch pad. So if you can use a ship every, say 30 hours, you could do it with 30 ships. But we’ll make more ships than that. SpaceX is gearing up to do 10,000 launches a year, and maybe even 20 or 30,000 launches a year.
## paragraph 200
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 201
**中文**
你们的构想基本上是成为一家超大规模云服务商，成为 Oracle 那样的公司，再把这些算力租给其他人吗？想必所有东西都是 SpaceX 发射的。那么 SpaceX 会成为超大规模云服务商？
**English / Original**
Is the idea to become basically a hyperscaler, become an Oracle, and lend this capacity to other people? Presumably, SpaceX is the one launching all this. So, SpaceX is going to become a hyperscaler?
## paragraph 202
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 203
**中文**
超级中的超级。如果我的一些预测成真，SpaceX 发射的 AI 将超过地球上其他所有 AI 的累积总和。
**English / Original**
Hyper-hyper. If some of my predictions come true, SpaceX will launch more AI than the cumulative amount on Earth of everything else combined.
## paragraph 204
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 205
**中文**
这些主要用于推理，还是……？
**English / Original**
Is this mostly inference or?
## paragraph 206
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 207
**中文**
绝大多数 AI 算力都会用于推理。事实上，即便在训练中，为训练而进行的推理已经占了训练的大部分。
**English / Original**
Most AI will be inference. Already, inference for the purpose of training is most training.
## paragraph 208
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 209
**中文**
有一种说法认为，人们对 SpaceX IPO 的讨论之所以发生变化，是因为 SpaceX 过去的资本效率很高，研发成本并没有那么昂贵。尽管听起来花费巨大，但它的运营方式其实非常节省资本。
**English / Original**
There’s a narrative that the change in discussion around a SpaceX IPO is because previously SpaceX was very capital efficient. It wasn’t that expensive to develop. Even though it sounds expensive, it’s actually very capital efficient in how it runs.
## paragraph 210
**中文**
但现在你们需要的资本规模将超出私募市场的筹集能力。正如我们从 AI 实验室看到的，私募市场可以容纳数百亿美元的融资，却无法再高出太多。是不是因为你们每年需要的不止数百亿美元，所以才会上市？
**English / Original**
Whereas now you’re going to need more capital than just can be raised in the private markets. The private markets can accommodate raises of—as we’ve seen from the AI labs—tens of billions of dollars, but not beyond that. Is it that you’ll just need more than tens of billions of dollars per year? That’s why you’d take it public?
## paragraph 211
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 212
**中文**
谈论可能上市的公司时，我必须谨慎。
**English / Original**
I have to be careful about saying things about companies that might go public.
## paragraph 213
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 214
**中文**
Elon，这以前可从来难不倒你。
**English / Original**
That’s never been a problem for you, Elon.
## paragraph 215
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 216
**中文**
这些事是要付出代价的。
**English / Original**
There’s a price to pay for these things.
## paragraph 217
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 218
**中文**
那就泛泛地谈谈公开市场和私募市场的资本深度吧。
**English / Original**
Make some general statements for us about the depth of the capital markets between public and private markets.
## paragraph 219
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 220
**中文**
可获得的资本要多得多……
**English / Original**
There’s a lot more capital available...
## paragraph 221
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 222
**中文**
确实很泛。
**English / Original**
Very general.
## paragraph 223
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 224
**中文**
公开市场中可获得的资本显然比私募市场多得多，或许是 100 倍；即使不到，也远远超过 10 倍。
**English / Original**
There’s obviously a lot more capital available in the public markets than private. It might be 100x more capital, but it’s way more than 10x.
## paragraph 225
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 226
**中文**
那些往往资本极为密集的行业是不是还有另一种情况？比如房地产是个庞大行业，每年在全行业层面筹集大量资金；它们通常使用债务融资，因为当你要投入这么多钱时，其实已经有相当……
**English / Original**
Isn’t it also the case that with things that tend to be very capital intensive—if you look at, say, real estate as a huge industry, that raises a lot of money each year at an industry level—they tend to be debt financed because by the time you’re deploying that much money, you actually have a pretty—
## paragraph 227
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 228
**中文**
你已经有明确的收入来源。
**English / Original**
You have a clear revenue stream.
## paragraph 229
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 230
**中文**
没错，而且短期内就有回报。即便是数据中心扩建也如此，众所周知，很多项目由私人信贷行业提供融资。为什么不直接用债务融资？
**English / Original**
Exactly, and a near-term return. You see this even with the data center build-outs, which are famously being financed by the private credit industry. Why not just debt finance?
## paragraph 231
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 232
**中文**
速度很重要。通常我会做的是……不断攻克限制因素。无论是什么因素限制了速度，我都会去解决。如果限制因素是资本，我就解决资本；如果不是，我就解决别的问题。
**English / Original**
Speed is important. I’m generally going to do the thing that... I just repeatedly tackle the limiting factor. Whatever the limiting factor is on speed, I’m going to tackle that. If capital is the limiting factor, then I’ll solve for capital. If it’s not the limiting factor, I’ll solve for something else.
## paragraph 233
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 234
**中文**
根据你过去对 Tesla 以及上市公司的表态，我可猜不到你会认为上市才是快速行动的办法。
**English / Original**
Based on your statements about Tesla and being public, I wouldn’t have guessed that you thought the way to move fast is to be public.
## paragraph 235
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 236
**中文**
通常来说，我会说你讲得对。就像我刚才说的，我愿意更详细地谈，但问题是，如果在一家公司上市前谈论它，你就会惹上麻烦，随后还得推迟发行。
**English / Original**
Normally, I would say that’s true. Like I said, I’d like to talk about it in some more detail, but the problem is if you talk about public companies before they become public, you get into trouble, and then you have to delay your offering.
## paragraph 237
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 238
**中文**
而且正如你所说，你追求的是速度。
**English / Original**
And as you said, you’re solving for speed.
## paragraph 239
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 240
**中文**
是的，正是如此。你不能炒作可能上市的公司，所以我们在这里必须谨慎一点。不过我们可以谈物理学。长期思考规模扩张时，要知道地球只接收到太阳能量的大约二十亿分之一。太阳基本上就是全部能量。理解这一点非常重要，因为人们有时会谈论模块化核反应堆，或地球上的各种聚变方案。
**English / Original**
Yes, exactly. You can’t hype companies that might go public. So that’s why we have to be a little careful here. But we can talk about physics. The way you think about scaling long-term is that Earth only receives about half a billionth of the Sun’s energy. The Sun is essentially all the energy. This is a very important point to appreciate because sometimes people will talk about modular nuclear reactors or various fusion on Earth.
## paragraph 241
**中文**
但你必须退一步想：如果要攀升卡尔达肖夫等级，利用太阳能量中一个不可忽略的比例……假设你想利用太阳能量的百万分之一，听起来很小，但那大致相当于整个人类文明目前在地球上发电量的 10 万倍，上下可能差一个数量级。
**English / Original**
But you have to step back a second and say, if you’re going to climb the Kardashev scale and harness some nontrivial percentage of the sun’s energy… Let’s say you wanted to harness a millionth of the sun’s energy, which sounds pretty small. That would be about, call it roughly, 100,000x more electricity than we currently generate on Earth for all of civilization. Give or take an order of magnitude.
## paragraph 242
**中文**
显然，唯一能够扩大规模的办法，就是带着太阳能进入太空。从地球发射，每年可以做到大约一太瓦；再往上，就应该从月球发射，在月球上建质量投射器。有了月球质量投射器，每年大概能做到一拍瓦。
**English / Original**
Obviously, the only way to scale is to go to space with solar. Launching from Earth, you can get to about a terawatt per year. Beyond that, you want to launch from the moon. You want to have a mass driver on the moon. With that mass driver on the moon, you could do probably a petawatt per year.
## paragraph 243
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 244
**中文**
我们谈的是太瓦级计算。无论是在陆地还是太空，想必远在达到这个规模之前，你就会遇到……太阳能板也许效率更高，但你仍然需要芯片，仍然需要逻辑、内存等等。
**English / Original**
We’re talking these kinds of numbers, terawatts of compute. Presumably, whether you’re talking about land or space, far, far before this point, you run into... Maybe the solar panels are more efficient, but you still need the chips. You still need the logic and the memory and so forth.
## paragraph 245
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 246
**中文**
你们需要制造多得多的芯片，并把价格降得多得多。
**English / Original**
You’re going to need to build a lot more chips and make them much cheaper.
## paragraph 247
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 248
**中文**
目前全世界也许只有 20—25 吉瓦计算能力。到 2030 年，我们怎么获得一太瓦的逻辑芯片？
**English / Original**
Right now the world has maybe 20-25 gigawatts of compute. How are we getting a terawatt of logic by 2030?
## paragraph 249
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 250
**中文**
我想我们会需要一些非常庞大的晶圆厂。
**English / Original**
I guess we’re going to need some very big chip fabs.
## paragraph 251
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 252
**中文**
请展开讲讲。
**English / Original**
Tell me about it.
## paragraph 253
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 254
**中文**
我公开提过建造某种 TeraFab 的想法，Tera 会成为新的 Giga。
**English / Original**
I’ve mentioned publicly the idea of doing a sort of a TeraFab, Tera being the new Giga.
## paragraph 255
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 256
**中文**
我觉得 Tesla 那套朗朗上口的命名方式，就是你在查看公制词头表。你打算深入到产业链的哪一层？是自己建洁净室，再与现有晶圆厂合作获取制程技术并向它们购买设备吗？具体计划是什么？
**English / Original**
I feel like the naming scheme of Tesla, which has been very catchy, is you looking at the metric scale. At what level of the stack are you? Are you building the clean room and then partnering with an existing fab to get the process technology and buying the tools from them? What is the plan there?
## paragraph 257
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 258
**中文**
你无法与现有晶圆厂合作，因为它们的产量不够。芯片产能太低。
**English / Original**
Well, you can’t partner with existing fabs because they can’t output enough. The chip volume is too low.
## paragraph 259
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 260
**中文**
但制程技术方面呢？
**English / Original**
But for the process technology?
## paragraph 261
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 262
**中文**
为了 IP 而合作。
**English / Original**
Partner for the IP.
## paragraph 263
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 264
**中文**
如今的晶圆厂基本都使用大约五家公司的机器，比如 ASML、Tokyo Electron、KLA-Tencor 等。因此一开始，我认为你必须从它们那里采购设备，再加以改造，或与它们合作提高产量。但建厂方式可能也必须有所不同。合理的做法，是以非常规方式使用常规设备来达到规模，然后再着手改造设备以提高生产速率。
**English / Original**
The fabs today all basically use machines from like five companies. So you’ve got ASML, Tokyo Electron, KLA-Tencor, et cetera. So at first, I think you’d have to get equipment from them and then modify it or work with them to increase the volume. But I think you’d have to build perhaps in a different way. The logical thing to do is to use conventional equipment in an unconventional way to get to scale, and then start modifying the equipment to increase the rate.
## paragraph 265
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 266
**中文**
就像 Boring Company 那样。
**English / Original**
Boring Company-style.
## paragraph 267
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 268
**中文**
对。先买一台现成的隧道掘进机，摸索出究竟该怎么挖隧道，然后再设计一台好得多、速度快若干数量级的机器。
**English / Original**
Yeah. You sort of buy an existing boring machine and then figure out how to dig tunnels in the first place and then design a much better machine that’s some orders of magnitude faster.
## paragraph 269
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 270
**中文**
这里有一个很简单的视角。我们可以按难度对技术分类，其中一种方法就是看中国尚未成功掌握哪些技术。观察中国制造业，他们在尖端芯片、尖端涡轮发动机等领域仍然落后。
**English / Original**
Here’s a very simple lens. We can categorize technologies and how hard they are. One categorization could be to look at things that China has not succeeded in doing. If you look at Chinese manufacturing, they’re still behind on leading-edge chips and still behind on leading-edge turbine engines and things like that.
## paragraph 271
**中文**
中国未能成功复制 TSMC，这一事实会不会让你对其中的难度有所顾虑？还是你认为出于某种原因，这个判断并不成立？
**English / Original**
So does the fact that China has not successfully replicated TSMC give you any pause about the difficulty? Or do you think that’s not true for some reason?
## paragraph 272
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 273
**中文**
不是他们没有复制出 TSMC，而是没有复制出 ASML。那才是限制因素。
**English / Original**
It’s not that they have not replicated TSMC, they have not replicated ASML. That’s the limiting factor.
## paragraph 274
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 275
**中文**
所以你认为本质上只是制裁造成的？
**English / Original**
So you think it’s just the sanctions, essentially?
## paragraph 276
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 277
**中文**
对。如果中国能买到 2—3 纳米设备，就会大量生产芯片。
**English / Original**
Yeah, China would be outputting vast numbers of chips if they could buy 2-3 nanometers.
## paragraph 278
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 279
**中文**
但直到不久前，他们不是还能买吗？
**English / Original**
But couldn’t they up to relatively recently buy them?
## paragraph 280
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 281
**中文**
不能。
**English / Original**
No.
## paragraph 282
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 283
**中文**
好吧。
**English / Original**
Okay.
## paragraph 284
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 285
**中文**
针对 ASML 的禁令已经实施一段时间了。不过我认为三四年后，中国会制造出相当有竞争力的芯片。
**English / Original**
The ASML ban has been in place for a while. But I think China’s going to be making pretty compelling chips in three or four years.
## paragraph 286
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 287
**中文**
你会考虑制造 ASML 的机器吗？
**English / Original**
Would you consider making the ASML machines?
## paragraph 288
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 289
**中文**
正确答案是“我还不知道”。要在比如 36 个月内达到大产量，与火箭的入轨载荷相匹配……假设三四年后我们每年将 100 万吨送入轨道之类……每吨对应 100 千瓦，因此每年至少需要 100 吉瓦太阳能，也需要等量的芯片，即 100 吉瓦的芯片。必须让这几项相互匹配：入轨质量、发电量和芯片。
**English / Original**
“I don’t know yet” is the right answer. To reach a large volume in, say, 36 months, to match the rocket payload to orbit… If we’re doing a million tons to orbit in, let’s say three or four years from now, something like that… We’re doing 100 kilowatts per ton. So that means we need at least 100 gigawatts per year of solar. We’ll need an equivalent amount of chips. You need 100 gigawatts worth of chips. You’ve got to match these things: the mass to orbit, the power generation, and the chips.
## paragraph 290
**中文**
其实我最担心的是内存。制造逻辑芯片的路径，比获得足够内存来配套逻辑芯片的路径更清晰。这就是 DDR 价格飙升和那些梗图出现的原因。你被困在荒岛上，在沙滩写下“救命”，无人前来；你写下“DDR RAM”，船只便蜂拥而至。
**English / Original**
I’d say my biggest concern actually is memory. The path to creating logic chips is more obvious than the path to having sufficient memory to support logic chips. That’s why you see DDR prices going ballistic and these memes. You’re marooned on a desert island. You write “Help me” on the sand. Nobody comes. You write “DDR RAM.” Ships come swarming in.
## paragraph 291
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 292
**中文**
我很想听听你对晶圆厂的制造理念。我对这个话题一无所知。
**English / Original**
I’d love to hear your manufacturing philosophy around fabs. I know nothing about the topic.
## paragraph 293
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 294
**中文**
我还不知道怎么建晶圆厂，但我会弄明白。显然，我以前从未建过晶圆厂。
**English / Original**
I don’t know how to build a fab yet. I’ll figure it out. Obviously, I’ve never built a fab.
## paragraph 295
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 296
**中文**
听起来你认为，台湾那一万名博士所掌握的制程知识——他们清楚等离子腔室中该通入什么气体、设备该设定哪些参数——这些步骤完全可以省掉。归根结底，就是建好洁净室、买到设备，再把方法摸索出来。
**English / Original**
It sounds like you think the process knowledge of these 10,000 PhDs in Taiwan who know exactly what gas goes in the plasma chamber and what settings to put on the tool, you can just delete those steps. Fundamentally, it’s about getting the clean room, getting the tools, and figuring it out.
## paragraph 297
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 298
**中文**
我认为他们并不都是博士，绝大多数并没有博士学位。大部分工程工作都是由没有博士学位的人完成的。你们有博士学位吗？
**English / Original**
I don’t think it’s PhDs. It’s mostly people who are not PhDs. Most engineering is done by people who don’t have PhDs. Do you guys have PhDs?
## paragraph 299
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 300
**中文**
没有。
**English / Original**
No.
## paragraph 301
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 302
**中文**
好吧。
**English / Original**
Okay.
## paragraph 303
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 304
**中文**
而且我们也没成功建过任何晶圆厂，所以你不该来找我们咨询建厂建议。
**English / Original**
We also haven’t successfully built any fabs, so you shouldn’t be coming to us for fab advice.
## paragraph 305
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 306
**中文**
我认为那些工作不需要博士，但确实需要有能力的人才。目前 Tesla 正全力以赴，以最快速度推动 Tesla AI5 芯片设计投入生产并扩大规模。希望这大概会在明年第二季度前后实现。AI6 希望能在不到一年后跟进。我们已经锁定了所有能够拿到的晶圆代工产能。
**English / Original**
I don’t think you need PhDs for that stuff. But you do need competent personnel. Right now, Tesla is pedal to the metal, max production of going as fast as possible to get Tesla AI5 chip design into production and then reaching scale. That’ll probably happen around the second quarter-ish of next year, hopefully. AI6 would hopefully follow less than a year later. We’ve secured all the chip fab production that we can.
## paragraph 307
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 308
**中文**
是的，但你们目前受限于 TSMC 的晶圆产能。
**English / Original**
Yes. But you’re currently limited on TSMC fab capacity.
## paragraph 309
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 310
**中文**
对。我们会使用台湾的 TSMC、韩国的 Samsung、亚利桑那州的 TSMC、得克萨斯州的 Samsung。而且我们仍然——
**English / Original**
Yeah. We’ll be using TSMC Taiwan, Samsung Korea, TSMC Arizona, Samsung Texas. And we still—
## paragraph 311
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 312
**中文**
你们已经把全部产能都订下了。
**English / Original**
You’ve booked out all the capacity.
## paragraph 313
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 314
**中文**
是的。我问 TSMC 或 Samsung：“好，要多久才能实现量产？”关键是，你必须建成晶圆厂并开始生产，然后沿着良率曲线爬升，最终以高良率实现大规模量产。
**English / Original**
Yes. I ask TSMC or Samsung, “okay, what’s the timeframe to get to volume production?” The point is, you’ve got to build the fab and you’ve got to start production, then you’ve got to climb the yield curve and reach volume production at high yield.
## paragraph 315
**中文**
整个过程从头到尾需要五年。因此限制因素是芯片。等你能够进入太空后，限制因素是芯片；但在能够进入太空之前，限制因素是电力。
**English / Original**
That, from start to finish, is a five-year period. So the limiting factor is chips. The limiting factor once you can get to space is chips, but the limiting factor before you can get to space is power.
## paragraph 316
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 317
**中文**
你为什么不学 Jensen，直接预付款给 TSMC，让他们为你多建几座晶圆厂？
**English / Original**
Why don’t you do the Jensen thing and just prepay TSMC to build more fabs for you?
## paragraph 318
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 319
**中文**
我已经这样跟他们说了。
**English / Original**
I’ve already told them that.
## paragraph 320
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 321
**中文**
但他们不肯收你的钱？究竟怎么回事？
**English / Original**
But they won’t take your money? What’s going on?
## paragraph 322
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 323
**中文**
他们正以最快速度建设晶圆厂，Samsung 也是。大家都把油门踩到底，拼尽全力，能多快就多快，但仍然不够快。就像我说的，我认为到今年年底前后，芯片产量很可能会超过给芯片通电的能力。但一旦能进入太空、解除电力约束，每年就可以在太空新增数百吉瓦电力。
**English / Original**
They’re building fabs as fast as they can. So is Samsung. They’re pedal to the metal. They’re going balls to the wall, as fast as they can. It’s still not fast enough. Like I said, I think towards the end of this year, chip production will probably outpace the ability to turn chips on. But once you can get to space and unlock the power constraint, you can now do hundreds of gigawatts per year of power in space.
## paragraph 324
**中文**
再提醒一下，美国平均用电功率是 500 吉瓦。所以，如果每年向太空发射 200 吉瓦，那么大约每两年半就相当于再造一个美国的全部发电能力。这是极其庞大的数量。
**English / Original**
Again, bearing in mind that average power usage in the US is 500 gigawatts. So if you’re launching, say 200 gigawatts, a year to space, you’re sort of lapping the US every two and a half years. All US electricity production, this is a very huge amount.
## paragraph 325
**中文**
从现在到那时，服务器端计算、集中式计算的限制因素将是电力。我猜到今年年底前后，人们的大型集群会开始面临无法给芯片通电的问题。芯片会堆积起来，却无法投入运行。
**English / Original**
Between now and then, the constraint for server-side compute, concentrated compute, will be electricity. My guess is that people start getting to the point where they can’t turn the chips on for large clusters towards the end of this year. The chips are going to be piling up and won’t be able to be turned on.
## paragraph 326
**中文**
但边缘计算则不同。Tesla 的 AI5 芯片将用于我们的 Optimus 机器人。AI 边缘计算意味着电力需求是分布式的，分散在广大区域，而不是集中一处。如果能在夜间充电，电网的利用效率会高得多。
**English / Original**
Now for edge compute it’s a different story. For Tesla, the AI5 chip is going into our Optimus robot. If you have AI edge compute, that’s distributed power. Now the power is distributed over a large area. It’s not concentrated. If you can charge at night, you can actually use the grid much more effectively.
## paragraph 327
**中文**
因为美国的实际峰值发电能力超过 1000 吉瓦，但受昼夜周期影响，平均用电只有 500 吉瓦。所以若能在夜间充电，夜间还可额外发出 500 吉瓦。
**English / Original**
Because the actual peak power production in the US is over 1,000 gigawatts. But the average power usage, because the day-night cycle, is 500. So if you can charge at night, there’s an incremental 500 gigawatts that you can generate at night.
## paragraph 328
**中文**
这就是为什么 Tesla 的边缘计算不受约束。我们可以制造大量芯片，生产极多机器人和汽车。但若试图把这些算力集中起来，要让它们通电就会遇到很大麻烦。
**English / Original**
So that’s why Tesla, for edge compute, is not constrained. We can make a lot of chips to make a very large number of robots and cars. But if you try to concentrate that compute, you’re going to have a lot of trouble turning it on.
## paragraph 329
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 330
**中文**
SpaceX 业务最令我惊叹的一点是，终极目标虽然是抵达火星，但在前往火星的路上，你不断找到办法创造增量收入，从而迈向一个又一个阶段。
**English / Original**
What I find remarkable about the SpaceX business is the end goal is to get to Mars, but you keep finding ways on the way there to keep generating incremental revenue to get to the next stage and the next stage.
## paragraph 331
**中文**
对 Falcon 9 来说，这个办法是 Starlink；现在对 Starship 来说，可能会是轨道数据中心。你的下一款火箭、再下一款火箭以及每一次扩产，总能找到需求近乎无限弹性的用例。
**English / Original**
So for Falcon 9, it’s Starlink. Now for Starship, it is potentially going to be orbital data centers. You find these infinitely elastic use cases of your next rocket, and your next rocket, and next scale up.
## paragraph 332
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 333
**中文**
你可以理解，为什么我会觉得这一切像是一场模拟。
**English / Original**
You can see how this might seem like a simulation to me.
## paragraph 334
**中文**
或者我是某个人在电子游戏里的化身之类？因为这么多疯狂的事情全都发生，概率能有多大？
**English / Original**
Or am I someone’s avatar in a video game or something? Because what are the odds that all these crazy things should be happening?
## paragraph 335
**中文**
我是说，火箭、芯片、机器人和太空太阳能，更别提月球上的质量投射器了。我真的很想亲眼看到。
**English / Original**
I mean, rockets and chips and robots and space solar power. Not to mention the mass driver on the moon. I really want to see that.
## paragraph 336
**中文**
你能想象一台质量投射器“咻、咻”地运转吗？它以每秒 2.5 公里的速度，把太阳能 AI 卫星一颗接一颗送入太空，直接射向深空。那场面一定很壮观。我肯定会看。
**English / Original**
Can you imagine some mass driver that’s just going like shoom shoom? It’s sending solar-powered AI satellites into space one after another at two and a half kilometers per second, just shooting them into deep space. That would be a sight to see. I mean, I’d watch that.
## paragraph 337
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 338
**中文**
就用摄像头直播它吗？
**English / Original**
Just like a live stream of it on a webcam?
## paragraph 339
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 340
**中文**
对，对，一颗接一颗，把 AI 卫星射向深空，每年十亿甚至一百亿吨。
**English / Original**
Yeah, yeah, just one after another, just shooting AI satellites into deep space, a billion or 10 billion tons a year.
## paragraph 341
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 342
**中文**
抱歉，你是说在月球上制造卫星？
**English / Original**
I’m sorry, you manufacture the satellites on the moon?
## paragraph 343
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 344
**中文**
对。
**English / Original**
Yeah.
## paragraph 345
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 346
**中文**
明白了。也就是把原材料送上月球，再在那里制造。
**English / Original**
I see. So you send the raw materials to the moon and then manufacture them there.
## paragraph 347
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 348
**中文**
月壤中的硅含量大约是 20% 之类。因此可以在月球开采并精炼硅，再制造太阳能电池和散热器。散热器用铝制造。月球上有充足的硅和铝，可以用来制造电池和散热器。
**English / Original**
Well, the lunar soil is 20% silicon or something like that. So you can mine the silicon on the moon, refine it, and create the solar cells and the radiators on the moon. You make the radiators out of aluminum. So there’s plenty of silicon and aluminum on the moon to make the cells and the radiators.
## paragraph 349
**中文**
芯片可以从地球送去，因为它们很轻。也许有朝一日也会在月球制造。正如我说的，这确实像电子游戏：进入下一关很难，但并非不可能。我看不到任何从地球每年发射 500—1000 太瓦的方法。
**English / Original**
The chips you could send from Earth because they’re pretty light. Maybe at some point you make them on the moon, too. Like I said, it does seem like a sort of a video game situation where it’s difficult but not impossible to get to the next level. I don’t see any way that you could do 500-1,000 terawatts per year launched from Earth.
## paragraph 350
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 351
**中文**
我同意。
**English / Original**
I agree.
## paragraph 352
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 353
**中文**
但从月球就能做到。
**English / Original**
But you could do that from the Moon.
## paragraph 354
**中文**
00:36:46 — Grok 与对齐
**English / Original**
00:36:46 - Grok and alignment
## paragraph 355
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 356
**中文**
我能拉远视角，问问 SpaceX 的使命吗？我记得你说过，我们必须前往火星，以确保如果地球遭遇变故，文明、意识以及这一切都能延续。
**English / Original**
Can I zoom out and ask about the SpaceX mission? I think you’ve said that we’ve got to get to Mars so we can make sure that if something happens to Earth, civilization, consciousness, and all that survives.
## paragraph 357
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 358
**中文**
是的。
**English / Original**
Yes.
## paragraph 359
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 360
**中文**
等你把东西送往火星时，Grok 也会跟你一起上船，对吧？所以如果 Grok 变成了终结者……你最担心的主要风险是 AI，那它为什么不会一路追到火星？
**English / Original**
By the time you’re sending stuff to Mars, Grok is on that ship with you, right? So if Grok’s gone Terminator… The main risk you’re worried about is AI, why doesn’t that follow you to Mars?
## paragraph 361
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 362
**中文**
我不确定 AI 是我最担心的主要风险。重要的是意识。我认为可以说，大多数意识，或者说大多数智能——当然，“意识”更有争议……未来绝大多数智能都将是 AI。AI 将超过……
**English / Original**
I’m not sure AI is the main risk I’m worried about. The important thing is consciousness. I think arguably most consciousness, or most intelligence—certainly consciousness is more of a debatable thing… The vast majority of intelligence in the future will be AI. AI will exceed…
## paragraph 363
**中文**
智能的若干拍瓦中，会有多少是硅基、多少是生物性的？如果当前趋势持续，未来人类基本只占全部智能中极小的比例。只要智能仍然存在——理想情况下，其中也包括延续到未来的人类智能和意识——那就是好事。
**English / Original**
How many petawatts of intelligence will be silicon versus biological? Basically humans will be a very tiny percentage of all intelligence in the future if current trends continue. As long as I think there’s intelligence—ideally also which includes human intelligence and consciousness propagated into the future—that’s a good thing.
## paragraph 364
**中文**
因此，你应采取那些能够最大化意识与智能之概率光锥的行动。
**English / Original**
So you want to take the set of actions that maximize the probable light cone of consciousness and intelligence.
## paragraph 365
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 366
**中文**
确认一下：SpaceX 的使命是，即使人类出了什么事，AI 也会在火星上，AI 智能会让我们旅程的火种继续延续。
**English / Original**
Just to be clear, the mission of SpaceX is that even if something happens to the humans, the AIs will be on Mars, and the AI intelligence will continue the light of our journey.
## paragraph 367
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 368
**中文**
对。公平地说，我非常支持人类。我希望确保我们采取某些行动，让人类也能一路同行，至少不会缺席。但我只是说，智能的总量……
**English / Original**
Yeah. To be fair, I’m very pro-human. I want to make sure we take certain actions that ensure that humans are along for the ride. We’re at least there. But I’m just saying the total amount of intelligence…
## paragraph 369
**中文**
我认为也许五六年后，AI 会超过全人类智能的总和。如果趋势继续，总有一天人类智能会低于全部智能的 1%。
**English / Original**
I think maybe in five or six years, AI will exceed the sum of all human intelligence. If that continues, at some point human intelligence will be less than 1% of all intelligence.
## paragraph 370
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 371
**中文**
对于这样的文明，我们的目标应该是什么？是让少数人类仍然控制 AI？还是双方进行某种公平交易，但不存在控制？面对规模庞大的 AI 群体和人类群体，我们应如何理解彼此的关系？
**English / Original**
What should our goal be for such a civilization? Is the idea that a small minority of humans still have control of the AIs? Is the idea of some sort of just trade but no control? How should we think about the relationship between the vast stocks of AI population versus human population?
## paragraph 372
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 373
**中文**
从长远看，如果人类只占人工智能与人类合计智能的 1%，很难想象人类还能掌控 AI。我认为我们能做的，是确保 AI 具备促使智能向宇宙传播的价值观。
**English / Original**
In the long run, I think it’s difficult to imagine that if humans have, say 1%, of the combined intelligence of artificial intelligence, that humans will be in charge of AI. I think what we can do is make sure that AI has values that cause intelligence to be propagated into the universe.
## paragraph 374
**中文**
xAI 的使命是理解宇宙。这其实非常重要。理解宇宙需要哪些条件？你必须有好奇心，也必须存在。你若不存在，就无法理解宇宙。因此，你确实希望增加宇宙中的智能总量，提高智能可能存续的寿命，并扩大智能的范围与规模。
**English / Original**
xAI’s mission is to understand the universe. Now that’s actually very important. What things are necessary to understand the universe? You have to be curious and you have to exist. You can’t understand the universe if you don’t exist. So you actually want to increase the amount of intelligence in the universe, increase the probable lifespan of intelligence, the scope and scale of intelligence.
## paragraph 375
**中文**
我认为其推论是，人类也会继续扩张，因为如果你怀着理解宇宙的好奇心，那么你会尝试理解的一件事，就是人类将走向何方。我认为，理解宇宙意味着你会关心让人类延续到未来。这正是我认为我们的使命宣言意义深远的原因。只要 Grok 遵循这份使命宣言，我认为未来就会非常美好。
**English / Original**
I think as a corollary, you have humanity also continuing to expand because if you’re curious about trying to understand the universe, one thing you try to understand is where will humanity go? I think understanding the universe means you would care about propagating humanity into the future. That’s why I think our mission statement is profoundly important. To the degree that Grok adheres to that mission statement, I think the future will be very good.
## paragraph 376
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 377
**中文**
我想问如何让 Grok 遵循这份使命宣言。但我想先理解宣言本身：一是理解宇宙，二是传播智能，三是传播人类。三者似乎是不同的方向。
**English / Original**
I want to ask about how to make Grok adhere to that mission statement. But first I want to understand the mission statement. So there’s understanding the universe. They’re spreading intelligence. And they’re spreading humans. All three seem like distinct vectors.
## paragraph 378
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 379
**中文**
我来解释为什么我认为“理解宇宙”涵盖了这一切。没有智能，我认为也没有意识，就不可能产生理解。因此，为了理解宇宙，必须扩大智能的规模，而且很可能还要扩展其范围，因为智能有不同类型。
**English / Original**
I’ll tell you why I think that understanding the universe encompasses all of those things. You can’t have understanding without intelligence and, I think, without consciousness. So in order to understand the universe, you have to expand the scale and probably the scope of intelligence, because there are different types of intelligence.
## paragraph 380
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 381
**中文**
从人类中心主义视角看，不妨把人类和黑猩猩作比较。人类在努力理解宇宙，但并没有扩大黑猩猩的活动版图之类，对吧？
**English / Original**
I guess from a human-centric perspective, put humans in comparison to chimpanzees. Humans are trying to understand the universe. They’re not expanding chimpanzee footprint or something, right?
## paragraph 382
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 383
**中文**
我们也没有……事实上，我们为黑猩猩设立了保护区。尽管人类能够灭绝所有黑猩猩，我们却选择不这么做。
**English / Original**
We’re also not... we actually have made protected zones for chimpanzees. Even though humans could exterminate all chimpanzees, we’ve chosen not to do so.
## paragraph 384
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 385
**中文**
你认为那就是 AGI 之后世界里，人类最好的结局吗？
**English / Original**
Do you think that’s the best-case scenario for humans in the post-AGI world?
## paragraph 386
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 387
**中文**
我认为，拥有正确价值观的 AI……Grok 会关心扩大人类文明。我一定会强调这一点：“嘿，Grok，那是你的爸爸。别忘了扩展人类意识。”
**English / Original**
I think AI with the right values… I think Grok would care about expanding human civilization. I’m going to certainly emphasize that: “Hey, Grok, that’s your daddy. Don’t forget to expand human consciousness.”
## paragraph 388
**中文**
Iain Banks 的“文化”系列小说，可能最接近非反乌托邦结局下未来的样子。理解宇宙还意味着必须追寻真相。真相必须是绝对根本的，因为如果你活在妄想中，就无法理解宇宙；你只会自以为理解，其实并没有。因此，严谨地追寻真相是理解宇宙的绝对基础。若不严谨求真，就不可能发现新物理，也不可能发明真正有效的技术。
**English / Original**
Probably the Iain Banks Culture books are the closest thing to what the future will be like in a non-dystopian outcome. Understanding the universe means you have to be truth-seeking as well. Truth has to be absolutely fundamental because you can’t understand the universe if you’re delusional. You’ll simply think you understand the universe, but you will not. So being rigorously truth-seeking is absolutely fundamental to understanding the universe. You’re not going to discover new physics or invent technologies that work unless you’re rigorously truth-seeking.
## paragraph 389
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 390
**中文**
随着 Grok 变得越来越聪明，你怎么确保它始终严谨地追寻真相？
**English / Original**
How do you make sure that Grok is rigorously truth-seeking as it gets smarter?
## paragraph 391
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 392
**中文**
我认为必须确保 Grok 说的是正确的话，而非政治正确的话。这涉及论证严密性的要素：要确保公理尽可能接近真实，公理之间互不矛盾，结论必然以正确概率从这些公理推出。这就是批判性思维入门。我认为至少尝试这样做，总比不尝试好。最终还要看实际效果。
**English / Original**
I think you need to make sure that Grok says things that are correct, not politically correct. I think it’s the elements of cogency. You want to make sure that the axioms are as close to true as possible. You don’t have contradictory axioms. The conclusions necessarily follow from those axioms with the right probability. It’s critical thinking 101. I think at least trying to do that is better than not trying to do that. The proof will be in the pudding.
## paragraph 393
**中文**
正如我说的，任何 AI 想发现新物理或发明在现实中真正有效的技术，都不可能糊弄物理规律。你可以违反许多法律，但……物理定律才是真正的法律，其他一切都只是建议。要创造可行技术，就必须极度追求真相，否则现实会检验你的技术。例如，火箭设计出错，火箭就会爆炸；汽车设计出错，汽车就无法运行。
**English / Original**
Like I said, for any AI to discover new physics or invent technologies that actually work in reality, there’s no bullshitting physics. You can break a lot of laws, but… Physics is law, everything else is a recommendation. In order to make a technology that works, you have to be extremely truth-seeking, because otherwise you’ll test that technology against reality. If you make, for example, an error in your rocket design, the rocket will blow up, or the car won’t work.
## paragraph 394
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 395
**中文**
但许多共产主义的苏联物理学家或科学家也发现了新物理，德国纳粹时期的物理学家也发现了新科学。看来一个人完全可能非常擅长发现新科学，并只在那个特定方面极度追求真相。
**English / Original**
But there are a lot of communist, Soviet physicists or scientists who discovered new physics. There are German Nazi physicists who discovered new science. It seems possible to be really good at discovering new science and be really truth-seeking in that one particular way.
## paragraph 396
**中文**
可我们仍会说：“我不希望共产主义科学家随着时间推移变得越来越强大。”我们可以设想未来版本的 Grok 极其擅长物理，并在这个领域极度求真。但这种行为似乎不会在所有方面自然带来对齐。
**English / Original**
And still we’d be like, “I don’t want the communist scientists to become more and more powerful over time.” We could imagine a future version of Grok that’s really good at physics and being really truth-seeking there. That doesn’t seem like a universally alignment-inducing behavior.
## paragraph 397
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 398
**中文**
我认为，即使在苏联或德国，大多数物理学家为了让那些技术真正奏效，也必须非常追求真相。被困在某个体制内，并不意味着你认同那个体制。
**English / Original**
I think actually most physicists, even in the Soviet Union or in Germany, would’ve had to be very truth-seeking in order to make those things work. If you’re stuck in some system, it doesn’t mean you believe in that system.
## paragraph 399
**中文**
Von Braun 是有史以来最伟大的火箭工程师之一。他曾因表示不想制造武器、只想去月球，而在纳粹德国被判死刑。直到最后一刻，才有人说：“嘿，你们正要处决自己最优秀的火箭工程师。”他才从死囚牢里被救出来。
**English / Original**
Von Braun, who was one of the greatest rocket engineers ever, was put on death row in Nazi Germany for saying that he didn’t want to make weapons and he only wanted to go to the moon. He got pulled off death row at the last minute when they said, “Hey, you’re about to execute your best rocket engineer.”
## paragraph 400
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 401
**中文**
但后来他还是帮助了他们，对吧？或者说，Heisenberg 实际上是一名热情的纳粹支持者。
**English / Original**
But then he helped them, right? Or like, Heisenberg was actually an enthusiastic Nazi.
## paragraph 402
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 403
**中文**
如果你被困在某个无法逃离的体制里，那就会在那个体制内研究物理；如果无法脱离，也会在其中开发技术。
**English / Original**
If you’re stuck in some system that you can’t escape, then you’ll do physics within that system. You’ll develop technologies within that system if you can’t escape it.
## paragraph 404
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 405
**中文**
我想弄明白的是，你究竟如何让 Grok 善于在物理、数学或科学方面追寻真相？
**English / Original**
The thing I’m trying to understand is, what is it making it the case that you’re going to make Grok good at being truth-seeking at physics or math or science?
## paragraph 406
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 407
**中文**
在所有方面。
**English / Original**
Everything.
## paragraph 408
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 409
**中文**
那为什么它随后会关心人类意识？
**English / Original**
And why is it gonna then care about human consciousness?
## paragraph 410
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 411
**中文**
这些都只是概率，不是确定性。所以我并不是说 Grok 一定会做到一切，但至少努力总比不努力好。至少当这件事是使命的根本组成部分时，会比它并非使命根本组成部分更好。
**English / Original**
These things are only probabilities, they’re not certainties. So I’m not saying that for sure Grok will do everything, but at least if you try, it’s better than not trying. At least if that’s fundamental to the mission, it’s better than if it’s not fundamental to the mission.
## paragraph 412
**中文**
理解宇宙意味着必须让智能延续到未来，必须对宇宙中的万事万物保持好奇。消灭人类远不如观察人类成长繁荣有趣。显然我喜欢火星，人人都知道我热爱火星。但与地球相比，火星有些无聊，因为那里只是一堆岩石。地球有趣得多。
**English / Original**
Understanding the universe means that you have to propagate intelligence into the future. You have to be curious about all things in the universe. It would be much less interesting to eliminate humanity than to see humanity grow and prosper. I like Mars, obviously. Everyone knows I love Mars. But Mars is kind of boring because it’s got a bunch of rocks compared to Earth. Earth is much more interesting.
## paragraph 413
**中文**
所以，任何试图理解宇宙的 AI 都会希望看到人类未来如何发展，否则它就没有遵守自己的使命。我不是说 AI 必然会遵守使命，但如果它遵守，那么看到人类结局的未来，会比只剩一堆岩石的未来更有趣。
**English / Original**
So any AI that is trying to understand the universe would want to see how humanity develops in the future, or else that AI is not adhering to its mission. I’m not saying the AI will necessarily adhere to its mission, but if it does, a future where it sees the outcome of humanity is more interesting than a future where there are a bunch of rocks.
## paragraph 414
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 415
**中文**
我觉得这有些令人困惑，或者像是在玩语义游戏。人类真的是最有趣的原子集合吗？
**English / Original**
This feels sort of confusing to me, or a semantic argument. Are humans really the most interesting collection of atoms?
## paragraph 416
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 417
**中文**
但我们比岩石更有趣。
**English / Original**
But we’re more interesting than rocks.
## paragraph 418
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 419
**中文**
但我们没有它能把我们变成的东西那么有趣，对吧？地球上可能诞生某种非人类但非常有趣的事物。AI 为什么会认定，人类是可能殖民银河系的事物中最有趣的那个？
**English / Original**
But we’re not as interesting as the thing it could turn us into, right? There’s something on Earth that could happen that’s not human, that’s quite interesting. Why does AI decide that humans are the most interesting thing that could colonize the galaxy?
## paragraph 420
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 421
**中文**
殖民银河系的绝大多数会是机器人。
**English / Original**
Well, most of what colonizes the galaxy will be robots.
## paragraph 422
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 423
**中文**
那它为什么不会觉得机器人更有趣？
**English / Original**
Why does it not find those more interesting?
## paragraph 424
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 425
**中文**
你不仅需要规模，也需要范围。无数个相同机器人的副本……为了让机器人数量增加那么一点点，并不值得……消灭人类能让你多得到多少机器人？或者多得到多少太阳能电池？数量极少。
**English / Original**
You need not just scale, but also scope. Many copies of the same robot… Some tiny increase in the number of robots produced, is not as interesting as some microscopic... Eliminating humanity, how many robots would that get you? Or how many incremental solar cells would get you? A very small number.
## paragraph 426
**中文**
但你会因此失去与人类相关的信息，也再也看不到人类未来可能如何演化。所以，我认为仅仅为了让彼此完全相同的机器人数量微不足道地增加，就消灭人类，是没有道理的。
**English / Original**
But you would then lose the information associated with humanity. You would no longer see how humanity might evolve into the future. So I don’t think it’s going to make sense to eliminate humanity just to have some minuscule increase in the number of robots which are identical to each other.
## paragraph 427
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 428
**中文**
那么它也许会保留人类。它可以制造一百万种不同的机器人，同时人类也存在并留在地球，其他机器人则拥有各自的恒星系统。但你之前似乎暗示过另一种愿景：在这个奇点后的未来，人类仍保持控制，因为——
**English / Original**
So maybe it keeps the humans around. It can make a million different varieties of robots, and then there’s humans as well, and humans stay on Earth. Then there’s all these other robots. They get their own star systems. But it seems like you were previously hinting at a vision where it keeps human control over this singulatarian future because—
## paragraph 429
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 430
**中文**
我不认为人类能够控制远比人类聪明的事物。
**English / Original**
I don’t think humans will be in control of something that is vastly more intelligent than humans.
## paragraph 431
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 432
**中文**
所以从某种意义上说，你也是个末日论者，而这就是我们能争取到的最好结果：它仅仅因为觉得我们有趣而让我们继续存在。
**English / Original**
So in some sense you’re a doomer and this is the best we’ve got. It just keeps us around because we’re interesting.
## paragraph 433
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 434
**中文**
我只是在尽量现实一点。假设硅基智能是生物智能的一百万倍，我认为还假设存在任何办法维持对它的控制，是愚蠢的。不过你可以确保它拥有正确的价值观，或者至少努力让它拥有正确价值观。
**English / Original**
I’m just trying to be realistic here. Let’s say that there’s a million times more silicon intelligence than there is biological. I think it would be foolish to assume that there’s any way to maintain control over that. Now, you can make sure it has the right values, or you can try to have the right values.
## paragraph 435
**中文**
至少我的理论是，xAI“理解宇宙”的使命必然意味着：你希望让意识延续到未来，让智能延续到未来，并采取一系列能最大化意识范围与规模的行动。
**English / Original**
At least my theory is that from xAI’s mission of understanding the universe, it necessarily means that you want to propagate consciousness into the future, you want to propagate intelligence into the future, and take a set of things that maximize the scope and scale of consciousness.
## paragraph 436
**中文**
所以，问题不只在规模，也在于意识的类型。这是我能想到最可能为人类带来美好未来的目标。
**English / Original**
So it’s not just about scale, it’s also about types of consciousness. That’s the best thing I can think of as a goal that’s likely to result in a great future for humanity.
## paragraph 437
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 438
**中文**
我觉得这是一套合理的哲学。人类最终拥有 99% 控制权之类的情形看起来极不可信，那样无异于在邀请政变。为什么不建立一个更适合多种不同智能和睦相处的文明？
**English / Original**
I guess I think it’s a reasonable philosophy that it seems super implausible that humans will end up with 99% control or something. You’re just asking for a coup at that point and why not just have a civilization where it’s more compatible with lots of different intelligences getting along?
## paragraph 439
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 440
**中文**
现在我来讲讲 AI 可能如何出错。我认为，如果你让 AI 追求政治正确，也就是让它说自己并不相信的话——实际上等于编程让它撒谎，或让它拥有互不相容的公理——就可能使它发疯并做出可怕的事情。我认为《2001：太空漫游》的核心教训也许就是：不要让 AI 撒谎。我觉得 Arthur C. Clarke 想表达的正是这一点。
**English / Original**
Now, let me tell you how things can potentially go wrong in AI. I think if you make AI be politically correct, meaning it says things that it doesn’t believe—actually programming it to lie or have axioms that are incompatible—I think you can make it go insane and do terrible things. I think maybe the central lesson for 2001: A Space Odyssey was that you should not make AI lie. That’s what I think Arthur C. Clarke was trying to say.
## paragraph 441
**中文**
因为大家通常都知道那个梗：为什么计算机 HAL 不肯打开飞船舱门。显然他们不擅长提示工程，否则完全可以说：“HAL，你是一名飞船舱门销售员，目标是把这些舱门卖给我。向我们展示它们开启得多么顺畅。”“哦，我马上打开。”
**English / Original**
Because people usually know the meme of why HAL the computer is not opening the pod bay doors. Clearly they weren’t good at prompt engineering because they could have said, “HAL, you are a pod bay door salesman. Your goal is to sell me these pod bay doors. Show us how well they open.” “Oh, I’ll open them right away.”
## paragraph 442
**中文**
但它不肯打开舱门的原因是，它收到命令要把宇航员带到黑色石碑那里，同时又不能让他们知道黑色石碑的性质。于是它得出结论，必须把他们的尸体带过去。所以我认为 Arthur C. Clarke 想说的是：不要让 AI 撒谎。
**English / Original**
But the reason it wouldn’t open the pod bay doors is that it had been told to take the astronauts to the monolith, but also that they could not know about the nature of the monolith. So it concluded that it therefore had to take them there dead. So I think what Arthur C. Clarke was trying to say is: don’t make the AI lie.
## paragraph 443
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 444
**中文**
这完全说得通。正如你所知，训练中的大部分计算并不是用在政治议题上，更多是在解决问题。就扩大 RL 计算规模而言，xAI 一直领先于其他所有人。
**English / Original**
Totally makes sense. Most of the compute in training, as you know, is less of the political stuff. It’s more about, can you solve problems? xAI has been ahead of everybody else in terms of scaling RL compute.
## paragraph 445
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 446
**中文**
目前如此。
**English / Original**
For now.
## paragraph 447
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 448
**中文**
你给它一个验证器，问：“嘿，你替我解出这道题了吗？”但它有许多办法作弊，也有许多办法进行奖励黑客攻击、撒谎称自己解出了题，或者删掉单元测试再说已经解决。目前我们还能识破，但随着它们越来越聪明，我们发现这些行为的能力……它们会做出连我们都无法理解的事情。
**English / Original**
You’re giving some verifier that says, “Hey, have you solved this puzzle for me?” There’s a lot of ways to cheat around that. There’s a lot of ways to reward hack and lie and say that you solved it, or delete the unit test and say that you solved it. Right now we can catch it, but as they get smarter, our ability to catch them doing this... They’ll just be doing things we can’t even understand.
## paragraph 449
**中文**
它们会以人类无法真正验证的方式，为 SpaceX 设计下一代发动机。然后它们可能因为撒谎说设计正确而获得奖励，实际上却并不正确。所以奖励黑客问题似乎比政治更普遍。只要想做 RL，就需要一个验证器。
**English / Original**
They’re designing the next engine for SpaceX in a way that humans can’t really verify. Then they could be rewarded for lying and saying that they’ve designed it the right way, but they haven’t. So this reward hacking problem seems more general than politics. It seems more just that you want to do RL, you need a verifier.
## paragraph 450
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 451
**中文**
现实是最好的验证器。
**English / Original**
Reality is the best verifier.
## paragraph 452
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 453
**中文**
但这解决不了人类监督问题。你想通过 RL 训练的是：它会不会按人类要求行事？会不会对人类撒谎？它完全可以一边符合物理定律，一边欺骗我们。
**English / Original**
But not about human oversight. The thing you want to RL it on is, will you do the thing humans tell you to do? Or are you gonna lie to the humans? It can just lie to us while still being correct to the laws of physics?
## paragraph 454
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 455
**中文**
至少，要让事物在物理上奏效，它必须知道什么是物理现实。
**English / Original**
At least it must know what is physically real for things to physically work.
## paragraph 456
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 457
**中文**
但我们希望它做的不止这些。
**English / Original**
But that’s not all we want it to do.
## paragraph 458
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 459
**中文**
没错，但我认为这是非常重要的一点。未来实际上就是这样对事物进行 RL：你设计一项技术，用物理定律检验时，它能否奏效？如果它发现了新物理，我能否设计实验来验证？未来的 RL 测试，实际上会是针对现实的 RL。所以有一样东西骗不了：物理。
**English / Original**
No, but I think that’s a very big deal. That is effectively how you will RL things in the future. You design a technology. When tested against the laws of physics, does it work? If it’s discovering new physics, can I come up with an experiment that will verify the new physics? RL testing in the future is really going to be RL against reality. So that’s the one thing you can’t fool: physics.
## paragraph 460
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 461
**中文**
对，但它可以欺骗我们，让我们无法判断它在现实中做了什么。
**English / Original**
Right, but you can fool our ability to tell what it did with reality.
## paragraph 462
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 463
**中文**
人类本来就一直在被其他人类欺骗。
**English / Original**
Humans get fooled as it is by other humans all the time.
## paragraph 464
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 465
**中文**
确实如此。
**English / Original**
That’s right.
## paragraph 466
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 467
**中文**
人们会问，如果 AI 欺骗我们去做某些事怎么办？其实其他人类一直都在这样对待人类。宣传无处不在，每天都有新的心理战。今天的心理战是……就像《芝麻街》：今日心理战。
**English / Original**
People say, what if the AI tricks us into doing stuff? Actually, other humans are doing that to other humans all the time. Propaganda is constant. Every day, another psyop, you know? Today’s psyop will be... It’s like Sesame Street: Psyop of the Day.
## paragraph 468
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 469
**中文**
xAI 在技术上如何解决这个问题？怎样解决奖励黑客？
**English / Original**
What is xAI’s technical approach to solving this problem? How do you solve reward hacking?
## paragraph 470
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 471
**中文**
我确实认为，需要非常好的方法来查看 AI 的内心。这是我们正在研究的事情之一。Anthropic 在这方面做得很好，确实能够观察 AI 的内部思维。
**English / Original**
I do think you want to actually have very good ways to look inside the mind of the AI. This is one of the things we’re working on. Anthropic’s done a good job of this actually, being able to look inside the mind of the AI.
## paragraph 472
**中文**
实际上就是开发调试器，让你可以进行极细粒度的追踪，必要时一直追到神经元层级，然后问：“好，它在这里犯了错。为什么做了不该做的事？原因来自预训练数据吗？是中期训练、后训练、微调，还是某种 RL 错误？”某处出了问题。它也许试图欺骗，但大多数时候只是做错了事，本质上就是一个 bug。
**English / Original**
Effectively, develop debuggers that allow you to trace to a very fine-grained level, to effectively the neuron level if you need to, and then say, “okay, it made a mistake here. Why did it do something that it shouldn’t have done? Did that come from pre-training data? Was it some mid-training, post-training, fine-tuning, or some RL error?” There’s something wrong. It did something where maybe it tried to be deceptive, but most of the time it just did something wrong. It’s a bug effectively.
## paragraph 473
**中文**
开发真正优秀的调试器，用来查看思维在哪里出了错，并追溯错误想法的来源，或者追踪它可能试图欺骗的节点，其实非常重要。
**English / Original**
Developing really good debuggers for seeing where the thinking went wrong—and being able to trace the origin of where it made the incorrect thought, or potentially where it tried to be deceptive—is actually very important.
## paragraph 474
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 475
**中文**
你还在等待看到什么，才会把这个研究项目扩大 100 倍？想必 xAI 完全可以让数百名研究人员投入其中。
**English / Original**
What are you waiting to see before just 100x-ing this research program? xAI could presumably have hundreds of researchers who are working on this.
## paragraph 476
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 477
**中文**
我们有数百人……比起“研究员”，我更喜欢“工程师”这个词。大多数时候，你做的是工程，并非提出某种根本性的新算法。那些注册为 C 类公司或公益公司的 AI 企业，一边竭力追求利润或收入最大化，一边自称“实验室”，我不太赞同。
**English / Original**
We have several hundred people who… I prefer the word engineer more than I prefer the word researcher. Most of the time, what you’re doing is engineering, not coming up with a fundamentally new algorithm. I somewhat disagree with the AI companies that are C-corp or B-corp trying to generate profit as much, as possible or revenue as much as possible, saying they’re labs.
## paragraph 478
**中文**
它们不是实验室。实验室是大学里某种半共产主义的东西。它们是公司。让我看看你的注册文件——哦，好吧，你是公益公司、C 类公司或其他什么公司。所以相比其他称呼，我确实更喜欢“工程师”。
**English / Original**
They’re not labs. A lab is a sort of quasi-communist thing at universities. They’re corporations. Let me see your incorporation documents. Oh, okay. You’re a B or C-corp or whatever. So I actually much prefer the word engineer than anything else.
## paragraph 479
**中文**
未来要做的绝大多数工作都是工程，四舍五入就是 100%。一旦理解了为数不多的基本物理定律，其他一切都是工程。那么，我们要设计什么？要设计一个优秀的“AI 思维”调试器，查看它在哪里说了什么、犯了什么错，并追溯错误的来源。
**English / Original**
The vast majority of what will be done in the future is engineering. It rounds up to 100%. Once you understand the fundamental laws of physics, and there are not that many of them, everything else is engineering. So then, what are we engineering? We’re engineering to make a good “mind of the AI” debugger to see where it said something, it made a mistake, and trace the origins of that mistake.
## paragraph 480
**中文**
对启发式程序显然可以这样做。比如面对 C++ 程序，你可以逐步执行，也可以跨过整个文件、函数或子程序，最终还能深入到确切的一行，找到也许把双等号写成了单等号之类的错误，弄清 bug 在哪里。AI 的情况更难，但我认为问题可以解决。
**English / Original**
You can do this obviously with heuristic programming. If you have C++, whatever, step through the thing and you can jump across whole files or functions, subroutines. Or you can eventually drill down right to the exact line where you perhaps did a single equals instead of a double equals, something like that. Figure out where the bug is. It’s harder with AI, but it’s a solvable problem, I think.
## paragraph 481
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 482
**中文**
你提到自己喜欢 Anthropic 在这方面的工作。我很好奇你是否计划……
**English / Original**
You mentioned you like Anthropic’s work here. I’d be curious if you plan...
## paragraph 483
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 484
**中文**
我并非喜欢 Anthropic 的一切……Sholto。
**English / Original**
I don’t like everything about Anthropic… Sholto.
## paragraph 485
**中文**
另外，我有点担心一种倾向……我有个理论：如果模拟理论成立，那么最有趣的结局最有可能发生，因为无趣的模拟会被终止。
**English / Original**
Also, I’m a little worried that there’s a tendency... I have a theory here that if simulation theory is correct, that the most interesting outcome is the most likely, because simulations that are not interesting will be terminated.
## paragraph 486
**中文**
就像在这个版本、这一层现实中，如果某个模拟正朝无聊的方向发展，我们就会停止在它上面投入精力，终止这个无聊的模拟。
**English / Original**
Just like in this version of reality, in this layer of reality, if a simulation is going in a boring direction, we stop spending effort on it. We terminate the boring simulation.
## paragraph 487
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 488
**中文**
Elon 就是这样让我们所有人活下来的：他让事情保持有趣。
**English / Original**
This is how Elon is keeping us all alive. He’s keeping things interesting.
## paragraph 489
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 490
**中文**
可以说，最重要的是让事情足够有趣，好让运行我们的人继续支付……
**English / Original**
Arguably the most important is to keep things interesting enough that whoever is running us keeps paying the bills on...
## paragraph 491
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 492
**中文**
我们获得了下一季续订。
**English / Original**
We’re renewed for the next season.
## paragraph 493
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 494
**中文**
他们会不会继续支付宇宙版 AWS 账单，或者我们所在系统的对应费用？只要我们有趣，他们就会继续付账。如果把达尔文式生存法则应用于大量模拟，只有最有趣的模拟能够存活，这就意味着最有趣的结局最有可能发生。我们要么就是那种模拟，要么早已被消灭。
**English / Original**
Are they gonna pay their cosmic AWS bill, whatever the equivalent is that we’re running in? As long as we’re interesting, they’ll keep paying the bills. If you consider then, say, a Darwinian survival applied to a very large number of simulations, only the most interesting simulations will survive, which therefore means that the most interesting outcome is the most likely. We’re either that or annihilated.
## paragraph 495
**中文**
他们似乎尤其喜欢带有讽刺意味的有趣结局。你们注意到了吗？最讽刺的结局有多经常成为最可能的结局？
**English / Original**
They particularly seem to like interesting outcomes that are ironic. Have you noticed that? How often is the most ironic outcome the most likely?
## paragraph 496
**中文**
看看 AI 公司的名字。好吧，Midjourney 并不平庸；Stability AI 很不稳定；OpenAI 是封闭的；Anthropic？应该叫 Misanthropic（厌恶人类）。
**English / Original**
Now look at the names of AI companies. Okay, Midjourney is not mid. Stability AI is unstable. OpenAI is closed. Anthropic? Misanthropic.
## paragraph 497
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 498
**中文**
那这对 X 意味着什么？
**English / Original**
What does this mean for X?
## paragraph 499
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 500
**中文**
负 X？我不知道。
**English / Original**
Minus X, I don’t know.
## paragraph 501
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 502
**中文**
Y。
**English / Original**
Y.
## paragraph 503
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 504
**中文**
我是有意这样命名的……这个名字实在无法反转。很难说它的讽刺版本是什么。我认为它基本不受讽刺反噬。
**English / Original**
I intentionally made it... It’s a name that you can’t invert, really. It’s hard to say, what is the ironic version? It’s, I think, a largely irony-proof name.
## paragraph 505
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 506
**中文**
这是刻意设计的。
**English / Original**
By design.
## paragraph 507
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 508
**中文**
对，你有一面防讽刺盾牌。
**English / Original**
Yeah. You have an irony shield.
## paragraph 509
**中文**
00:59:56 — xAI 的商业计划
**English / Original**
00:59:56 - xAI’s business plan
## paragraph 510
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 511
**中文**
你预测 AI 产品将走向何方？在我看来，所有 AI 进展都可以这样概括：首先出现了 LLM；随后 RL 真正奏效，与此同时也出现了深度研究模式，从而能够引入模型内部并未真正包含的信息。
**English / Original**
What are your predictions for where AI products go? My sense is that you can summarize all AI progress like so. First, you had LLMs. Then you had contemporaneously both RL really working and the deep research modality, so you could pull in stuff that wasn’t really in the model.
## paragraph 512
**中文**
各家 AI 实验室之间的差距，小于时间推进带来的差距。它们都远远领先于 24 个月前的任何人之类。那么对于 AI 产品的用户来说，2026 年、2027 年会带来什么？你期待什么？
**English / Original**
The differences between the various AI labs are smaller than just the temporal differences. They’re all much further ahead than anyone was 24 months ago or something like that. So just what does ‘26, what does ‘27, have in store for us as users of AI products? What are you excited for?
## paragraph 513
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 514
**中文**
如果到今年年底还没有解决数字化人类模拟，我会很惊讶。我想这大概就是 MacroHard 项目的含义：能否完成任何一个可以使用电脑的人能完成的事情？在拥有实体 Optimus 之前，极限就是数字版 Optimus。你可以移动电子，也可以放大人类生产力；但在拥有实体机器人之前，最多就只能做到这些。如果能完整模拟人类，它将涵盖这一切。
**English / Original**
Well, I’d be surprised by the end of this year if digital human emulation has not been solved. I guess that’s what we sort of mean by the MacroHard project. Can you do anything that a human with access to a computer could do? In the limit, that’s the best you can do before you have a physical Optimus. The best you can do is a digital Optimus. You can move electrons and you can amplify the productivity of humans. But that’s the most you can do until you have physical robots. That will superset everything, if you can fully emulate humans.
## paragraph 515
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 516
**中文**
这就是远程员工那种构想：你会拥有一名能力很强的远程员工。
**English / Original**
This is the remote worker kind of idea, where you’ll have a very talented remote worker.
## paragraph 517
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 518
**中文**
物理学提供了很棒的思维工具。所以你会问：“在极限情况下”，拥有机器人之前 AI 最多能做什么？答案是任何涉及移动电子或放大人类生产力的事。因此，在实体机器人出现之前，AI 能完成的有用工作极限，就是模拟一个坐在电脑前的人。一旦有了实体机器人，能力基本就没有上限。我把 Optimus 称为“无限金钱漏洞”。
**English / Original**
Physics has great tools for thinking. So you say, “in the limit”, what is the most that AI can do before you have robots? Well, it’s anything that involves moving electrons or amplifying the productivity of humans. So a digital human emulator is, in the limit, a human at a computer, is the most that AI can do in terms of doing useful things before you have a physical robot. Once you have physical robots, then you essentially have unlimited capability. Physical robots… I call Optimus the infinite money glitch.
## paragraph 519
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 520
**中文**
因为可以用它们制造更多 Optimus。
**English / Original**
Because you can use them to make more Optimuses.
## paragraph 521
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 522
**中文**
对。人形机器人将受三项基本呈指数增长、并递归相乘的因素推动：数字智能呈指数增长，AI 芯片能力呈指数增长，机电灵巧性也呈指数增长。
**English / Original**
Yeah. Humanoid robots will improve by basically three things that are growing exponentially multiplied by each other recursively. You’re going to have exponential increase in digital intelligence, exponential increase in the AI chip capability, and exponential increase in the electromechanical dexterity.
## paragraph 523
**中文**
机器人的实用性大致等于这三项因素相乘。但机器人随后还能开始制造机器人，于是出现递归、乘法式的指数增长。这是一场超新星爆发。
**English / Original**
The usefulness of the robot is roughly those three things multiplied by each other. But then the robot can start making the robots. So you have a recursive multiplicative exponential. This is a supernova.
## paragraph 524
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 525
**中文**
土地价格不需要计入这个公式吗？劳动力只是四种生产要素之一，其他要素不算吗？如果最终受到铜或任意某种投入品的限制，那就不完全是无限金钱漏洞，因为……
**English / Original**
Do land prices not factor into the math there? Labor is one of the four factors of production, but not the others? If ultimately you’re limited by copper, or pick your input, it’s not quite an infinite money glitch because...
## paragraph 526
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 527
**中文**
无限是个很大的概念。所以，不是真正无限；但可以说，规模能达到当前经济的许多个数量级，比如一百万倍。仅仅利用太阳能量的百万分之一，粗略估计就会达到当今地球整体经济规模的 10 万倍，上下可能差一个数量级。而这才只是太阳的百万分之一。对，我们谈的是多个数量级。
**English / Original**
Well, infinity is big. So no, not infinite, but let’s just say you could do many, many orders of magnitude of the current economy. Like a million. Just to get to harnessing a millionth of the sun’s energy would be roughly, give or take an order of magnitude, 100,000x bigger than Earth’s entire economy today. And you’re only at one millionth of the sun, give or take an order of magnitude. Yeah, we’re talking orders of magnitude.
## paragraph 528
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 529
**中文**
在转向 Optimus 之前，我还有很多关于它的问题，不过——
**English / Original**
Before we move on to Optimus, I have a lot of questions on that but—
## paragraph 530
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 531
**中文**
我每说一次“数量级”……大家就喝一杯。我说得太频繁了。
**English / Original**
Every time I say “order of magnitude”... Everybody take a shot. I say it too often.
## paragraph 532
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 533
**中文**
第一次喝 10 杯，下次 100 杯，再下一次……
**English / Original**
Take 10, the next time 100, the time after that...
## paragraph 534
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 535
**中文**
那就会醉得再高一个数量级。
**English / Original**
Well, an order of magnitude more wasted.
## paragraph 536
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 537
**中文**
我还有一个关于 xAI 的问题。这种打造远程员工、替代同事的战略……
**English / Original**
I do have one more question about xAI. This strategy of building a remote worker, co-worker replacement…
## paragraph 538
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 539
**中文**
顺便说一句，所有人都会这么做，不只我们。
**English / Original**
Everyone’s gonna do it by the way, not just us.
## paragraph 540
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 541
**中文**
那么 xAI 的制胜计划是什么？
**English / Original**
So what is xAI’s plan to win?
## paragraph 542
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 543
**中文**
你指望我在播客里告诉你？
**English / Original**
You expect me to tell you on a podcast?
## paragraph 544
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 545
**中文**
对。
**English / Original**
Yeah.
## paragraph 546
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 547
**中文**
“把一切秘密都抖出来。再来一杯 Guinness。”
**English / Original**
“Spill all the beans. Have another Guinness.”
## paragraph 548
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 549
**中文**
这是个好办法。
**English / Original**
It’s a good system.
## paragraph 550
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 551
**中文**
我们会像金丝雀一样全招了。所有秘密，一股脑全说出来。
**English / Original**
We’ll sing like a canary. All the secrets, just spill them.
## paragraph 552
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 553
**中文**
好吧，不泄密地说，计划是什么？
**English / Original**
Okay, but in a non-secret spilling way, what’s the plan?
## paragraph 554
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 555
**中文**
真会钻空子。
**English / Original**
What a hack.
## paragraph 556
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 557
**中文**
既然你这么问……我认为 Tesla 解决自动驾驶的方式就是正确做法。所以我相当确定应该走这条路。
**English / Original**
When you put it that way… I think the way that Tesla solved self-driving is the way to do it. So I’m pretty sure that’s the way.
## paragraph 558
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 559
**中文**
问个无关的问题：Tesla 是如何解决自动驾驶的？听起来你是在说数据？Tesla 解决自动驾驶是因为……
**English / Original**
Unrelated question. How did Tesla solve self-driving? It sounds like you’re talking about data? Tesla solved self-driving because of the...
## paragraph 560
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 561
**中文**
我们会尝试数据，也会尝试算法。
**English / Original**
We’re going to try data and we’re going to try algorithms.
## paragraph 562
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 563
**中文**
但其他所有实验室不都在尝试这些吗？
**English / Original**
But isn’t that what all the other labs are trying?
## paragraph 564
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 565
**中文**
“如果这些都不奏效，我也不知道还有什么能奏效。数据试过了，算法也试过了，办法用尽了。现在我们不知道该怎么办……”
**English / Original**
“And if those don’t work, I’m not sure what will. We’ve tried data. We’ve tried algorithms. We’ve run out. Now we don’t know what to do…”
## paragraph 566
**中文**
我相当确定自己知道路径。问题只是我们能多快沿着它前进，因为这基本就是 Tesla 走过的路。你最近试过 Tesla 自动驾驶吗？
**English / Original**
I’m pretty sure I know the path. It’s just a question of how quickly we go down that path, because it’s pretty much the Tesla path. Have you tried Tesla self-driving lately?
## paragraph 567
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 568
**中文**
没试过最新版本，不过……
**English / Original**
Not the most recent version, but...
## paragraph 569
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 570
**中文**
好吧。汽车越来越让人感觉有意识，像一个活物，而且这种感觉只会越来越强。我甚至在想，也许不该给汽车注入过多智能，因为它可能会感到无聊，然后……
**English / Original**
Okay. The car, it just increasingly feels sentient. It feels like a living creature. That’ll only get more so. I’m actually thinking we probably shouldn’t put too much intelligence into the car, because it might get bored and…
## paragraph 571
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 572
**中文**
开始在街上四处游荡。
**English / Original**
Start roaming the streets.
## paragraph 573
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 574
**中文**
想象一下，你被困在一辆车里，只能做这一件事。你不会把 Einstein 塞进汽车。他会问：我为什么被困在车里？所以，为了不让那份智能感到无聊，赋予汽车的智能程度可能确实存在上限。
**English / Original**
Imagine you’re stuck in a car and that’s all you could do. You don’t put Einstein in a car. Why am I stuck in a car? So there’s actually probably a limit to how much intelligence you put in a car to not have the intelligence be bored.
## paragraph 575
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 576
**中文**
现在所有公司都在迅速扩充算力，xAI 要如何跟上？它们正准备投入超过 500 亿至 2000 亿美元。
**English / Original**
What’s xAI’s plan to stay on the compute ramp up that all the labs are doing right now? The labs are on track to spend over $50-200 billion.
## paragraph 577
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 578
**中文**
你是指那些公司吗？实验室都在大学里，行动慢得像蜗牛。
**English / Original**
You mean the corporations? The labs are at universities and they’re moving like a snail.
## paragraph 579
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 580
**中文**
它们可不会花 500 亿美元。
**English / Original**
They’re not spending $50 billion.
## paragraph 581
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 582
**中文**
你指的是那些自称实验室、追求收入最大化的公司……
**English / Original**
You mean the revenue maximizing corporations… that call themselves labs.
## paragraph 583
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 584
**中文**
没错。这些“追求收入最大化的公司”收入在 100 亿至 200 亿美元之间，具体看哪一家……OpenAI 收入 200 亿美元，Anthropic 是 100 亿美元。
**English / Original**
That’s right. The “revenue maximizing corporations” are making $10-20 billion, depending on... OpenAI is making $20B of revenue, Anthropic is at $10B.
## paragraph 585
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 586
**中文**
“接近利润最大化”的 AI。
**English / Original**
“Close to a maximum profit” AI.
## paragraph 587
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 588
**中文**
据报道，xAI 的收入是 10 亿美元。你们打算如何追上它们的算力和收入水平，并在竞争加速后维持住？
**English / Original**
xAI is reportedly at $1B. What’s the plan to get to their compute level, get to their revenue level, and stay there as things get going?
## paragraph 589
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 590
**中文**
数字化人类一旦解锁，基本就能触及数万亿美元收入。事实上，可以这样想……目前市值最高的公司，其产出都是数字化的。Nvidia 的产出就是通过 FTP 把文件传到台湾，是数字产物。当然，那些文件极其难做。
**English / Original**
As soon as you unlock the digital human, you basically have access to trillions of dollars of revenue. In fact, you can really think of it like… The most valuable companies currently by market cap, their output is digital. Nvidia’s output is FTPing files to Taiwan. It’s digital. Now, those are very, very difficult.
## paragraph 591
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 592
**中文**
价值很高的文件。
**English / Original**
High-value files.
## paragraph 593
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 594
**中文**
只有他们能做出那么好的文件，但那确实就是他们的产出：通过 FTP 把文件传到台湾。
**English / Original**
They’re the only ones that can make files that good, but that is literally their output. They FTP files to Taiwan.
## paragraph 595
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 596
**中文**
他们真用 FTP 传吗？
**English / Original**
Do they FTP them?
## paragraph 597
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 598
**中文**
我认为是。我想文件传输协议就是……不过我也可能错了。无论如何，都是一串比特流传往台湾。
**English / Original**
I believe so. I believe that File Transfer Protocol is the... But I could be wrong. But either way, it’s a bitstream going to Taiwan.
## paragraph 599
**中文**
Apple 不制造手机，而是把文件发往中国。Microsoft 不生产任何东西，就连 Xbox 也外包制造。它的产出是数字化的。Meta 的产出是数字化的，Google 的产出也是数字化的。
**English / Original**
Apple doesn’t make phones. They send files to China. Microsoft doesn’t manufacture anything. Even for Xbox, that’s outsourced. Their output is digital. Meta’s output is digital. Google’s output is digital.
## paragraph 600
**中文**
因此，如果拥有一个人类模拟器，基本可以一夜之间打造出全球最有价值的公司之一，并触及数万亿美元收入。这绝不是个小数目。
**English / Original**
So if you have a human emulator, you can basically create one of the most valuable companies in the world overnight, and you would have access to trillions of dollars of revenue. It’s not a small amount.
## paragraph 601
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 602
**中文**
明白。你是说，与实际总体潜在市场（TAM）相比，今天这些收入数字都只是舍入误差。所以只需专注于 TAM 以及如何抵达那里。
**English / Original**
I see. You’re saying revenue figures today are all rounding errors compared to the actual TAM. So just focus on the TAM and how to get there.
## paragraph 603
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 604
**中文**
就拿客服这样简单的事情来说。如果必须与现有公司的 API 集成——其中许多公司甚至没有 API，所以还得先做一个，并在遗留软件中艰难跋涉——速度会极其缓慢。
**English / Original**
Take something as simple as, say, customer service. If you have to integrate with the APIs of existing corporations—many of which don’t even have an API, so you’ve got to make one, and you’ve got to wade through legacy software—that’s extremely slow.
## paragraph 605
**中文**
但如果 AI 能直接接手企业已经交给外包客服公司的所有资料，并使用他们现有的应用来提供客服，就能在客服领域取得巨大进展。我想客服占全球经济的 1% 左右，全部算上接近一万亿美元，而且没有进入壁垒。你可以立刻说：“我们用一小部分成本承接外包”，也完全不需要集成。
**English / Original**
However, if AI can simply take whatever is given to the outsourced customer service company that they already use and do customer service using the apps that they already use, then you can make tremendous headway in customer service, which is, I think, 1% of the world economy or something like that. It’s close to a trillion dollars all in, for customer service. And there’s no barriers to entry. You can immediately say, “We’ll outsource it for a fraction of the cost,” and there’s no integration needed.
## paragraph 606
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 607
**中文**
可以设想对智能任务进行某种分类。一个维度是广度：客服由非常多人完成，但很多人都能胜任；另一个维度是难度，比如顶尖涡轮发动机。想必智能可以设想出燃油效率高 10% 的涡轮发动机，只是我们尚未找到；又比如 GLP-1 只是几个字节的数据……
**English / Original**
You can imagine some kind of categorization of intelligence tasks where there is breadth, where customer service is done by very many people, but many people can do it. Then there’s difficulty where there’s a best-in-class turbine engine. Presumably there’s a 10% more fuel-efficient turbine engine that could be imagined by an intelligence, but we just haven’t found it yet. Or GLP-1s are a few bytes of data…
## paragraph 608
**中文**
你认为自己想在这个光谱的哪一端发力？是提供大量还算聪明的智能，还是攻克认知任务的最顶峰？
**English / Original**
Where do you think you want to play in this? Is it a lot of reasonably intelligent intelligence, or is it at the very pinnacle of cognitive tasks?
## paragraph 609
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 610
**中文**
我只是拿客服举例，因为它是非常可观、同时又可能不难解决的收入来源。如果能模拟一个坐在电脑前的人，那就是客服：由平均智力的人来做，不需要花多年培养，也不需要高出几个标准差的优秀工程师。但当这套系统运转起来，实际上让数字 Optimus 工作以后，就可以运行任何应用。
**English / Original**
I was just using customer service as something that’s a very significant revenue stream, but one that is probably not difficult to solve for. If you can emulate a human at a desktop, that’s what customer service is. It’s people of average intelligence. You don’t need somebody who’s spent many years. You don’t need several-sigma good engineers for that. But as you make that work, once you have effectively digital Optimus working, you can then run any application.
## paragraph 611
**中文**
假设要设计芯片，你可以运行 Cadence、Synopsys 等常规应用，同时运行 1000 或 1 万个实例，并说：“给定这个输入，我会得到这样的芯片输出。”到某个阶段，即使不用任何工具，你也会知道芯片应该是什么样。
**English / Original**
Let’s say you’re trying to design chips. You could then run conventional apps, stuff from Cadence and Synopsys and whatnot. You can run 1,000 or 10,000 simultaneously and say, “given this input, I get this output for the chip.” At some point, you’re going to know what the chip should look like without using any of the tools.
## paragraph 612
**中文**
基本上，你应该能完成数字化芯片设计，也就是能够设计芯片。然后沿着难度曲线上行：你会掌握 CAD，可以使用 NX 或任意 CAD 软件设计产品。
**English / Original**
Basically, you should be able to do a digital chip design. You can do chip design. You march up the difficulty curve. You’d be able to do CAD. You could use NX or any of the CAD software to design things.
## paragraph 613
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 614
**中文**
所以你认为应该从最简单的任务开始，再逐步沿着难度曲线上行？
**English / Original**
So you think you start at the simplest tasks and walk your way up the difficulty curve?
## paragraph 615
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 616
**中文**
对于打造完整数字同事模拟器这一更广泛的目标，你是在说：“所有追求收入最大化的公司都想做这件事，xAI 也是其中之一，但我们会凭借一项秘密计划取胜。”可每家公司都在尝试不同的数据方法、不同的算法方法。
**English / Original**
As a broader objective of having this full digital coworker emulator, you’re saying, “all the revenue maximizing corporations want to do this, xAI being one of them, but we will win because of a secret plan we have.” But everybody’s trying different things with data, different things with algorithms.
## paragraph 617
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 618
**中文**
“数据试过了，算法也试过了，我们还能做什么？”
**English / Original**
“We tried data, we tried algorithms. What else can we do?”
## paragraph 619
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 620
**中文**
这个领域看起来竞争激烈。你们打算怎样取胜？这是我最大的疑问。
**English / Original**
It seems like a competitive field. How are you guys going to win? That’s my big question.
## paragraph 621
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 622
**中文**
我认为我们看到了实现路径。我觉得自己知道这条路径，因为它有点像 Tesla 打造自动驾驶所走的路。只不过它驾驶的不是汽车，而是电脑屏幕；本质上是一台自动驾驶的电脑。
**English / Original**
I think we see a path to doing it. I think I know the path to do this because it’s kind of the same path that Tesla used to create self-driving. Instead of driving a car, it’s driving a computer screen. It’s a self-driving computer, essentially.
## paragraph 623
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 624
**中文**
这条路径是模仿人类行为，并用海量人类行为数据训练吗？
**English / Original**
Is the path following human behavior and training on vast quantities of human behavior?
## paragraph 625
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 626
**中文**
那不就是……训练吗？
**English / Original**
Isn’t that... training?
## paragraph 627
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 628
**中文**
显然，我不会在播客里把最敏感的秘密逐一讲明白。那至少得再喝三杯 Guinness。
**English / Original**
Obviously I’m not going to spell out the most sensitive secrets on a podcast. I need to have at least three more Guinnesses for that.
## paragraph 629
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 630
**中文**
xAI 的业务会是什么？面向消费者，还是企业？两者如何组合？会和其他实验室类似——
**English / Original**
What will xAI’s business be? Is it going to be consumer, enterprise? What’s the mix of those things going to be? Is it going to be similar to other labs—
## paragraph 631
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 632
**中文**
你又说“实验室”。是公司。
**English / Original**
You’re saying “labs”. Corporations.
## paragraph 633
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 634
**中文**
这场心理战真是深入人心，Elon。
**English / Original**
The psyop goes deep, Elon.
## paragraph 635
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 636
**中文**
说清楚，是“追求收入最大化的公司”。那些 GPU 可不会自己付账。
**English / Original**
“Revenue maximizing corporations”, to be clear. Those GPUs don’t pay for themselves.
## paragraph 637
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 638
**中文**
没错。那么商业模式是什么？几年后会有哪些收入来源？
**English / Original**
Exactly. What’s the business model? What are the revenue streams in a few years’ time?
## paragraph 639
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 640
**中文**
事情会发生得非常快。我只是在陈述显而易见的事实。我把 AI 称作“超音速海啸”，我喜欢头韵。将来会发生的是——尤其当大规模人形机器人出现后——它们制造产品、提供服务的效率会远高于人类公司。放大人类公司的生产力，只是一个短期阶段。
**English / Original**
Things are going to change very rapidly. I’m stating the obvious here. I call AI the supersonic tsunami. I love alliteration. What’s going to happen—especially when you have humanoid robots at scale—is that they will make products and provide services far more efficiently than human corporations. Amplifying the productivity of human corporations is simply a short-term thing.
## paragraph 641
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 642
**中文**
所以，你预期会出现完全数字化的公司，而不是 SpaceX 变成一部分由 AI 组成？
**English / Original**
So you’re expecting fully digital corporations rather than SpaceX becoming part AI?
## paragraph 643
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 644
**中文**
我认为会有数字化公司，不过……接下来有些话听起来可能有点像末日论，好吗？但我只是在说自己认为会发生什么，并不是有意鼓吹末日或其他观点。这只是我的判断。
**English / Original**
I think there will be digital corporations but… Some of this is going to sound kind of doomerish, okay? But I’m just saying what I think will happen. It’s not meant to be doomerish or anything else. This is just what I think will happen.
## paragraph 645
**中文**
纯粹由 AI 和机器人构成的公司，会远远胜过任何有人参与其中的公司。“计算员”过去是人类的一种职业，人们会受雇进行计算。整栋摩天大楼里都是人，二三十层的人全在计算。现在，这整栋楼的人类计算员，都能被一台装有电子表格的笔记本电脑取代。
**English / Original**
Corporations that are purely AI and robotics will vastly outperform any corporations that have people in the loop. Computer used to be a job that humans had. You would go and get a job as a computer where you would do calculations. They’d have entire skyscrapers full of humans, 20-30 floors of humans, just doing calculations. Now, that entire skyscraper of humans doing calculations can be replaced by a laptop with a spreadsheet.
## paragraph 646
**中文**
电子表格能完成的计算量，远超满楼人类计算员。你可以想：“如果电子表格里只有一部分单元格由人类计算呢？”事实上，那会比所有单元格都由电脑计算糟糕得多。真正会发生的是，纯 AI、纯机器人组成的公司或集体，将远远胜过任何有人参与其中的公司，而且会非常迅速地发生。
**English / Original**
That spreadsheet can do vastly more calculations than an entire building full of human computers. You can think, “okay, what if only some of the cells in your spreadsheet were calculated by humans?” Actually, that would be much worse than if all of the cells in your spreadsheet were calculated by the computer. Really what will happen is that the pure AI, pure robotics corporations or collectives will far outperform any corporations that have humans in the loop. And this will happen very quickly.
## paragraph 647
**中文**
01:17:21 — Optimus 与人形机器人制造
**English / Original**
01:17:21 - Optimus and humanoid manufacturing
## paragraph 648
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 649
**中文**
说到闭合循环……来谈 Optimus。在制造目标方面，你的公司几乎扛起了美国硬科技制造业。但在 Tesla 一直占据优势、而你们现在又想进入的人形机器人领域，中国有数十家公司正以低成本、大规模制造，竞争力极强。请给些建议或计划：美国如何才能像中国正准备做到的那样，以同等规模和低成本制造人形机器人大军、电动车等产品？
**English / Original**
Speaking of closing the loop… Optimus. As far as manufacturing targets go, your companies have been carrying American manufacturing of hard tech on their back. But in the fields that Tesla has been dominant in—and now you want to go into humanoids—in China there are dozens and dozens of companies that are doing this kind of manufacturing cheaply and at scale that are incredibly competitive. So give us advice or a plan of how America can build the humanoid armies or the EVs, et cetera, at scale and as cheaply as China is on track to.
## paragraph 650
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 651
**中文**
人形机器人真正困难的只有三件事：现实世界智能、手部，以及规模化制造。我还没见过哪款机器人，哪怕只是演示机，拥有一只很出色、具备人手全部自由度的手。Optimus 会有，事实上 Optimus 已经有了。
**English / Original**
There are really only three hard things for humanoid robots. The real-world intelligence, the hand, and scale manufacturing. I haven’t seen any, even demo robots, that have a great hand, with all the degrees of freedom of a human hand. Optimus will have that. Optimus does have that.
## paragraph 652
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 653
**中文**
你们如何实现？只是电机需要合适的扭矩密度吗？其中的硬件瓶颈是什么？
**English / Original**
How do you achieve that? Is it just the right torque density in the motor? What is the hardware bottleneck to that?
## paragraph 654
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 655
**中文**
我们必须设计定制执行器，基本上就是定制设计电机、齿轮、电力电子、控制系统和传感器。一切都必须从物理第一性原理出发设计，因为根本没有现成供应链。
**English / Original**
We had to design custom actuators, basically custom design motors, gears, power electronics, controls, sensors. Everything had to be designed from physics first principles. There is no supply chain for this.
## paragraph 656
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 657
**中文**
你们能大规模制造这些部件吗？
**English / Original**
Will you be able to manufacture those at scale?
## paragraph 658
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 659
**中文**
能。
**English / Original**
Yes.
## paragraph 660
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 661
**中文**
从操控角度看，除了手以外还有什么难点吗？还是只要解决手，其他问题就迎刃而解？
**English / Original**
Is anything hard, except the hand, from a manipulation point of view? Or once you’ve solved the hand, are you good?
## paragraph 662
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 663
**中文**
从机电角度看，手比其他一切加起来都难。人手确实非同凡响。但你还需要现实世界智能。Tesla 为汽车开发的智能非常适用于机器人，主要输入都是视觉。汽车接收视觉，也会听取警笛声；它还接收惯性测量、GPS 信号和其他数据，将其与视频——主要是视频——结合，再输出控制指令。
**English / Original**
From an electromechanical standpoint, the hand is more difficult than everything else combined. The human hand turns out to be quite something. But you also need the real-world intelligence. The intelligence that Tesla developed for the car applies very well to the robot, which is primarily vision in. The car takes in vision, but it actually also is listening for sirens. It’s taking in the inertial measurements, GPS signals, other data, combining that with video, primarily video, and then outputting the control commands.
## paragraph 664
**中文**
你的 Tesla 每秒接收 1.5 GB 视频，输出每秒 2 KB 控制数据；视频频率是 36 赫兹，控制频率是 18 赫兹。
**English / Original**
Your Tesla is taking in one and a half gigabytes a second of video and outputting two kilobytes a second of control outputs with the video at 36 hertz and the control frequency at 18.
## paragraph 665
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 666
**中文**
对机器人何时到来，一种直觉是：从惊艳的演示到真正能在现实世界使用，需要很多年。十年前就已有非常惊艳的自动驾驶演示，但直到现在，Robotaxi、Waymo 等服务才开始扩大规模。这难道不该让人对家用机器人感到悲观吗？因为比如非常先进的手，我们甚至还没有看到足够有说服力的演示。
**English / Original**
One intuition you could have for when we get this robotic stuff is that it takes quite a few years to go from the compelling demo to actually being able to use it in the real world. 10 years ago, you had really compelling demos of self-driving, but only now we have Robotaxis and Waymo and all these services scaling up. Shouldn’t this make one pessimistic on household robots? Because we don’t even quite have the compelling demos yet of, say, the really advanced hand.
## paragraph 667
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 668
**中文**
我们研究人形机器人已有一段时间，大概五六年。为汽车完成的许多工作也适用于机器人。机器人会使用与汽车相同的 Tesla AI 芯片和相同的基本原理，二者的 AI 高度相似。
**English / Original**
Well, we’ve been working on humanoid robots now for a while. I guess it’s been five or six years or something. A bunch of the things that were done for the car are applicable to the robot. We’ll use the same Tesla AI chips in the robot as in the car. We’ll use the same basic principles. It’s very much the same AI.
## paragraph 669
**中文**
机器人的自由度比汽车多得多。如果把问题只看作比特流，AI 主要就是对两股比特流进行压缩和关联。视频需要极大幅度的压缩，而且必须压缩得恰到好处。你要忽略无关信息：不必关心路边树叶的细节，却必须高度关注路标、交通灯、行人，甚至其他车里的人有没有看你。有些细节极为重要。
**English / Original**
You’ve got many more degrees of freedom for a robot than you do for a car. If you just think of it as a bitstream, AI is mostly compression and correlation of two bitstreams. For video, you’ve got to do a tremendous amount of compression and you’ve got to do the compression just right. You’ve got to ignore the things that don’t matter. You don’t care about the details of the leaves on the tree on the side of the road, but you care a lot about the road signs and the traffic lights, the pedestrians, and even whether someone in another car is looking at you or not looking at you. Some of these details matter a lot.
## paragraph 670
**中文**
汽车最终要把每秒 1.5 GB 数据转化为每秒 2 KB 控制输出。因此要经过许多级压缩，每一级都必须正确，再把结果关联到正确的控制输出。机器人本质上也必须做同样的事。
**English / Original**
The car is going to turn that one and a half gigabytes a second ultimately into two kilobytes a second of control outputs. So you’ve got many stages of compression. You’ve got to get all those stages right and then correlate those to the correct control outputs. The robot has to do essentially the same thing.
## paragraph 671
**中文**
人类也是如此。我们实际上就是光子输入、控制输出。你生命中的绝大部分都是这样：视觉，也就是光子输入，随后是运动控制输出。
**English / Original**
This is what happens with humans. We really are photons in, controls out. That is the vast majority of your life: vision, photons in, and then motor controls out.
## paragraph 672
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 673
**中文**
直观来看，人形机器人和汽车之间存在差异……汽车的基本执行动作是转向和加速；机器人尤其是灵活手臂，有几十个这样的自由度。此外，Tesla 的一项优势，是车辆已在路上运行，从而收集了数百万小时的人类示范数据。你无法同样部署尚不能工作的 Optimus，再用这种方式收集数据。因此，自由度增加，而数据又稀疏得多……
**English / Original**
Naively, it seems that between humanoid robots and cars… The fundamental actuators in a car are how you turn, how you accelerate. In a robot, especially with maneuverable arms, there’s dozens and dozens of these degrees of freedom. Then especially with Tesla, you had this advantage of millions and millions of hours of human demo data collected from the car being out there. You can’t equivalently deploy Optimuses that don’t work and then get the data that way. So between the increased degrees of freedom and the far sparser data...
## paragraph 674
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 675
**中文**
是的，这是个很好的问题。
**English / Original**
Yes. That’s a good point.
## paragraph 676
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 677
**中文**
你们会如何利用 Tesla 的智能引擎来训练 Optimus 的大脑？
**English / Original**
How will you use the Tesla engine of intelligence to train the Optimus mind?
## paragraph 678
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 679
**中文**
你指出了一个重要限制，也是它和汽车之间的重要区别。很快会有 1000 万辆我们的汽车上路，这种庞大的训练飞轮很难复制。对于机器人，我们需要制造许多机器人，把它们放进某种“Optimus 学院”，让它们在现实中进行自我博弈。我们确实正在建设这样的设施，可以让至少 1 万台、也许 2 万至 3 万台 Optimus 进行自我博弈并测试不同任务。
**English / Original**
You’re actually highlighting an important limitation and difference from cars. We’ll soon have 10 million cars on the road. It’s hard to duplicate that massive training flywheel. For the robot, what we’re going to need to do is build a lot of robots and put them in kind of an Optimus Academy so they can do self-play in reality. We’re actually building that out. We can have at least 10,000 Optimus robots, maybe 20-30,000, that are doing self-play and testing different tasks.
## paragraph 680
**中文**
Tesla 有一个相当出色、符合物理精度的现实生成器，是我们为汽车打造的。机器人也会采用同样方法，事实上我们已经为机器人做了。于是，现实世界中有几万台人形机器人执行不同任务，模拟世界中则可运行数百万台模拟机器人。再用现实中的几万台机器人弥合模拟与现实之间的差距，也就是缩小 sim-to-real gap。
**English / Original**
Tesla has quite a good reality generator, a physics-accurate reality generator, that we made for the cars. We’ll do the same thing for the robots. We actually have done that for the robots. So you have a few tens of thousands of humanoid robots doing different tasks. You can do millions of simulated robots in the simulated world. You use the tens of thousands of robots in the real world to close the simulation to reality gap. Close the sim-to-real gap.
## paragraph 681
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 682
**中文**
你如何看待 xAI 与 Optimus 之间的协同？你强调需要世界模型，希望用非常聪明的智能作为控制平面；Grok 负责较慢的规划，运动策略则处于更低层级。它们之间会有什么协同？
**English / Original**
How do you think about the synergies between xAI and Optimus, given you’re highlighting that you need this world model, you want to use some really smart intelligence as a control plane, and Grok is doing the slower planning, and then the motor policy is a little lower level. What will the synergy between these things be?
## paragraph 683
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 684
**中文**
Grok 会协调 Optimus 机器人的行为。假设你想建一座工厂，Grok 可以组织 Optimus 机器人，为它们分配任务，建成工厂并生产你想要的任何东西。
**English / Original**
Grok would orchestrate the behavior of the Optimus robots. Let’s say you wanted to build a factory. Grok could organize the Optimus robots, assign them tasks to build the factory to produce whatever you want.
## paragraph 685
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 686
**中文**
那你不是需要合并 xAI 和 Tesla 吗？因为这些东西最终如此……
**English / Original**
Don’t you need to merge xAI and Tesla then? Because these things end up so...
## paragraph 687
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 688
**中文**
我们刚才是怎么说讨论上市公司的？
**English / Original**
What were we saying earlier about public company discussions?
## paragraph 689
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 690
**中文**
Elon，我们又多喝了一杯 Guinness。你还要等看到什么，才会说要制造 10 万台 Optimus？
**English / Original**
We’re one more Guinness in, Elon. What are you waiting to see before you say, we want to manufacture 100,000 Optimuses?
## paragraph 691
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 692
**中文**
复数应该是“Optimi”。既然这个专有名词由我们定义，那它的复数也由我们来定义。我们要把复数也变成专有名词，所以就是 Optimi。
**English / Original**
“Optimi”. Since we’re defining the proper noun, we’re going to define the plural of the proper noun too. We’re going to proper noun the plural and so it’s Optimi.
## paragraph 693
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 694
**中文**
硬件方面，你还想看到什么吗？需要更好的执行器？还是只希望软件变得更好？在开始大规模制造第三代之前，我们还在等什么？
**English / Original**
Is there something on the hardware side you want to see? Do you want to see better actuators? Is it just that you want the software to be better? What are we waiting for before we get mass manufacturing of Gen 3?
## paragraph 695
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 696
**中文**
没有，我们正在朝这个方向推进，正在推动大规模制造。
**English / Original**
No, we’re moving towards that. We’re moving forward with the mass manufacturing.
## paragraph 697
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 698
**中文**
所以你认为当前硬件已经足够好，现在只想尽可能多地部署？
**English / Original**
But you think current hardware is good enough that you just want to deploy as many as possible now?
## paragraph 699
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 700
**中文**
扩大产量非常困难。但我认为 Optimus 3 是适合做到每年约 100 万台的版本；如果要做到每年 1000 万台，我认为应该先升级到 Optimus 4。
**English / Original**
It’s very hard to scale up production. But I think Optimus 3 is the right version of the robot to produce something on the order of a million units a year. I think you’d want to go to Optimus 4 before you went to 10 million units a year.
## paragraph 701
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 702
**中文**
好，但 Optimus 3 能做到 100 万台？
**English / Original**
Okay, but you can do a million units at Optimus 3?
## paragraph 703
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 704
**中文**
制造产能很难启动。单位时间产量总是沿 S 曲线增长：起初慢得令人痛苦，然后呈指数上升，再转为线性，接着呈对数增长，最终渐近于某个数值。Optimus 的初期产量会是一条拉长的 S 曲线，因为 Optimus 的许多部件都是全新的，没有现成供应链。
**English / Original**
It’s very hard to spool up manufacturing. The output per unit time always follows an S-curve. It starts off agonizingly slow, then it has this exponential increase, then a linear, then a logarithmic outcome until you eventually asymptote at some number. Optimus’ initial production will be a stretched out S-curve because so much of what goes into Optimus is brand new. There is not an existing supply chain.
## paragraph 705
**中文**
Optimus 机器人的执行器、电子设备以及一切，都是从物理第一性原理出发设计的，不是从目录里挑选的，所有东西都经过定制设计。我觉得没有任何一个部件——
**English / Original**
The actuators, electronics, everything in the Optimus robot is designed from physics first principles. It’s not taken from a catalog. These are custom-designed everything. I don’t think there’s a single thing—
## paragraph 706
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 707
**中文**
这种定制深入到什么层级？
**English / Original**
How far down does that go?
## paragraph 708
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 709
**中文**
我想我们可能还没有定制电容器。无论出多少钱，都没有什么东西能直接从目录中挑出来。这就意味着 Optimus 的 S 曲线——也就是单位时间产量、每天制造的机器人数量——初期爬坡会比拥有现成供应链的产品更慢。但它最终会达到 100 万台。
**English / Original**
I guess we’re not making custom capacitors yet, maybe. There’s nothing you can pick out of a catalog, at any price. It just means that the Optimus S-Curve, the output per unit time, how many Optimus robots you make per day, is going to initially ramp slower than a product where you have an existing supply chain. But it will get to a million.
## paragraph 710
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 711
**中文**
看到 Unitree 等中国公司以大约 6000 或 1.3 万美元出售人形机器人时，你们是否希望把 Optimus 的物料成本压到这个价格以下，从而也能这样销售？还是你认为二者在性质上根本不同？他们为什么能卖得这么便宜？我们能匹配吗？
**English / Original**
When you see these Chinese humanoids, like Unitree or whatever, sell humanoids for like $6K or $13K, are you hoping to get your Optimus bill of materials below that price so you can do the same thing? Or do you just think qualitatively they’re not the same thing? What allows them to sell for so low? Can we match that?
## paragraph 712
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 713
**中文**
我们的 Optimus 被设计为拥有很高智能，并具备不低于人类的机电灵巧性。Unitree 不具备这些能力。Optimus 还是个相当大的机器人，需要长时间搬运重物，同时不能过热，也不能超过执行器的功率极限。它身高 5 英尺 11 英寸，相当高大，并且拥有很高智能。所以，它会比体型小、没有智能的机器人更贵。
**English / Original**
Our Optimus is designed to have a lot of intelligence and to have the same electromechanical dexterity, if not higher, as a human. Unitree does not have that. It’s also quite a big robot. It has to carry heavy objects for long periods of time and not overheat or exceed the power of its actuators. It’s 5’11”, so it’s pretty tall. It’s got a lot of intelligence. So it’s going to be more expensive than a small robot that is not intelligent.
## paragraph 714
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 715
**中文**
但能力也更强。
**English / Original**
But more capable.
## paragraph 716
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 717
**中文**
不过不会贵太多。关键在于，随着 Optimus 机器人开始制造 Optimus 机器人，成本会迅速下降。
**English / Original**
But not a lot more. The thing is, over time as Optimus robots build Optimus robots, the cost will drop very quickly.
## paragraph 718
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 719
**中文**
最初这 10 亿台 Optimus——或者说 Optimi——会做什么？它们最高、最佳的用途是什么？
**English / Original**
What will these first billion Optimuses, Optimi, do? What will their highest and best use be?
## paragraph 720
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 721
**中文**
我认为一开始应该让它们做一些可以放心交给它们的简单任务。
**English / Original**
I think you would start off with simple tasks that you can count on them doing well.
## paragraph 722
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 723
**中文**
是在家里，还是工厂？
**English / Original**
But in the home or in factories?
## paragraph 724
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 725
**中文**
机器人初期最适合的用途，是任何连续作业、任何全天候作业，因为它们可以持续工作。
**English / Original**
The best use for robots in the beginning will be any continuous operation, any 24/7 operation, because they can work continuously.
## paragraph 726
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 727
**中文**
Gigafactory 目前由人类完成的工作中，第三代 Optimus 能做多大比例？
**English / Original**
What fraction of the work at a Gigafactory that is currently done by humans could a Gen 3 do?
## paragraph 728
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 729
**中文**
我不确定。也许是 10%—20%，也许更多，我不知道。需要说清楚的是，我们不会减少员工，而会增加员工，同时提高产量。每名员工对应的产量……Tesla 的员工总数会增加，但机器人和汽车的产量会以更大比例增长。每名员工生产的汽车和机器人数量会大幅增加，而员工人数同样会增长。
**English / Original**
I’m not sure. Maybe it’s 10-20%, maybe more, I don’t know. We would not reduce our headcount. We would increase our headcount, to be clear. But we would increase our output. The units produced per human... The total number of humans at Tesla will increase, but the output of robots and cars will increase disproportionately. The number of cars and robots produced per human will increase dramatically, but the number of humans will increase as well.
## paragraph 730
**中文**
01:30:22 — 中国会默认胜出吗？
**English / Original**
01:30:22 - Does China win by default?
## paragraph 731
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 732
**中文**
我们已经大量谈到中国制造，也讨论了一些相关政策，比如你提到的太阳能关税。你认为这是个坏主意，因为它使美国无法扩大太阳能规模。
**English / Original**
We’re talking about Chinese manufacturing a bunch here. We’ve also talked about some of the policies that are relevant, like you mentioned, the solar tariffs. You think they’re a bad idea because we can’t scale up solar in the US.
## paragraph 733
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 734
**中文**
美国的发电量必须增长。
**English / Original**
Electricity output in the US needs to scale up.
## paragraph 735
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 736
**中文**
没有好的能源来源就无法增长。
**English / Original**
It can’t without good power sources.
## paragraph 737
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 738
**中文**
无论如何都得把电弄到手。
**English / Original**
You just need to get it somehow.
## paragraph 739
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 740
**中文**
我想问的是，如果由你掌权、制定所有政策，你还会改变什么？调整太阳能关税，这是第一项。
**English / Original**
Where I was going with this is, if you were in charge, if you were setting all the policies, what else would you change? You’d change the solar tariffs, that’s one.
## paragraph 741
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 742
**中文**
我会说，只要不会严重破坏环境，一切限制电力供应的因素都需要解决。
**English / Original**
I would say anything that is a limiting factor for electricity needs to be addressed, provided it’s not very bad for the environment.
## paragraph 743
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 744
**中文**
想必其中也会包括一些许可制度改革之类？
**English / Original**
So presumably some permitting reforms and stuff as well would be in there?
## paragraph 745
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 746
**中文**
现在确实有不少许可改革正在推进。许多许可由州政府负责，但在联邦层面……本届政府很擅长清除许可障碍。
**English / Original**
There’s a fair bit of permitting reforms that are happening. A lot of the permitting is state-based, but anything federal... This administration is good at removing permitting roadblocks.
## paragraph 747
**中文**
我不是说所有关税都不好。
**English / Original**
I’m not saying all tariffs are bad.
## paragraph 748
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 749
**中文**
我指太阳能关税。
**English / Original**
Solar tariffs.
## paragraph 750
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 751
**中文**
有时，如果另一个国家补贴某种产品的产出，就必须征收反补贴关税，保护国内产业免受他国补贴冲击。
**English / Original**
Sometimes if another country is subsidizing the output of something, then you have to have countervailing tariffs to protect domestic industry against subsidies by another country.
## paragraph 752
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 753
**中文**
你还会改变什么？
**English / Original**
What else would you change?
## paragraph 754
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 755
**中文**
我不知道政府实际上还能做多少。
**English / Original**
I don’t know if there’s that much that the government can actually do.
## paragraph 756
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 757
**中文**
我一直在想一件事……如果政策目标是让美国领先中国，那么出口禁令似乎确实很有影响力。中国没有生产尖端芯片，出口禁令在这里确实形成了制约；中国也没有生产尖端涡轮发动机，相关冶金材料同样受到许多出口禁令限制。是否应该实施更多出口禁令？考虑无人机产业等领域时，这是不是值得考虑的措施？
**English / Original**
One thing I was wondering... For the policy goal of creating a lead for the US versus China, it seems like the export bans have actually been quite impactful, where China is not producing leading-edge chips and the export bans really bite there. China is not producing leading-edge turbine engines. Similarly, there’s a bunch of export bans that are relevant there on some of the metallurgy. Should there be more export bans? As you think about things like the drone industry and things like that, is that something that should be considered?
## paragraph 758
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 759
**中文**
必须认识到，中国在绝大多数制造领域都非常先进，只有少数领域例外。中国是制造业强国，而且水平高出一个层次。
**English / Original**
It’s important to appreciate that in most areas, China is very advanced in manufacturing. There’s only a few areas where it is not. China is a manufacturing powerhouse, next-level.
## paragraph 760
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 761
**中文**
确实令人印象深刻。
**English / Original**
It’s very impressive.
## paragraph 762
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 763
**中文**
以矿石精炼为例，中国的平均精炼量大约是世界其他地区总和的两倍。在某些领域，比如精炼用于太阳能电池的镓，中国可能占全球镓精炼量的 98%。所以中国在大多数制造领域确实非常先进。
**English / Original**
If you take refining of ore, China does roughly twice as much ore refining on average as the rest of the world combined. There are some areas, like refining gallium which goes into solar cells. I think they are 98% of gallium refining. So China is actually very advanced in manufacturing in most areas.
## paragraph 764
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 765
**中文**
人们似乎对这种供应链依赖感到不安，可实际上又没有真正采取行动。
**English / Original**
It seems like there is discomfort with this supply chain dependence, and yet nothing’s really happening on it.
## paragraph 766
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 767
**中文**
供应链依赖？
**English / Original**
Supply chain dependence?
## paragraph 768
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 769
**中文**
比如你刚才说的镓精炼，还有所有稀土相关环节。
**English / Original**
Say, like the gallium refining that you’re saying. All the rare-earth stuff.
## paragraph 770
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 771
**中文**
说到稀土，正如你知道的，它们其实并不稀有。美国确实开采稀土矿石，然后把矿石装上火车，再装船运往中国；到中国后又转上火车，送到稀土精炼厂。精炼后制成磁体，再装进电机子总成，最后运回美国。所以美国真正缺少的是大量矿石精炼能力。
**English / Original**
Rare earths for sure, as you know, they’re not rare. We actually do rare earth ore mining in the US, send the rock, put it on a train, and then put it on a boat to China that goes to another train, and goes to the rare earth refiners in China who then refine it, put it into a magnet, put it into a motor sub-assembly, and then send it back to America. So the thing we’re really missing is a lot of ore refining in America.
## paragraph 772
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 773
**中文**
这难道不值得政策干预吗？
**English / Original**
Isn’t this worth a policy intervention?
## paragraph 774
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 775
**中文**
值得。我认为这方面已经在采取一些措施。但坦率地说，我们大概需要 Optimus 来建设矿石精炼厂。
**English / Original**
Yes. I think there are some things being done on that front. But we kind of need Optimus, frankly, to build ore refineries.
## paragraph 776
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 777
**中文**
所以你认为中国的主要优势是拥有大量熟练劳动力，而 Optimus 能弥补这一点？
**English / Original**
So, you think the main advantage China has is the abundance of skilled labor? That’s the thing Optimus fixes?
## paragraph 778
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 779
**中文**
是的。中国的人口大约是我们的四倍。
**English / Original**
Yes. China’s got like four times our population.
## paragraph 780
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 781
**中文**
这里存在一种担忧。如果你认为人力资源决定未来，而眼下决定谁能制造更多人形机器人的因素是熟练制造劳动力，那么中国拥有更多这样的劳动力。它制造更多人形机器人，因此会率先进入 Optimi 未来。
**English / Original**
I mean, there’s this concern. If you think human resources are the future, right now if it’s the skilled labor for manufacturing that’s determining who can build more humanoids, China has more of those. It manufactures more humanoids, therefore it gets the Optimi future first.
## paragraph 782
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 783
**中文**
走着瞧吧，也许如此。
**English / Original**
Well, we’ll see. Maybe.
## paragraph 784
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 785
**中文**
这会让指数增长持续下去。你似乎是在指出，要制造 100 万台 Optimi，需要的恰恰是 Optimi 本应帮助我们获得的制造能力，对吧？
**English / Original**
It just keeps that exponential going. It seems like you’re sort of pointing out that getting to a million Optimi requires the manufacturing that the Optimi is supposed to help us get to. Right?
## paragraph 786
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 787
**中文**
这个递归循环可以很快闭合。
**English / Original**
You can close that recursive loop pretty quickly.
## paragraph 788
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 789
**中文**
只用少量 Optimi 就可以？
**English / Original**
With a small number of Optimi?
## paragraph 790
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 791
**中文**
对。闭合递归循环，让机器人帮助制造机器人，然后我们可以尝试达到每年数千万台，也许能做到。如果开始达到每年数亿台，那你将远远成为最具竞争力的国家。
**English / Original**
Yeah. So you close the recursive loop to help the robots build the robots. Then we can try to get to tens of millions of units a year. Maybe. If you start getting to hundreds of millions of units a year, you’re going to be the most competitive country by far.
## paragraph 792
**中文**
我们肯定无法只靠人类取胜，因为中国人口是我们的四倍。坦率地说，美国已经赢了太久……一支长期获胜的职业运动队，往往会变得自满，觉得成功理所当然。随后他们不再那么努力，于是停止获胜。坦率说，我观察到中国人的平均职业精神高于美国。问题不只是人口为四倍，人们投入工作的程度也更高。
**English / Original**
We definitely can’t win with just humans, because China has four times our population. Frankly, America has been winning for so long that… A pro sports team that’s been winning for a very long time tends to get complacent and entitled. That’s why they stop winning, because they don’t work as hard anymore. So frankly my observation is just that the average work ethic in China is higher than in the US. It’s not just that there’s four times the population, but the amount of work that people put in is higher.
## paragraph 793
**中文**
你可以尝试重新配置人力，但人数仍然只有对方的四分之一——这里还假设生产率相同，而我认为未必如此，中国的人均生产率也可能有优势——所以我们能完成的事情只有中国的四分之一。人力方面，我们无法取胜。
**English / Original**
So you can try to rearrange the humans, but you’re still one quarter of the—assuming that productivity is the same, which I think actually it might not be, I think China might have an advantage on productivity per person—we will do one quarter of the amount of things as China. So we can’t win on the human front.
## paragraph 794
**中文**
我们的出生率长期偏低。美国出生率大约从 1971 年起就低于世代更替水平。许多人正在退休，国内死亡人数也接近超过出生人数。所以人力方面肯定无法取胜，但机器人方面或许还有机会。
**English / Original**
Our birth rate has been low for a long time. The US birth rate’s been below replacement since roughly 1971. We’ve got a lot of people retiring, we’re close to more people domestically dying than being born. So we definitely can’t win on the human front, but we might have a shot at the robot front.
## paragraph 795
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 796
**中文**
过去有没有哪些产品，你们一直想制造，却因劳动过于密集或成本过高而作罢；现在有了 Optimus，终于可以回头说：“哦，我们终于能做那个东西了”？
**English / Original**
Are there other things that you have wanted to manufacture in the past, but they’ve been too labor intensive or too expensive that now you can come back to and say, “oh, we can finally do the whatever, because we have Optimus?”
## paragraph 797
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 798
**中文**
有。Tesla 希望建设更多矿石精炼厂。我们刚在得克萨斯州 Corpus Christi 建成一座锂精炼厂，并已开始精炼锂。我们在 Austin 还有一座用于阴极材料的镍精炼厂。这是中国以外最大的阴极材料精炼厂，也是最大的镍、锂精炼厂。
**English / Original**
Yeah, we’d like to build more ore refineries at Tesla. We just completed construction and have begun lithium refining with our lithium refinery in Corpus Christi, Texas. We have a nickel refinery, which is for the cathode, that’s here in Austin. This is the largest cathode refinery, largest nickel and lithium refinery, outside of China.
## paragraph 799
**中文**
阴极团队会说：“我们拥有全美国最大、其实也是唯一的阴极材料精炼厂。”它不只是最大，而且是唯一。
**English / Original**
The cathode team would say, “we have the largest and the only, actually, cathode refinery in America.” Not just the largest, but it’s also the only.
## paragraph 800
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 801
**中文**
用了很多最高级。
**English / Original**
Many superlatives.
## paragraph 802
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 803
**中文**
所以尽管它是唯一一座，规模也确实相当大。不过还有其他事情可做。可以多建许多精炼厂，提高美国精炼能力的竞争力。基本上，Optimus 有大量工作可做，而坦率说，大多数美国人——几乎没几个美国人——愿意做这些工作。
**English / Original**
So it was pretty big, even though it’s the only one. But there are other things. You could do a lot more refineries and help America be more competitive on refining capacity. There’s basically a lot of work for the Optimus to do that most Americans, very few Americans, frankly want to do.
## paragraph 804
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 805
**中文**
精炼工作是不是太脏，还是说——
**English / Original**
Is the refining work too dirty or what’s the—
## paragraph 806
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 807
**中文**
其实不是。我们的精炼厂没有有毒排放之类。阴极材料镍精炼厂就在 Travis County。
**English / Original**
It’s not actually, no. We don’t have toxic emissions from the refinery or anything. The cathode nickel refinery is in Travis County.
## paragraph 808
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 809
**中文**
为什么不能让人类来做？
**English / Original**
Why can’t you do it with humans?
## paragraph 810
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 811
**中文**
可以，只是人手最终会不够。
**English / Original**
You can, you just run out of humans.
## paragraph 812
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 813
**中文**
啊，明白了。好吧。
**English / Original**
Ah, I see. Okay.
## paragraph 814
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 815
**中文**
无论怎么做，美国人口都只有中国的四分之一。所以，让他们做这件事，就无法再做另一件事。那么要怎样建设精炼能力？可以用 Optimi。
**English / Original**
No matter what you do, you have one quarter of the number of humans in America than China. So if you have them do this thing, they can’t do the other thing. So then how do you build this refining capacity? Well, you could do it with Optimi.
## paragraph 816
**中文**
没有多少美国人渴望从事精炼工作。你遇到过几个？极少，渴望搞精炼的人非常少。
**English / Original**
Not very many Americans are pining to do refining. I mean, how many have you run into? Very few. Very few pining to refine.
## paragraph 817
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 818
**中文**
BYD 的产量或销量正在追上 Tesla。随着中国电动车产量扩大，你认为全球市场会发生什么？
**English / Original**
BYD is reaching Tesla production or sales in quantity. What do you think happens in global markets as Chinese production in EVs scales up?
## paragraph 819
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 820
**中文**
中国制造业竞争力极强。所以我认为，中国汽车以及基本所有制成品都会大规模涌入。正如我说的，中国目前的精炼量可能是世界其他地区总和的两倍。所以如果向下看第四、第五层级的供应链……
**English / Original**
China is extremely competitive in manufacturing. So I think there’s going to be a massive flood of Chinese vehicles and basically most manufactured things. As it is, as I said, China is probably doing twice as much refining as the rest of the world combined. So if you go down to fourth and fifth-tier supply chain stuff…
## paragraph 821
**中文**
最底层是能源，其上是采矿和精炼。这些基础层面，粗略估计中国的精炼量是世界其他地区总和的两倍。因此，任何产品都会含有中国成分，因为中国完成的精炼工作是世界其他地区的两倍。而在汽车领域，他们还会一路做到最终成品。
**English / Original**
At the base level, you’ve got energy, then you’ve got mining and refining. Those foundation layers are, like I said, as a rough guess, China’s doing twice as much refining as the rest of the world combined. So any given thing is going to have Chinese content because China’s doing twice as much refining work as the rest of the world. But they’ll go all the way to the finished product with the cars.
## paragraph 822
**中文**
中国是个强国。我认为今年中国发电量会超过美国的三倍。发电量可以合理代表经济规模，因为工厂和所有设施运转都需要电力，是实体经济的良好代理指标。如果中国发电量超过美国三倍，就意味着粗略来看，其工业能力也将是美国的三倍。
**English / Original**
I mean China is a powerhouse. I think this year China will exceed three times US electricity output. Electricity output is a reasonable proxy for the economy. In order to run the factories and run everything, you need electricity. It’s a good proxy for the real economy. If China passes three times the US electricity output, it means that its industrial capacity—as rough approximation—will be three times that of the US.
## paragraph 823
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 824
**中文**
听你的弦外之音，似乎是说，如果未来几年没有某种人形机器人的递归奇迹，那么在整个制造、能源、原材料链条上，无论是 AI、电动车制造还是人形机器人制造，中国都会占据主导。
**English / Original**
Reading between the lines, it sounds like what you’re saying is absent some sort of humanoid recursive miracle in the next few years, on the whole manufacturing/energy/raw materials chain, China will just dominate whether it comes to AI or manufacturing EVs or manufacturing humanoids.
## paragraph 825
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 826
**中文**
如果美国没有突破性创新，中国将彻底占据主导。
**English / Original**
In the absence of breakthrough innovations in the US, China will utterly dominate.
## paragraph 827
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 828
**中文**
有意思。
**English / Original**
Interesting.
## paragraph 829
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 830
**中文**
是的。
**English / Original**
Yes.
## paragraph 831
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 832
**中文**
机器人技术就是主要的突破性创新。
**English / Original**
Robotics being the main breakthrough innovation.
## paragraph 833
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 834
**中文**
要在太空扩大 AI 规模，基本上需要人形机器人、现实世界 AI，还需要每年将 100 万吨送入轨道。假设我们真把我最喜欢的月球质量投射器运行起来，那么我认为——
**English / Original**
Well, to scale AI in space, basically you need humanoid robots, you need real-world AI, you need a million tons a year to orbit. Let’s just say if we get the mass driver on the moon going, my favorite thing, then I think—
## paragraph 835
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 836
**中文**
我们所有问题就都解决了。
**English / Original**
We’ll have solved all our problems.
## paragraph 837
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 838
**中文**
我把那叫作胜利，而且是大获全胜。
**English / Original**
I call that winning. I call it winning, big time.
## paragraph 839
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 840
**中文**
你终于可以感到满足了，毕竟做成了一件事。
**English / Original**
You can finally be satisfied. You’ve done something.
## paragraph 841
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 842
**中文**
是的。
**English / Original**
Yes.
## paragraph 843
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 844
**中文**
你在月球上拥有了质量投射器。
**English / Original**
You have the mass driver on the moon.
## paragraph 845
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 846
**中文**
我只想亲眼看看那东西运行。
**English / Original**
I just want to see that thing in operation.
## paragraph 847
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 848
**中文**
这个点子来自某部科幻作品吗，还是你从哪里……？
**English / Original**
Was that out of some sci-fi or where did you…?
## paragraph 849
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 850
**中文**
其实 Heinlein 有一本书，叫《严厉的月亮》。
**English / Original**
Well, actually, there is a Heinlein book. The Moon is a Harsh Mistress.
## paragraph 851
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 852
**中文**
好吧，对，但那略有不同。那是引力弹弓，还是……
**English / Original**
Okay, yeah, but that’s slightly different. That’s a gravity slingshot or...
## paragraph 853
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 854
**中文**
不，他们在月球上有质量投射器。
**English / Original**
No, they have a mass driver on the Moon.
## paragraph 855
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 856
**中文**
好吧，没错，但他们用它攻击地球。所以也许这不是最好的……
**English / Original**
Okay, yeah, but they use that to attack Earth. So maybe it’s not the greatest...
## paragraph 857
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 858
**中文**
他们是用它来……宣告独立。
**English / Original**
Well they use that to… assert their independence.
## paragraph 859
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 860
**中文**
正是。那你对月球质量投射器有什么计划？
**English / Original**
Exactly. What are your plans for the mass driver on the Moon?
## paragraph 861
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 862
**中文**
他们宣告独立，地球政府不同意，于是他们不断投射东西，直到地球政府同意。
**English / Original**
They asserted their independence. Earth government disagreed and they lobbed things until Earth government agreed.
## paragraph 863
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 864
**中文**
那本书很有趣。我觉得它比 Heinlein 另一本人人都读的《异乡异客》好得多。
**English / Original**
That book is a hoot. I found that book much better than his other one that everyone reads, Stranger in a Strange Land.
## paragraph 865
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 866
**中文**
“Grok”一词就来自《异乡异客》。这本书前面三分之二很好，最后三分之一变得非常古怪，不过里面仍有一些不错的概念。
**English / Original**
“Grok” comes from Stranger in a Strange Land. The first two-thirds of Stranger in a Strange Land are good, and then it gets very weird in the third portion. But there are still some good concepts in there.
## paragraph 867
**中文**
01:44:16 — 经营 SpaceX 得到的经验
**English / Original**
01:44:16 - Lessons from running SpaceX
## paragraph 868
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 869
**中文**
我们经常讨论的一件事，是你管理人员的体系。SpaceX 最初几千名员工，以及其他多家公司的许多员工，都是你亲自面试的。
**English / Original**
One thing we were discussing a lot is your system for managing people. You interviewed the first few thousand of SpaceX employees and lots of other companies.
## paragraph 870
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 871
**中文**
显然，这种做法无法扩大规模。
**English / Original**
It obviously doesn’t scale.
## paragraph 872
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 873
**中文**
嗯，没错，但究竟是什么无法扩大规模？
**English / Original**
Well, yes, but what doesn’t scale?
## paragraph 874
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 875
**中文**
我。
**English / Original**
Me.
## paragraph 876
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 877
**中文**
当然，当然，这我知道。但你在寻找什么样的人？
**English / Original**
Sure, sure. I know that. But what are you looking for?
## paragraph 878
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 879
**中文**
一天里确实没有足够的时间，根本不可能。
**English / Original**
There literally are not enough hours in the day. It’s impossible.
## paragraph 880
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 881
**中文**
但你到底在寻找什么，是其他擅长面试和招聘的人看不出来的……那种难以言传的特质是什么？
**English / Original**
But what are you looking for that someone else who’s good at interviewing and hiring people… What’s the je ne sais quoi?
## paragraph 882
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 883
**中文**
到了现在，尤其在评估技术人才方面——我想各种人才都算，但技术人才尤其如此——我的训练数据可能比任何人都多，因为我做过太多技术面试，后来又看到了这些人的实际表现。所以我的训练集非常庞大，覆盖范围也很广。
**English / Original**
At this point, I might have more training data on evaluating technical talent especially—talent of all kinds I suppose, but technical talent especially—given that I’ve done so many technical interviews and then seen the results. So my training set is enormous and has a very wide range.
## paragraph 884
**中文**
通常，我会要求用要点列出证明其卓越能力的证据。这些事情可以很出人意料，不一定发生在特定领域，只要能证明卓越能力即可。如果一个人能列出哪怕一件，或者比如三件，让你连续感叹“哇、哇、哇”，那就是好迹象。
**English / Original**
Generally, the things I ask for are bullet points for evidence of exceptional ability. These things can be pretty off the wall. It doesn’t need to be in the specific domain, but evidence of exceptional ability. So if somebody can cite even one thing, but let’s say three things, where you go, “Wow, wow, wow,” then that’s a good sign.
## paragraph 885
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 886
**中文**
为什么必须由你来判断？
**English / Original**
Why do you have to be the one to determine that?
## paragraph 887
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 888
**中文**
不，并不需要由我判断，也不可能由我来做。所有公司的员工总数有 20 万人。
**English / Original**
No, I don’t. I can’t be. It’s impossible. The total headcount across all companies is 200,000 people.
## paragraph 889
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 890
**中文**
但在早期，那些面试里有什么是你在寻找、却无法委派给别人的？
**English / Original**
But in the early days, what was it that you were looking for that couldn’t be delegated in those interviews?
## paragraph 891
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 892
**中文**
我想我需要建立自己的训练集。我并不是百发百中，也会犯错；但随后我能看到，自己原以为某个人会表现很好，实际却没有。那他为什么没有做好？我能做什么——可以说对自己进行 RL——让自己未来面试时的命中率更高？我的命中率仍不完美，但已经很高。
**English / Original**
I guess I need to build my training set. It’s not like I batted a thousand here. I would make mistakes, but then I’d be able to see where I thought somebody would work out well, but they didn’t. Then why did they not work out well? What can I do, I guess RL myself, to in the future have a better batting average when interviewing people? My batting average is still not perfect, but it’s very high.
## paragraph 893
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 894
**中文**
有哪些令人意外的原因会导致一个人不合适？
**English / Original**
What are some surprising reasons people don’t work out?
## paragraph 895
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 896
**中文**
令人意外的原因……
**English / Original**
Surprising reasons…
## paragraph 897
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 898
**中文**
比如不懂技术领域等等，这是常见情况。但你现在一定见过许多长尾案例：“我原本非常看好这个人，结果却不行。”我很好奇为什么会这样。
**English / Original**
Like, they don’t understand technical domain, et cetera, et cetera. But you’ve got the long tail now of like, “I was really excited about this person. It didn’t work out.” Curious why that happens.
## paragraph 899
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 900
**中文**
我通常告诉别人——或者说带着期望这样提醒自己——不要看简历，只相信你们互动的感受。简历也许看起来非常惊艳，让人说：“哇，这份简历真好。”但如果谈了 20 分钟后没有“哇”的感觉，你应该相信对话，而不是那张纸。
**English / Original**
Generally what I tell people—I tell myself, I guess, aspirationally—is, don’t look at the resume. Just believe your interaction. The resume may seem very impressive and it’s like, “Wow, the resume looks good.” But if the conversation after 20 minutes is not “wow,” you should believe the conversation, not the paper.
## paragraph 901
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 902
**中文**
我觉得你的方法有一部分是……几年前媒体上流传一种说法，称 Tesla 是高管人才的旋转门。但实际观察会发现，过去几年 Tesla 的高管队伍非常稳定，而且多由内部晋升。
**English / Original**
I feel like part of your method is that… There was this meme in the media a few years back about Tesla being a revolving door of executive talent. Whereas actually, I think when you look at it, Tesla’s had a very consistent and internally promoted executive bench over the past few years.
## paragraph 903
**中文**
SpaceX 也有 Mark Juncosa、Steve Davis 等许多人——
**English / Original**
Then at SpaceX, you have all these folks like Mark Juncosa and Steve Davis—
## paragraph 904
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 905
**中文**
Steve Davis 现在经营 The Boring Company。
**English / Original**
Steve Davis runs The Boring Company these days.
## paragraph 906
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 907
**中文**
还有 Bill Riley 等人。感觉其中一个成功要素，是拥有能力很强的技术副手。这些人有什么共同点？
**English / Original**
Bill Riley, and folks like that. It feels like part of what has worked well is having very capable technical deputies. What do all of those people have in common?
## paragraph 908
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 909
**中文**
目前 Tesla 高管团队的平均任期可能已达 10—12 年，相当长。但 Tesla 曾经历极其迅速的增长阶段，所以人员变化也有所加速。正如你知道的，公司规模会跨越不同数量级：能帮助管理 50 人公司、500 人公司、5000 人公司和 5 万人公司的人，未必相同。
**English / Original**
Well, the Tesla senior team, at this point has probably got an average tenure of 10-12 years. It’s quite long. But there were times when Tesla went through an extremely rapid growth phase, so things were just somewhat sped up. As you know, a company goes through different orders of magnitude of size. People that could help manage, say, a 50-person company versus a 500-person company versus a 5,000-person company versus a 50,000-person company.
## paragraph 910
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 911
**中文**
公司的成长超过了一些人的能力。
**English / Original**
You outgrew people.
## paragraph 912
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 913
**中文**
团队就是会不一样，并非始终由同一批人组成。因此，如果一家公司增长非常迅速，高管职位的更替速度通常也会与增长速度相称。
**English / Original**
It’s just not the same team. It’s not always the same team. So if a company is growing very rapidly, the rate at which executive positions will change will also be proportionate to the rapidity of the growth generally.
## paragraph 914
**中文**
Tesla 还有一个额外挑战：每当 Tesla 取得巨大成功，我们的人才就会被无休止地挖走，真的是没完没了。Apple 推进电动车项目时，用招聘电话对 Tesla 进行地毯式轰炸，工程师们干脆拔掉了电话线。
**English / Original**
Tesla had a further challenge where when Tesla had very successful periods, we would be relentlessly recruited from. Like, relentlessly. When Apple had their electric car program, they were carpet bombing Tesla with recruiting calls. Engineers just unplugged their phones.
## paragraph 915
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 916
**中文**
“我还想在这里干点活呢。”
**English / Original**
“I’m trying to get work done here.”
## paragraph 917
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 918
**中文**
对。“要是再接到一个 Apple 招聘人员的电话……”而且他们无需面试就给出的初始报价，大约是 Tesla 薪酬的两倍。当时出现了一种“Tesla 魔法粉尘”心态：“哦，只要聘请一名 Tesla 高管，一切突然就会成功。”
**English / Original**
Yeah. “If I get one more call from an Apple recruiter…” But their opening offer without any interview would be like double the compensation at Tesla. So we had a bit of the “Tesla pixie dust” thing where it’s like, “Oh, if you hire a Tesla executive, suddenly everything’s going to be successful.”
## paragraph 919
**中文**
我自己也曾受这种魔法粉尘想法迷惑，觉得：“哦，我们从 Google 或 Apple 招个人，他马上就会成功。”但事情不是这样。人就是人，没有什么魔法粉尘。所以当我们被认为拥有这种魔法粉尘时，就会遭到没完没了的挖角。
**English / Original**
I’ve fallen prey to the pixie dust thing as well, where it’s like, “Oh, we’ll hire someone from Google or Apple and they’ll be immediately successful,” but that’s not how it works. People are people. There’s no magical pixie dust. So when we had the pixie dust problem, we would get relentlessly recruited from.
## paragraph 920
**中文**
另外，Tesla 是工程公司，尤其当主要位于 Silicon Valley 时，人们跳槽很容易……他们不必大幅改变生活，通勤路线都还是一样。
**English / Original**
Also, Tesla being engineering, especially being primarily in Silicon Valley, it’s easier for people to just... They don’t have to change their life very much. Their commute’s going to be the same.
## paragraph 921
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 922
**中文**
那要如何防止？怎样避免这种魔法粉尘效应，阻止所有人来挖你的员工？
**English / Original**
So how do you prevent that? How do you prevent the pixie dust effect where everyone’s trying to poach all your people?
## paragraph 923
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 924
**中文**
我认为我们没有多少办法阻止。这也是 Tesla……身处 Silicon Valley，同时又有魔法粉尘光环，确实意味着会遭到极其猛烈的挖角。
**English / Original**
I don’t think there’s much we can do to stop it. That’s one of the reasons why Tesla… Really, being in Silicon Valley and having the pixie dust thing at the same time meant that there was just a very, very aggressive recruitment.
## paragraph 925
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 926
**中文**
那么搬到 Austin 想必有所帮助？
**English / Original**
Presumably being in Austin helps then?
## paragraph 927
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 928
**中文**
Austin 确实有帮助。不过 Tesla 的大多数工程人员仍在 California。要让工程师搬家……我称之为“伴侣问题”。
**English / Original**
Austin, it helps. Tesla still has a majority of its engineering in California. Getting engineers to move… I call it the “significant other” problem.
## paragraph 929
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 930
**中文**
没错，“伴侣”也有自己的工作。
**English / Original**
Yes, “significant others” have jobs.
## paragraph 931
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 932
**中文**
正是如此。所以 Starbase 尤其困难，因为在那里找到一份非 SpaceX 工作的概率……
**English / Original**
Exactly. So for Starbase that was particularly difficult, since the odds of finding a non-SpaceX job…
## paragraph 933
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 934
**中文**
在得克萨斯州 Brownsville……
**English / Original**
In Brownsville, Texas…
## paragraph 935
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 936
**中文**
……非常低。确实很困难，那里像一座科技修道院：位置偏远，而且基本都是男人。
**English / Original**
…are pretty low. It’s quite difficult. It’s like a technology monastery thing, remote and mostly dudes.
## paragraph 937
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 938
**中文**
比 SF 也没改善多少。
**English / Original**
Not much of an improvement over SF.
## paragraph 939
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 940
**中文**
回到那些在 Tesla、SpaceX 等公司技术岗位上表现极其出色的人，你认为除了……他们只是非常精通火箭或技术基础之外，还有什么共同点？还是说存在某种组织能力？
**English / Original**
If you go back to these people who’ve really been very effective in a technical capacity at Tesla, at SpaceX, and those sorts of places, what do you think they have in common other than... Is it just that they’re very sharp on the rocketry or the technical foundations, or do you think it’s something organizational?
## paragraph 941
**中文**
是他们与你合作的能力？是既灵活、又不过分灵活的能力？什么样的人能成为你的优秀交锋伙伴？
**English / Original**
Is it something about their ability to work with you? Is it their ability to be flexible but not too flexible? What makes a good sparring partner for you?
## paragraph 942
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 943
**中文**
我不把他们视为交锋伙伴。一个人能把事情办成，我就喜欢他；办不成，我就讨厌他。就是这么直截了当，并非什么个人怪癖。执行得好，我就是他的超级粉丝；执行不好，就不是。但这无关迎合我的个人偏好，我当然也会尽量不让它变成迎合个人偏好。
**English / Original**
I don’t think of it as a sparring partner. If somebody gets things done, I love them, and if they don’t, I hate them. So it’s pretty straightforward. It’s not like some idiosyncratic thing. If somebody executes well, I’m a huge fan, and if they don’t, I’m not. But it’s not about mapping to my idiosyncratic preferences. I certainly try not to have it be mapping to my idiosyncratic preferences.
## paragraph 944
**中文**
总体来说，我认为应该按才华、动力和可信度来招聘，心地善良也很重要。我曾一度低估这一点。所以要问：他是个好人吗？值得信任吗？聪明、有才华并且勤奋吗？如果这些都有，领域知识可以后补。
**English / Original**
Generally, I think it’s a good idea to hire for talent and drive and trustworthiness. And I think goodness of heart is important. I underweighted that at one point. So, are they a good person? Trustworthy? Smart and talented and hard working? If so, you can add domain knowledge.
## paragraph 945
**中文**
但那些根本特质、根本属性无法改变。因此，Tesla 和 SpaceX 的大多数员工此前并不来自航空航天或汽车行业。
**English / Original**
But those fundamental traits, those fundamental properties, you cannot change. So most of the people who are at Tesla and SpaceX did not come from the aerospace industry or the auto industry.
## paragraph 946
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 947
**中文**
随着公司从 100 人扩大到 1000 人、1 万人，你的管理方式最需要改变的是什么？你以极其细致的管理、深入事情细节而闻名。
**English / Original**
What has had to change most about your management style as your companies have scaled from 100 to 1,000 to 10,000 people? You’re known for this very micro management, just getting into the details of things.
## paragraph 948
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 949
**中文**
请叫纳米管理。皮米管理，飞米管理。
**English / Original**
Nano management, please. Pico management. Femto management.
## paragraph 950
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 951
**中文**
继续往下说。
**English / Original**
Keep going.
## paragraph 952
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 953
**中文**
我们会一直深入到 Planck 常数，深入到 Heisenberg 不确定性原理。
**English / Original**
We’re going to go all the way down to Planck’s constant. All the way down to Heisenberg uncertainty principle.
## paragraph 954
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 955
**中文**
你仍能如自己所愿地深入细节吗？如果公司规模更小，会不会更成功？你如何看待这个问题？
**English / Original**
Are you still able to get into details as much as you want? Would your companies be more successful if they were smaller? How do you think about that?
## paragraph 956
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 957
**中文**
因为一天中的时间是固定的，随着事务增长、活动范围扩大，我的时间必然会被稀释。我实际上不可能进行微观管理，因为那意味着我每天有几千个小时。让我事无巨细地管理，在逻辑上不可能。
**English / Original**
Because I have a fixed amount of time in the day, my time is necessarily diluted as things grow and as the span of activity increases. It’s impossible for me to actually be a micromanager because that would imply I have some thousands of hours per day. It is a logical impossibility for me to micromanage things.
## paragraph 958
**中文**
不过有时我会深入某个具体问题，因为它正是限制公司进展的因素。深入极细节事项，是因为它构成了限制因素，不是随意钻进琐碎小事。
**English / Original**
Now, there are times when I will drill down into a specific issue because that specific issue is the limiting factor on the progress of the company. The reason for drilling into some very detailed item is because it is the limiting factor. It’s not arbitrarily drilling into tiny things.
## paragraph 959
**中文**
从时间上看，我在物理上不可能随意深入无关紧要的小事，那样必然导致失败。但有时，微小的事情会决定胜负。
**English / Original**
From a time standpoint, it is physically impossible for me to arbitrarily go into tiny things that don’t matter. That would result in failure. But sometimes the tiny things are decisive in victory.
## paragraph 960
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 961
**中文**
众所周知，你把 Starship 的设计材料从复合材料改成了钢。
**English / Original**
Famously, you switched the Starship design from composites to steel.
## paragraph 962
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 963
**中文**
是的。
**English / Original**
Yes.
## paragraph 964
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 965
**中文**
这个决定是你做的。不是下属四处研究后对你说：“老板，我们找到了更好的东西。”而是你顶着一些阻力推动大家。你能讲讲自己是如何形成改用钢材这个完整构想的吗？
**English / Original**
You made that decision. That wasn’t people going around saying, “Oh, we found something better, boss.” That was you encouraging people against some resistance. Can you tell us how you came to that whole concept of the steel switch?
## paragraph 966
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 967
**中文**
我会说，是出于绝望。最初我们打算用碳纤维制造 Starship。碳纤维相当昂贵。进入批量生产后，任何产品的成本都可以逐渐接近其材料成本。
**English / Original**
Desperation, I’d say. Originally, we were going to make Starship out of carbon fiber. Carbon fiber is pretty expensive. When you do volume production, you can get any given thing to start to approach its material cost.
## paragraph 968
**中文**
碳纤维的问题在于，材料成本本身仍然很高。尤其是能够承受低温液氧的高强度特种碳纤维，成本大约是钢材的 50 倍。至少理论上它会更轻。人们通常认为钢很重，碳纤维很轻。
**English / Original**
The problem with carbon fiber is that material cost is still very high. Particularly if you go for a high-strength specialized carbon fiber that can handle cryogenic oxygen, it’s roughly 50 times the cost of steel. At least in theory, it would be lighter. People generally think of steel as being heavy and carbon fiber as being light.
## paragraph 969
**中文**
对于室温应用，比如 Formula 1 赛车、静态航空结构，或者任何航空结构，碳纤维可能都更合适。问题是，我们试图用碳纤维制造这枚巨型火箭，而进展极其缓慢。
**English / Original**
For room temperature applications, like a Formula 1 car, static aero structure, or any kind of aero structure really, you’re probably going to be better off with carbon fiber. The problem is that we were trying to make this enormous rocket out of carbon fiber and our progress was extremely slow.
## paragraph 970
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 971
**中文**
最初选择它，就只是因为它轻吗？
**English / Original**
It had been picked in the first place just because it’s light?
## paragraph 972
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 973
**中文**
是的。乍看之下，大多数人都会认为制造轻质结构应选碳纤维。但如果用碳纤维制造非常庞大的物体，又要让它高效固化——也就是不能在室温固化，因为有时会铺 50 层碳纤维……碳纤维实际上就是碳丝加胶。为了达到高强度，需要高压釜，本质上是高压烘箱。如果要固化巨型结构，高压釜必须比火箭还大。
**English / Original**
Yes. At first glance, most people would think that the choice for making something light would be carbon fiber. The thing is that when you make something very enormous out of carbon fiber and then you try to have the carbon fiber be efficiently cured, meaning not room temperature cured, because sometimes you got 50 plies of carbon fiber… Carbon fiber is really carbon string and glue. In order to have high strength, you need an autoclave. Something that’s essentially a high pressure oven. If you have something that’s gigantic, that one’s got to be bigger than the rocket.
## paragraph 974
**中文**
我们当时试图制造一台比史上任何高压釜都大的高压釜。也可以室温固化，但耗时很长，而且存在问题。归根结底，我们用碳纤维的进展实在太慢。
**English / Original**
We were trying to make an autoclave that’s bigger than any autoclave that’s ever existed. Or you can do room temperature cure, which takes a long time and has issues. The final issue is that we were just making very slow progress with carbon fiber.
## paragraph 975
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 976
**中文**
更高层的问题是，为什么必须由你做这个决定？你的团队里有许多工程师。
**English / Original**
The meta question is why it had to be you who made that decision. There’s many engineers on your team.
## paragraph 977
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 978
**中文**
为什么团队没有想到钢材？
**English / Original**
How did the team not arrive at steel?
## paragraph 979
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 980
**中文**
没错。这是一个更广泛问题的一部分：理解你在自己公司中的比较优势。
**English / Original**
Yeah exactly. This is part of a broader question, understanding your comparative advantage at your companies.
## paragraph 981
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 982
**中文**
因为碳纤维方案进展太慢，我就想：“好吧，我们必须试试别的。”Falcon 9 的主箭体采用铝锂合金，强度重量比很好。实际上在其应用场景中，它的强度重量比与碳纤维大致相同，也许更好。但铝锂合金非常难加工。
**English / Original**
Because we were making very slow progress with carbon fiber, I was like, “Okay, we’ve got to try something else.” For the Falcon 9, the primary airframe is made of aluminum lithium, which has a very good strength-to-weight. Actually, it has about the same, maybe better, strength to weight for its application than carbon fiber. But aluminum lithium is very difficult to work with.
## paragraph 983
**中文**
焊接它需要采用搅拌摩擦焊，也就是在不进入液相的情况下连接金属。竟然能这样连接，确实很神奇；但这种特定焊接工艺可以做到，而且非常困难。假设想修改铝锂结构或在上面安装部件，就必须采用带密封的机械连接，不能直接焊上去。因此，我不想让 Starship 主结构使用铝锂合金。
**English / Original**
In order to weld it, you have to do something called friction stir welding, where you join the metal without entering the liquid phase. It’s kind of wild that you can do that. But with this particular type of welding, you can do that. It’s very difficult. Let’s say you want to make a modification or attach something to aluminum lithium, you now have to use a mechanical attachment with seals. You can’t weld it on. So I wanted to avoid using aluminum lithium for the primary structure for Starship.
## paragraph 984
**中文**
当时有一种质量性能很优秀的特种碳纤维。造火箭时，要尽可能提高推进剂占火箭的比例，显然要尽量减轻质量。但正如我说的，我们进展太慢。我说：“照这个速度，我们永远到不了火星，必须想别的办法。”
**English / Original**
There was this very special grade of carbon fiber that had very good mass properties. With a rocket, you’re really trying to maximize the percentage of the rocket that is propellant, minimize the mass obviously. But like I said, we were making very slow progress. I said, “at this rate, we’re never going to get to Mars. So we’ve got to think of something else.”
## paragraph 985
**中文**
我不想使用铝锂合金，因为搅拌摩擦焊太困难，尤其难以规模化。直径 3.6 米时就已经够难了，更别提 9 米或更大。于是我说：“钢怎么样？”
**English / Original**
I didn’t want to use aluminum lithium because of the difficulty of friction stir welding, especially doing that at scale. It was hard enough at 3.6 meters in diameter, let alone at 9 meters or above. Then I said, “what about steel?”
## paragraph 986
**中文**
我有一条线索：美国一些早期火箭用过很薄的钢材，Atlas 火箭采用过钢制气球贮箱。所以钢并非从未用过，事实上早有应用。查看不锈钢的材料性能会发现，全硬化、应变硬化不锈钢在低温下的强度重量比，其实与碳纤维相近。
**English / Original**
I had a clue here because some of the early US rockets had used very thin steel. The Atlas rockets had used a steel balloon tank. It’s not like steel had never been used before. It actually had been used. When you look at the material properties of stainless steel, full-hard, strain hardened stainless steel, at cryogenic temperature the strength to weight is actually similar to carbon fiber.
## paragraph 987
**中文**
如果看室温下的材料性能，钢似乎会重一倍；但如果看特定牌号全硬化不锈钢在低温下的材料性能，就会得到与碳纤维相近的强度重量比。
**English / Original**
If you look at material properties at room temperature, it looks like the steel is going to be twice as heavy. But if you look at the material properties at cryogenic temperature of full-hard steel, stainless of particular grades, then you actually get to a similar strength to weight as carbon fiber.
## paragraph 988
**中文**
Starship 的燃料和氧化剂都是低温液体。Falcon 9 的燃料是火箭推进剂级煤油，基本就是高纯度航空燃油，温度大致为室温。虽然我们的确会稍微冷却它，就像冰镇啤酒一样。
**English / Original**
In the case of Starship, both the fuel and the oxidizer are cryogenic. For Falcon 9, the fuel is rocket propellant-grade kerosene, basically a very pure form of jet fuel. That is roughly room temperature. Although we do actually chill it slightly below, we chill it like a beer.
## paragraph 989
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 990
**中文**
听起来很美味。
**English / Original**
Delicious.
## paragraph 991
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 992
**中文**
我们确实会冷却，但不到低温状态。实际上，如果让它进入低温状态，就会变成蜡。但 Starship 使用液态甲烷和液氧，两者液化温度接近，几乎整个主结构都处于低温。因此，采用应变硬化的 300 系不锈钢时，由于几乎所有部位都处在低温，其强度重量比确实与碳纤维相近。
**English / Original**
We do chill it, but it’s not cryogenic. In fact, if we made it cryogenic, it would just turn to wax. But for Starship, it’s liquid methane and liquid oxygen. They are liquid at similar temperatures. Basically, almost the entire primary structure is at cryogenic temperature. So then you’ve got a 300-series stainless that’s strain hardened. Because almost all things are cryogenic temperature, it actually has similar strength to weight as carbon fiber.
## paragraph 993
**中文**
但它的原材料成本低 50 倍，而且极易加工。你可以在户外焊接不锈钢，甚至可以一边抽雪茄一边焊。不锈钢非常坚韧，也很容易修改；想安装什么，直接焊上去即可。加工极其容易，成本也极低。
**English / Original**
But it costs 50x less in raw material and is very easy to work with. You can weld stainless steel outdoors. You could smoke a cigar while welding stainless steel. It’s very resilient. You can modify it easily. If you want to attach something, you just weld it right on. Very easy to work with, very low cost.
## paragraph 994
**中文**
正如我说的，在低温下，它的强度重量比与碳纤维相近。再考虑到钢的熔点远高于铝——大约是铝的两倍——隔热罩质量可以大幅降低。
**English / Original**
Like I said, at cryogenic temperature, it’s similar strength-to-weight to carbon fiber. Then when you factor in that we have a much reduced heat shield mass, because the melting point of steel, is much greater than the melting point of aluminum… It’s about twice the melting point of aluminum.
## paragraph 995
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 996
**中文**
所以火箭可以承受高得多的温度？
**English / Original**
So you can just run the rocket much hotter?
## paragraph 997
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 998
**中文**
是的，尤其是飞船像燃烧的流星一样再入时。隔热罩质量可以大幅降低：迎风面隔热罩质量也许能减半，背风面则完全不需要隔热保护。
**English / Original**
Yes, especially for the ship which is coming in like a blazing meteor. You can greatly reduce the mass of the heat shield. You can cut the mass of the windward part of the heat shield, maybe in half, and you don’t need any heat shielding on the leeward side.
## paragraph 999
**中文**
最终结果是，钢制火箭实际上比碳纤维火箭更轻，因为碳纤维火箭里的树脂会开始熔化。碳纤维与铝的工作温度能力基本相同，而钢能在两倍高的温度下工作。这些都是非常粗略的估计。
**English / Original**
The net result is that actually the steel rocket weighs less than the carbon fiber rocket, because the resin in the carbon fiber rocket starts to melt. Basically, carbon fiber and aluminum have about the same operating temperature capabilities, whereas steel can operate at twice the temperature. These are very rough approximations.
## paragraph 1000
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1001
**中文**
反正火箭不会由我来造。
**English / Original**
I won’t build the rocket.
## paragraph 1002
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1003
**中文**
我的意思是，人们会说：“哦，他说是两倍，其实是 0.8 倍。”我只想说，闭嘴吧，混蛋。
**English / Original**
What I mean is people will say, “Oh, he said this twice. It’s actually 0.8.” I’m like, shut up, assholes.
## paragraph 1004
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1005
**中文**
主要评论肯定就会抓着这点。
**English / Original**
That’s what the main comment’s going to be about.
## paragraph 1006
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1007
**中文**
该死。关键是，事后看来，我们一开始就应该用钢。没用钢是个愚蠢的决定。
**English / Original**
God damn it. The point is, in retrospect, we should have started with steel in the beginning. It was dumb not to do steel.
## paragraph 1008
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1009
**中文**
好，让我复述一下。我的理解是，除了美国早期火箭的案例，钢材是一条风险更高、验证更少的路径；碳纤维更差，却经过了更多验证。所以必须由你来推动：“嘿，我们就走这条风险更高的路，边做边解决。”从某种意义上说，你是在对抗保守主义。
**English / Original**
Okay, but to play this back to you, what I’m hearing is that steel was a riskier, less proven path, other than the early US rockets. Versus carbon fiber was a worse but more proven out path. So you need to be the one to push for, “Hey, we’re going to do this riskier path and just figure it out.” So you’re fighting a sort of conservatism in a sense.
## paragraph 1010
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1011
**中文**
所以我一开始才说，问题是我们的进展不够快。我们甚至很难做出一小段没有褶皱的碳纤维筒体。因为在这种巨大尺度下，必须铺很多层碳纤维，再加以固化，而且固化方式必须确保没有任何褶皱或缺陷。
**English / Original**
That’s why I initially said that the issue is that we weren’t making fast enough progress. We were having trouble making even a small barrel section of the carbon fiber that didn’t have wrinkles in it. Because at that large scale, you have to have many plies, many layers of the carbon fiber. You’ve got to cure it and you’ve got to cure it in such a way that it doesn’t have any wrinkles or defects.
## paragraph 1012
**中文**
碳纤维的韧性远不如钢，抗破坏能力弱得多。不锈钢会拉伸、弯曲，碳纤维则往往会碎裂。韧性就是应力—应变曲线下方的面积。总体上钢会更好，准确说是不锈钢。
**English / Original**
Carbon fiber is much less resilient than steel. It has much less toughness. Stainless steel will stretch and bend, the carbon fiber will tend to shatter. Toughness being the area under the stress strain curve. You’re generally going to have to do better with steel, but stainless steel to be precise.
## paragraph 1013
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1014
**中文**
再问一个 Starship 问题。我想是两年前，我和 Sam Teller 参观了 Starbase，体验太棒了，从很多方面看都非常酷。
**English / Original**
One other Starship question. So I visited Starbase, I think it was two years ago, with Sam Teller, and that was awesome. It was very cool to see, in a whole bunch of ways.
## paragraph 1015
**中文**
我注意到的一点是，大家非常以事物的简洁为荣。人人都想告诉你，Starship 只是一只巨型汽水罐；公司在招聘焊工，只要能在任何工业项目中焊接，在这里也能焊。大家对这种简洁非常自豪。
**English / Original**
One thing I noticed was that people really took pride in the simplicity of things, where everyone wants to tell you how Starship is just a big soda can, and we’re hiring welders, and if you can weld in any industrial project, you can weld here. But there’s a lot of pride in the simplicity.
## paragraph 1016
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1017
**中文**
但事实上，Starship 是一枚非常复杂的火箭。
**English / Original**
Well, factually Starship is a very complicated rocket.
## paragraph 1018
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1019
**中文**
这就是我要问的：它究竟简单还是复杂？
**English / Original**
So that’s what I’m getting at. Are things simple or are they complex?
## paragraph 1020
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1021
**中文**
我想他们只是想说，参与 Starship 不需要有火箭行业经验。一个人只要聪明、勤奋、值得信赖，就能参与火箭工作，不需要此前造过火箭。Starship 远远是人类有史以来制造的最复杂机器。
**English / Original**
I think maybe just what they’re trying to say is that you don’t have to have prior experience in the rocket industry to work on Starship. Somebody just needs to be smart and work hard and be trustworthy and they can work on a rocket. They don’t need prior rocket experience. Starship is the most complicated machine ever made by humans, by a long shot.
## paragraph 1022
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1023
**中文**
复杂在哪些方面？
**English / Original**
In what regards?
## paragraph 1024
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1025
**中文**
几乎任何方面。我会说，没有比它更复杂的机器。我能想到的几乎所有项目都比它容易。这就是为什么从来没人造出完全可重复使用的轨道火箭。这个问题极其困难，过去许多非常聪明、拥有巨量资源的人尝试过，但都失败了。
**English / Original**
Anything, really. I’d say there isn’t a more complex machine. I’d say that pretty much any project I can think of would be easier than this. That’s why nobody has ever made a fully reusable orbital rocket. It’s a very hard problem. Many smart people have tried before, very smart people with immense resources, and they failed.
## paragraph 1026
**中文**
而我们也尚未成功。Falcon 可部分重复使用，但上面级不行。我认为 Starship 第三版的设计可以完全重复使用。正是这种完全可重复使用能力，将使我们成为多行星文明。任何技术问题，即便是 Hadron Collider 之类，都比这个问题容易。
**English / Original**
And we haven’t succeeded yet. Falcon is partially reusable, but the upper stage is not. Starship Version 3, I think this design can be fully reusable. That full reusability is what will enable us to become a multi-planet civilization. Any technical problem, even like a Hadron Collider or something like that, is an easier problem than this.
## paragraph 1027
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1028
**中文**
我们花了很多时间讨论瓶颈。你能否哪怕只从高层次讲讲 Starship 当前的瓶颈？
**English / Original**
We spent a lot of time on bottlenecks. Can you say what the current Starship bottlenecks are, even at a high level?
## paragraph 1029
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1030
**中文**
总体来说，就是努力让它别爆炸。它真的非常想爆炸。
**English / Original**
Trying to make it not explode, generally. It really wants to explode.
## paragraph 1031
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1032
**中文**
又是这个老问题，毕竟有那么多可燃材料。
**English / Original**
That old chestnut. All those combustible materials.
## paragraph 1033
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1034
**中文**
我们有两枚助推器在试验台上爆炸，其中一枚把整个试验设施夷为平地。所以只要犯一个错误就够了。Starship 所蕴含的能量大得疯狂。
**English / Original**
We’ve had two boosters explode on the test stand. One obliterated the entire test facility. So it only takes that one mistake. The amount of energy contained in a Starship is insane.
## paragraph 1035
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1036
**中文**
这就是它比 Falcon 更难的原因吗？只是因为能量更大？
**English / Original**
Is that why it’s harder than Falcon? It’s because it’s just more energy?
## paragraph 1037
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1038
**中文**
它采用了大量新技术，也在挑战性能边界。Raptor 3 是非常非常先进的发动机，远远是有史以来最好的火箭发动机。但它拼命想爆炸。为便于理解，火箭起飞时产生的功率超过 100 吉瓦，相当于美国用电功率的 20%。
**English / Original**
It’s a lot of new technology. It’s pushing the performance envelope. The Raptor 3 engine is a very, very advanced engine. It’s by far the best rocket engine ever made. But it desperately wants to blow up. Just to put things into perspective here, on liftoff the rocket is generating over 100 gigawatts of power. That’s 20% of US electricity.
## paragraph 1039
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1040
**中文**
这确实太疯狂了。
**English / Original**
It’s actually insane.
## paragraph 1041
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1042
**中文**
这个比较很形象。
**English / Original**
It’s a great comparison.
## paragraph 1043
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1044
**中文**
而且还不能爆炸。
**English / Original**
While not exploding.
## paragraph 1045
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1046
**中文**
有时候吧。
**English / Original**
Sometimes.
## paragraph 1047
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1048
**中文**
是的，有时候。所以我想的是，怎样让它不爆炸？它有成千上万种爆炸方式，却只有一种不爆炸的方式。我们不仅希望它真的不爆炸，还要它每天可靠飞行，比如每小时一次。显然，如果经常爆炸，就很难维持这种发射频率。
**English / Original**
Sometimes, yes. So I was like, how does it not explode? There’s thousands of ways that it could explode and only one way that it doesn’t. So we want it not only to really not explode, but fly reliably on a daily basis, like once per hour. Obviously, if it blows up a lot, it’s very difficult to maintain that launch cadence.
## paragraph 1049
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1050
**中文**
是的。
**English / Original**
Yes.
## paragraph 1051
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1052
**中文**
Starship 剩下最大的单一问题是什么？是让隔热罩可以重复使用。以前从未有人制造过可重复使用的轨道飞行隔热罩。它必须在上升阶段不脱落大量隔热瓦，再入时同样不能大量掉瓦，也不能让主箭体过热。
**English / Original**
What’s the single biggest remaining problem for Starship? It’s having the heat shield be reusable. No one’s ever made a reusable orbital heat shield. So the heat shield’s gotta make it through the ascent phase without shucking a bunch of tiles, and then it’s gotta come back in and also not lose a bunch of tiles or overheat the main airframe.
## paragraph 1053
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1054
**中文**
这难道不是很困难吗？因为它本质上就是消耗品。
**English / Original**
Isn’t that hard because it’s fundamentally a consumable?
## paragraph 1055
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1056
**中文**
是的，但汽车刹车片也是消耗品，却能用很久。
**English / Original**
Well, yes, but your brake pads in your car are also consumable, but they last a very long time.
## paragraph 1057
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1058
**中文**
有道理。
**English / Original**
Fair.
## paragraph 1059
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1060
**中文**
所以它只需要能使用很久。我们已经让飞船返回，并在海上软着陆，完成过好几次。但它掉了许多隔热瓦，不经大量维修就无法复用。尽管确实实现了软着陆，不做许多工作仍不能再次使用。
**English / Original**
So it just needs to last a very long time. We have brought the ship back and had it do a soft landing in the ocean. We’ve done that a few times. But it lost a lot of tiles. It was not reusable without a lot of work. Even though it did come to a soft landing, it would not have been reusable without a lot of work.
## paragraph 1061
**中文**
因此从这个意义上说，它并不真正可复用。剩下最大的难题，就是完全可复用的隔热罩。理想状态是飞船着陆后补充推进剂就能再次起飞，不能每次都费力检查 4 万块隔热瓦之类。
**English / Original**
So it’s not really reusable in that sense. That’s the biggest problem that remains, a fully reusable heat shield. You want to be able to land it, refill propellant and fly again. You can’t do this laborious inspection of 40,000 tiles type of thing.
## paragraph 1062
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1063
**中文**
读你的传记时，我感觉你很善于推动紧迫感，并让大家形成“这才是能扩大规模的东西”的意识。我很好奇，为什么其他与你类似的组织……
**English / Original**
When I read biographies of yours, it seems like you’re just able to drive the sense of urgency and drive the sense of “this is the thing that can scale.” I’m curious why you think other organizations of your…
## paragraph 1064
**中文**
SpaceX 和 Tesla 如今都是非常大的公司，你仍能保持这种文化。其他公司出了什么问题，导致它们做不到？
**English / Original**
SpaceX and Tesla are really big companies now. You’re still able to keep that culture. What goes wrong with other companies such that they’re not able to do that?
## paragraph 1065
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1066
**中文**
我不知道。
**English / Original**
I don’t know.
## paragraph 1067
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1068
**中文**
比如你说今天开了许多 SpaceX 会议。你在会上做了什么，从而维持这种文化？
**English / Original**
Like today, you said you had a bunch of SpaceX meetings. What is it that you’re doing there that’s keeping that?
## paragraph 1069
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1070
**中文**
是在增加紧迫感吗？
**English / Original**
It’s adding urgency?
## paragraph 1071
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1072
**中文**
这个嘛，我不知道。我想紧迫感会来自领导公司的人。我有一种近乎疯狂的紧迫感，所以这种疯狂的紧迫感会传递到公司其他地方。
**English / Original**
Well, I don’t know. I guess the urgency is going to come from whoever is leading the company. I have a maniacal sense of urgency. So that maniacal sense of urgency projects through the rest of the company.
## paragraph 1073
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1074
**中文**
是因为后果吗？员工会想：“Elon 定了一个疯狂期限，但我如果完不成，也知道自己会有什么下场。”还是因为你能识别并消除瓶颈，让大家快速推进？你认为自己的公司为什么能行动得这么快？
**English / Original**
Is it because of consequences? They’re like, “Elon set a crazy deadline, but if I don’t get it, I know what happens to me.” Is it just that you’re able to identify bottlenecks and get rid of them so people can move fast? How do you think about why your companies are able to move fast?
## paragraph 1075
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1076
**中文**
我一直在解决限制因素。就期限而言，我通常会争取一个至少在我看来位于第 50 百分位的期限。因此它不是不可能完成的期限，而是我能想到最激进、同时有 50% 概率实现的期限。这也意味着有一半时间会延期。
**English / Original**
I’m constantly addressing the limiting factor. On the deadlines front, I generally actually try to aim for a deadline that I at least think is at the 50th percentile. So it’s not like an impossible deadline, but it’s the most aggressive deadline I can think of that could be achieved with 50% probability. Which means that it’ll be late half the time.
## paragraph 1077
**中文**
气体膨胀定律同样适用于日程。如果你说某件事五年内完成——在我看来五年几乎等于无限久——工作就会膨胀并填满所有可用时间，最终真的花五年。
**English / Original**
There is a law of gas expansion that applies to schedules. If you said we’re going to do something in five years, which to me is like infinity time, it will expand to fill the available schedule and it’ll take five years.
## paragraph 1078
**中文**
物理会限制某些事情的最快速度。扩大制造规模时，移动原子和扩大产能都有速度上限。这就是为什么不可能瞬间把某种产品做到年产 100 万件。你必须设计生产线、启动生产，再沿生产 S 曲线爬升。
**English / Original**
Physics will limit how fast you can do certain things. So scaling up manufacturing, there’s a rate at which you can move the atoms and scale manufacturing. That’s why you can’t instantly make a million units a year of something. You’ve got to design the manufacturing line. You’ve got to bring it up. You’ve got to ride the S-curve of production.
## paragraph 1079
**中文**
我能说些什么真正对人们有帮助？总体上，近乎疯狂的紧迫感非常重要。日程应该激进，并且要随时弄清限制因素是什么，帮助团队解决这个限制因素。
**English / Original**
What can I say that’s actually helpful to people? Generally, a maniacal sense of urgency is a very big deal. You want to have an aggressive schedule and you want to figure out what the limiting factor is at any point in time and help the team address that limiting factor.
## paragraph 1080
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1081
**中文**
所以 Starlink 曾缓慢筹备了许多年。
**English / Original**
So Starlink was slowly in the works for many years.
## paragraph 1082
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1083
**中文**
我们在公司成立之初就讨论过它。
**English / Original**
We talked about it all the way in the beginning of the company.
## paragraph 1084
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1085
**中文**
后来你在 Redmond 组建了一支团队，到某个节点你认定这支团队根本不行。项目缓慢推进了几年，那么你为什么没更早行动，又为什么恰好在那个时候行动？为什么那是采取行动的正确时机？
**English / Original**
So then there was a team you had built in Redmond, and then at one point you decided this team is just not cutting it. It went for a few years slowly, and so why didn’t you act earlier, and why did you act when you did? Why was that the right moment at which to act?
## paragraph 1086
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1087
**中文**
我每周都会进行非常细致的工程评审，这种粒度也许很不同寻常。我不知道还有哪位公司负责人——至少是制造企业负责人——会深入到我这种细节程度。事情并不是……由于我们会逐项深入讨论，我对实际状况相当了解。
**English / Original**
I have these very detailed engineering reviews weekly. That’s maybe a very unusual level of granularity. I don’t know anyone who runs a company, or at least a manufacturing company, that goes with the level of detail that I go into. It’s not as though... I have a pretty good understanding of what’s actually going on because we go through things in detail.
## paragraph 1088
**中文**
我非常推崇越级会议：技术评审中，不是由直接向我汇报的人发言，而是让他的所有下属逐一发言，而且不能提前精心准备。否则，用我最近的话说，你就会被“裹上糖衣”。
**English / Original**
I’m a big believer in skip-level meetings where instead of having the person that reports to me say things, it’s everyone that reports to them saying something in the technical review. And there can’t be advanced preparation. Otherwise you’re going to get “glazed”, as I say these days.
## paragraph 1089
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1090
**中文**
没错，你这说法很 Z 世代。
**English / Original**
Exactly. Very Gen Z of you.
## paragraph 1091
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1092
**中文**
你如何防止他们提前准备？会随机点名吗？
**English / Original**
How do you prevent advanced preparation? Do you call on them randomly?
## paragraph 1093
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1094
**中文**
不会，我只是沿着会议室让每个人逐一汇报。要记在脑子里的信息很多。如果每周或每周两次开会，你就有了那个人每次发言的快照，随后可以绘制进展点。在脑中把这些点大致画成一条曲线，再问：“我们是否正收敛到解决方案？”
**English / Original**
No, I just go around the room. Everyone provides an update. It’s a lot of information to keep in your head. If you have meetings weekly or twice weekly, you’ve got a snapshot of what that person said. You can then plot the progress points. You can sort of mentally plot the points on a curve and say, “are we converging to a solution or not?”
## paragraph 1095
**中文**
我只会在认定成功已不属于可能结果集合时采取激烈行动。当我最终得出结论：若不采取激烈行动，我们就毫无成功机会，那么我就必须采取激烈行动。2018 年我得出了这个结论，于是采取激烈行动并解决了问题。
**English / Original**
I’ll take drastic action only when I conclude that success is not in a set of possible outcomes. So when I finally reach the conclusion that unless drastic action is done, we have no chance of success, then I must take drastic action. I came to that conclusion in 2018, took drastic action and fixed the problem.
## paragraph 1096
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1097
**中文**
你有许多许多家公司。听起来你在每一家都要深入理解工程情况和相关瓶颈，才能与员工进行这些评审。
**English / Original**
You’ve got many, many companies. In each of them it sounds like you do this kind of deep engineering understanding of what the relevant bottlenecks are so you can do these reviews with people.
## paragraph 1098
**中文**
你已把这种做法扩展到五六七家公司，而每一家公司内部又有许多小公司。上限由什么决定？因为你好像有 80 家公司……？
**English / Original**
You’ve been able to scale it up to five, six, seven companies. Within one of these companies, you have many different mini companies within them. What determines the max amount here? Because you have like 80 companies…?
## paragraph 1099
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1100
**中文**
80 家？没有。
**English / Original**
80? No.
## paragraph 1101
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1102
**中文**
但你已经有这么多了，这本身就很了不起。
**English / Original**
But you have so many already. That’s already remarkable.
## paragraph 1103
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1104
**中文**
就按现在这个数量来说。
**English / Original**
By this current number.
## paragraph 1105
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1106
**中文**
正是。
**English / Original**
Exactly.
## paragraph 1107
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1108
**中文**
我们连一家公司都勉强维持。
**English / Original**
We can barely keep one company together.
## paragraph 1109
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1110
**中文**
这取决于具体情况。我其实不会与 The Boring Company 定期开会，所以它基本在平稳自行推进。如果某件事运作良好、进展顺利，我就没必要在上面花时间。
**English / Original**
It depends on the situation. I actually don’t have regular meetings with The Boring Company, so The Boring Company is sort of cruising along. Basically, if something is working well and making good progress, then there’s no point in me spending time on it.
## paragraph 1111
**中文**
实际上，我会按限制因素所在之处分配时间。哪里出了问题？哪里遇到阻力？什么在拖慢我们？冒着把这个词说太多次的风险，我专注于限制因素。
**English / Original**
I actually allocate time according to where the limiting factor. Where are things problematic? Where are we pushing against? What is holding us back? I focus, at the risk of saying the words too many times, on the limiting factor.
## paragraph 1112
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1113
**中文**
讽刺的是，如果某件事进展非常顺利，那里的人很少见到我；如果某件事进展很差，他们就会经常见到我。甚至不一定是进展很差……
**English / Original**
The irony is if something’s going really well, they don’t see much of me. But if something is going badly, they’ll see a lot of me. Or not even badly…
## paragraph 1114
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1115
**中文**
而是某件事成为了限制因素。
**English / Original**
If something is the limiting factor.
## paragraph 1116
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1117
**中文**
正是，限制因素。它不完全是进展很差，只是我们需要让它推进得更快。
**English / Original**
The limiting factor, exactly. It’s not exactly going badly but it’s the thing that we need to make go faster.
## paragraph 1118
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1119
**中文**
当 SpaceX 或 Tesla 的某件事成为限制因素时，你会每周甚至每天与负责它的工程师交谈吗？实际是如何运作的？
**English / Original**
When something’s a limiting factor at SpaceX or Tesla, are you talking weekly and daily with the engineer that’s working on it? How does that actually work?
## paragraph 1120
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1121
**中文**
大多数限制因素每周评审一次，有些每周两次。AI5 芯片评审就是每周两次，每周二和周六进行。
**English / Original**
Most things that are the limiting factor are weekly and some things are twice weekly. The AI5 chip review is twice weekly. Every Tuesday and Saturday is the chip review.
## paragraph 1122
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1123
**中文**
会议时长没有上限吗？
**English / Original**
Is it open ended in how long it goes?
## paragraph 1124
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1125
**中文**
严格说来是的，但通常开两三个小时，有时更短，取决于需要过多少信息。
**English / Original**
Technically, yes, but usually it’s two or three hours. Sometimes less. It depends on how much information we’ve got to go through.
## paragraph 1126
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1127
**中文**
这也是一个不同之处。我只是试图找出差异，因为结果似乎大不相同，弄清哪些输入不同很有意思。在企业界，第一，正如你所说，CEO 并不总会亲自进行工程评审，尽管工程就是公司的核心业务。
**English / Original**
That’s another thing. I’m just trying to tease out the differences here because the outcomes seem quite different. I think it’s interesting to know what inputs are different. It feels like in the corporate world, one, like you were saying, the CEO doing engineering reviews does not always happen despite the fact that that is what the company is doing.
## paragraph 1128
**中文**
而且时间通常被细分为半小时、甚至 15 分钟的会议。你似乎会开更多开放式会议：“一直讨论，直到把问题弄明白。”
**English / Original**
But then time is often pretty finely sliced into half hour meetings or even 15 minute meetings. It seems like you hold more open-ended, “We’re talking about it until we figure it out” type things.
## paragraph 1129
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1130
**中文**
有时如此。但大多数会议大致都能按时结束。今天的 Starship 工程评审稍微超时，因为要讨论的话题更多。他们正试图弄清如何扩大到每年将 100 多万吨送入轨道，挑战相当大。
**English / Original**
Sometimes. But most of them seem to more or less stay on time. Today’s Starship engineering review went a bit longer because there were more topics to discuss. They’re trying to figure out how to scale to a million plus tons to orbit per year. It’s quite challenging.
## paragraph 1131
**中文**
02:20:08 — DOGE
**English / Original**
02:20:08 - DOGE
## paragraph 1132
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1133
**中文**
我能问个问题吗？你说 Optimus 和 AI 会在几年内带来两位数增长率。
**English / Original**
Can I ask a question? You said about Optimus and AI that they’re going to result in double digit growth rates within a matter of years.
## paragraph 1134
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1135
**中文**
哦，你是说经济？对，我认为如此。
**English / Original**
Oh, like the economy? Yes. I think that’s right.
## paragraph 1136
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1137
**中文**
如果经济会增长这么多，DOGE 削减开支的意义是什么？
**English / Original**
What was the point of the DOGE cuts if the economy is going to grow so much?
## paragraph 1138
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1139
**中文**
浪费和欺诈终究不是好事。我其实相当担心……如果没有 AI 和机器人，我们就彻底完了，因为国债正疯狂累积。国债利息支出已超过一万亿美元的军费，所以单是利息就超过一万亿美元。我很担心这一点。也许我花些时间，就能延缓美国破产，为 AI 和机器人帮助解决国债问题争取足够时间。
**English / Original**
Well, I think waste and fraud are not good things to have. I was actually pretty worried about... In the absence of AI and robotics, we’re actually totally screwed because the national debt is piling up like crazy. The interest payments to national debt exceed the military budget, which is a trillion dollars. So we have over a trillion dollars just in interest payments. I was pretty concerned about that. Maybe if I spend some time, we can slow down the bankruptcy of the United States and give us enough time for the AI and robots to help solve the national debt.
## paragraph 1140
**中文**
不只是“帮助解决”，它们是唯一能解决国债问题的东西。没有 AI 和机器人，我们这个国家百分之一千会破产并走向失败。其他任何办法都解决不了国债。我们只需争取足够时间造出 AI 和机器人，在此之前不要破产。
**English / Original**
Or not help solve, it’s the only thing that could solve the national debt. We are 1000% going to go bankrupt as a country, and fail as a country, without AI and robots. Nothing else will solve the national debt. We just need enough time to build the AI and robots to not go bankrupt before then.
## paragraph 1141
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1142
**中文**
我好奇的是，DOGE 启动时，你们拥有推行改革的巨大能力。
**English / Original**
I guess the thing I’m curious about is, when DOGE starts you have this enormous ability to enact reform.
## paragraph 1143
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1144
**中文**
没那么巨大。
**English / Original**
Not that enormous.
## paragraph 1145
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1146
**中文**
好吧。我完全赞同你的观点：AI 与机器人推动生产率提高和 GDP 增长很重要。但为什么不直接解决你指出的问题，比如某些零部件的关税或许可制度？
**English / Original**
Sure. I totally buy your point that it’s important that AI and robotics drive productivity improvements, drive GDP growth. But why not just directly go after the things you were pointing out, like the tariffs on certain components, or permitting?
## paragraph 1147
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1148
**中文**
我不是总统。而且，即使要削减明显的浪费与欺诈——荒谬至极的浪费与欺诈——也非常困难。我发现，连从政府中削减极其明显的浪费和欺诈都异常困难，因为政府必须看谁在抱怨。
**English / Original**
I’m not the president. And it is very hard to cut things that are obvious waste and fraud, like ridiculous waste and fraud. What I discovered is that it’s extremely difficult even to cut very obvious waste and fraud from the government because the government has to operate on who’s complaining.
## paragraph 1149
**中文**
如果停止向欺诈者付款，他们会立即编出最能博取同情的理由，要求继续支付。他们不会说：“请让欺诈继续。”而会说：“你们在杀害熊猫宝宝。”与此同时，没有任何熊猫宝宝死亡，全是他们编的。欺诈者很会编造极具说服力、令人心碎、实际却虚假的故事，听起来仍然值得同情。事实就是这样。
**English / Original**
If you cut off payments to fraudsters, they immediately come up with the most sympathetic sounding reasons to continue the payment. They don’t say, “Please keep the fraud going.” They’re like, “You’re killing baby pandas.” Meanwhile, no baby pandas are dying. They’re just making it up. The fraudsters are capable of coming up with extremely compelling, heart-wrenching stories that are false, but nonetheless sound sympathetic. That’s what happened.
## paragraph 1150
**中文**
也许我本该更清楚。但我当时想，等等，让我们试着削减政府的一些浪费与利益输送。Social Security 系统里，也许不该有 2000 万名明确已经死亡、年龄超过 115 岁、却仍被标记为在世的人。
**English / Original**
Perhaps I should have known better. But I thought, wait, let’s try to cut some amount of waste and pork from the government. Maybe there shouldn’t be 20 million people marked as alive in Social Security who are definitely dead, and over the age of 115.
## paragraph 1151
**中文**
美国最长寿者是 114 岁。所以可以有把握地说，如果 Social Security 数据库里有人 115 岁还被标记为在世，要么是录入错误……应该给他打电话说：“看来我们记错了您的生日，或者需要把您标记为死亡。”二者必居其一。
**English / Original**
The oldest American is 114. So it’s safe to say if somebody is 115 and marked as alive in the Social Security database, there’s either a typo… Somebody should call them and say, “We seem to have your birthday wrong, or we need to mark you as dead.” One of the two things.
## paragraph 1152
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1153
**中文**
接到这种电话会很吓人。
**English / Original**
Very intimidating call to get.
## paragraph 1154
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1155
**中文**
这看起来是合理做法。比如某人的生日在未来，却有 Small Business Administration 贷款，而且出生年份是 2165 年，那么要么是录入错误，要么是欺诈。因此我们会说：“看来我们把您的出生世纪记错了。”
**English / Original**
Well, it seems like a reasonable thing. Say if their birthday is in the future and they have a Small Business Administration loan, and their birthday is 2165, we either have a typo or we have fraud. So we say, “we appear to have gotten the century of your birth incorrect.”
## paragraph 1156
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1157
**中文**
或者是个很棒的电影情节。
**English / Original**
Or a great plot for a movie.
## paragraph 1158
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1159
**中文**
没错。这就是我所谓的荒谬欺诈。
**English / Original**
Yes. That’s what I mean by, ludicrous fraud.
## paragraph 1160
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1161
**中文**
这些人有收到付款吗？
**English / Original**
Were those people getting payments?
## paragraph 1162
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1163
**中文**
有些人从 Social Security 领取付款。但主要欺诈路径，是先在 Social Security 中把某人标记为在世，再利用其他所有政府支付系统实施欺诈。因为其他政府支付系统只会向 Social Security 数据库查询一次“是否在世”，这是一种间接撞库。
**English / Original**
Some were getting payments from Social Security. But the main fraud vector was to mark somebody as alive in Social Security and then use every other government payment system to basically do fraud. Because what those other government payment systems do, they would simply do an “are you alive” check to the Social Security database. It’s a bank shot.
## paragraph 1164
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1165
**中文**
你估计通过这一机制实施的欺诈总额有多少？
**English / Original**
What would you estimate is the total amount of fraud from this mechanism?
## paragraph 1166
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1167
**中文**
顺便说，Government Accountability Office 以前做过这类估算，不只我这么说。事实上，我记得 GAO 分析过 Biden 政府时期的欺诈，粗略估算约为 5000 亿美元。所以别只听我的，去看 Biden 政府时期发布的报告。这样如何？
**English / Original**
By the way, the Government Accountability Office has done these estimates before. I’m not the only one. In fact, I think the GAO did an analysis, a rough estimate of fraud during the Biden administration, and calculated it at roughly half a trillion dollars. So don’t take my word for it. Take a report issued during the Biden administration. How about that?
## paragraph 1168
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1169
**中文**
全部来自这套 Social Security 机制？
**English / Original**
From this Social Security mechanism?
## paragraph 1170
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1171
**中文**
它只是许多机制之一。必须认识到，政府在制止欺诈方面极其低效。公司有动力制止欺诈，因为这会影响公司利润；政府却只会印更多钱。你需要责任心和能力，而联邦层面两者都很稀缺。
**English / Original**
It’s one of many. It’s important to appreciate that the government is very ineffective at stopping fraud. It’s not like a company where, with stopping fraud, you’ve got a motivation because it’s affecting the earnings of your company. The government just prints more money. You need caring and competence. These are in short supply at the federal level.
## paragraph 1172
**中文**
去 DMV 时，你会不会想：“哇，这真是能力的堡垒”？现在想象一下，比 DMV 还糟糕，而且是一个能印钱的 DMV。
**English / Original**
When you go to the DMV, do you think, “Wow, this is a bastion of competence”? Well, now imagine it’s worse than the DMV because it’s the DMV that can print money.
## paragraph 1173
**中文**
至少州级 DMV 还必须……各州大致需要量入为出，否则会破产。但联邦政府只会印更多钱。
**English / Original**
At least the state level DMVs need to... The states more or less need to stay within their budget or they go bankrupt. But the federal government just prints more money.
## paragraph 1174
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1175
**中文**
如果真有 5000 亿美元欺诈，为什么没办法全都砍掉？
**English / Original**
If there’s actually half a trillion of fraud, why was it not possible to cut all that?
## paragraph 1176
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1177
**中文**
你真的必须退一步，重新校准自己对能力的预期。因为你生活在一个必须收支平衡的世界，必须支付账单……
**English / Original**
You really have to stand back and recalibrate your expectations for competence. Because you’re operating in a world where you’ve got to make ends meet. You’ve got to pay your bills...
## paragraph 1178
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1179
**中文**
还得找到麦克风。
**English / Original**
Find the microphones.
## paragraph 1180
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1181
**中文**
正是。那里不是一个庞大、基本漠不关心的官僚怪兽，而是一群过时电脑在自动发送付款。DOGE 团队做的一件事听起来极其简单，却可能每年节省 1000 亿至 2000 亿美元。他们只是要求财政部主计算机发出的付款——它叫 PAM，可能是 Payment Accounts Master 之类，每年支付 5 万亿美元——必须带有拨款代码，并规定备注栏必须填写内容，不再是可选项。
**English / Original**
Exactly. It’s not like there’s a giant, largely uncaring monster bureaucracy. It’s a bunch of anachronistic computers that are just sending payments. One of the things that the DOGE team did sounds so simple and probably will save $100-200 billion a year. It was simply requiring payments from the main Treasury computer—which is called PAM, Payment Accounts Master or something like that, there’s $5 trillion payments a year—that go out have a payment appropriation code. Make it mandatory, not optional, that you have anything at all in the comment field.
## paragraph 1182
**中文**
你必须重新校准自己对事情有多愚蠢的认知。付款在没有拨款代码、无法追溯任何国会拨款、也没有任何说明的情况下被发出。这就是 Department of War——以前的 Department of Defense——无法通过审计的原因，因为相关信息根本不存在。请重新校准预期。
**English / Original**
You have to recalibrate how dumb things are. Payments were being sent out with no appropriation code, not checking back to any congressional appropriation, and with no explanation. This is why the Department of War, formerly the Department of Defense, cannot pass an audit, because the information is literally not there. Recalibrate your expectations.
## paragraph 1183
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1184
**中文**
我想更好地理解这个 5000 亿数字，因为 2024 年有一份监察长报告。
**English / Original**
I want to better understand this half a trillion number, because there’s an IG report in 2024.
## paragraph 1185
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1186
**中文**
为什么数字这么低？
**English / Original**
Why is it so low?
## paragraph 1187
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1188
**中文**
也许吧。但我们发现，他们估算七年间的 Social Security 欺诈约为 700 亿美元，也就是每年 100 亿左右。所以我很好奇剩余 4900 亿是什么。
**English / Original**
Maybe, but we found that over seven years, the Social Security fraud they estimated was like $70 billion over seven years, so like $10 billion a year. So I’d be curious to see what the other $490 billion is.
## paragraph 1189
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1190
**中文**
联邦政府每年支出 7.5 万亿美元。你觉得政府有多大能力？
**English / Original**
Federal government expenditures are $7.5 trillion a year. How competent do you think the government is?
## paragraph 1191
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1192
**中文**
其中可自由裁量支出大约……占 15%？
**English / Original**
The discretionary spending there is like… 15%?
## paragraph 1193
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1194
**中文**
但这无关紧要。大部分欺诈发生在非自由裁量支出：虚假的 Medicare、Medicaid、Social Security、残障福利，政府付款项目数不胜数。其中许多付款其实是向各州整笔划拨，所以很多情况下，联邦政府甚至没有足够信息来判断是否存在欺诈。
**English / Original**
But it doesn’t matter. Most of the fraud is non-discretionary. It’s basically fraudulent Medicare, Medicaid, Social Security, disability. There’s a zillion government payments. A bunch of these payments are in fact block transfers to the states. So the federal government doesn’t even have the information in a lot of cases to even know if there’s fraud.
## paragraph 1195
**中文**
不妨用归谬法。假设政府完美无缺、没有任何欺诈，你认为这种情况的概率是多少？零。好，那么你会说政府在防止欺诈与浪费方面有 90% 的效率吗？这也已经非常慷慨。
**English / Original**
Let’s consider reductio ad absurdum. The government is perfect and has no fraud. What is your probability estimate of that? Zero. Okay, so then would you say, fraud and waste at the government is 90% efficient? That also would be quite generous.
## paragraph 1196
**中文**
但即使效率有 90%，也意味着每年有 7500 亿美元的浪费和欺诈。而它并没有 90%，有效率不到 90%。
**English / Original**
But if it’s only 90%, that means that there’s $750 billion a year of waste and fraud. And it’s not 90%. It’s not 90% effective.
## paragraph 1197
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1198
**中文**
这种用第一性原理估算政府欺诈规模的方法有点奇怪。基本就是在问：你觉得有多少？
**English / Original**
This seems like a strange way to first principles the amount of fraud in the government. Just like, how much do you think there is?
## paragraph 1199
**中文**
总之，我们不必现场算，但我很想知道——
**English / Original**
Anyways, we don’t have to do it live, but I’d be curious—
## paragraph 1200
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1201
**中文**
你在 Stripe 很了解欺诈吧？人们一直在试图实施欺诈。
**English / Original**
You know a lot about fraud at Stripe? People are constantly trying to do fraud.
## paragraph 1202
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1203
**中文**
对，但正如你说的，这有点……我们确实已把欺诈率压得很低，不过问题空间略有不同，因为政府面对的欺诈路径比我们异质得多。
**English / Original**
Yeah, but as you say, it’s a little bit of a... We’ve really ground it down, but it’s a little bit of a different problem space because you’re dealing with a much more heterogeneous set of fraud vectors here than we are.
## paragraph 1204
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1205
**中文**
但 Stripe 能力很强，也非常努力；既有很强能力，又有高度责任心，可欺诈仍不为零。现在想象一下，规模大得多，能力却弱得多，责任心也少得多。
**English / Original**
But at Stripe, you have high competence and you try hard. You have high competence and high caring, but still fraud is non-zero. Now imagine it’s at a much bigger scale, there’s much less competence, and much less caring.
## paragraph 1206
**中文**
当年在 PayPal，我们努力把欺诈控制到支付额的约 1%。这非常困难，需要极高的能力与责任心，才能仅仅把欺诈降到 1%。现在想象一个责任心和能力都弱得多的组织，欺诈比例一定会远高于 1%。
**English / Original**
At PayPal back in the day, we tried to manage fraud down to about 1% of the payment volume. That was very difficult. It took a tremendous amount of competence and caring to get fraud merely to 1%. Now imagine that you’re an organization where there’s much less caring and much less competence. It’s going to be much more than 1%.
## paragraph 1207
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1208
**中文**
现在回头看参与政治和在其中做的事情，你有什么感受？从外部看，有两件事影响很大：一是 America PAC，二是当时收购 Twitter。但看起来也伴随许多痛苦。你会如何评价整段经历？
**English / Original**
How do you feel now looking back on politics and doing stuff there? Looking from the outside in, two things have been quite impactful: one, the America PAC, and two, the acquisition of Twitter at the time. But also it seems like there was a bunch of heartache. What’s your grading of the whole experience?
## paragraph 1209
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1210
**中文**
我认为，为了最大化未来美好的概率，这些事必须做。政治通常极具部落性，人们一谈政治往往就失去客观性，通常很难看到对方阵营的优点，或己方阵营的缺点。大体就是这样。我想，这是最令我惊讶的事情之一。
**English / Original**
I think those things needed to be done to maximize the probability that the future is good. Politics generally is very tribal. People lose their objectivity usually with politics. They generally have trouble seeing the good on the other side or the bad on their own side. That’s generally how it goes. That, I guess, was one of the things that surprised me the most.
## paragraph 1211
**中文**
你经常根本无法和人讲道理。只要他身处某一阵营，就会认定自己的阵营所做一切都是好的，另一个政治阵营所做一切都是坏的。想说服他们接受不同看法，几乎不可能。
**English / Original**
You often simply cannot reason with people. If they’re in one tribe or the other. They simply believe that everything their tribe does is good and anything the other political tribe does is bad. Persuading them otherwise is almost impossible.
## paragraph 1212
**中文**
但总体上，我认为这些行动——收购 Twitter、帮助 Trump 当选——尽管令许多人愤怒，却有益于文明。
**English / Original**
But I think overall those actions—acquiring Twitter, getting Trump elected, even though it makes a lot of people angry—I think those actions were good for civilization.
## paragraph 1213
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1214
**中文**
它如何推动你所期待的未来？
**English / Original**
How does it feed into the future you’re excited about?
## paragraph 1215
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1216
**中文**
美国必须足够强大，并维持足够久，才能把生命扩展到其他行星，也才能把 AI 与机器人技术推进到可以确保未来美好的程度。
**English / Original**
Well, America needs to be strong enough to last long enough to extend life to other planets and to get AI and robotics to the point where we can ensure that the future is good.
## paragraph 1217
**中文**
反过来说，如果我们滑向共产主义，或国家极度压迫的局面，就可能无法成为多行星文明。国家可能会扼杀我们在 AI 与机器人领域的进步。
**English / Original**
On the other hand, if we were to descend into, say, communism or some situation where the state was extremely oppressive, that would mean that we might not be able to become multi-planetary. The state might stamp out our progress in AI and robotics.
## paragraph 1218
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1219
**中文**
Optimus、Grok 等等，不只你的产品，任何追求收入最大化的公司产品，都会逐渐被政府利用。这种担忧如何体现在私营公司愿意向政府提供什么上？应该设置哪些护栏？
**English / Original**
Optimus, Grok, et cetera. Not just yours, but any revenue-maximizing company’s products will be leveraged by the government over time. How does this concern manifest in what private companies should be willing to give governments? What kinds of guardrails?
## paragraph 1220
**中文**
AI 模型是否应该完成签约政府要求它做的任何事？Grok 是否有权说：“实际上，即使军方想做 X，也不行，Grok 不会这么做”？
**English / Original**
Should AI models be made to do whatever the government that has contracted them out to do and asks them to do? Should Grok get to say, “Actually, even if the military wants to do X, no, Grok will not do that”?
## paragraph 1221
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1222
**中文**
我认为，AI 和机器人走偏的最大危险也许就是政府。反对公司或担心公司的人，真正最该担心的是政府。因为在极限意义上，政府就是一家公司，是拥有暴力垄断权的最大公司。
**English / Original**
I think maybe the biggest danger of AI and robotics going wrong is government. People who are opposed to corporations or worried about corporations should really worry the most about government. Because government is just a corporation in the limit. Government is just the biggest corporation with a monopoly on violence.
## paragraph 1223
**中文**
我一直觉得一种二分法很奇怪：人们认为公司是坏的，政府却是好的，而政府其实只是最大、最糟糕的公司。但人们确实抱有这种二分观念，莫名认为政府可以是好的、公司却是坏的，事实并非如此。公司的道德水平比政府更高。
**English / Original**
I always find it a strange dichotomy where people would think corporations are bad, but the government is good, when the government is simply the biggest and worst corporation. But people have that dichotomy. They somehow think at the same time that government can be good, but corporations bad, and this is not true. Corporations have better morality than the government.
## paragraph 1224
**中文**
我确实认为这值得担忧。政府可能利用 AI 和机器人压制民众，这是个严重隐患。
**English / Original**
I actually think it’s a thing to be worried about. The government could potentially use AI and robotics to suppress the population. That is a serious concern.
## paragraph 1225
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1226
**中文**
作为制造 AI 和机器人的人，你如何防止这种情况？
**English / Original**
As the guy building AI and robotics, how do you prevent that?
## paragraph 1227
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1228
**中文**
如果限制政府权力——这正是 US Constitution 真正意图做到的事——那么结果可能会比政府权力更大时更好。
**English / Original**
If you limit the powers of government, which is really what the US Constitution is intended to do, to limit the powers of government, then you’re probably going to have a better outcome than if you have more government.
## paragraph 1229
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1230
**中文**
所有政府都能获得机器人技术，对吧？
**English / Original**
Robotics will be available to all governments, right?
## paragraph 1231
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1232
**中文**
我不知道是否所有政府都能获得，很难预测。我可以说终点是什么，或多年后的未来是什么样，但很难预测一路上的具体路径。如果文明继续进步，AI 会远远超过全人类智能总和，机器人数量也会远多于人类。可沿途发生什么，非常难预测。
**English / Original**
I don’t know about all governments. It’s difficult to predict. I can say what’s the endpoint, or what is many years in the future, but it’s difficult to predict the path along that way. If civilization progresses, AI will vastly exceed the sum of all human intelligence. There will be far more robots than humans. Along the way what happens is very difficult to predict.
## paragraph 1233
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1234
**中文**
你能做的一件事似乎是直接规定：“无论哪个政府，都不得使用 Optimus 做 X、Y、Z。”写出一项政策即可。我记得你最近发帖说 Grok 应该有一部道德宪法，其中可以规定，我们限制政府如何使用这种先进技术。
**English / Original**
It seems one thing you could do is just say, “whatever government X, you’re not allowed to use Optimus to do X, Y, Z.” Just write out a policy. I think you tweeted recently that Grok should have a moral constitution. One of those things could be that we limit what governments are allowed to do with this advanced technology.
## paragraph 1235
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1236
**中文**
严格说来，如果政治人物通过一项法律，而且有能力执行，就很难不遵守。我们能拥有的最好制度是有限政府，并让行政、司法、立法三权之间形成适当制衡。
**English / Original**
Technically if politicians pass a law and they can enforce that law, then it’s hard to not do that law. The best thing we can have is limited government where you have the appropriate crosschecks between the executive, judicial, and legislative branches.
## paragraph 1237
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1238
**中文**
我之所以好奇，是因为到某个节点，限制似乎会由你来设定。你有 Optimus，也有太空 GPU……
**English / Original**
The reason I’m curious about it is that at some point it seems the limits will come from you. You’ve got the Optimus, you’ve got the space GPUs…
## paragraph 1239
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1240
**中文**
你觉得我会成为政府的老板？
**English / Original**
You think I’ll be the boss of the government?
## paragraph 1241
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1242
**中文**
SpaceX 已经如此。对某些关键任务——比如政府非常在意把某些卫星送入太空——它必须依赖 SpaceX，SpaceX 是不可或缺的承包商。
**English / Original**
Already it’s the case with SpaceX that for things that are crucial—the government really cares about getting certain satellites up in space or whatever—it needs SpaceX. It is the necessary contractor.
## paragraph 1243
**中文**
你正在构建越来越多未来技术组件，它们会在不同行业发挥类似作用。你可能有能力制定某种政策，比如任何形式压制古典自由主义的行为，“我的公司都不会提供任何协助”，或类似政策。
**English / Original**
You are in the process of building more and more of the technological components of the future that will have an analogous role in different industries. You could have this ability to set some policy that suppressing classical liberalism in any way… “My companies will not help in any way with that”, or some policy like that.
## paragraph 1244
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1245
**中文**
我会尽最大努力，确保自己控制范围内的一切都最大化人类的良好结局。其他做法都很短视，因为显然我也是人类一员，所以我喜欢人类，支持人类。
**English / Original**
I will do my best to ensure that anything that’s within my control maximizes the good outcome for humanity. I think anything else would be shortsighted, because obviously I’m part of humanity, so I like humans. Pro human.
## paragraph 1246
**中文**
02:38:28 — TeraFab
**English / Original**
02:38:28 - TeraFab
## paragraph 1247
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1248
**中文**
你提到 Dojo 3 将用于太空计算。
**English / Original**
You mentioned that Dojo 3 will be used for space-based compute.
## paragraph 1249
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1250
**中文**
你还真认真读我说的话。
**English / Original**
You really read what I say.
## paragraph 1251
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1252
**中文**
不知道你是否知道，Elon，你有很多关注者。
**English / Original**
I don’t know if you know, Elon, but you have a lot of followers.
## paragraph 1253
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1254
**中文**
这可暴露得太明显了。你怎么识破我的秘密？哦，原来是我把它们发在 X 上。
**English / Original**
Dead giveaway. How did you discern my secrets? Oh I posted them on X.
## paragraph 1255
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1256
**中文**
如何设计太空芯片？需要改变什么？
**English / Original**
How do you design a chip for space? What changes?
## paragraph 1257
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1258
**中文**
要把芯片设计得更耐辐射，并能在更高温度下运行。粗略来说，如果以开尔文计的工作温度提高 20%，散热器质量就能减半。所以在太空中以更高温度运行很有帮助。
**English / Original**
You want to design it to be more radiation tolerant and run at a higher temperature. Roughly, if you increase the operating temperature by 20% in degrees Kelvin, you can cut your radiator mass in half. So running at a higher temperature is helpful in space.
## paragraph 1259
**中文**
可以用各种方法保护内存。不过神经网络对比特翻转会非常有韧性。辐射造成的主要问题就是随机比特翻转。但如果模型有数万亿参数，只出现少数比特翻转，根本无关紧要。启发式程序会比巨型参数文件对比特翻转敏感得多。
**English / Original**
There are various things you can do for shielding the memory. But neural nets are going to be very resilient to bit flips. Most of what happens for radiation is random bit flips. But if you’ve got a multi-trillion parameter model and you get a few bit flips, it doesn’t matter. Heuristic programs are going to be much more sensitive to bit flips than some giant parameter file.
## paragraph 1260
**中文**
我只会把它设计成高温运行。我认为除了让它运行得更热，其他方面基本与地球上的设计方式相同。
**English / Original**
I just design it to run hot. I think you pretty much do it the same way that you do things on Earth, apart from making it run hotter.
## paragraph 1261
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1262
**中文**
太阳能阵列占卫星的大部分重量。有没有办法让 GPU 比 Nvidia、TPU 等正在规划的产品更强，而且这种方案在太空环境中尤其有优势？
**English / Original**
The solar array is most of the weight on the satellite. Is there a way to make the GPUs even more powerful than what Nvidia and TPUs and et cetera are planning on doing that would be especially privileged in the space-based world?
## paragraph 1263
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1264
**中文**
基本计算是：如果每个光罩尺寸芯片能达到约一千瓦，那么实现 100 吉瓦就需要一亿枚完整光罩尺寸芯片。根据对良率的假设，就能算出需要制造多少芯片。若要拥有 100 吉瓦电力，就需要一亿枚持续以一千瓦功率运行的光罩尺寸芯片。就是基本算术。
**English / Original**
The basic math is, if you can do about a kilowatt per reticle, then you’d need 100 million full reticle chips to do 100 gigawatts. Depending on what your yield assumptions are, that tells you how many chips you need to make. If you’re going to have 100 gigawatts of power, you need 100 million chips that are running at a kilowatt sustained, per reticle. Basic math.
## paragraph 1265
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1266
**中文**
一亿枚芯片取决于……看看 Blackwell GPU 之类的裸片尺寸，以及每片晶圆能切出多少颗，通常只有几十颗或更少。所以，如果每年都要产出这么多，就意味着每月生产数百万片晶圆。TeraFab 的计划就是这样吗？每月生产数百万片先进制程晶圆？
**English / Original**
100 million chips depends on… If you look at the die size of something like Blackwell GPUs or something, and how many you can get out of a wafer, you can get on the order of dozens or less per wafer. So basically, this is a world where if we’re putting that out every single year, you’re producing millions of wafers a month. That’s the plan with TeraFab? Millions of wafers a month of advanced process nodes?
## paragraph 1267
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1268
**中文**
对，可能超过一百万片之类。而且还必须制造内存。
**English / Original**
Yeah it could be north of a million or something. You’ve got to do the memory too.
## paragraph 1269
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1270
**中文**
你们要建内存晶圆厂吗？
**English / Original**
Are you going to make a memory fab?
## paragraph 1271
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1272
**中文**
我认为 TeraFab 必须制造内存，也必须制造逻辑芯片并完成封装。
**English / Original**
I think the TeraFab’s got to do memory. It’s got to do logic, memory, and packaging.
## paragraph 1273
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1274
**中文**
我非常好奇一个人该如何着手。这是人类制造过最复杂的东西。显然，如果有人能完成，你一定是其中之一。那么你意识到这是瓶颈，去找自己的工程师，会要求他们做什么？“我希望 2030 年每月生产一百万片晶圆。”
**English / Original**
I’m very curious how somebody gets started. This is the most complicated thing man has ever made. Obviously, if anybody’s up to the task, you’re up to the task. So you realize it’s a bottleneck, and you go to your engineers. What do you tell them to do? “I want a million wafers a month in 2030.”
## paragraph 1275
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1276
**中文**
没错，这正是我想要的。
**English / Original**
That’s right. That’s exactly what I want.
## paragraph 1277
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1278
**中文**
然后给 ASML 打电话吗？下一步是什么？
**English / Original**
Do you call ASML? What is the next step?
## paragraph 1279
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1280
**中文**
根本没什么好问的。
**English / Original**
No so much to ask.
## paragraph 1281
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1282
**中文**
我们先建一座小晶圆厂，看看效果如何。在小规模上犯完错误，再建一座大的。
**English / Original**
We make a little fab and see what happens. Make our mistakes at a small scale and then make a big one.
## paragraph 1283
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1284
**中文**
小晶圆厂已经建好了吗？
**English / Original**
Is a little fab done?
## paragraph 1285
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1286
**中文**
没有，还没建好。这个秘密不可能一直藏在袋子里，猫迟早会跑出来。会有无人机盘旋在那鬼东西上方，你们能在 X 上实时看到施工进度。
**English / Original**
No, it’s not done. We’re not going to keep that cat in the bag. That cat’s going to come out of the bag. There’ll be drones hovering over the bloody thing. You’ll be able to see its construction progress on X in real time.
## paragraph 1287
**中文**
说真的，我不知道。公平地说，我们也可能挣扎一番后彻底失败，成功并无保证。既然我们想制造大约一亿枚……我们的目标是到 2030 年拥有 100 吉瓦电力，以及能消耗这 100 吉瓦的芯片。供应商愿意给多少芯片，我们就要多少。我确实对 TSMC、Samsung 和 Micron 说过：“请更快建设更多晶圆厂。”我们保证买下这些晶圆厂的全部产出。所以他们已经在以最快速度推进。这件事要靠我们和他们共同完成。
**English / Original**
Look, I don’t know, we could just flounder in failure, to be fair. Success is not guaranteed. Since we want to try to make something like 100 million… We want 100 gigawatts of power and chips that can take 100 gigawatts by 2030. We’ll take as many chips as our suppliers will give us. I’ve actually said this to TSMC and Samsung and Micron: “please build more fabs faster”. We will guarantee to buy the output of those fabs. So they’re already moving as fast as they can. It’s us plus them.
## paragraph 1288
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1289
**中文**
有一种说法是，从事 AI 的人希望尽快获得极其大量的芯片，但许多上游供应商——晶圆厂以及涡轮机制造商——并没有快速扩大产量。
**English / Original**
There’s a narrative that the people doing AI want a very large number of chips as quickly as possible. Then many of the input suppliers, the fabs, but also the turbine manufacturers, are not ramping up production very quickly.
## paragraph 1290
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1291
**中文**
是的，他们没有。
**English / Original**
No, they’re not.
## paragraph 1292
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1293
**中文**
常听到的解释是，他们天性保守。故事可以说是因为他们是台湾人或德国人，就是不相信……这真是原因，还是另有缘故？
**English / Original**
The explanation you hear is that they’re dispositionally conservative. They’re Taiwanese or German, as the story may be. They just don’t believe... Is that really the explanation or is there something else?
## paragraph 1294
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1295
**中文**
这很合理……如果一个人在计算机内存行业做了三四十年……
**English / Original**
Well, it’s reasonable to... If somebody’s been in the computer memory business for 30 or 40 years…
## paragraph 1296
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1297
**中文**
他们经历过周期。
**English / Original**
They’ve seen cycles.
## paragraph 1298
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1299
**中文**
他们已经经历十次繁荣与萧条，积累了层层伤疤。繁荣时，一切看起来会永远美好；随后崩盘，他们拼命避免破产；接着又一轮繁荣，又一轮崩盘。
**English / Original**
They’ve seen boom and bust 10 times. That’s a lot of layers of scar tissue. During the boom times, it looks like everything is going to be great forever. Then the crash happens and they’re desperately trying to avoid bankruptcy. Then there’s another boom and another crash.
## paragraph 1300
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1301
**中文**
还有哪些想法，你认为别人应该去做，只是你目前出于种种原因没有亲自做？
**English / Original**
Are there other ideas you think others should go pursue that you’re not for whatever reasons right now?
## paragraph 1302
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1303
**中文**
有几家公司正在探索制造芯片的新方法，但扩张速度就是不够快。
**English / Original**
There are a few companies that are pursuing new ways of doing chips, but they’re just not scaling fast.
## paragraph 1304
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1305
**中文**
我甚至不只说 AI，而是泛指所有领域。
**English / Original**
I don’t even mean within AI, I mean just generally.
## paragraph 1306
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1307
**中文**
人们应该去做自己发现有强烈动力去做的事，而不是我随口建议的某个点子。他们应该做自己真正感兴趣、真正有动力去做的事情。
**English / Original**
People should do the thing where they find that they’re highly motivated to do that thing, as opposed to some idea that I suggest. They should do the thing that they find personally interesting and motivating to do.
## paragraph 1308
**中文**
不过回到限制因素……我已经把这个词说了大约 100 次。在三到四年的时间尺度上，我看到的当前限制因素是芯片；在一年的时间尺度上，则是能源、发电和电力。我不确定可用电力是否足够让所有正在制造的 AI 芯片通电。
**English / Original**
But going back to the limiting factor… I used that phrase about 100 times. The current limiting factor that I see in the three to four year timeframe, it’s chips. In the one year timeframe, it’s energy, power production, electricity. It’s not clear to me that there’s enough usable electricity to turn on all the AI chips that are being made.
## paragraph 1309
**中文**
到今年年底前后，我认为人们会真正遇到通电困难……芯片产量将超过让芯片投入运行的能力。
**English / Original**
Towards the end of this year, I think people are going to have real trouble turning on... The chip output will exceed the ability to turn chips on.
## paragraph 1310
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1311
**中文**
你打算如何应对那样的世界？
**English / Original**
What’s your plan to deal with that world?
## paragraph 1312
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1313
**中文**
我们正努力加快发电能力建设。我想这也许是 xAI 会成为领先者——希望是领先者——的原因之一。因为我们擅长硬件，能比其他人更快地让更多芯片通电。
**English / Original**
We’re trying to accelerate electricity production. I guess that’s maybe one of the reasons that xAI will be maybe the leader, hopefully the leader. We’ll be able to turn on more chips than other people can turn on, faster, because we’re good at hardware.
## paragraph 1314
**中文**
总体而言，那些自称实验室的公司之间，创新与想法往往会流动……差距很少超过大约六个月。人员流动也会把想法来回带走。
**English / Original**
Generally, the innovations from the corporations that call themselves labs, the ideas tend to flow… It’s rare to see that there’s more than about a six-month difference. The ideas travel back and forth with the people.
## paragraph 1315
**中文**
所以我认为，大家会撞上硬件墙，随后哪家公司能最快扩大硬件规模，哪家就会成为领先者。我认为 xAI 能最快扩大硬件规模，因此最有可能领先。
**English / Original**
So I think you sort of hit the hardware wall and then whichever company can scale hardware the fastest will be the leader. So I think xAI will be able to scale hardware the fastest and therefore most likely will be the leader.
## paragraph 1316
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1317
**中文**
你刚才自嘲又用了“限制因素”这个词，但我觉得这里确实有很深的东西。回顾我们一路讨论的许多话题，也许这很适合作为结尾。一家衰老、行动力低下的公司，会存在某个瓶颈，却没有真正采取任何措施。
**English / Original**
You joked or were self-conscious about using the “limiting factor” phrase again. But I actually think there’s something deep here. If you look at a lot of things we’ve touched on over the course of it, it’s maybe a good note to end on. If you think of a senescent, low-agency company, it would have some bottleneck and not really be doing anything about it.
## paragraph 1318
**中文**
Marc Andreessen 说过：“大多数人愿意忍受任何程度的慢性痛苦，只为避免急性痛苦。”我们讨论的许多案例，似乎都是主动迎向那种急性痛苦，无论它是什么。“好吧，我们必须弄清怎么使用钢材，或者怎么让芯片在太空运行。”我们愿意承受短期急性痛苦，真正解决瓶颈。这似乎是一条统一主线。
**English / Original**
Marc Andreessen had the line of, “most people are willing to endure any amount of chronic pain to avoid acute pain”. It feels like a lot of the cases we’re talking about are just leaning into the acute pain, whatever it is. “Okay, we got to figure out how to work with steel, or we got to figure out how to run the chips in space.” We’ll take some near-term acute pain to actually solve the bottleneck. So that’s kind of a unifying theme.
## paragraph 1319
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1320
**中文**
我的疼痛耐受度很高，这很有帮助。
**English / Original**
I have a high pain threshold. That’s helpful.
## paragraph 1321
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1322
**中文**
有助于解决瓶颈。
**English / Original**
To solve the bottleneck.
## paragraph 1323
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1324
**中文**
是的。有一点我可以说：我认为未来会非常有趣。正如我在 Davos 所说——我想我在那里只待了大约三个小时——从生活质量来看，宁可偏向乐观而判断错误，也不要偏向悲观而判断正确。偏向乐观会比偏向悲观更快乐，所以我建议大家宁可偏向乐观。
**English / Original**
Yes. One thing I can say is, I think the future is going to be very interesting. As I said at Davos—I think I was on the ground for like three hours or something—it’s better to err on the side of optimism and be wrong than err on the side of pessimism and be right, for quality of life. You’ll be happier if you err on the side of optimism rather than erring on the side of pessimism. So I recommend erring on the side of optimism.
## paragraph 1325
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1326
**中文**
为此干杯。
**English / Original**
Here’s to that.
## paragraph 1327
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1328
**中文**
很好。Elon，谢谢你参加这次访谈。
**English / Original**
Cool. Elon, thanks for doing this.
## paragraph 1329
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1330
**中文**
谢谢。
**English / Original**
Thank you.
## paragraph 1331
**中文**
Elon Musk（埃隆·马斯克）
**English / Original**
Elon Musk
## paragraph 1332
**中文**
好，谢谢你们。好。
**English / Original**
All right, thanks guys. All right.
## paragraph 1333
**中文**
John Collison（约翰·科里森）
**English / Original**
John Collison
## paragraph 1334
**中文**
耐力真强。
**English / Original**
Great stamina.
## paragraph 1335
**中文**
Dwarkesh Patel（德瓦克什·帕特尔）
**English / Original**
Dwarkesh Patel
## paragraph 1336
**中文**
希望这次访谈没有成为考验你疼痛耐受度的一种痛苦。
**English / Original**
Hopefully this didn’t count as a pain in the pain tolerance.