Now I'm sure we can do a bigger round of applause for one of the greatest two leaders of our history.
我相信我們可以為這兩位歷史上最偉大的領袖之一，獻上更熱烈的掌聲。
00:43
Let's go ahead.
我們開始吧。
00:46
So we're talking about I lost count.
我們正在談論的，我都數不清了。
00:50
You know, $7 to $8 trillion worth of market cap, comp, I lost count.
你知道，價值 7 到 8 兆美元的市值，電腦（comp），我都數不清了。
00:56
But right now we're here to celebrate a historic moment.
但現在我們在此慶祝一個歷史性的時刻。
01:01
A moment that yesterday during the dinner, and thank you for joining us under the patronage of the honorable president and his Royal Highness, the conference, Musaidi, where we have the pleasure to hear firsthand.
This is the greatest alliance between the Kingdom of Saudi Arabia and the United States, where we have joined hands and you have helped us build our energy-based economy, fueling and energizing the industrial age,
and now fast forward going to the intelligence age, where we can fuel AI factories, robotics, EVs, and all of the rest.
現在快速前進到智慧時代，我們可以為人工智慧工廠、機器人、電動車以及所有其他事物提供動力。
01:43
Speaking of that, let's start with you, Elon, if you don't mind Jensen, feel free to chime in.
說到這個，讓我們從您開始，伊隆，如果您不介意的話，黃仁勳，請隨時插話。
01:48
You have a big fascination of something all of us have admired, first order thinking, which Jensen sometimes calls first order scaling, which is an opportunity for how you have dropped the cost of batteries from 1,000 for kilowatt-hour
您對我們所有人都欽佩的一件事有著濃厚的興趣，那就是第一性原理思考，黃仁勳有時稱之為一階擴展（first order scaling），這是一個機會，讓您如何將電池的成本從每千瓦時 1,000 美元
02:05
through sub-hundred bucks, and right now you're doing the same thing with robotics for actuators with servo rotors and motors.
降至 100 美元以下，而現在您正在為機器人的致動器、伺服轉子和馬達做同樣的事情。
02:16
So I want to hear from you, how do you manage to always disrupt every single industry with that thinking?
所以我想聽聽您的看法，您是如何設法總是用這種思維顛覆每一個行業的？
02:24
Well, it's mostly not disruption, it's creation.
嗯，這大部分不是顛覆，而是創造。
02:29
So with, say, SpaceX, with reusable rockets, they really weren't any reusable rockets.
所以，以 SpaceX 為例，透過可重複使用的火箭，它們實際上原本並沒有任何可重複使用的火箭。
02:39
But the essence of getting, of revolutionizing space travel is reusability.
但要實現太空旅行的精髓，革命性的太空旅行就是可重複使用。
02:45
If you throw the rocket away every time, the cost of access to space is extremely high.
如果你每次都丟掉火箭，那麼進入太空的成本將會非常高。
02:51
With respect to electric cars, there weren't any electric cars when we started making them, really.
關於電動汽車，當我們開始製造時，真的沒有任何電動汽車。
02:57
You couldn't buy any to the best of my knowledge.
據我所知，你買不到任何一輛。
03:01
So with Tesla, we wanted to make electric cars compelling and affordable, that was the goal.
所以，特斯拉的目標是製造引人注目且價格實惠的電動汽車。
03:08
The, you know, with respect to humanoid robotics, there are no useful humanoid robotics robots at this point.
在人形機器人方面，現階段沒有任何有用的ти人形機器人。
03:15
There are sort of gimmicks, but there are no actually useful humanoid robots.
有一些噱頭，但沒有真正實用的人形機器人。
03:20
And I think Tesla's gonna make the first actually useful humanoid robots.
我認為特斯拉將會製造出第一批真正實用的人形機器人。
03:25
And this will be quite a revolution.
這將是一場巨大的革命。
03:27
And I think something that everyone will want, because I always think of, like, who wouldn't want their own personal C3PO RTD2?
我認為每個人都會想要，因為我總是想到，誰不想要自己的個人 C3PO 或 R2D2 呢？
03:36
Oh, yeah?
哦，是嗎？
03:37
Of course, everyone will want one, right?
當然，每個人都會想要一個，對吧？
03:41
And then there would be many in industry providing products and services.
然後，工業界將會有許多公司提供產品和服務。
03:45
This is why I say that humanoid robots will be the biggest industry, or the biggest product ever.
這就是為什麼我說人形機器人將會是最大的產業，或有史以來最大的產品。
03:51
Bigger than cell phones or anything else, because everyone's gonna want one, and, or maybe more than one, and there'll be many in industry.
比手機或其他任何東西都大，因為每個人都會想要一個，或者可能不止一個，而且工業界將會有許多相關產品。
04:01
I use one RTD2 in C3PO's body.
我在 C3PO 的身體裡使用了一個 R2D2。
04:05
Yeah.
是的。
04:06
There you go.
就這樣。
04:07
Well, I mean, a humanoid robot will be better than RTD2 in C3PO combined times 10.
嗯，我的意思是，一個ти人形機器人將會比 C3PO 和 R2D2 加起來還要好十倍。
04:12
Yeah.
是的。
04:13
So the, and, you know, people often talk about sort of eliminating poverty and that kind of thing.
所以，你知道，人們經常談論消除貧困這類事情。
04:21
But really the, how long have they been talking about that?
但他們到底談論多久了？
04:25
There's lots of talk, you know, there's lots of NGOs sort of trying to do these things, but really not succeeding.
有很多討論，你知道，有很多非政府組織（NGOs）試圖做這些事，但實際上並未成功。
04:34
And, you know, the evidence speaks to herself.
而且，你知道，事實勝於雄辯。
04:38
But, but AI and humanoid robots will actually eliminate poverty.
但是，AI 和人形機器人將真正消除貧困。
04:43
And Tesla won't be the only one that makes them.
而且特斯拉不會是唯一製造它們的公司。
04:44
I think Tesla will pioneer this, but there will be many other companies that make humanoid robots.
我認為特斯拉會開創這條路，但會有許多其他公司製造人形機器人。
04:49
But there's only basically one way to make everyone wealthy, and that is AI and robotics.
但讓所有人都富裕起來基本上只有一種方法，那就是 AI 和機器人技術。
04:56
And we can't talk about robotics without AI factories.
而談到機器人技術，就不能不談 AI 工廠。
05:00
And yesterday was such a historic day for the two nations, but also for all of us, where we celebrate the AI strategic partnership with the US, signed witness by the honorable president and his Royal Highness about how we are committing
昨天對這兩個國家，也對我們所有人來說，都是歷史性的一天，我們慶祝與美國的 AI 戰略夥伴關係，由尊敬的總統殿下見證簽署，承諾我們將投入資本、能源和土地，來推動 AI-美國生態系統，建立推理節點、訓練節點，並成為 AI 應用最廣泛的國家。
05:17
our capital energy land to energize the AI-US ecosystem to be able to build inference, node training nodes, and to be the most AI-enabled nation.
隨著這項宣布，吉姆辛普森（Jim Simpson），請告訴我們 AI 工廠的下一步是什麼？
05:30
With that announcement, tell me what's next in AI factories, Jim Simpson.
有一個很棒的故事，講述沙烏地阿拉伯如何興建 AI 煉油廠，以及如何興建 AI 工廠，或者說從石油煉油廠轉型為 AI 工廠。
05:34
There's a beautiful story about how Saudi Arabia's building AI refineries and how building AI factories, or oil refineries to AI factories.
我很喜歡這個說法。
05:43
I love that.
我說過 AI 是一種基礎設施，原因在於，我們當然是從技術的角度來理解 AI，以及它如何顛覆每個行業。
05:44
I've said that AI is an infrastructure, and the reason for that is, of course, we understand AI from the perspective of the technology and how it's revolutionizing every industry.
數位智慧當然可以應用到每個領域。
05:55
Digital intelligence, of course, has applications into every field.
所以它將被每家公司、每個行業、每個國家使用。
06:01
And so it's gonna be used by every company, every industry, every country.
從這個角度來看，它是基礎性的，因此是基礎設施的一部分。
06:05
In that way, it's foundational, and therefore it's part of infrastructure.
從計算機科學的角度來看，AI 的新穎之處在於，過去的計算方式主要是基於檢索的計算。
06:10
What is new about AI, from a computer science perspective, is that the way computing was done in the past was largely retrieval-based computing.
Somebody typed in a story, or somebody created a piece of art, or came up with four versions of a digital ad, or it's all pre-built by somebody, which is then using a system to retrieve the appropriate version for you.
Hadoop and many of the frameworks and operating systems that have passed all designed to retrieve the appropriate information for you.
Hadoop 以及許多已過時的框架和作業系統，全都是為了為您檢索合適的資訊而設計的。
06:50
But today, software is going to be generated in real time.
但如今，軟體將會即時生成。
06:54
It's generative.
這是生成式的。
06:55
Based on the context, based on the circumstance, based on who you are, based on the problem you asked, based on your prompt, it will generate unique content for you.
它會根據情境、狀況、您的身份、您提出的問題以及您的提示，為您生成獨特的內容。
07:04
Every single time for everybody, it's unique.
對每個人來說，每一次都是獨一無二的。
07:07
When you use grok, every time you use it, it's different.
當您使用 Grok 時，每次使用體驗都會不同。
07:09
It's based on the prompt that you give it, and based on the circumstance.
這取決於您給它的提示以及當時的情境。
07:15
And so therefore, it used to be retrieval-based.
因此，過去是基於檢索的。
07:19
Today, it's generative.
如今則是生成式的。
07:21
And if it's generative, then, and every time is different, then you need AI factories all over the world to generate the content in real time.
既然是生成式的，而且每次都不一樣，那麼您就需要遍佈全球的 AI 工廠來即時生成內容。
07:30
Which is the reason why you need AI factories.
這就是為什麼您需要 AI 工廠的原因。
07:31
And this is a unique way of doing computation, but the benefit, of course, is that everything is pre-conceived and pre-documented, and it's contextually sensible and therefore intelligent.
So AI factories and robotics, and we heard it yesterday from his Royal Highness, his vision, how to augment our workforce with roughly tens of millions of robotics to be able to infuse the next wave of productivity and progress.
因此，關於 AI 工廠和機器人，我們昨天從殿下那裡聽到了他的願景，即如何利用大約數千萬台機器人來擴充我們的勞動力，以注入下一波的生產力和進步。
08:00
But this carries a lot of folks here when it comes to the future of jobs.
但這讓在座的很多人對未來的工作感到憂心。
08:04
So let's hear about your thoughts, Elon and Denson on that.
那麼，我們來聽聽伊隆（Elon）和丹森（Denson）對此的看法。
08:09
Sure, well, if you say, like, in the long term, where will things end up, long term, I don't know what long term is, maybe it's 10, 20 years, something like that, for me, that's long term.
Yeah, I mean, it'll be like playing sports or a video game or something like that.
是的，我的意思是，那會像是玩運動或電玩遊戲之類的事情。
08:38
If you want to work, you know, in the same way, like you can go to the store and just buy some vegetables, or you could grow vegetables in your backyard.
如果你想工作，你知道，就像你可以去商店買些蔬菜，或者你可以在自家後院種菜一樣。
08:49
It's much harder to grow vegetables in your backyard because some people still do it, because they like growing vegetables.
在自家後院種菜要困難得多，但有些人仍然這麼做，因為他們喜歡種菜。
08:55
That will be what work is like, optional.
那就是工作會變成什麼樣子，可選擇的。
08:59
And between now and then, there's actually a lot of work to get to that point.
而從現在到那時之間，實際上還有很多工作要做才能達到那個境界。
09:04
I'd always recommend people read in banks, culture books to get a sensible, what a probable, positive AI future is like.
我總是建議人們閱讀班克斯的「文化」系列書籍，以了解一個合理、可能且正面的AI未來會是什麼樣子。
09:16
And interestingly, in those books, money is no longer, doesn't exist.
有趣的是，在那些書中，金錢已不復存在。
09:21
It's kind of interesting.
這有點意思。
09:22
And my guess is, if you go out long enough, assuming there's a continued improvement in AI and robotics, which seems likely, the money will stop being relevant at some point in the future.
Now, there will still be constraints on power, like electricity and mass.
現在，電力和質量等能量方面仍會存在限制。
09:44
The fundamental physics elements will still be constraints.
基本的物理元素仍將是限制。
09:50
But I think at some point, currency becomes irrelevant.
但我認為在某個時刻，貨幣將變得無關緊要。
09:58
Jensen, any thoughts?
黃仁勳，有什麼想法嗎？
10:05
By the way, the NVIDIA earnings call is later today.
順帶一提，輝達的財報電話會議在今天稍晚舉行。
10:09
And by the way, since currency's relevant.
順帶一提，既然貨幣有其意義。
10:13
Cheers.
乾杯。
10:14
Cheers.
乾杯。
10:18
Elon just wants to share with you the breaking news.
伊隆只是想與你分享突發新聞。
10:26
The two of us who like to share some breaking news.
我們兩個人喜歡分享一些突發消息。
10:30
Listen, I would say there's different horizons you could look at.
聽著，我認為可以從不同的時間跨度來看這件事。
10:36
Everybody's jobs will be different.
每個人的工作都會有所不同。
10:38
That I think that's for sure.
我認為這是肯定的。
10:40
How the students learn will be different, how people do their work will be different, obviously, because a lot of the things that we do mundanely or arduously or very difficultly are going to be done very simply.
And so we're gonna be more productive from that sense.
從這個意義上說，我們的生產力將會提升。
10:58
One of the things that I will say is that for most people or a company, if your life becomes more productive, and if the things that you're doing with great difficulty become simpler, it is very likely, because you have so many ideas,
It is my guess that Elon will be busier as a result of AI.
我猜，伊隆將會因為 AI 而變得更忙。
11:20
I'm gonna be busier as a result of AI.
因為 AI，我反而會更忙。
11:22
And the reason for that is because we have so many ideas we wanna pursue, so many things that we still have in our backlog inside our company that we can go pursue.
原因在於，我們有太多想實現的想法，公司內部的待辦事項清單裡也還有許多可以執行的事。
11:30
If we were more productive, we can get to those things faster.
如果我們的效率提升，就能更快完成這些事。
11:35
And so in the near term, I would say that there's every evidence that we will be more productive and yet still be busier because we have so many ideas.
所以短期來看，我認為有充分的證據顯示，雖然我們的效率會提升，但因為我們有太多想法，我們反而會更忙。
11:43
One thing that I will say, give you some evidence, is that, and I was just telling Elon about this earlier, radiology, for example, has largely been converted to AI-driven radiology, and there's some really great companies doing that.
我想舉個例子來證明，我剛才也跟伊隆提過，例如放射科，現在大部分已經轉為 AI 驅動的放射科，而且有幾家非常棒的公司在做這件事。
11:59
And the surprising thing is the prediction that all radiologists would be the first jobs to go was exactly the opposite.
令人驚訝的是，原本預測放射科醫師會是第一個消失的工作，結果卻完全相反。
12:08
The trend shows that there are more radiologists being hired now as a result of AI.
趨勢顯示，因為 AI 的關係，現在反而聘用了更多的放射科醫師。
12:14
And the reason for that, if you take a step back, it's because the goal of a radiologist is not to study the images.
如果我們退一步思考，原因在於放射科醫師的目標並不是研究影像。
12:20
The goal of a radiologist is to diagnose a disease.
放射科醫師的目標是診斷疾病。
12:24
Now the studying of the images became so productive, they could study more in images, study more modalities, spend more time with the patients.
現在，影像判讀的效率大幅提升，他們可以分析更多影像、更多檢查模式，也能花更多時間與病人相處。
12:33
And as a result, they were actually accepting more patients or doing more radiology all around the world.
結果，他們實際上在全世界接受了更多病患，或執行了更多的放射檢查。
12:38
We're doing a better job of diagnosing disease.
我們在疾病診斷方面做得更好。
12:40
And so that's kind of the near-term outcome of AI and productivity.
這就是 AI 帶來的短期成果與生產力提升。
12:48
And we'll see what happens long-term.
長期的影響我們拭目以待。
12:53
When currency doesn't matter anymore, just let me know right before.
當貨幣不再重要時，請在那之前告訴我一聲。
12:59
You'll see it coming.
你會看到它來臨。
13:02
Let's see it coming.
讓我們一起見證它的到來。
13:04
We text the offense.
我們簡訊通知了防守方。
13:05
Yeah, we do.
是的，我們做了。
13:06
Yeah, just text it off.
對，就用簡訊通知。
13:08
Let me know more.
請再多告訴我一些。
13:08
I kind of agree with both of you, because if you look at every technological trend, every general-purpose technology has been net-new positive for the globe, for humanity, and so forth.
So one of them is Professor Omar Yagi, who's the first American Saudi to win a Nobel Prize in creating new chemistry.
其中一位是 Omar Yagi 教授，他是首位獲得諾貝爾獎的沙裔美國人，致力於創造新化學。
14:19
And the way he has done that, he has leveraged you AI accelerators and models like Grog to be able to create new chemistry when it comes to metal organic frameworks.
他之所以能做到這點，是利用了 AI 加速器和像 Grog 這樣的模型，來在金屬有機骨架（metal organic frameworks）領域創造出新化學。
14:30
Those are metal ions.
這些是金屬離子。
14:32
They're positively charged with organic linkers to be able to effectively create a sponge with 0.33 nanometers, pores, to capture water from air, and also to capture carbon dioxide.
The second story has also to do with AI, accelerated by NVIDIA, and with models like Grog, which is nano-pom, which is effectively creating a nano robot 500 nanometers by 1,000 nanometers to be able to do gene editing,
leveraging the CRISPR technology to take out sick ill cell disease.
利用 CRISPR 技術來移除致病的細胞。
15:08
Now, in both these instances, they originated 20 years ago in research, but AI was able to really accelerate the outcomes and the outputs such that we can move into new value pools.
現在，在這兩個案例中，它們都起源於 20 年前的研究，但 AI 能夠真正加速成果與輸出，讓我們得以邁入新的價值池。
15:21
So I think with every technological trend, humanity is going to always manage to shift to new value pools when it comes to workforce and productivity.
所以我認為，面對每一波科技趨勢，人類總能設法在勞動力與生產力方面，轉向新的價值池。
15:30
But we have some great announcements to talk about here today.
但今天我們在這裡有一些重磅消息要宣布。
15:33
Let's begin with you, Elon, the things that we're doing with XAI.
讓我們從你開始，Elon，談談我們在 xAI 方面的進展。
15:37
Yeah, we're excited to announce that we're doing a 500 megawatt, I mean, yeah, 500, sorry.
Yeah, yeah, yeah, yeah, it's 500 gigawatt won't have to wait.
是是是，是 500 吉瓦（gigawatt），不用等太久。
16:01
So that'll be eight Brazilian trillion dollars.
那大概會是 8 個巴西兆（Brazilian trillion）美元。
16:06
Not bad.
不錯喔。
16:12
So, yeah, we're doing XAI and King of Saudi Arabia are doing a...
所以，是的，我們正在與 xAI 以及沙特阿拉伯國王合作進行一項...
16:20
500 megawatt starting with 50 megawatts phase one and we're doing it with NVIDIA.
500 百萬瓦的計畫，第一階段從 50 百萬瓦開始，我們正與 NVIDIA 合作。
16:25
Yeah.
是的。
16:25
Congratulations to the human team, to target team, such a fantastic job.
恭喜人類團隊，目標團隊（target team），做得太棒了。
16:31
Jensen, I think we're also doing some great announcements this week.
Jensen，我想我們這週也要宣布一些很棒的消息。
16:35
We are?
我們有嗎？
16:36
Yeah.
是的。
16:37
We're announcing all kinds of things.
我們將宣布各種事情。
16:40
Our partnership with Humane is going incredibly well.
我們與 Humane 的合作進展得非常順利。
16:43
First of all, we worked together to get this company started and off the ground and just got an incredible customer with Elon.
首先，我們共同努力讓這家公司起步並順利運營，並且剛剛獲得了 Elon 這位令人難以置信的客戶。
16:53
Could you imagine a startup company, approximately zero billion dollars in revenues now going to build a data center for Elon?
你能想像一家營收近乎為零的初創公司，現在要為 Elon 建造數據中心嗎？
17:01
500 megawatts is gigantic.
500 兆瓦（megawatts）是非常龐大的規模。
17:04
This company is off the charts right away.
這家公司的表現立刻一飛沖天。
17:07
In addition to that, we're working AWS, as you know, is also coming...
除此之外，我們正在與 AWS 合作，如你所知，他們也將加入...
17:13
Congratulations to the humane team with AWS, starting with 100 megawatts with a gigawatt ambition and counting.
恭喜 Humane team 與 AWS 合作，從 100 兆瓦起步，目標是達到吉瓦（gigawatt）級別，並且數字還在持續增加。
17:22
So AWS is also coming to Humane.
所以 AWS 也將加入 Humane。
17:24
We're working with Humane on omniverse digital twins.
我們正與 Humane 合作開發 Omniverse 數位孿生（digital twins）。
17:29
As you know, that AI is not just...
如你所知，AI 不僅僅是...
17:32
well, just agentic AI and chatbots and cognitive AI is incredibly important to the world.
嗯，僅僅是代理型 AI（agentic AI）、聊天機器人和認知 AI 對世界就已經非常重要。
17:39
But AI applies to everything, chemicals and proteins and genes and physics and fluid dynamics and particles and, of course, robotics and activation.
但 AI 應用於萬物，包括化學物質、蛋白質、基因、物理學、流體動力學、粒子，當然還有機器人學和激活性能（activation）。
17:48
And we created this world called Omniverse where robots can learn how to be good robots.
我們創建了這個名為 Omniverse 的世界，讓機器人可以在其中學習如何成為好的機器人。
17:53
And it's physically based.
而它是基於物理原理的。
17:56
It obeys the laws of physics.
它遵循物理定律。
17:57
And so robots can learn in these environments.
因此，機器人可以在這些環境中學習。
18:00
And we're working with Humane to apply Omniverse to all kinds of digital factories and robotics and warehouses and things like that.
我們正與 Humane 合作，將 Omniverse 應用於各種數位工廠、機器人、倉庫等領域。
18:08
And so that's another.
這就是另一個例子。
18:10
We're also working in Saudi Arabia to build supercomputers to simulate quantum computers.
我們也正在沙烏地阿拉伯合作建造超級電腦，用來模擬量子電腦。
18:16
And using our computers to be the controller and the error correction, one quantum error correction requires an enormous amount of computation.
並利用我們的電腦作為控制器和進行錯誤校正，單是一項量子錯誤校正就需要龐大的運算量。
18:26
And so we're doing a lot of great work there, too.
因此，我們在那裡也做了很多出色的工作。
18:28
So a big partnership with Humane.
這是一項與 Humane 的重大合作。
18:30
They're off the charts, off the ground and off the charts at the same time.
它們的表現不僅突破天際，更已實際落地，成績斐然。
18:34
This is how we walk the talk in the Kingdom of Saudi Arabia in partnership with the U.S.
這就是我們與美國攜手合作，在沙烏地阿拉伯王國落實承諾的方式。
18:40
Yesterday, the president and his role line has announced the AI strategic framework and partnership.
昨天，總統及其團隊宣布了 AI 戰略框架與合作夥伴關係。
18:45
Today, we're going big with Elon and Jensen.
今天，我們將與 Elon 和 Jensen 展開重大合作。
18:49
So thank you for those opportunities.
感謝各位提供的這些機會。
18:55
Now, they told me I have time for two last questions.
現在，他們告訴我還有時間回答最後兩個問題。
18:59
So last night at the dinner, I got a number of questions because it seems that the schedule leaked.
昨晚的晚宴上，很多人向我提問，似乎是議程外流了。
19:06
And everybody was giving me hints about the last two questions I'm going to do.
大家都在暗示我最後兩個問題會問什麼。
19:11
So the first one was for you, Elon.
所以第一個問題是問你的，伊隆。
19:13
And there's a big one for you, Jensen.
還有一個大問題是問你的，黃仁勳。
19:17
So prepare for that one.
所以你要做好準備。
19:19
AI in space.
太空中的 AI。
19:21
Is that possible?
這有可能嗎？
19:23
Yes.
有可能。
19:24
If civilization continues, which probably will, then AI in space is inevitable.
如果文明延續下去——很可能會——那麼太空中的 AI 就是不可避免的。
19:32
You know, we just have to, like, preface that, you know.
我們得先做個前置說明。
19:38
We shouldn't take civilization for granted.
我們不該把文明視為理所當然。
19:40
We need to make sure to take care to ensure that civilization hasn't an upward arc.
我們需要謹慎確保文明是朝著向上的軌跡發展。
19:45
I mean, any student of history knows that civilization does not always have an upward arc and, in fact, civilizations have life cycles.
我的意思是，任何學過歷史的人都知道，文明並非總是向上發展，事實上，文明是有生命週期的。
19:54
So hopefully we are in a strong upward arc.
所以希望我們正處於強勁的上升期。
19:56
I think we are for now, but we don't want to take that for granted or be complacent.
我認為目前是，但我們不該視為理所當然，也不該自滿。
20:01
But the way to think of AI in space is that in order to achieve any meaningful percentage of a Kardashev 2-scale civilization where you're using even a millionth of the sun's energy you must have solar powered AI satellites in deep space.
但思考太空 AI 的方式是，為了達到卡爾達肖夫二級文明（Kardashev 2-scale civilization）的任何有意義比例——即便只是利用太陽能量的百萬分之一——我們必須在深空部署太陽能供電的 AI 衛星。
20:22
So once you realize, like, once you think in terms of a Kardashev 2-scale civilization, which is what percentage of the sun's energy are you turning it to useful work, then it becomes obvious that space is overwhelmingly
Earth only receives roughly 1-2 billionth of the sun's energy.
地球只接收到大約太陽能量的十億分之一到二。
20:51
So if you want to have something that is, say, a million times more energy than Earth could possibly produce, you must go into space.
所以，如果你想要擁有比地球可能產生的能量高出一百萬倍的東西，你就必須進入太空。
21:02
And so this is where it's kind of handy to have a space company, I guess.
這就是我認為擁有一家太空公司會派上用場的地方。
21:10
Sell the book.
把書賣掉吧。
21:11
Easier to cool chips in space, too.
在太空中冷卻晶片也更容易。
21:13
Easier to cool chips in space.
在太空中冷卻晶片更容易。
21:15
Yes, there's definitely no water in space, so you're going to have to do something that doesn't involve water.
是的，太空中絕對沒有水，所以你必須做不涉及水的事情。
21:20
Just hang out.
就待在那裡（散熱）。
21:21
Well, it's just got to radiate.
嗯，它只需要進行輻射散熱。
21:23
That's right.
沒錯。
21:26
So my estimate is that actually that the cost of electricity, like the cost effectiveness of AI in space will be overwhelmingly better than AI on the ground.
所以我的估計是，實際上在太空中運行 AI 的電力成本，或者說成本效益，將會比地面上的 AI 壓倒性地更好。
21:41
So long before you exhaust potential energy sources on Earth, long before, meaning I think even perhaps in the four or five-year timeframe, the lowest-cost way to do AI compute will be with solar-powered AI satellites.
所以早在你耗盡地球上的潛在能源之前——所謂「早」，我認為甚至可能在四到五年的時間範圍內——進行 AI 運算成本最低的方式將是使用太陽能供電的 AI 衛星。
22:00
So I'd say not more than five years from now.
所以我認為從現在起不會超過五年。
22:04
Wow.
哇。
22:05
And just look at the supercomputers we're building together.
來看看我們正在共同打造的超級電腦吧。
22:08
Let's say each one of the racks is two tons.
假設每個機櫃重達兩噸。
22:11
Out of that two tons, 1.95 of it is probably for cooling.
在這兩噸中，大概有 1.95 噸是用於散熱的。
22:15
Right.
沒錯。
22:16
Just imagine how tiny that little supercomputer is, right, each one of these GB300 racks would just be a little tiny thing.
想像一下那個小小的超級電腦有多麼微小，沒錯，這些 GB300 機櫃中的每一個都只會是個小小的東西。
22:23
And just electricity generation is already becoming a challenge.
光是發電就已經逐漸成為一個挑戰。
22:27
So if you start doing any kind of scaling for both electricity generation and cooling, you realize, okay, space is incredibly compelling.
所以如果你開始規劃發電與冷卻的規模，你會發現，空間的考量至關重要。
22:38
So, like let's say you wanted to do, I don't know, 200 or 300 gigawatts per year of AI compute.
舉例來說，假設你想每年處理 200 或 300 吉瓦（gigawatts）的 AI 運算量。
22:49
It's very difficult to do that on Earth.
在地球上這非常難以實現。
22:52
So the U.S. electricity usage, last time I checked, was around 460 gigawatts per year average usage.
根據我上次的了解，美國的年平均用電量大約是 460 吉瓦。
23:02
So something like say, you know, if you're doing 300 gigawatts a year, that would be like two-thirds of U.S. electricity production per year.
所以如果你每年要處理 300 吉瓦，這大約佔了美國年發電量的三分之二。
23:12
There's no way you're building power plants at that level.
你根本不可能興建這麼大規模的發電廠。
23:16
And then if you take it up to say a terawatt per year, impossible, like you have to do that in space.
如果再往上提升到每年 1 太瓦（terawatt），那是不可能的，你必須在太空中進行。
23:22
There just is no way to do a terawatt per year on Earth.
在地球上根本沒有辦法每年供應 1 太瓦的電力。
23:29
And in space, you've got continuous solar.
而在太空中，你有持續不斷的太陽能。
23:34
You actually don't need batteries because it's always sunny in space.
你其實不需要電池，因為太空中永遠是白天。
23:41
And the solar panels actually become cheaper because you don't need glass or framing.
而且太陽能板實際上會變得更便宜，因為不需要玻璃或框架。
23:46
And the cooling is just radiative.
冷卻則只需靠輻射散熱。
23:48
So that's why I think- That's the dream.
所以這就是為什麼我認為——那是理想的境界。
23:52
Yes.
沒錯。
23:52
That's the dream.
那就是理想的境界。
23:54
So Jensen, everybody last night was asking me, and I'm mindful it's Ernie's call for you today.
老實說，昨晚大家都在問我，而且我知道今天是 Ernie 要請教你。
24:00
So I'm going to say this delicately.
所以我會委婉地說。
24:02
Everybody has been asking me to ask you, are we going to have an AI bubble?
大家都一直要我問你，我們會有 AI 泡沫嗎？
24:07
That's the last question.
這是最後一個問題。
24:09
All right, let me just tell you what we see.
好吧，讓我告訴你我們觀察到的現象。
24:14
Okay, so I think it's really important when you look at what's happening around the world and go back to first principles of what's happening in computer science and computing.
所以我認為，當你審視全球正在發生的事情，並回歸到計算機科學與運算領域的根本原則時，這一點非常重要。
24:23
There are three things that's happening.
目前正發生三件事。
24:24
The first thing is that we all know that Moore's law has run its course and the ability, the demand for computing versus the amount of computation we can get out of general purpose computing is really challenging.
第一件事是，眾所周知摩爾定律已走到盡頭，運算需求與通用運算所能提供的算力之間，正面臨極大挑戰。
24:36
And so the world's been moving to accelerated computing for some time.
因此，全球早已開始轉向加速運算。
24:39
We've been pushing this now for some over 20 years.
我們推動這項技術至今已超過 20 年。
24:42
Let me give you one statistic.
讓我給你一個統計數字。
24:43
I was just at supercomputing.
我不久前才參加了超級運算大會。
24:45
Six years ago, CPUs were 90% of the world's supercomputers.
六年前，CPU 佔了全球超級電腦的 90%。
24:53
Top 500 supercomputers.
也就是前 500 大的超級電腦。
24:54
Six years ago.
六年前的情況。
24:56
This year, less than 15% went from 90% to 10% and meanwhile accelerated computing went from the other way.
今年，這個比例不到 15%，從 90% 降至 10%，與此同時，加速運算的占比則剛好相反。
25:04
10% to now 90%.
從 10% 增長到現在的 90%。
25:06
Okay, so you're seeing that inflection point, the transition in high performance computing from general purpose computing to accelerated computing.
所以你看到了這個轉捩點，也就是高效能運算從通用運算轉向加速運算的過程。
25:14
Well, one of the most data intensive, one of the most intensive computation things that the world does in cloud is data processing.
而雲端中，資料處理是數據最密集、運算強度最高的應用之一。
25:20
Several hundred billion dollars of computation is done on just raw data processing.
僅原始資料處理的運算支出就高達數千億美元。
25:26
Nothing to do with AI.
這與 AI 無關。
25:28
Just SQL processing, data frames, everybody's names, address, their sex, their age, where they live, how much money they make.
只是 SQL 處理、資料框架，包含每個人的姓名、地址、性別、年齡、居住地、收入等資訊。
25:36
All of that sits into a data frame.
所有這些都包含在一個資料框架中。
25:38
That data frame drives the world today.
這個資料框架推動著今日的世界。
25:40
Whether it's in banking or whether it's in credit cards or, of course, e-commerce and everything from ad recommendation, everything is driven off of that data frame.
The world that's the internet is so gigantic without a recommender system that the little tiny phone of us will have no chance of ever seeing the right information.
If you take that into consideration, you'll come to the conclusion that, in fact, what is left over to fuel that revolutionary agentic AI is not only substantially less than you thought, and all of it justified.
如果你把這點納入考量，就會得出結論：實際上，能用來驅動那種革命性代理式 AI 的資源，不僅遠比你想像的要少，而且每一項都合情合理。
27:25
Well, I was just informed by the team that my boss and your bosses is going to talk next, the honorable president and his royal highness, the conference, and hence we ran out of time.
嗯，我剛剛接到團隊通知，我的老闆和各位的老闆接下來要發言，就是尊敬的主席和殿下，大會因此時間已到。
27:36
But in essence, this is so much love for you, Elon, and Jensen.
但歸根結底，這對你、Elon 和 Jensen 充滿了深厚的愛。
27:46
But this, in essence, is a 92 alliance that shifted from energy to digital to the intelligence age, powered by pioneers such as Elon and Jensen to serve humanity and create, on a net new basis,
At [503], Muskstates: 'Mypredictionisthatworkwillbeoptional.'
在 [503]，馬斯克表示：「我的預測是，工作將變得可有可無。」
5WhydidthedemandforradiologistsincreasedespiteAIadvancements?為什麼儘管 AI 進步，放射科醫師的需求反而增加了？WhydidthedemandforradiologistsincreasedespiteAIadvancements?
為什麼儘管 AI 進步，放射科醫師的需求反而增加了？
✅ 正確！❌ 錯誤，正確答案是 C
At [744], HuangexplainsthatAIallowsradiologiststostudymoreimagesandspendmoretimewithpatients, increasingthroughput.
在 [744]，黃仁勳解釋 AI 讓放射科醫師能分析更多影像並花更多時間與病人相處，從而增加了處理量。
6WhatisthemainreasonJensenHuanggivesforwhytheAIindustryisnotabubble?黃仁勳認為 AI 產業不是泡沫的主要原因是什麼？WhatisthemainreasonJensenHuanggivesforwhytheAIindustryisnotabubble?
黃仁勳認為 AI 產業不是泡沫的主要原因是什麼？
✅ 正確！❌ 錯誤，正確答案是 B
At [1463], HuangexplainsthatMoore'sLawhasended, necessitatingashifttoacceleratedcomputing (GPUs) tomeetdemand.
