Yesterday I was at the Tesla shareholder meeting and the energy was great.

昨天我參加了特斯拉股東大會，現場氣氛非常熱烈。

Elon had announced the day prior that the two most important resolutions had been approved by shareholders.

埃隆在前一天宣佈，股東們已經準許了兩項最重要的決議。

So I won't go into that today but in this video I want to focus on two specific things.

所以今天我就不多說了，但在這段視頻中，我想重點談談兩件具體的事情。

First I want to talk about the unique position of Tesla to make the best humanoid robot and number two why being first matters with humanoid robots.

首先，我想談談特斯拉在製造最佳仿人機器人方面的獨特地位；其次，為什麼說在仿人機器人領域，先行者很重要。

All right in order for a company to make a successful humanoid robot I think it comes down to two main factors.

我認為，一家公司要製造出成功的仿人機器人，主要取決於兩個因素。

The first is the humanoid robot needs to be capable of doing many useful things that humans are able to do.

首先，仿人機器人需要能夠做許多人類能夠做的有用的事情。

If a robot is not able to do that then it lacks value and has no compelling features to really offer people.

如果機器人不能做到這一點，那麼它就缺乏價值，也就沒有真正能為人們提供的令人信服的功能。

The second thing is the humanoid robot needs to be affordable.

其次，人形機器人的價格必須低廉。

It makes no sense for a humanoid robot to cost millions of dollars because not many people can afford it.

一個仿人機器人要花費數百萬美元是沒有意義的，因為沒有多少人買得起。

It's much easier to hire humans at that price.

這個價格僱人要容易得多。

So in order to get to that point where a humanoid robot can be affordable and capable it takes a complete reinventing of every part and angle of a humanoid robot.

是以，要實現人形機器人的經濟性和功能性，就需要對人形機器人的每個部分和角度進行徹底改造。

You see typical robots that are made in the past haven't been good enough.

過去製造的典型機器人還不夠好。

They haven't been able to do many of the things that humans are able to do on a regular daily basis.

它們還不能做許多人類日常能做的事情。

So in order to make a truly capable humanoid robot Tesla needs to reinvent the humanoid robot hardware, the body.

是以，為了製造出真正有能力的仿人機器人，特斯拉需要重塑仿人機器人的硬件--身體。

They need to make the hardware with exceptionally better motors, gearboxes, sensors, power electronics, the whole robotic hardware system.

他們需要用更好的電機、變速箱、傳感器、電力電子設備以及整個機器人硬件系統來製造硬件。

The hardware itself needs to be capable of doing many of the things that humans are able to physically do like folding clothes, picking up tools.

硬件本身需要能夠完成許多人類身體能夠完成的事情，比如摺疊衣服、拿起工具等。

What Tesla has found is that there's no out-of-the-box solution or parts available for that type or standard of humanoid robot.

特斯拉發現，這種類型或標準的仿人機器人沒有現成的解決方案或零部件。

So Tesla has had to go to the drawing board and get their best electrical engineers, mechanical engineers, and other engineers involved in creating from scratch new motors, new gearboxes, sensors, power electronics, everything in the robot.

是以，特斯拉不得不回到繪圖板前，讓他們最好的電氣工程師、機械工程師和其他工程師參與進來，從頭開始創造新的電機、新的變速箱、傳感器、電力電子設備，以及機器人中的一切。

They've had to design from scratch every part of this new humanoid optimist robot.

他們不得不從頭開始設計這款新型人形樂觀主義機器人的每一個部件。

Further, Tesla has a huge lead with the AI computer.

此外，特斯拉在人工智能計算機方面也擁有巨大的領先優勢。

They've had their FSD computer in their cars since 2017 or so with hardware 2, then hardware 3, now hardware 4.

從 2017 年左右開始，他們就在汽車上安裝了 FSD 電腦，先是硬件 2，然後是硬件 3，現在是硬件 4。

Soon in a couple years it'll be hardware or AI5.

再過幾年，就會變成硬件或 AI5 了。

That inference computer in Tesla's cars is super powerful yet it's super efficient.

特斯拉汽車中的推理計算機功能強大，但效率極高。

It's exactly what's but also for a humanoid robot.

這正是人形機器人所需要的。

In other words, Tesla has a multi-year lead with this super important part of the humanoid robot which is the hardware brains.

換句話說，特斯拉在人形機器人的這一超級重要部分（即硬件大腦）上擁有多年的領先優勢。

So you've got the hardware body that is super difficult and Tesla has all of the pieces already in place to make, develop, and reinvent those parts to bring these parts together for the humanoid robot body.

是以，你已經擁有了難度極高的硬件機身，而特斯拉已經具備了製造、開發和重塑這些部件的所有條件，可以將這些部件整合到人形機器人機身中。

And then you've got AI hardware brain which is the AI inference computer and Tesla has a multi-year lead on that.

人工智能硬件大腦是人工智能推理計算機，特斯拉在這方面已領先多年。

And then you've got the AI brain itself or the software and that's the AI model which involves a robot knowing how to navigate to see things, learn things, and do things.

然後是人工智能大腦本身或軟件，這就是人工智能模型，它涉及機器人如何導航，如何看東西、學東西和做事情。

And Tesla has a multi-year lead on this as well through their full self-driving training.

而特斯拉通過全自動駕駛培訓，在這方面也領先了多年。

Now Tesla is applying a lot of that real-world AI experience into a humanoid robot.

現在，特斯拉正在將大量現實世界中的人工智能經驗應用到人形機器人中。

Tesla also has a lot of experience with manufacturing.

特斯拉在製造方面也擁有豐富的經驗。

They're one of the best manufacturers in the world.

他們是世界上最好的製造商之一。

They know how to design for manufacturing and how to manufacture at high volume.

他們知道如何進行生產設計，如何進行大批量生產。

You've got all these pieces that Tesla has uniquely and I think it's this unique combination that no other company in the world has right now.

特斯拉擁有所有這些獨特的元素，我認為這是目前世界上任何其他公司都不具備的獨特組合。

On top of that, you've got a visionary leader who gets it, who understands the potential of a humanoid robot.

除此之外，你還需要一位有遠見的領導者，他了解人形機器人的潛力。

He understands what it takes to make the best humanoid robot out and he has a uniquely powerful skill to focus and push his team forward toward aggressive goals.

他深知如何才能製造出最好的仿人機器人，而且他擁有獨特的強大技能，能夠集中精力，推動團隊朝著積極的目標前進。

When you put all of this together, you see that Tesla is in a unique position to make the best humanoid robot out there.

將所有這些結合在一起，你就會發現特斯拉在製造最好的仿人機器人方面擁有得天獨厚的優勢。

Alright, so the second point I want to dive into is why being first matters with humanoid robots.

好了，我想深入探討的第二點是，對於仿人機器人來說，為什麼 "先行 "很重要。

So the first reason I think is if you're first, you'll get more data.

所以，我認為第一個原因是，如果你是第一個，你就會得到更多的數據。

In other words, you'll have more robots out there and you're able to improve your AI models more into your AI models and make them better.

換句話說，你將會有更多的機器人出現，你也能更多地改進你的人工智能模型，讓它們變得更好。

Once you have millions of humanoid robots out there in people's homes and workplaces and you have all this data being collected, that can significantly improve how fast your AI models are becoming more capable.

一旦有數百萬個仿人機器人出現在人們的家中和工作場所，並收集到所有這些數據，就能大大提高人工智能模型的能力。

This is an advantage if you're first to the market.

如果您是第一次進入市場，這將是一個優勢。

Second, you're able to bring down the price of your product through economies of scale.

其次，你可以通過規模經濟降低產品價格。

If you're able to sell more of the product, the price per item will go down because you're able to utilize economies of scale.

如果你能賣出更多的產品，每件產品的價格就會下降，因為你能利用規模經濟。

As a result, if the price is cheaper, you're going to be able to sell more of them because demand will be higher.

是以，如果價格便宜，你就能賣出更多的產品，因為需求量會更大。

And then you're going to be able to cross over into profitability more quickly.

這樣，你就能更快地實現盈利。

This also causes big problems for every other company in the humanoid robot market because they're going to need a lot of volume in order to do the same thing.

這也給仿人機器人市場上的其他公司帶來了很大的麻煩，因為他們需要大量的產品才能實現同樣的目標。

Now you've already entered and taken your stake in the market.

現在，您已經進入市場並獲得了自己的股份。

In other words, the market isn't empty anymore.

換句話說，市場不再空虛。

And if you're dominating that market, other companies are going to either have to make a better humanoid robot than you, which is difficult because you have economies of scale already from being first and your humanoid robot is cheaper.

如果你在這個市場上佔據主導地位，其他公司要麼必須製造出比你更好的仿人機器人，而這是很難做到的，因為你已經率先實現了規模經濟，而且你的仿人機器人更便宜。

If they can't do that, then they're going to have to subsidize their introduction of the robot and lose a lot of money for a long time until they can get those economies of scale.

如果做不到這一點，他們就不得不為引進機器人提供補貼，並在很長一段時間內損失大量資金，直到獲得規模經濟效益。

Even that is difficult because how do you make a robot better than yours if you have the data from millions of robots on the field making your AI models better?

即使是這樣也很難，因為如果你擁有數百萬個機器人的數據，你如何讓機器人比你的機器人更好？

The other option is companies can try to focus on a more narrow use case of a robot rather than being a generalized humanoid robot.

另一種選擇是，公司可以嘗試將重點放在機器人更狹窄的使用案例上，而不是成為一個通用的仿人機器人。

They can have maybe some specialized functionality, but then that's a smaller market.

它們或許可以有一些專門的功能，但那是一個較小的市場。

I mean it could lead to profitability because maybe that narrow use robot can do certain things better and maybe it's cheaper than a generalized humanoid robot.

我的意思是，這可能會帶來利潤，因為也許這種用途狹窄的機器人能更好地完成某些任務，也許它比通用人形機器人更便宜。

However, it's a smaller market and it might be slower to reach economies of scale in a slower market.

但這是一個較小的市場，在一個較慢的市場中，要實現規模經濟可能會比較慢。

A generalized humanoid robot like Optimus can learn quite fast, so it could actually make these other narrow use cases perhaps irrelevant over time as it becomes more capable.

像 "擎天柱 "這樣的通用型仿人機器人學習速度相當快，是以，隨著時間的推移，隨著它能力的增強，這些其他狹窄的使用案例也許就不再重要了。

All right, a third thing about being first to humanoid robots is that you have mind share and marketing.

好了，首先進入仿人機器人領域的第三點是，你擁有心智共享和市場營銷。

If you're first to the market and you're first to really getting millions of humanoid robots out there, everyone will associate humanoid robots basically with your company, with your product.

如果你是第一個進入市場，第一個真正讓數以百萬計的仿人機器人出現在市場上，那麼每個人都會把仿人機器人與你的公司、你的產品聯繫在一起。

I think when you put all of these things together, they matter.

我認為，當你把所有這些事情放在一起時，它們都很重要。

For example, getting more data to make your AI models better, to make your product better, getting your price down, getting to economies of scale, basically taking over the market before others get in.

例如，獲取更多數據，使人工智能模型更好，使產品更好，降低價格，實現規模經濟，基本上在別人進入之前佔領市場。

Then other companies, they basically have to beat your product in capability and price, but that's extremely, extremely difficult for a startup to do in a field where there's a big first mover advantage.

然後，其他公司基本上必須在能力和價格上擊敗你的產品，但這對於一家初創公司來說，在一個有很大先發優勢的領域是非常非常困難的。

The first mover product is getting better quickly and the price is coming down as well.

先行者產品的性能正在迅速提高，價格也在下降。

Ultimately, I think the most important thing is to continue that pace of innovation because that pace of innovation matters the most ultimately.

歸根結底，我認為最重要的是繼續保持這種創新步伐，因為這種創新步伐最終是最重要的。

It doesn't matter if you're first to market, but you're slow with innovation after that.

如果你率先進入市場，但之後創新緩慢，這並不重要。

The others will catch up and will outpace you, but if you're first to innovation, quickly iterating, then you're going to be a formidable company that's extremely difficult to beat.

其他公司會迎頭趕上並超過你，但如果你率先創新，快速迭代，那麼你將成為一家極難被打敗的強大公司。

All in all, I think this is part of what I think Elon Musk means by opening up a new book at Tesla.

總而言之，我認為這就是埃隆-馬斯克在特斯拉開新書的部分含義。

The first book was Electric Vehicles and getting into Tesla Energy and Sustainable Energy.

第一本書是《電動汽車和特斯拉能源與可持續能源》。

This next book, however, I think is about Robotaxi and Optimus.

不過，我認為下一本書是關於機器人大喜和擎天柱的。

It's about these applications of people's lives and the world.

這是關於人們生活和世界的這些應用。

I think at the center of a lot of this is Tesla shareholders reaffirming their commitment and their approval of Elon Musk as the leader of Tesla.

我認為，這一切的核心是特斯拉股東重申了他們的承諾，以及對埃隆-馬斯克作為特斯拉領導者的認可。

I think the key thing that Elon brings to Tesla is that he is an accelerant of innovation.

我認為，埃隆給特斯拉帶來的關鍵因素是，他是創新的加速器。

Even he used that phrase at the shareholder meeting.

甚至他在股東大會上也用了這句話。

What that means is that when there's something that needs to improve, that needs to be developed or invented, or a new product has to come to life, Elon knows exactly how to get there.

這意味著，當有什麼東西需要改進、需要開發或發明，或者有什麼新產品需要問世時，埃隆知道如何去做。

He knows how to recruit the right talent.

他知道如何招募合適的人才。

He knows how to focus on the right things.

他知道如何專注於正確的事情。

He knows how to push his teams forward to the right goals.

他知道如何推動他的球隊朝著正確的目標前進。

He understands and gets how innovation works.

他理解並掌握了創新的運作方式。

This really matters because leadership is key in order to make all of this come together.

這一點非常重要，因為領導力是實現這一切的關鍵。

It's interesting and exciting because the humanoid robot is probably the biggest market out there.

這很有趣，也很令人興奮，因為仿人機器人可能是最大的市場。

The humanoid robot is probably the most substantial product, at least I could think of right now.

人形機器人可能是最有分量的產品，至少我現在能想到。

If things go well, I mean obviously there are risks involved, but if things go well, then Tesla I think has definitely a bright future with Robotaxi and Optimus humanoid robot.

如果一切順利，我的意思是，顯然會有風險，但如果一切順利，我認為特斯拉的 Robotaxi 和 Optimus 人形機器人一定會有光明的未來。

All right, hope that's been helpful and we'll see you guys in my next video.

好了，希望這對你們有所幫助，我們下期視頻再見。

I'll go ahead and end with the final three minutes or so of the shareholder meeting.

我將以股東大會的最後三分鐘左右結束髮言。

A great clip by Elon Musk diving into some of these points that I share in this video regarding Tesla's position to develop this humanoid robot and what it means in terms of potential.

埃隆-馬斯克（Elon Musk）在這段視頻中深入探討了其中的一些要點，我在這段視頻中分享了特斯拉開發這種仿人機器人的立場以及它在潛力方面的意義。

Tesla is going to be by far the leader in that.

在這方面，特斯拉將遙遙領先。

You're seeing a lot of robot startups, but I think it's actually very challenging to do Optimus as a robot startup because what we found to make Optimus work, we've had to design from first principles, from scratch, every part of electronics, the communication system, everything had to be done from scratch.

你們看到了很多機器人初創公司，但我認為，作為一家機器人初創公司，Optimus 實際上非常具有挑戰性，因為我們發現，要讓 Optimus 運行起來，我們必須從頭開始設計，從第一原理開始設計，每一部分電子設備、通信系統，一切都必須從頭開始。

We found that there's basically nothing, there's no supply chain.

我們發現基本上什麼都沒有，沒有供應鏈。

So even though there are many electric motors made in the world, there's no supply chain for the types of motors and sensors and gearboxes that are needed for a humanoid robot.

是以，儘管世界上有許多電動馬達，但仿人機器人所需的馬達、傳感器和變速箱類型卻沒有供應鏈。

Because it requires so much ground-up design, designing every motor, gearbox, sensor, power electronics from scratch, it's very hard for a startup to, if not impossible, for a startup to replicate that.

因為這需要大量的基礎設計，從零開始設計每一個電機、變速箱、傳感器和電力電子設備，對於初創公司來說，即使不是不可能，也是很難複製的。

But at Tesla, we have the world's best electrical engineering.

但在特斯拉，我們擁有世界上最好的電氣工程。

I think we've got the world's best mechanical engineering for gearboxes and for electric motors, power electronics.

我認為，我們擁有世界上最好的齒輪箱機械工程、電機和電力電子設備。

We have the resources to do that.

我們擁有這樣做的資源。

It applies quite well.

它非常適用。

And then you also have to have the brain.

然後，你還得有大腦。

You need a power-efficient inference computer, which we've got for the car, and we'll be using an Optimus.

你需要一臺省電的推理計算機，我們已經為汽車配備了這種計算機，我們將使用 Optimus。

You need to be the best in real world AI, and Tesla's the best in real world AI.

你需要成為現實世界中人工智能的佼佼者，而特斯拉就是現實世界中人工智能的佼佼者。

So you need all of these, you need a very strong hand of cards in order to make a compelling robot.

是以，你需要所有這些，需要一手非常強大的牌，才能製作出令人信服的機器人。

And then you also need to be very good at scale manufacturing.

然後，你還需要非常擅長規模化生產。

So in order to have the robot not cost like hundreds of thousands of dollars, in order to make it cost ten or twenty thousand dollars, you actually need to design for manufacturing and be very good at manufacturing.

是以，為了讓機器人的成本不至於高達數十萬美元，為了讓它的成本僅為一兩萬美元，你實際上需要為製造而設計，並且非常擅長製造。

In my experience, prototypes are easy compared to volume manufacturing.

根據我的經驗，與批量生產相比，原型製作很容易。

Prototypes are easy, production is hard, relatively speaking.

相對而言，原型容易，生產難。

So Tesla has the production capability, it has the engineering capability, and it has the AI hardware and software capability.

是以，特斯拉擁有生產能力、工程能力以及人工智能硬件和軟件能力。

And even the most optimistic estimates that I've seen for Optimus, the optimist, I think undercount the magnitude of what this robot will be able to do.

即使是我看到的對 "擎天柱 "最樂觀的估計，我認為也低估了這個機器人的能力。

As I said at the beginning of the presentation, I agree with the Arken best analysis that autonomous transport is called a five to seven trillion dollar market cap situation.

正如我在演講開始時所說的，我同意阿肯最佳分析的觀點，即自主運輸被稱為五到七萬億美元的市值。

Optimus, I think, is a twenty-five, literally twenty-five trillion dollar market cap situation.

我認為，Optimus 的市值高達 25 萬億美元。

Now, I don't want to trivialize what's necessary to get there.

現在，我不想輕視達到這一目標所需的一切。

It's an immense amount of work that is required to get there, super difficult.

要達到這個目標，需要付出巨大的努力，超級困難。

But we are moving very fast down that road, we're going to make it happen.

但我們正沿著這條路快速前進，我們一定會實現這一目標。

So, thank you.

所以，謝謝你。