马克·扎克伯格谈人工智能泡沫与Meta新款显示眼镜 | ACCESS

我会的，是的。

I'm gonna yeah.

我会戴上它们。

I'll put them on.

是的。

Yeah.

来吧。

Come on.

是的。

Yeah.

它们的另一点是

That's Another

它们的另一个特点是，你根本看不出来它们有显示屏，即使

thing about them is you can't you can't tell they have a display, like, even when

是的。

Yeah.

在它们的过渡过程中，甚至在它们的传感器上。

In their transitions and even when their senses.

因此，这实际上是这项技术的一个重要部分，即光泄漏是一些类型波导的特性。

So that's actually an important part of the technology, is is light leakage is a feature of some types of waveguides.

所以基本上，你会遇到这些权衡，你希望它们非常高效。

And so basically, you get these trade offs where you want them to be very efficient.

有些波导需要注入大量的光，才能让任何东西显示出来。

There are waveguides where you have to pump just a ton of light through them in order to get anything to show up.

但另一些波导则会有不同的瑕疵，而且通常都是负面的。

But then some waveguides have just different artifacts, and usually in a bad way.

光线会照射到它们，你会看到各种彩虹状的瑕疵——另一个我们认为非常糟糕的瑕疵是隐私问题：当你看着的人基本上能

It's like the light will catch them, you'll see all kind of like rainbowing Another or artifact that we think is pretty bad because it's a privacy issue is if the person who you're looking at can basically

看到它。

See it.

是的。

Yeah.

看到什么

See what

最糟糕的情况是，他们能看见你在说什么。

Well, you very worst version of it would be if they could see what you're saying.

但我认为另一种仍然不太被社会接受的情况是，他们能察觉到你在看什么东西。

But I think another version that I think is still not that socially acceptable is if they can see that you're looking at something at all.

是的。

Yeah.

对吧？

Right?

所以我认为，我们在这项设计中真正引以为傲、并投入了大量精力的一点是：对你们来说，显示画面非常明亮，但对你所注视的人而言，他们根本无法察觉你在做什么。

So I think that that's one of the things that we're really proud of in the design here, and that we put a lot of energy into is the displays are super bright to you, and to the person that you're looking at, they can not even really tell that you're doing anything.

我认为，要让这项技术被社会接受，这一点非常重要。

I think that that's an important thing for having it to be socially acceptable.

设计时让显示内容略微偏向一侧。

Design them so that way when the display comes up, it's offset slightly to the side.

我们不希望妨碍你正在做的事情。

We don't want to block what you're doing.

眼镜的一个重要原则是，技术必须退到幕后。

An important principle for the glasses is the technology needs to get out of the way.

从根本上说，这是你一整天大部分时间都会戴在脸上的东西。

Fundamentally, it's like this is something that you're going be wearing on your face for a lot of the day.

我们的设计特别考虑了室内外使用，配合变色镜片，户外使用时效果非常好，就像太阳镜一样。

Mean, we designed these specifically to be both indoor and outdoor, that with the transition lenses, they work really well, sunglasses outdoors.

但事实上，你一整天的大部分时间并不会使用技术，至少不会使用视觉交互，对吧？

But the reality is that most of the day, you're not going to be using technology, or at least visual yet, right?

所以你可能只是在听音乐之类的，但当我们希望你与某物互动时，它就应该出现。

So maybe you'll be listening to music or something, but we want When you are interacting with something, it should show up.

它应该稍微偏向一侧。

It should kind of be off to the side.

如果你不与它互动，它需要迅速消失。

If you don't interact with it, it needs to get out of the way really quickly.

这是整个技术的一个非常重要的原则。

That's a really important principle of the whole technology.

这个通过手环的唤醒手势，只需轻点一下就能让显示消失。

This wake gesture with the band where you can just tap quickly to make the display go away.

是的。

Yeah.

非常非常微妙。

Very, very subtle.

而且，我想再谈谈这个手环。

And, again, I wanna get into the band.

这个手环本身就很特别。

The band is wild in its own.

对吧？

Right?

是的。

Yeah.

从我的演示中，最让我印象深刻的是，有些功能只有在这种形态下才能实现。

The thing that really stood out to me from my demo from these was some things that you only could do in a form factor like this.

因为，说实话，发短信很酷，但还有这个实时字幕功能，当时我在一个满是人的房间里，大家都开始大声说话。

Because, I mean, the texting is cool, but, like, there's this live captions thing where I was in a room with a bunch of people and they all started talking really loudly.

只要我看着某个人，它就会实时字幕显示他们说的话，同时屏蔽掉其他所有声音。

And I if I just looked at someone, it would live caption what they were saying and tune out everything else.

这简直像超级听力。

It's like super hearing.

嗯哼。

Mhmm.

而且你还在用它来做语言翻译。

And then you're also doing that with language translation.

是的。

Yep.

所以你可以实现实时翻译，理想情况下，两个人都戴着眼镜才能获得完整体验，但其实你并不一定需要对方也戴，你只需要听到或看到他们说的话被翻译成你的语言即可。

So you can do real time I mean, ideally, both people are wearing the glasses to get the full experience, but you don't actually need to have the other you could just hear what they're saying in your language or see it.

是的

Yeah.

这相当惊人，这也说明了这种形态能实现什么。

That's pretty wild, and that speaks to, like, what just this form factor could do.

我很想知道，确实有这一点。

I'm curious, like, there's that.

在你看来，这种形态还能做哪些智能手机做不到的独特事情？

Are there other things that this form factor uniquely can do in your mind that a smartphone can't?

我的意思是，所有与人工智能相关的事情，是的。

I mean, the all the things around AI Yeah.

你希望有一个AI能够了解你周围的状况。

Where you have an AI that basically you wanna have context around what's going on with you.

对吧？

Right?

所以，如果你想要一个能看见你所见、听见你所听、能全天候被动与你交流，并能上下文情境化地向你展示信息的AI，这确实是手机做不到的。

So, like, if you want an AI that can see what you see, hear what you hear, can just kind of talk to you passively throughout the day, and then can show you information contextually, that's just not something that a phone can do.

我的意思是，从技术上讲，你当然可以这样拿着手机走来走去，但其实没人会这么做

I mean, I guess, technically, you could walk around holding a phone like this, but you can't really do

没人会这么做。

No one does it.

你做不到的。

You can't.

那些演示早就存在了，我总是想，我真不想这样拿着手机。

Those demos have existed forever, I'm always like, I don't want to So hold my phone

我认为这实际上会成为最主要的一个用途。

I think that's actually going to be the main one.

而且我觉得所有的实时AI功能都很有趣。

And I think all the Live AI stuff, it's interesting.

它会带来一种不同的体验。

It takes on a different feel.

我们甚至在没有屏幕的经典RayBan眼镜上也实现了实时AI功能。

So we have Live AI in the RayBans without a display too, the classic RayBans.

而对于这一点，它只是音频形式的实时AI。

And for that, it's audio only Live AI.

所以当你独自做事情时，它非常有用。

So it's really helpful for when you're doing something kind of by yourself.

比如你在做饭的时候，它可以通过视频观察你正在做什么，你可以问它你应该怎么做，或者它能给你一些建议，这都很棒。

If you're cooking or something, then it can just give you It's watching what you're doing with the video, and you can ask questions about what you should be doing, or it can give you tips, that's all great.

但当你参与其他类型的对话时，它就不太有用。

But it's not really useful when you're in another kind of conversation.

我通过一些思想实验，也通过实际佩戴这些设备观察到的是：我们每天会经历数十次对话，而在每次对话中，我通常都有五件事想后续跟进。

And the thing that I've observed with the thought experiment I've run, but also just wearing these is, so we go through the world, we have dozens of conversations a day, and in every conversation I usually have five things I want to follow-up on.

也许它让我想起某件事该去做了，或者让我想起一个我想联系的人。

Maybe there's some It reminded me that I should go do a thing, or it reminded me of a person who I wanted to talk to.

或者当我跟某人交谈时，他们提出了一些听起来不太对的假设，我想去核实或凭直觉判断一下。

Or maybe I'm talking to someone, and maybe they assert some assumption that doesn't quite sound right, and I want to fact check it or gut check it.

我认为，借助实时AI，你可以让一个AI在后台运行，经常为你处理事务，然后将这些上下文带回——无论是异步地在对话结束后离线处理，还是在对话进行中，能即时获得更多信息也是非常有用的。

These are all things that I think with live AI, you can have this AI that's running in the background and that goes and often does work for you, and then can bring that context back, whether it's asynchronously offline when you're done with the conversation, or sometimes when you're in the middle of a conversation, it's just useful to have more context right then.

你平时是怎么使用这些设备的？

How have you been using these?

你团队里有人说你经常通过它们发短信。

Someone on your team was saying you text a lot through them.

是的。

Yeah.

嗯，我是个爱发短信的人。

I well, I'm a texter.

我基本上是通过短信来管理公司的。

I, like, run the company through text messages.

当你问起哪些事情是手机上做不到的，我想你当然可以在手机上发短信，我们每天都会发几十次。

So so and when you were asking about what, what can you do that you can't do on a phone, I guess you can obviously text on a phone, we all do it dozens of times a day.

但很多时候，发短信在社交场合是不合适的。

But I think that there's a lot of times where it's just not socially acceptable to send a text message.

比如说，你正在和别人交谈，却想问对方一个问题或获取一些信息。

And so let's say you're in the middle of a conversation, you want to ask someone a question or get some information.

我经常遇到这种情况。

I have this all the time.

我正在和某人进行一对一的对话，这时候我想问另一个人一个问题，获取一些背景信息，以便能更好地问眼前这个人对这件事的看法，但我不会在对话中途掏出手机。

I'm having a one on one conversation with someone, and I'm like, I wanted to ask someone this, or wanted to ask someone else a question to pull some context so I can ask this person what they think about it, but I'm not going to pull out my phone in the middle of a conversation.

用这个的话，实际上非常快。

With this, it's actually just super quick.

只需五秒钟就能发条消息，然后马上获得背景信息。

Can just send a message in five seconds, get the context back.

我发现，这真的大大提升了你正在进行的对话质量。

It actually just really improves the conversations that you're having, I find.

对我来说，这正是我认为Zoom相比面对面交流更胜一筹的地方——你可以稍微多任务处理一下。

To me, this is the one thing that I think is basically better about Zoom than in person conversations is that you can sort of multitask a little bit.

对吧？

Right?

就像很多情况一样。

It's like Or a lot.

但在几乎所有其他方面，它都比面对面的实体对话更糟糕。

It's it's it's worse in basically, like, every other way than than kind of an in person physical conversation.

但我认为有用的一点是，你可以从对话直接转去问别人一个问题。

But the one thing that I that I think is useful is you can go from having a conversation to basically asking someone else a question.

这并不一定分散注意力，而是起到补充作用，对吧？

It's not necessarily distracting, it's additive, right?

否则你的选择就是：先进行一场对话，然后去询问别人，再回去重新联系另一个人，再来一次完整的对话。

Because otherwise your option is like, all right, you have a conversation, then you go check-in with someone else, then you have to go back and call the other person back and have a whole second conversation.

所以它经常能直接跳过这些繁琐的步骤。

So it just short circuits these things all the time.

现在我觉得，这种做法能把线上对话的这一优势带入到面对面的对话中。

And now I think this kind of brings the best part of that into physical conversations.

你基本上能全身心投入到对话里。

Well, you basically feel present in the in the conversation.

你可以随时调取你需要的任何信息。

You can pull in whatever information you need.

这非常非常有帮助。

It's super, super helpful.

是的。

Yeah.

这个产品真正令人震撼的地方是这个手环。

A real, like, holy shit thing about this product is the band.

我一开始以为它和你们去年展示的Orion是同一种东西，当时我就有这种感觉。

I I thought that because it was with Orion, the demo you guys had last year too, and I thought it at the time.

我觉得这个手环有些特别之处。

I was like, there's something special with this band.

你们称它为神经手环。

And you're calling it's a neural band.

对吗？

Is that right?

你们就是这么叫它的吗？

That's what you're calling it?

这是因为这是一个神经接口。

It's a because it's a neural interface.

它是一个神经接口，它是

It's a neural It's

所以，它感觉就像在读取你的思维。

So like, it feels like it's reading your mind.

它其实并没有那样做，但是

It's not doing that, but

它并没有读取你的思维。

It's not reading your mind.

你实际上是在通过你的肌肉神经系统发送信号，它在你做出动作之前就能捕捉到这些信号，它基本上能捕捉到这些微小动作，从而让你无论手在什么位置都能控制界面。

What you're doing is you're sending signals through your muscular nervous system that it actually picks up before you even make movements, but it basically picks up these micro gestures, and that allows you to control your interface no matter where your hand is.

所以它并不是通过视觉进行手部追踪之类的操作。

So it's not doing hand tracking visually or anything like that.

你可以把手放在身体侧面。

You can have your hand by your side.

你可以把手放在背后，或者插在夹克口袋里，都没问题。

You can have your hand behind your back, whatever in your jacket pocket, and it's fine.

这些手势非常细微。

And the gestures are really subtle.

对吧？

Right?

所以，就像，我只需要做这个动作就能调出界面。

So, like, it's like, this is all I need to do to bring it up.

我的意思是，这个动作就能调出MedAI。

I mean, this brings up MedAI.

我真的很喜欢音乐那个功能。

So and I really like the music one.

你有没有试过

Did did you

我没试过那个。

I didn't try that one.

没有。

No.

哦，所以当你听音乐时调整音量的方式是，嗯，你只是稍微……

Oh, so the way when you're listening to music that you adjust the volume Uh-huh.

就是你只是

Is you just kind of

哦，那个旋钮吗？

Oh, the dial thing?

是的。

Yeah.

你假装有个旋钮，然后轻轻转动它。

You pretend that there's a dial, and you just turn the dial.

我用过类似的手势来放大和做其他操作。

I did that with the zoom in and something too.

你也可以在照片上这么做。

You could do it you could do it on on photos too.

是的。

Yeah.

当你这么做的时候，感觉就像《少数派报告》里的场景，就像在现实生活中一样。

That's like it feels like Minority Report when you do that It's like in real life.

这是一个很好的界面。

It's a it's a good interface.

是的。

Yeah.

我认为这就是你的意思。

I think it's what you mean.

但，没错，这不像《少数派报告》里那些奇怪的部分。

But, yeah, it's like Not not the weird part of Minority Report.

它确实是，像是其中一部分。

It's yeah, it's like part of

《少数派报告》里汤姆·克鲁斯在做手势，就是那样，是的。

Minority Tom Cruise is doing the hands, like Yeah.

这感觉就像科幻片。

It's it feels like sci fi.

而且是的。

And Yeah.

我在想，为什么是这个乐队？你们为什么选择这个乐队作为这个的输入？

And I'm wondering why this band like, why did you land on this band as the input for this?

因为人们一直在尝试为这种眼镜寻找输入方式，通常是语音或手势之类的。

Because people have been trying to figure out input for glasses like these forever, and it's usually voice or hand gestures or something.

但我是不会在地铁里，对着空中比划的。

But it's like, I'm not gonna be in the subway, like, gesturing out into space.

所以好吧。

So okay.

我觉得那些也会很有用，但我不认为它们是完整的。

I think that those are going to be useful too, but I don't think that they're I don't think that they're complete.

对吧？

Right?

所以语音显然会很重要。

So voice is is obviously gonna be important.

人们会与MedAI对话。

People talk to MedAI.

他们会进行语音通话。

They do voice calls.

你知道的，你们也会进行视频聊天。

They you know, you do video chats.

所以语音会是一个重要的部分，但现实是，我们很多时候都在他人身边。

So voice is gonna be a big thing, but I think the reality is that a lot of the time we're around other people.

而对于我们真正想攻克的使用场景——我认为这也是我们在手机上最频繁、最重要的使用场景——那就是发消息。

And for the use case that we really wanted to nail, which I actually think is the most frequent and most important use case that we do on our phones, is messaging.

所以如果你想完美实现这一点，就需要让消息进来时不会干扰你，不会出现在你视野正中央，而是静静地在那里。

So if you wanna nail that, what you need is the ability for a message to come in, to not be distracting, not be like center in your field of view, but just be there.

然后你需要一种方式，无论身处何种情境，都能在五到十秒内快速回应，这种方式不会打断你当前的互动，符合社交礼仪，也不会显得无礼。

And then you need a way in whatever situation you're in to be able to quickly respond in like five or ten seconds in a way that is not interruptive to your current interaction and is socially acceptable and doesn't feel rude.

然后你就到了手势操作这一部分。

And then you get to, okay, hand gestures.

我的意思是，确实会有一些有用的应用，比如《少数派报告》里，他做了不少这样的动作。

I mean, yeah, I think that there are going to be useful things for, yeah, I mean, in Minority Report, he's doing a fair amount of that.

但对于游戏之类的事情，我觉得你会用到这些手势。

But for gaming and things like that, I think you'll do that.

但就像你说的，你不可能走在街上也这样比划。

But like you said, you're not gonna walk down the street like that.

对吧？

Right?

我的意思是，那样做有点

Mean, that's that's kind of

太滑稽了。

Pretty goofy.

是的。

Yeah.

看起来很奇怪。

It looks it looks weird.

你的手臂会累。

Your arms get tired.

是的。

Yeah.

你知道，从它不奏效的原因来看，前者远比后者更重要。

You know, it's much more the former than the latter in terms of reason why it doesn't work.

是的。

Yeah.

但后者也是对的。

But the latter is also true.

所以我们需要一种基本上是安静且微妙的方案。

So we needed something that was basically silent and subtle.

所以问题来了，这有几个可选方案。

So the questions are, there are a few options for that.

人们正在研究的一种方式基本上是耳语，对吧？

One that people are working on is basically whispering, right?

所以你可以通过亚听觉方式捕捉到声音，或者使用摄像头观察你的嘴唇进行唇读。

So you can sub audibly even pick up on the sound, or you could have some camera that can look at your mouth and do lip reading.

在会议中这仍然相当奇怪。

That's still pretty weird in a meeting.

我同意。

I agree.

是的。

Yeah.

我同意。

I agree.

尽管它是静音的，但它没能通过我对隐蔽性的标准，所以我认为你需要转向神经接口。

So it didn't pass my bar for subtlety, even though it is silent in So think you need to go for the neural interface.

神经接口的另一个优点是，你可以实现极高的带宽输入。

And the other thing that's nice about the neural interface is you can get really high bandwidth input too.

所以这不像现在的智能手表，你基本上只能移动手臂，它能识别一两个手势，但带宽非常低。

So it's not like With smartwatches today, you basically You can move your arm, and it can pick up a gesture or two, but it's a very low bandwidth.

它能做的事情并不多。

There aren't that many things that it can do.

你需要一种能够读取肌肉信号的设备。

You need something that can basically be reading the muscle signal.

这样你就可以非常微妙地控制它。

So that way you can just control it very subtly.

而这个设备就能做到这一点。

And and this can do that.

我的意思是，我已经能了，虽然我们还没那么优化？

I mean, I can I can I mean, already, I I haven't, like you know, we're not that optimized?

我觉得自动纠错会变得更好，但目前我已经能每分钟输入大约30个单词了。

I think we're gonna get the autocorrect a lot better, but I'm already, I think, around, like, 30 words a minute type.

真的吗？

Really?

是的

Yeah.

不

No.

是的

Yeah.

是的

Yeah.

天啊，我是说，以它目前的形态，这个手环能实现多先进的功能？

Man, I mean, how how advanced do you think the band gets in its current form factor in terms of what it can do?

我觉得会相当厉害。

I think quite a bit.

我的意思是，现在你已经有传感器可以捕捉肌肉信号了。

I mean, basically, today well, I guess you have the sensors which can pick up the signals from your muscles.

但除此之外，这本质上就是一个AI机器学习问题，就是要能准确理解你动作的意图。

But then on top of that, it's basically just an AI machine learning problem to be able to pick up what you mean by the thing.

而现在，它并没有特别个性化。

And right now, it's not particularly personalized.

你刚拿出来时，它需要适应某些手势，而你之前从未使用过它。

You get it out of the box, and it needs to work with certain gestures, and you've never used it before.

所以它现在的工作方式是，比如你正在做的这个动作——这其实不算大动作，但相比未来它应该达到的精度，还是太大了。

So it works with these, like you're doing kind of It's like, I mean, this isn't a big gesture, but I mean, this is much bigger than what it needs to be in the future.

至于神经文字输入，你可以想象自己拿着一支微型铅笔，只是在写每个字母。

And then for kind of the neural text entry, you're basically You can kind of think about it as if you have a mini pencil and you're just writing out the letter.

但随着时间推移，AI应该学会你书写每个字母的模式，你就能做出越来越细微、几乎看不见的动作，而系统会逐渐识别出这些就是你表达该字母或输入的方式。

But over time, what should happen is that the AI learns your pattern for how you write each letter, And you should be able to write, make increasingly subtle and invisible motions that it basically learns are your way of doing that letter or that input or whatever.

我认为未来的版本中，这些动作会变得极其细微，你实际上只是让肌肉相互对抗，完全不产生可见的移动，而系统仍能捕捉到。

And I think the future version of this is that the motions just get really subtle, and you're effectively just firing muscles in opposition to each other, and making no visible movement at all, and it picks it

上。

up.

所以，本质上就是通过手腕实现个性化自动补全。

So personalized auto complete via your wrist, basically.

是的，速度非常快，因为如果你没有移动，就不会有实际物理移动和收回所带来的延迟。我认为这方面的上限非常高。

Yeah, so super fast, because if you're not moving, there's no latency from actually having to actually physically move and then retract after So making a I think the upper bound of that is very high.

另一点是，这仅仅针对打字这一种输入方式，但我认为还有许多其他维度可以利用它来与各种事物交互，比如控制空间中的一个虚拟手来操作用户界面。

The other thing is and that's just for typing, which is one modality, but I think you can kind of there are all these different dimensions where you can use it to input into things, and you could control a a hand in space that is like operating a UI.

对吧？

Right?

我的意思是，它能做的事情多种多样，我相信随着时间推移，这些应用会变得非常有趣。

I mean, it's like there's all kinds of different things that it can do that I think will just be really interesting to get into over time.

我们基本上是为眼镜开发了神经带，但我实际上认为，神经带本身最终可能成为一个独立的平台——

And we basically, we invented the neural band to work with the glasses, but I actually think that the neural band could end up being a platform on its own-

我同意。

I agree.

一旦它变得更成熟，就可以用来与你所有的电子设备和装置交互，完成各种不同的操作。

To basically just interact with all of your electronics and devices and do all kinds of different things once we of get it to be a little bit more mature.

所以你可以为它开发一个API，理论上可以接入智能家居之类的系统。

So you could have an API for it that could theoretically plug into a smart home or something like that.

那太疯狂了。

That'd be wild.

是的。

Yeah.

是的。

Yeah.

是的。

Yeah.

是的。

Yeah.

这些设备的价格也比我预期的要低。

The the price point for these is also lower than I expected.

只要800美元左右。

It's only, like, $800.

是的。

Yeah.

这些产品是面向谁的？是给早期采用者的吗？

Who are these for from a like, is this an early adopter thing?

你关心前沿技术吗？

Like, you care about the cutting edge?

这正是为你准备的。

Like, this is for you.

你们没生产太多这种产品，对吧？

You're not making a ton of these.

对吧？

Right?

我猜这不会成为你们的一个大规模产品。

I I assume this is not, like, a mat gonna be a massive thing for you.

更多是为了观察人们如何使用这项技术。

It's more to to see how people use this technology.

我觉得这将成为未来的重要组成部分。

What's I I think that this is going to be a big part of the future.

我的观点是，今天有大约二十亿人每天佩戴眼镜来矫正视力。

I mean, my view is that there's between a 2,000,000,000 people who wear glasses on a daily basis today for vision correction.

在五到七年后，是否有可能绝大多数这类眼镜都以某种形式变成智能眼镜？

Is there a world where in five or seven years the vast majority of those glasses are in AI glasses in some capacity?

我认为这就像iPhone刚推出时，人人都在用翻盖手机，只是时间问题，最终都会变成智能手机。

I think that that's it's kind of like when the iPhone came out and everyone had flip phones, and it's just a matter of time before they all become smartphones.

我认为这些都会变成智能眼镜。

I think that these are all going to become AI glasses.

问题是，世界上有八十亿人，不是二十亿到四十亿。

And the question is, all well, there are 8,000,000,000 people in the world, not one to two.

那么，会不会有更多人也佩戴眼镜呢？

So are there going to be a bunch of other people who also wear glasses?

我猜测会的。

I would guess yes.

我的意思是，还有更多人有时会戴太阳镜。

I mean, there are a lot more people who wear sunglasses some of the time.

所以，是的，我认为这是一个很大的领域。

So yeah, I mean, I think it's a big category.

而且这里有很多事情正在发生。

And there's a lot going on here.

我的意思是，当你在开发产品时，你会发现，第一版你构建的是你认为会很棒的东西，但你会收到很多反馈，同时第一版也并非完美无缺。

I mean, there's I think what you see when you're building products is that V1, you build what you think is going to be great, and then you get a lot of feedback, but you also didn't get everything exactly perfect in V1.

你知道，第二版和第三版通常会好很多。

You know, V2 and V3 just end up a lot better.

对吧？

Right?

我认为，Ray-Ban Ray-Ban Stories的第一版出来时，我们觉得它还不错，这并不是巧合。

It's not a it's not a coincidence, I think, that, you know, the first version of the Ray Ban's Ray Ban stories, and we thought it was good.

当我们推出第二代Ray-Ban Meta时，销量增长了五倍，因为产品得到了优化。

Then when we did the second version of Ray Ban Meta, I think it sold five times more, and it was just refined.

所以我认为，这里会有一种动态：你推出第一版，然后从中学习。

So I think that there's going to be some dynamic here where it's like you have the first version, you learn from it.

第二代产品要精致得多，软件也同样得到优化，而不仅仅是硬件，这种改进会不断累积，变得越来越好。

The second version is just a lot more kind of polished and take and and it's and the software gets polished too, not just the hardware, and that just kind of compounds and gets better and better.

而能够完全填充你视野的全息增强现实，也还在路上。

And full AR that fills your vision, that's still coming too.

是的。

Yeah.

我的意思是，我们正在研发所有这些技术，并希望让它们尽可能地亲民。

I mean, we're working on all this stuff, and we want to get it all to be as affordable as possible.

现实是，你往眼镜里塞入的技术越多，成本就越高，因为你加入了更多的组件。

The reality is that the more technology you want to cram into the glasses, the more expensive that is because you're putting more components in.

我们也希望眼镜尽可能轻薄，这需要通过微型化技术逐步实现。

We also want the glasses to be as thin as possible, and that's a process of miniaturization that happens.

同样，你塞入的技术越多，实现小型化就越困难。

And similarly, the more technology that you cram in, the harder it is to make smaller.

因此，我们能将这项技术微型化到什么程度，始终成立的一个原则是：如果你只放入一半的技术，就能做出更轻薄的眼镜。

So as much as we can miniaturize this technology, it will always be true that if you put half the technology in, you'll be able to make even thinner glasses.

然后有些人有不同的审美偏好，比如，幸运的是，现在厚框眼镜还挺流行的，但有些人还是想要像你这样的细框眼镜。

And then some people have different aesthetic preferences where they'll want I mean, like, fortunately, you know, thick glasses are kind of in style, but, you know, some people want thinner ones like yours.

是的。

Yeah.

你没法塞进太多电子元件。

You can't fit many electronics

嗯，现在我感觉不到。

in Well, now I don't feel

是的。

Yeah.

嗯，确实如此。

Well, it's yeah.

你可能

You may

将来不得不重新考虑你的审美选择。

have to rethink your aesthetic choices in the future.

但是，是的。

But Yeah.

你看，在定价方面，我们确实努力让它尽可能实惠。

See, on pricing, we I mean, we did we work on on getting it to be as affordable as possible.

我们的观点是，利润不会来自设备本身的高利润率。

And our view is that our profit margin isn't going to come from a large device profit margin.

而是来自人们随着时间推移使用AI和其他服务。

It's going to come from people using AI and the other services over time.

因为你需要为AI或其他服务支付订阅费吗？

Because you'll pay a subscription for the AI or something?

或者，是的，或者通过它进行消费，或者其他人们会做的各种事情。

Or, yeah, or use it and do commerce through it or whatever the different things are that people do.

我们不像苹果这样的公司，它们的利润主要或很大一部分来自硬件的高利润率。

We're not like a company like Apple whose primary or a large part of their margin comes from having a large margin on the hardware.

但总的来说，是的，我们尽量让它尽可能实惠。

But in general, yeah, we try to make it as affordable as possible.

我希望，如果我们再推出一款这样的产品，它能更加实惠，或者我们可以选择在保持价格不变的情况下，加入更多技术。

My hope is that if we build another one of these, hopefully it's even more affordable, or with this other choice that we can make is put even more technology in it and keep the price point there.

但是是的。

But Yeah.

我认为你会看到几个不同的价格区间。

I think you're gonna have a few different price points.

会有一种没有显示屏的普通AI眼镜，我认为它们的价格范围大概在300到500美元，甚至600美元左右，具体取决于设计风格和时尚程度。

There's gonna be the kind of standard AI glasses that don't have a display, and I think those will sort of range between you know, 300 to 500 or $600 depending on, you know, the aesthetic and and kind of how high fashion it is.

如果是一款非常时尚的款式，价格甚至可能超过600美元。

Maybe even more than 600 if you get something that's really high fashion.

但这是我们迄今为止从最初的雷朋到一些Oakley与Meta合作的先锋款所看到的价格范围，当然，还有各种其他情况。

But that's kind of the range that we've seen so far from the the early Ray Bans to to some of the Oakley meta vanguards with, like, all the Yeah.

比如一些定制化的产品，像带有处方镜片的光学镜片。

Kind of, you know, custom stuff in it, op like like, you know, optical lenses that can with a prescription.

是的。

Yeah.

然后还有一类像这样的，配备的不是全视野AR显示的显示屏。

Then there's a category like like this with the kind of a display that isn't kind of a full field of view AR display.

我认为这类产品的价格大概在1000美元左右，也许能再便宜一点。

And I think that that's going to be, yeah, on the order of $1,000 Maybe you get it a little better.

也许会更贵一些，但大致在这个范围内。

Maybe it's a little more, but you can call it in that area.

而当你进入全AR眼镜领域时，初期价格会更高一些。

And then I think when you get to the full AR glasses, that'll be somewhat more to start.

我认为人们会希望拥有整个产品组合，而长期目标是将这些技术尽可能融入更便宜、更轻薄的形态中，从而提供尽可能多的款式。

And I think people will just want to have the whole portfolio, then and the goal over time will be to get as much of that technology into as affordable and as thin of a form factor so you can just, like, have as many styles as possible.

所以你们有了这些新的Oakley眼镜，而你们与Elser或Luxottica的合作意味着你们可以利用他们其他品牌推出更多智能眼镜。

And so you've got these new Oakleys, and your deal with Elser or Luxottica means you can do other smart glasses with all their other brands.

对吧？

Right?

所以我认为这意味着，未来会有更多品牌搭载Meta的技术

So I think that implies, right, that there'll there will be future brands that have meta tech

我们非常希望去做。

We'd love to do it.

是的。

Yeah.

是的。

Yeah.

是的。

Yeah.

所以你认为这是在构建一个涵盖各种形态和价格点的星座式产品体系。

So you see it as, like, building out a kinda constellation of all these different form factors and price points.

而且

And

这就是目标。

That's the goal.

是的。

Yeah.

那其他这些不是眼镜的AI可穿戴设备呢？它们也在发展中，你怎么看？

What what about all these other AI wearables that aren't glasses that are happening?

比如那个朋友吊坠。

Like, there's the friend pendant.

我相信你已经见过。

I'm sure you've seen.

还有许多没有屏幕的、非眼镜类的设备。

There's all these, like, display less non glasses devices.

是的。

Yeah.

萨姆和约翰尼·伊沃似乎正在研发某种产品。

Sam and Johnny Ivor are apparently working on something.

人们对这个领域非常感兴趣。

There's a lot of interest in this.

我很想知道，自从近几年AI迅猛发展以来，除了眼镜之外，你是否也看到了其他机会？还是说你们的主要焦点仍然只是眼镜？

And I'm curious, is this something that since AI has really taken off in the last few years, you see opportunity there in addition to glasses, or are you still just the main focus is is glasses?

我们的主要焦点是眼镜，因为我认为眼镜是这种技术最好的形态，原因就是我们之前讨论过的那些。

Well, our main focus is glasses because I think glasses are the best form factor for this, for all the reasons that we talked about before.

我认为，任何你需要额外操作的东西都会分散你对周围物理世界的注意力，我不认为还有其他形态能像眼镜一样，看到你所见、听到你所听、全天候与你对话，并在你的视野中生成用户界面。

I I think that anything else that you have to kind of fiddle with takes your attention away from the physical world around I don't think that there's any other form factor that can see what you see, hear what you hear, talk to you throughout the day, and generate a UI in your vision.

还有增强现实这一部分，它能将物理世界和数字世界融合在一起。

And then there's the whole augmented reality part about blending the physical and digital world.

但我不确定。

But I don't know.

我的意思是，人们使用不同的电子产品。

I mean, people use different electronics.

所以，我不会说，未来世界上80亿人会做完全相同的事情。

So I mean, I'm not going to I mean, I certainly don't think that it's like in the future, all 8,000,000,000 people in the world are doing the exact same thing.

有些人更常使用手机。

I mean, some people use their phone more.

有些人则更常使用电脑。

Some people use a computer more.

有些人用iPad代替电脑，对吧？

Some people use an iPad instead of a computer, right?

有些人主要在电视上看视频。

Some people primarily watch videos on a TV.

有些人主要在手机上看视频。

Some people watch videos primarily on phones.

所以我认为它们会是不同的东西。

So I do think they're going to be different things.

我猜测眼镜会是最重要的。

My guess is that glasses will be the most important.

我觉得类似耳塞的东西也很有意思。

I think something like earbuds is kind of interesting too.

我的意思是，苹果显然在这一领域遥遥领先，凭借AirPods，我认为部分原因是他们做得很好，部分原因是他们通过捆绑、整合以及与手机协同的技术给自己带来了一些不公平的优势，而这些技术现在才刚开始开放，这很好，但以前我认为这使得其他人根本无法做出类似AirPods的产品。

I mean, Apple clearly is by far the leader on that with AirPods, I think partially because they did a good job and partially because I think they gave themselves some unfair advantages with how they bundle it and couple it and have technology that works with the phone that I guess they're now just starting to open up, which is great, but for a while, I think it just made it impossible for anyone else to build anything like the AirPods.

手表，我觉得挺有意思的，嗯。

Watches, I think, are interesting Mhmm.

在某些方面也是。

In some ways too.

我的意思是，这种

I mean, that kind

你不喜欢那种最近开始兴起的吊坠风潮吗？

You're not a fan of the, like, the pendant thing, the pendant trend that's kind of starting right now?

我的意思是，这算是一种潮流吗？

I mean, is it a trend?

我不知道。

Is it I don't know.

这还为时过早。

It's it's early.

我的意思是，现在有很多初创公司在做

I mean There's lot of startups doing

这些东西。

this stuff.

我的意思是，我觉得这是个有趣的想法。

I mean I think it's an interesting idea.

我不想显得太轻率地否定。

I I don't wanna be I don't wanna be too dismissive.

我的意思是，我认为不同的人会喜欢不同的东西，但我觉得眼镜可能会是最受欢迎的。

I mean, my my point is that I actually think that they're my guess is that different people are gonna like different things, but that glasses are gonna be the most popular.

今年在Kinect上没有推出新的Quest，我想知道你对如今Quest以及VR和混合现实这个品类的整体感受如何。

There was no new Quest this year at Kinect, and I'm curious how you're feeling about the Quest these days in VR mixed reality generally as a category.

看起来眼镜真的火起来了。

It seems like glasses are really taken off.

世界上显然有更多Quest设备在使用。

There's obviously a lot more quests out in the world.

它们已经卖出了

They've sold

很多

a lot

更多。

more.

但我很好奇，你知道，今年没有推出新品，而且我也想了解你对这个品类如今的看法。

But I'm curious, you know, not they're not being a new one this year and also just how you're feeling about it these days, like the category.

是的。

Yeah.

不。

No.

我的意思是，我觉得我们在这件事上正在取得进展。

I mean, I think we're we're making progress on it.

我的意思是，今年我们专注于Meta Horizon创作工具。

I mean, this year, what we focused on was the Meta Horizon creation tool.

因此，我们发布了Meta Horizon Studio和Meta Horizon Engine，这些是利用AI创建世界和内容的基础工具。

So we announced Meta Horizon Studio and Meta Horizon Engine, which are these basically foundational tools for creating worlds and content using AI.

这将有助于让人们在VR中创建更多内容，但我觉得这些技术也将会同样适用于AR。

And that's going to go towards making it so that people can create a lot more content in VR, but I think that that's also going to be the case in all that stuff, I think, should translate over to AR too.

我认为你可以在那里拥有大量这类内容。

I think a lot of this content you'll be able to have there.

我的意思是，透明眼镜可能没有VR那么沉浸，但你可以提供许多类似的全息体验。

I mean, glasses that are see through may not be quite as immersive as VR, but you can deliver a lot of the same kind of holographic experience.

然后，我认为这些功能也会出现在手机上，对吧？

And then I think a lot of these things will also end up showing up on phones, right?

我的意思是，AI带来了巨大的机会，比如你在浏览Instagram和Facebook的信息流时，每个故事都应该成为一个你可以跳入的独立世界。

I mean, it's think if there's this huge opportunity with AI to you know, you're browsing your feed on Instagram and Facebook, and, like, it's like each story should be its own world that you can jump into.

而且，你已经开始看到一些AI模型和谷歌最近推出的产品展现出这种可能性。

But, like, the the and you're starting to see some of this with some of the AI models and the stuff that Google has put out recently.

例如，这些都让人一窥未来的发展方向。

For example, are interesting glimpses of where that could go.

但我认为，要创造这类沉浸式体验，整个技术栈都需要重新思考。

But I think that there's this real sense that the whole stack for how you create those kinds of immersive experiences needs to get rethought.

这不再仅仅是人们以过去制作3D视频游戏的方式来做事情了。

It's not just going to be people doing things in the same way that they've created three d video games historically.

我的意思是，这是一个非常非常耗时的过程，工具也相当

I mean, that's a very, very intensive process where the tools are very

是的，混合的，是的。

Yeah, hybrid Yeah.

而且

And

我的意思是，是的。

mean Yeah.

是的，我的意思是，我的孩子们对编程和制作东西很感兴趣，我们试着构建不同的三维世界，但我认为有些东西对他们来说根本无法实现，对吧？

Yeah, I mean, so my kids are kind of into, into are programming and into making things, and we try to build different three d worlds, and I think some of the stuff is just intractable for them, right?

我的意思是，他们毕竟还是孩子，所以没关系。

It's like, mean, they're still kids, so it's fine.

我的意思是，但如果你有这种感觉。

Mean, but if it you sense.

把它提取成一个提示词，我的意思是，这样就变成了

Extract it to a prompt, I mean, then it becomes

我想通过那个，是的。

I think by the yeah.

所以这是关于Meta Horizon Studio的，我一直在用它，显然这并不是主要设计给八岁孩子使用的，但我的标准是：只要我能和我八岁的孩子一起做出一些不错的东西，那就很酷了。

So that's, but with the Meta Horizon Studio, which I've been playing with them, and obviously, this isn't primarily designed for an eight year old to be able to use, but my bar is if it's enough that I can make something good with my eight year old, then that's pretty cool.

你真的可以创造出所有这些有趣的东西，对吧？

You really can create all these interesting things, right?

你可以定义世界的动态机制，你想让世界变成什么样。

It's like you can define what the world dynamic is, what you want the world to be.

如果你想把东西放进世界里，你就可以做到。

If you want to put stuff in the world, you can do it.

如果你想用不同的材质来处理物体，或者想更换天空盒，你也可以做到。

If you want to texture things differently, if you want to change the skybox, you can do that.

所以我认为，这是一种完全不同的创作方式，它从根本上更加易用，从而释放出更多的创造力，会有更多有趣的世界和内容出现。

So I think it'd be a very different way of creating things that's fundamentally a lot more accessible, which will then unlock a lot more creativity, and there will just be a lot more interesting worlds and things to do.

我认为这将非常重要，不仅对VR和AR如此，而且我认为它将解锁无数体验，数十亿人可能首先在手机上接触到这些体验。这就是我们正在通过Meta Horizon Studio所做的事情。

And that, I think, is going to be important, not just for VR and AR, but I think it's going to unlock all these experiences that billions of people will probably first see on their phones at some So that's what we're doing with the Meta Horizon Studio work.

这是一种具有代理能力的AI流程，不同熟练程度的人都可以参与其中，创造出非常有趣的世界和沉浸式环境。

It's this kind of agentic AI flow where people at different levels of sophistication can go in and create really interesting worlds and immersive environments.

我觉得这很酷。

I think that's neat.

然后我们将它与Meta Horizon引擎结合在一起，这实际上是我们过去两年来一直开发的自定义图形渲染引擎。

Then we paired that with Meta Horizon Engine, which is basically this custom graphical rendering engine that we've been creating for two years now.

这个项目我们必须从零开始构建，因为之前我们使用的是Unity，它虽然很棒，但并不是为这种使用场景设计的。

It's this project that we've had to build it from the ground up because previously we were using Unity, which is great, but it's not really built for this use case.

我的意思是，大多数游戏加载时需要二十秒才能进入，这很正常，因为你需要加载整个三维世界，以便能够与之互动。

I mean, most games, you load a game, it takes twenty seconds to kind of get into the game, which makes sense because you're loading this whole three d world that you now need to be able to interact with.

但我们希望你所互动的世界能像在两个网页之间跳转，或在原生应用的两个界面之间切换那样迅速流畅。

But we want the worlds that you can interact with to feel more like jumping between two web pages or jumping between two screens within a native app that's really fast.

所以，那种需要二十秒才能将整个新世界加载进内存的做法是行不通的。

So the whole, okay, it has to take twenty seconds to page this whole new world into memory was not gonna cut it.

因此，我们重新构建了Meta Horizon引擎，从零开始打造这个图形引擎，使其能够以高并发方式渲染这类世界，支持虚拟形象系统、写实虚拟形象等所有功能，且加载时间只需几秒钟。

So we rebuilt, we basically built Meta Horizon Engine from scratch to be this graphics engine that can support rendering these kinds of worlds with high concurrency and the avatar system, the photorealistic avatars, and all this with just a few seconds of load time.

所以这更像一个网站，或者应用程序中的一个切换，正是这种体验让你在VR中可以轻松地在不同世界之间跳转。

So it's more like a website or just a transition in an app, and that's the kind of thing that'll make it so that when you're in VR, you could jump between worlds easily.

这不像是一种重大承诺或重大决定。

It's not like some big commitment or some big decision.

你可以自由探索，因为你不必等待二十或三十秒来加载下一个内容。

You can just feel free to explore because it's not like you're going to have to wait twenty or thirty seconds for the next thing to load.

你穿过传送门。

You walk through the portal.

你不喜欢它。

You don't like it.

你就从传送门往回走。

You walk back through the portal the other direction.

同样地，对于Facebook或Instagram这样的平台，能够查看一条帖子并直接进入一个世界，这需要极低的使用门槛。

And similarly, for things like within Facebook or Instagram, having the ability to see a post and jump into a world, that's something that needs to have very low friction to do.

因此，Meta Horizon引擎是这项核心技术。

So the Meta Horizon engine is this core piece of technology.

所以，我觉得在元宇宙方面，今年Connect大会上发布的更多是软件基础，而不是硬件。

So yeah, I'd say on the metaverse side, this year's announcements at Connect were more about the software foundations than hardware.

但你们仍然致力于硬件。

But you're still committed to the hardware.

是的，我的意思是，我们的硬件策略并不是每年都会推出一款新设备。

Yeah, I mean, the way that we do the hardware is we don't have it planned that there's a new device every single year.

当然。

Sure.

我们有多条产品线。

We have multiple device lines.

其中一条是高端产品线，我们会引入一些新技术，然后努力降低成本，使其更实惠。

There's sort of the higher end one where we introduce some new technology, and then we try to get it to be as affordable as possible.

我们推出了Quest 3，然后又推出了Quest 3s。

So we did Quest three, then we did Quest 3s.

但并不是每年都会推出一款新设备。

But it's not like every year there's one.

嗯。

Mhmm.

基本上，大多数年份这两条产品线中都会有一款新品推出，有时则会是空档年，我们主要专注于软件优化，为新的技术范式做准备。

It's basically, you know, most years there will be a new one of one of those two, and then sometimes there's like an off year where we're pretty much just tuning the the software to get ready for the new paradigm.

明白了。

Got it.

好的。

Okay.

所以Quest系列仍在继续。

So Quest is still going.

是的。

Yeah.

我们我们

I we're we're

我们专注于。

we're focused.

好的。

Okay.

是的。

Yeah.

我一直在保存这个。

I I've I've been been saving this.

我得问问你关于这个。

I've gotta ask you about this.

你知道现在对你的AI战略有着极大的兴趣，嗯。

You know there is a tremendous amount of interest in your AI strategy right now Mhmm.

老实说，这在科技行业是我见过的前所未有的。

Unlike anything I've seen in the tech industry, honestly.

或者说是整个AI领域。

Or AI overall.

整个AI领域。

AI overall.

我不知道是不是，但我认为，你夏天在招聘和超级智能方面所做的那些事情。

I don't know if it's But I think, like, what you've done over the summer with the hiring and the the superintelligence Mhmm.

你知道，你提出的使命以及所有这些。

You know, mission that you put out and all of that.

是的。

Yeah.

在这次对话中，我们一直在讨论AI与硬件的关系。

And we've been talking about AI as it relates to the hardware throughout this conversation.

而且，你知道，AI也是我演示的一部分。

And, you know, AI was a part of my demo.

但我想说，它并不是最核心的部分。

It wasn't, I would say, like, front and center.

而你目前的很多工作似乎都是在为它成为核心做准备。

And it seems like a lot of the work you're doing now is to get ready for when it will be.

我很想知道，我之前来参加演示时，有机会看到那个新实验室的团队在工作。

And I'm curious, you know, when I was here actually for the for the demo, I got to see the pod of the the new the new lab and see them at work.

他们在里面全力运转。

And they're in there cranking.

你能明显感觉到。

Like, you can tell.

我真的很想知道，也许我们可以从这里开始。

And I I would love to know, like maybe we can start here.

当你决定必须做出改变时，你是怎么想的？为什么选择这种方式？

Like, when you decided I need to I need to change things and and why you decided to go about it the way that you did.

因为我觉得这就是让人们惊叹的地方。

Because I think that's the thing that people were like, woah.

这太疯狂了。

Like, this is this is crazy.

是的。

Yeah.

我觉得如果

I think if

你身在其中时，感觉并没有那么疯狂，因为你知道，人才市场非常小。

you're on the inside, it doesn't feel as crazy because, you know, the talent market is very small.

你知道，如果你看数据的话，这其实是合乎逻辑的。

You know, it it's kind of rational if you look at the numbers.

但仅仅是这个策略。

But just the strategy.

比如，跟我讲讲，当你突然意识到的时候。

Like, walk me through, like, when you were like, okay.

我想做出改变。

I wanna make a change.

这就是我想做的。

This is what I wanna do.

跟我详细说说这个过程。

Walk me through that.

是的。

Yeah.

我的意思是，我认为人工智能和超级智能将是我们这一代最重要的技术。

I mean, this is an area that I just think it I think AI and superintelligence are going to be the most important technologies in our lifetime.

我觉得它如此重要，以至于需要一个独立的硬件平台，这也是我对眼镜如此兴奋的主要原因，因为我认为眼镜将成为为人们提供个人超级智能的最佳硬件类别。

I think it's so important that it demands its own hardware platform, which is a big part of why I'm so excited about glasses, because I think glasses are going to be the best hardware device category to provide personal super intelligence to people.

但我认为人工智能本身就是一件极其深刻的事情。

But I think AI is just this incredibly profound thing.

它将改变我们运营公司的方式。

It's going to change how we run the company.

它将改变所有公司的运营方式。

It's going to change how all companies run.

它将改变我们开发产品的方式。

It's going to change how we build products.

它将改变产品可能实现的形式，改变创作者的工作方式，从而改变内容的可能性、内容的组合，以及所有这些不同的方面。

It's going change what products are possible, Change how creators do their work, so change the content that's possible, the mix of content, all these different things.

因此，我认为如果想继续从事有趣的工作并推动世界进步，站在这一前沿至关重要。

So I think being on the frontier there is really critical if you want to continue just doing interesting work and pushing the world forward.

想想移动互联网时代，你并没有发明手机，但依然可以开发有趣的App。

Think just like with mobile, you didn't invent the mobile phone, you could still do interesting work building apps.

但我确实认为，在某种程度上，如果你能把软件与硬件体验结合起来，就能做更有趣的工作。

But I do think at some level you can do even more interesting work if you can both pair the software with the hardware experience.

当然。

Sure.

因此，我们坚定地致力于在构建超级智能的前沿领域工作。

So we are definitely committed to being at the frontier in building superintelligence.

我认为这将是最重要的技术，就像我刚才说的。

I think it's going be the most important technology, like I just said.

正因为如此，我特别关注确保我们打造一个整合紧密的项目。

And because of that, I just think we're very focused on making sure we build a weaving effort.

所以过去几年，我们启动了一个进展非常迅速的项目，对吧？

So over the last few years, we stood up an effort that was improving very quickly, right?

因此，LAMA 是一个不错的初始学术项目。

So LAMA was a good initial academic project.

LAMA 2 是该开源发布的一个不错的初始版本。

LAMA two was a good initial version of that as an open source release.

LAMA 3 比 LAMA 2 有了显著提升。

LAMA three was a big improvement over LAMA two.

随后，LAMA 4 在 LAMA 3.2 的基础上引入了一些重要改进，但我感觉我们并没有走在实现前沿突破、推动领域发展的正确轨道上。

And then LAMA four introduced some important improvements over LAMA 3.2, but I didn't feel like we were on the trajectory that we needed to be on to basically be at the frontier and be pushing the field forward.

我认为每个公司都会经历一段时期，觉得自己没有走在理想的轨道上，而这些正是你在人生或创业过程中必须做出的决策——真正的问题不在于你是否会感到自己偏离了目标，而在于你在这个时刻会怎么做。

And that was so I think every company at some point goes through periods where you're not on the trajectory that you want to be on something, and these are decisions that you get to make in your life or in building a company where the real question is not like, is there going be a moment where you feel like you're not on the track that you want to be on?

那么，在那一刻，你该怎么做？

It's what do you do in that moment?

因此，我决定我们应当退一步，建立一个新的实验室。

And so I just decided that we should take a step back and build a new lab.

我认为这一决定部分源于我对这项工作应有形态的思考。

And I think part of that was informed by the shape that I thought the effort should be.

我们非常注重人才密度。

We have this real focus on talent density.

我们的想法是，这本质上就像一个团队科学项目。

And the idea is that you really want to have This is like a group science project.

因此，你希望找到一个最小的团队，能够把整个项目都装进他们的头脑里。

So you want to have the smallest group of people who can fit the whole thing in their heads at once.

确实没多少人能做到这一点。

And there's not many people who can do that.

没错，但你也希望团队尽可能小。

No, but you also want the group to be as small as possible.

因此，公司里有些问题我们可以让更多人参与，即使每个人的边际生产力下降，我们仍然可以通过扩大团队规模来提升整体生产力。

So there are some problems that we're working on around the company where you can just have more people work on them, and even if the marginal productivity per person declines, you can just keep on scaling the net productivity of the effort.

我们的信息流和广告推荐系统就是一个有趣的例子，有很多人在不断测试系统改进方案，如果旁边那个人的实验效果不好，也不会太拖慢你的进度。

Our feed and ads recommendation is an interesting example of this, where we have a lot of people who are just testing different improvements to the systems, and if one guy's sitting next to you, if that guy's experiments don't work that well, it doesn't necessarily slow you down that much.

但我认为，构建这些语言模型并不是这样运作的。

But I think building these language models is not that way.

这是一个小团队协作的项目。

It's a small group effort.

你需要一个最小规模的团队，能够把整个项目牢记于心，并发挥出他们最好的水平。

You want the smallest group of people that can keep the whole thing in their head and do the best work that they can.

因此，这个团队中的每一个位置都极其宝贵且竞争激烈。

So each seat on that boat is incredibly precious and in high demand.

你也不希望有太多层级的管理结构，因为一旦进入管理层，他们的技术能力就会迅速退化。

You also don't want a lot of layers of hierarchy, because when someone gets into management, their technical skills start decaying pretty quickly.

即使他们几个月前还是独立研究员，但现在如果他们把所有时间都花在管理上，那么六个月或一年后，他们对技术细节的掌握可能还不如从前。

Even if they were an IC researcher a few months ago, now if they're spending all their time helping to manage, then, okay, after six months, a year, they might be less than the technical details than they were before.

所以我认为，关键在于组建一个规模相对较小、人才高度密集且组织结构非常扁平的团队。

So you kind of I think that there's this huge premium on just having a relatively small, extremely talent dense effort that is organized to be quite flat.

你对这个团队非常亲力亲为。

You're very hands on with this team.

是的。

Well, yeah.

我的意思是，我并不是一个AI科学家。

I mean, in the sense that I mean, I'm not like an AI scientist.

是的。

Yeah.

但我要说的是

But the thing that I

他们就坐在你旁边。

They're sitting near you.

我的意思是，很明显这是优先事项。

I mean, it's clear that this is, like, the priority.

是的。

Yeah.

所以我关注的一件事是，吸引世界上最好的人才加入团队。

So the the thing that that I'm focused on is, one, getting the very best people in the world to join the team.

我花了很多时间与这个领域的顶尖研究人员和人士会面，了解谁适合加入我们，以及谁正处于职业生涯中可以接受更好机会的阶段。

So I've spent a lot of time just meeting all of the top researchers and folks around the field and getting a sense for who I think would be good here and who might be at a point in their career where we can give them a better opportunity.

这是其中一部分。

That's one piece.

我另一个非常关注的重点是确保每位研究人员拥有的计算资源远超其他实验室，我认为我们目前在人均计算资源上已经远远领先于其他实验室。作为创始人兼CEO，加上我们有强大的商业模式支撑，我们的盈利非常可观。

Another thing that I'm very focused on is making sure that we have significantly higher compute per researcher than any other lab, which I think we are just way higher on compute per researcher than any other lab today, and as the founder and CEO, and because we have a strong business model that can support this, I mean, we make a lot of profit.

相当可观，是的。

A decent amount, yeah.

是的，这是一笔可观的收入。

Yeah, it's a reasonable amount.

你只要打个电话给黄仁勋，说‘请再给点GPU’就行。

You can just call up Jensen and be like, more GPUs, please.

没那么简单。

It's not that simple.

没那么简单。

It's not that simple.

但这是

But it's

我通常戴着眼镜给他发短信，但不行。

And I normally text him with my glasses, but no.