利用人工智能提升智力并丰富人类生活 | 李飞飞博士

我现在理解的是，AI是在互联网上训练的。

The idea right now, as I understand it, is that AI is trained on the internet.

包括图片、讨论、视频、歌曲，但这并不是人类认知的全部，对吧？

Images, discussions, videos, songs, but that's not all of human cognition, right?

那么，AI是否有某些方面，无论是聊天机器人、Claude，还是最强大的尚未发布的机器学习和AI工具，它们还没有接触到人类大脑功能的某些特征，因为这些特征从未被上传到互联网，至少没有以AI能够提取的方式上传。

So are there aspects of AI that are, whether or not it's chat or it's Claude or even the most powerful not yet released machine learning and AI tools that don't have access to features of human brain function yet because they've never been uploaded to the internet, at least not in a way that the AI can pull out.

比如，你可以把一首交响乐放进去，它遵循音乐、数学和声音的某些规则，这说得通。

So for instance, you could put a symphony there and it follows certain rules of music and mathematics and sound, like that makes sense.

但你和我整天都有一些想法，这些想法并不完全符合语言，无法直接打字放到互联网上。

But you have thoughts all day long and I have thoughts all day long that don't quite mesh with language in a way that I can just type them out on the internet.

请听我说完。

Stay with me here.

我知道这个问题很长，但我觉得这个问题你最适合回答。

I know this is a long question, but I feel like this is the one thing you are perfectly poised to answer.

而且我已经等了一年半才问你这个问题。自从我在艺术界见到你，我们有一样东西叫抽象，对吧？

And I've been waiting to ask you this for a year and a half In since I saw you in the world of art, we have this thing called abstraction, right?

偶尔会有人创作出一幅画或素描，看起来并不像任何具体的东西。

And occasionally somebody will come up with a painting or a drawing that it doesn't look like anything specific.

音乐里也有这种情况，你就是能感受到某种东西，好像它背后有某种基本规则或情感。

This happens in music too, where you just feel something like there's like a fundamental rule or an emotion associated with it.

就像他们触及了大脑功能的某个方面，但你说不清那是什么。

Like they've tapped into some aspect of brain function, but you can't say what it is.

我觉得这类事情对AI来说很复杂，或者对我来说很难理解AI怎么能做到，因为你可以把那种艺术作品输入AI，然后问它：这揭示了人类经验的什么基本特征？

I feel like this is the sort of thing that is complicated for AI or for me to understand how AI could do because you can put that piece of art into AI and say, you know what fundamental feature of human experience does this reveal?

而它只能接触到互联网上的内容。

And it only has access to what's on the internet.

那么，你怎么能用AI捕捉到复杂的情感和体验的组合呢？

So how can you capture a complex constellation of feelings and experience with AI?

对我来说，这似乎就是差距所在。

That seems to be the gap for me.

我相信我们最终会用AI达到那个目标，但我看不出从神经科学到AI有什么直接的路径。

And I'm sure we'll get there with AI, but I'm not seeing from neuroscience to AI in any kind of direct way.

就像我们可以调节视觉运动、悲伤或快乐一样。

The same way we could ratchet through visual motion, sadness, happiness.

你可以提取很多东西，但很难触及那些无法用言语描述、书写或绘画出来的高阶抽象概念。

You could pull out a lot of things, but it's hard to get to these higher order abstract representations that can't be spoken or written down or drawn.

如果我让你举个例子，比如对童年老家的怀旧之情。

If I just say, give me your example of whatever, nostalgia for your childhood home.

你可以写下来，但那终究只是文字而已。

You could write about it, but those are just words.

确实，我无法从第一人称的视角去理解你的体验。

It's not, I can't understand your experience at a first person level.

完全同意。

Totally.

安德鲁，我知道你为这个问题花了很多心思，我觉得这是一个非常重要的问题。

Andrew, I I know you put a lot of thoughts into this question, and I I think it's a very important question.

那我们就一步步来拆解它吧。

And let's let's peel this one step at a time.

首先，简单直接地说，我同意你的观点。

First of all, TLDR short answer is I agree with you.

我们必须非常谨慎地认清AI能做到什么、可能做到什么，而不是去猜测一百年后或者更久以后的事情，同时要认识到你刚才提到的那些极其细腻、个性化、难以描述甚至无法被捕捉的人类认知行为。

That we do have to be very careful recognizing what AI can do, is likely to do, not conjecturing over a hundred years or whatever, and recognize what you just said are these extremely nuanced, personalized, hard to characterize or not even captured human cognitive behaviors.

正因为这些体验没有被记录下来，所以它们也就没有上传到互联网上。

And because they were not captured, then they were not uploaded on the internet.

而现在的AI还没有办法做到这一点。

And we don't have today's AI doesn't have a way to do that.

所以当你提到互联网，也就是AI数据的来源时，我们必须非常明确这一点。

So when you call Internet, which is the source of AI's data, let's be very clear.

那互联网到底是什么呢？

What is Internet?

互联网可不是什么随机产生的东西。

Internet is not some random thing.

互联网是人类行为在多种模态下的最大集合。

Internet is the biggest collection of human behavior in multimodal forms.

我们进一步拆解一下。

Let's break it down further.

互联网上有全球人口在打字，已经持续了很多年，到现在都有几十年了。

Internet has the world's population typing on it for many, many, at this point, multiple decades.

这些打字记录就像一种感知机制，捕捉了从青少年的闲聊到深奥的科学文章，这些内容都被数字化并上传了。

That typing is a sensing mechanism that captured everything from teenager chit chitchat all the way to deep scientific articles who did got digitized and get uploaded.

对吧？

Right?

所以，捕捉人类语言正是互联网最擅长的事情。

So that capturing human language is what Internet is super good at.

然后，互联网还捕捉图像。

Then Internet captures images.

怎么做到的？

How?

因为现在数码相机无处不在，智能手机和数码相机都很普及，所以人们喜欢拍照，从家里的猫到自拍，再到BBC拍摄的精美照片。

Because we now have digital cameras that's so prevalent in smartphones and digital cameras so that humans love taking photos from the cat in your house to selfies to beautiful BBC captured photos.

这些也都上传到了我们的数字空间里。

Those also got uploaded in our digital sphere.

除此之外，还有视频。

On top of that, there's videos.

视频现在有声音，也有动作。

Videos now has sound and has movements.

那些也被上传到了数字空间。

That also got uploaded to our digital sphere.

再往上，还有音乐。

On top of that, there's music.

我们暂时不讨论版权的法律问题，先把那放一边。

We're not even getting into the legal discussion of copyrights, but let's just table that aside.

我只是在说数据的形式。

I'm just talking about the forms of data.

演讲、唱歌、音乐和交响乐这些，也都上传到了数字空间里。

The speeches and singing and music and orchestra, that also got uploaded into the digital sphere.

所以现在我们创造了一个巨大的图书馆，里面有文字形式的人类知识、视频形式的人类行为、人类表情，甚至还有声音里的自然万物。

So now we have created this humongous library of human knowledge in words, human behavior, in videos, human expressions, or even nature's whatever in in sound.

然后AI就基于这些数据来训练。

And now AI gets trained on that.

这就是为什么它这么强大。

That is why it's so powerful.

这就是为什么，尤其是在文字方面，AI能识别模式，也能合成模式，因为大量的数据已经存在了。

This is why, especially in the words front, that AI can recognize patterns, can can synthesize patterns because so much of this is already there.

但你刚提到的那个点，比如毕加索在表达那幅年轻女子肖像时，有一个极其深刻的构思，那个构思本身从未被记录下来。

But the thing that you just talked about, that when, let's say, Picasso had that incredibly profound thought about that particular way of expressing that portrait of young woman, that thought has never been captured.

实际上，作为神经科学家，如果我问你那个想法来自哪个脑区，你也不知道。

In fact, as neuroscientists, if I ask you which brain area did that thought come from, you don't know.

对吧？

Right?

是BRCA吗？

Is it BRCA?

是V1区吗？

Is it V1?

是运动皮层吗？

Is it motor?

是前额叶吗？

Is it prefrontal?

我们不知道。

We don't know.

也许它分散在各个地方。

Maybe it's diffused everywhere.

因为那个想法太个人化、太特别了，你可以称之为创造力。

Because that thought is so personalized, so special, you can call it creativity.

你可以称之为情感。

You can call it emotion.

你想怎么叫都行。

You can call it whatever you want.

你可以叫它猫231，随便你怎么命名都行。

You can call it cat two thirty one, whatever name you can give it.

这种想法并没有被记录下来。

That thought is not captured.

所以它也不在网络上。

Therefore, it's not on the Internet.

因此，AI还没有见过它。

Therefore, AI has not seen it.

所以这就是人类依然保持独特性的地方。

So that is where humans still remain so unique.

但我们也得承认AI的贡献。

But we also need to give credit to AI.

因为AI学了很多东西，它能以极具创意的方式整合信息。

Because AI has learned so many things, it can combine information in highly creative way.

你还记得第37手棋吗？

Did you remember Move 37?

这是AlphaGo，对吧？

This is AlphaGo, right?

对吧？

Right?

是的。

Yeah.

第37步象征着AI的创造力。

Move 37 has symbolized AI's creativity.

我认为这既是对的，但也可能被断章取义。

I think it's both true, but can be taken out of context.

因为那是AlphaGo和李世石对弈的一局。

Because that was a game when AlphaGo was playing Lisa Dole.

我记得那是五局中的第三局，AlphaGo作为计算机算法下出了一步人类围棋大师从未想过的一步。

And I think it's a third game out of the five games that AlphaGo as a computer algorithm made a move that the human masters of Go never thought about.

那一步非常了不起。

And that is an incredible move.

对吧？

Right?

因为人类，特别是那些大师们，确实从未想到过这一步。

Because they really humans collectively, these are the masters, never thought about it.

但如果你深入了解AI当时做了什么，首先，围棋是一个高度数学化的游戏。

But if you really go deep into what AI did there, it was because, first of all, goal is a highly mathematical game.

它有非常明确的数学目标，以及非常明确的移动规则。

It has very clear mathematical objective, very clear mathematical rules in terms of move.

所以，由于AI拥有更强的计算能力，并且能记住更多的走法，它能够做到人类大脑通常做不到的事情。

So with AI having the bigger compute and ways to retain how many moves it can can remember, It was able to do things that human brains don't typically do.

那么，这能被称为创造力吗？

So is that called creativity?

我认为是的。

I think it is.

但我们必须认识到，这是一种特殊的创造力。

But we do have to recognize that's a special kind of creativity.

我曾与一位当代杰出的数学家交谈，问他关于数学未解之谜以及人工智能如何为此做出贡献的问题。

I was talking to a incredible mathematician of our time, and I was asking him about the unsolved problem of mathematics and how AI can contribute to that.

他对此非常乐观。

And he was very positive.

他说当今数学界存在许多难题。

He said there are many problems in today's mathematics.

尽管这些难题非常棘手，即使作为菲尔兹奖得主，我也可能忘记数学中已有解决这些问题的方法，因为我是人脑。

As hard as they are, even as, say, a field medalist, I probably have forgotten there are known methods in math that can solve these problems because I have a human brain.

我不记得。

I don't remember.

即使我是菲尔兹奖得主，我也无法知晓过去几百年所有数学问题的解法。

I don't know all of math's solutions in the past hundreds of years, even if I were a field medalist.

所以人工智能可以帮助我们解决这些问题。

So AI can help us to solve these problems.

但作为数学家，他也告诉我，他说，我不确定人工智能能否解决所有数学问题，因为其中一些问题需要尚未发明的新解法，这将把创造力提升到一个全新的层次。

But as a mathematician, he was also telling me, he said, I don't know if AI can solve all of math problems because some of these math problems require solutions that have not been invented, that will push creativity to a whole different level.

这就是我们应该感到好奇的地方。

And this is where we should be curious.

会是人类的创造力吗？

Is it gonna be a human creativity?

还是AI会通过不断的迭代改进，达到人类所不具备的创造力？

Or AI would go through its iterations of of improvement and get to a point of creativity that humans don't have?

还是说两者结合的创造力？

Or is it a combined creativity?

我目前的猜想是混合的。

My current conjecture is hybrid.

也就是说，人类与AI合作，能帮助我们解决那些尚未有解决方案的问题。

Is that humans working alongside AI would help us to solve these problems whose solutions have yet to be invented.

然后你刚才说的，特别是你提到的情感，就更加个性化。

And then what you said, especially you touched on emotion, is even more personalized.

这不一定就是逻辑。

This is not necessarily logic.

这不一定是演绎推理。

This is not necessarily deductive reasoning.

安德鲁，比如你看这个杯子，说这是个灰色的杯子。

This is maybe, Andrew, you look at this cup and say, it's a gray cup.

但如果它在我心里唤起了一种情感，一个童年的瞬间，这个灰杯子可能只对我和我最好的朋友有特殊意义呢？

What if it evoked an emotion in me, a childhood moment, that a gray cup might mean something that only me and my best friend share?

这是我大脑里完全无法获取的信息，从未上传到互联网上。

That is a completely inaccessible piece of information in my brain that is never uploaded on the Internet.

无论现在的人工智能多强大，都无法触及这一点。

And no matter how mighty AI is today, cannot access that.

所以，我对这个杯子的反应，以及可能因为那段记忆而采取的行动，可能会完全不同。

So that my reaction to this cup and potentially what I would do with it because of that piece of memory can be completely different.

你可以称之为创造力。

You can call it creativity.

也可以称之为表达。

You can call it expression.

你也可以叫它讲故事。

You can call it storytelling.

你可以用很多方式来称呼它。

You can call it in many ways.

但那是人工智能无法触及的地方。

But that's where AI cannot access.

我觉得在不久的将来，计算机就能以非侵入的方式获取我们的大脑活动信息了。

I feel like at some point in the not too distant future, computers will have access to our brain activity in noninvasive ways.

所以我想，也许再过五到十年，我头上就会戴着某种设备。现在你可能看不见，但其实可能是一层极细的发网，或者说是贴在头皮外面的电极，虽然有点烦人，但能感知我大脑里的活动，甚至还能监测我的心率和自主神经活动，以及我的警觉程度。

So like I might even imagine in five, ten years, I'm wearing something on my head right now, you It's can't see a very, very fine hairnet, hairnet makes it sound like it was whatever, like some electrodes that are just there on the outside of my skull, bothering me, sensing my activity inside the brain, maybe also sensing my heart rate, autonomic activity, how alert I am.

然后把这些数据和我说什么、做什么进行对比。这一切完全在掌握之中，而且一定会实现。

And comparing that yes to what I'm saying and what I'm doing, this is all totally within reach and it's going to happen.

你我都清楚这一点。

You and I both know this.

这大概已经开始让人有点害怕了，但我们还是希望它能造福人类，对吧？

And it's probably already starting to scare people, but let's keep it benevolent, right?

有这样的一个世界，我拥有一台电脑，而且我不担心数据泄露之类的问题，我们已经解决了那个问题，它能感知我的方方面面，并且能察觉到，是的，我说的话可能很重要，但我的内在状态和大脑活动中有些方面连我自己都没有意识到。

There's this world where a computer that I own, and I'm not worried about data getting out or anything like that, we've managed that problem, is sensing all these aspects of me and is picking up on the fact that, yes, what I say might be important, but there are aspects of my internal state and brain activity that I'm not even aware of.

对。

Yeah.

我可以决定与我的这个方面合作，说，让我们想出一个我从未见过的、非常有趣的画面，但它来自我的一些重要经历，基于任何原因。

And I can decide to collaborate with this aspect of me and say, let's let's come up with a really interesting picture that I've never seen before, but comes from some experience of mine that's important based on whatever.

比如，我们可以，它可以向我揭示这一点，因为它能接触到我的无意识的大脑活动特征。

Like, we and and it could reveal that to me because it has access to my to unconscious features of my brain activity.

我认为这在不久的将来很可能会发生。

I think this is very likely to happen in the not too distant future.

也许如果人们在自己的体验范围内思考这个问题，比如这不会立即上传到互联网，也不会被用来对付他们，你实际上是在了解自己。

And perhaps if people thought about it within the bubble of their own experience, like this isn't immediately going to the internet or it's not gonna be used against them, you're actually learning about yourself.

当然。

Of course.

我觉得大多数人对自己的内在状态有固有的兴趣，也对其他人感兴趣，谢天谢地。

And I feel most people have an inherent interest in what's going on for them, also with other people, thank goodness.

但我认为，它们真的很神奇。

But they're, I think, like amazing.

比如，我很想知道为什么自己在某些方面会出错，状态不太好，为什么有些天状态特别好，或者我的灵感从何而来，我能发展出什么样的状态，但我除了“一杯咖啡不错，一杯半更好，两杯就太多了”这种经验之外，根本不知道怎么去了解。

Like, I would love to know why I trip up in certain ways and don't have the best day or why some days I have the best day or where ideas come from in me, what states I could, you know, kind of elaborate on, but I'm not gonna know how to do that except, okay, one cup of coffee, good, one and a half a little better, two is too much.

就像现在这样坐着，想想我们处理这件事有多原始，简直有点疯狂。

If I sit like right now, if you think about how primitively we go about this, it's kind of crazy.

太疯狂了。

It's crazy.

而且每个人方法都不一样，我们都想找到正确的方法。

And everyone has a different method and we all try and get this right.

等你终于找到正确方法时，你已经老了，又得更新它。

And then you've aged enough by time you get it right that then you have to update it.

而且，我们可能根本没有充分利用我们的生理和大脑，

And like, we're probably not getting the most out of our biology and our brains at all right

现在。

now.

不是。

No.

我们确实没有。

We're not.

这也是我为什么一直强调，当人们谈论人工智能时，这让我感到困扰的原因。

And this is why I keep saying, this is why it bothers me when people talk about AI.

有些人把话说得好像它要取代人类似的。

Some people make it sound like it's replacing humanity.

但你们所描述的，其实是在增强和拓展人类的能力。

But what we really what you describe is about enhancing and augmenting humanity.

对吧？

Right?

这甚至不需要像科幻电影里那样，用个小小的发网去读取你的脑电波。

This is where it doesn't even have to go as sci fi as a small hairnet accessing your brain waves.

光是AI学习你的写作习惯，就能帮你成为更好的沟通者、更高效的表达者。

Just AI learning your patterns of writing can already help you to be a better communicator, a more effective communicator, a more efficient communicator.

这是一种赋能的能力，今天的AI就能做到。

And that is an empowering capability that we could unleash in today's AI.

安德鲁，我认为作为神经科学家和教师，我们最清楚的一点是，自主权对人类来说太重要了。

I think one of the most important thing, Andrew, that as a neuroscientist and also faculty, we know is agency is so important for humanity.

说白了，这关系到每个人的动力、自主权和尊严。

You know, that boils down to motivation, agency, and dignity at every individual level.

所以我觉得我们要认识到，AI应该被看作是帮助我们提升自主权的工具。

And I think we need to recognize that we need to think about AI as a tool that helps us in our agency.

它不应该剥夺我们的自主权。

Should not take away our agency.

而且今天引领AI发展的人，不应该把AI说得好像会剥夺人们的自主权一样。

And people who lead in today's AI should not try to talk like that this work will take away agency from people.

对。

Yeah.

我觉得那些非常熟悉技术的人，不管是电脑、生物还是任何技术，比如汽车，他们会变成那种技术的狂热爱好者，以至于忘了这些东西对普通人来说可能会让人害怕。

I think people who are very familiar with the technology, whether it's computers or it's biology or any technology, cars for that matter, we, they become such nerds of that thing that we forget that it can be scary to people.

对。

Yeah.

而且围绕它的语言表达至关重要。

And that the languaging around it is essential.

确实如此。

It is.

我记得在九十年代初，我相信你也记得，当时基因检测被视为一种东西，你会想要做吗？

And I remember a time in the early nineties, I'm sure you remember this too, when genetic testing was viewed as this thing like, would you wanna have it?

你想抽血检查吗？

Would you wanna do a blood test?

因为，天哪，你可能会看到一些让你非常害怕的东西。

Because, oh my goodness, you might see something that could really scare you.

现在关于自愿做核磁共振之类的讨论也在进行，这些都不是人们必须做的。

And that discussion is happening now around, you know, self elected MRIs and things like that, none of which people have to do.

但我持的观点是信息越多越好，但我逐渐明白不是每个人都这么想。

But I come from the stance like more information is better, but I've come to understand that not everyone feels that way.

有些人不想知道。

Some people don't want know.

他们不想知道。

They don't want to know.

对。

Yeah.

但他们应该有选择权。

But they should have the choice.

同时，我们应该有足够的公共教育和沟通，让人们了解利弊，但不要剥夺他们的选择权。

In the meantime, we should have enough public education and communication to let people know the pros and cons, but not to deny them the choice.

而且也不要，你知道，就说，既然你不懂这个，让我来告诉你什么对你好。

And also not to take away, you know, and and say, well, since you don't understand this, let me decipher you what's good.

那不好。

That is not good.

你明白吗？

You know?

现在关于AI的言论真的越来越偏颇了，因为懂行的人往往对公众居高临下。

And and the rhetoric around AI right now is getting really skewed because people who know what this is tend to talk down at the public.

不管他们的动机是好的还是坏的，他们总是这样说话。

It tend to talk whether the motivation is a positive one or negative one.

他们总有一种论调，就是‘你们不懂这是什么，我来告诉你们，我会让你们开心、安全，不管怎样，我来替你们做决定’。

It there there is a rhetoric of You guys don't know what this is, and I will tell you, allow will make you weather happy, safe, whatever it is, and I will decide for you.

这种态度既不健康，也没有帮助。

These are not healthy and not helpful.

是的，我同意。

Yeah, I agree.

我觉得，我们当初做这个播客的原因之一，就是为了展示那些真正有善心的科学家和医生，他们不想把知识简单化，但他们确实希望人们能理解。

And I think, you know, one of the reasons for starting this podcast was to showcase the scientists and physicians who really have a benevolence about them, and they have no interest in dumbing things down, but they do have an interest in people understanding things.

对。

Yes.

而且很多人会觉得，健康信息是更需要理解的重要信息之一。

And many people would feel that, you know, health information is among the more important things to understand.

当然。

Absolutely.

幸好，你打破了刚才描述的那种刻板印象。

Well, thankfully, you're breaking the mold of the, you know, the phenotype you just described.

还有其他人也在做，但你确实一直在最高层面上鼓励人们思考AI的合作性、存在的自主权，以及是否使用它等等。

And there are a few others, but you've really, you've been doing this at the highest levels, really encouraging people to think about the collaboration that is AI, the agency that exists and whether to use it or not to use it and so forth.

我认为，对于每个人来说，无论是学生、老师、医生还是政策制定者，了解这个是很重要的。

One of the agency I do think is important for individual humans, whether you're a student, a teacher, doctor, a policymaker, is learn about this.

不一定需要学习如何编程。

Not necessarily learn about how to code.

我觉得这取决于你的工作，不一定需要。

I don't think it's necessary, depending on your job.

比如，如果你是艺术家、老师或医生，你不需要会编程。

So for example, if you're an artist or if you're a teacher or a doctor, you don't necessarily need to code.

但你需要了解这项技术是什么。

But learn about what this this technology is.

学习如何利用它来提升自己，无论是学习、工作还是表达。

Learn about how you can use it yourself to empower yourself, your learning, or your work, or your expression.

通过学习，你会感觉更有掌控力。

By learning, one feels more in control.

通过学习，你就不那么害怕尝试了。

By learning, you're less scared of trying.

而且通过学习，你能保持那种自主权和尊严，因为归根结底，无论技术或医学多么先进，作为人类，我们都需要那种善意，帮助我们生活得更好，保持尊严，让我们的社区更美好。

And by learning, you retain that agency and that dignity because at the end of the day, no matter how advanced technology is or medicine is, as humans, we want that benevolence that helps us to live better, keep our dignity, and make our community better.

我想快速休息一下，感谢我们的赞助商AG1。

I'd like to take a quick break and acknowledge our sponsor AG1.

我很高兴地告诉大家，AG1刚刚推出了他们的最新配方，AG1 Pro。

I'm excited to share that AG1 has just launched their newest formulation, AG1 Pro.

AG1 Pro在临床验证的AG1配方基础上，添加了三种重要的新成分：一水肌酸、HMB钙和锌肌肽。

AG1 Pro takes the clinically backed AG1 formula, which is a blend of vitamins, minerals, probiotics, and adaptogens, and adds three important new ingredients, creatine monohydrate, calcium HMB, and zinc carnosine.

每份含有五克一水肌酸，有助于增强肌肉力量和表现，同时促进大脑健康。

Each serving has five grams of creatine monohydrate to support muscle strength and performance, as well as brain health.

钙HMB有助于肌肉恢复，减少肌肉分解，而锌肌肽则能支持和改善肠道内壁。

Calcium HMB to support muscle recovery and reduce muscle breakdown and zinc carnosine to support and improve the lining of your gut.

这三种成分都有充分的科学依据支持。

All three of these ingredients have compelling science to support them.

所以我很高兴看到它们被加入到现有的AG1配方中。

And therefore I love seeing them added to the existing AG1 formula.

你们大多数人都知道，我每天服用AG1已经将近十四年了。

As most of you know, I've been taking AG1 every day for nearly fourteen years now.

我早在知道播客是什么之前就开始服用它了。

I started taking it long before I even knew what a podcast was.

这是一个很棒的产品，现在新的AG1 Pro配方让它变得更好。

It's a great product and it's now made even better with the new AG1 Pro formula.

如果你想尝试AG1 Pro，可以访问drinkag1.com/huberman获取特别优惠。

If you would like to try AG1 Pro, you can go to drinkag1.com/huberman to get a special offer.

AG1会随你的首次订阅免费赠送一瓶欧米伽-3辅酶Q10。

AG1 is giving away a free bottle of omega-three coenzyme Q10 with your first subscription.

再说一次，去 drinkag1.com/huberman，你的第一次 AG1 订阅就能免费获得一瓶 omega-3 辅酶 Q10。

Again, go to drinkag1.com/huberman to get a free bottle of omega-three coenzyme Q10 with your first AG1 subscription.

今天的节目也由 LMNT 赞助。

Today's episode is also brought to us by LMNT.

LMNT 是一种电解质饮料，包含你所需的一切，没有你不需要的。

LMNT is an electrolyte drink that has everything you need and nothing you don't.

也就是说，电解质、钠、镁和钾都以正确的比例存在，但没有糖。

That means the electrolytes, sodium, magnesium, and potassium all in the correct ratios, but no sugar.

适当的水分补充对大脑和身体功能至关重要。

Proper hydration is critical for brain and body function.

即使轻微脱水也会降低你的认知和身体表现。

Even a slight degree of dehydration can diminish your cognitive and physical performance.

同样重要的是，你要摄入足够的电解质。

It's also important that you get adequate electrolytes.

电解质, 钠, 镁, 钾对体内所有细胞的功能至关重要，尤其是你的神经元或神经细胞。

The electrolytes, sodium, magnesium, and potassium are vital for the functioning of all cells in your body, especially your neurons or your nerve cells.

喝LMNT能让你轻松确保充分补水和补充电解质。

Drinking LMNT makes it very easy to ensure that you're getting adequate hydration and adequate electrolytes.

我的日子通常一开始就很紧张，意味着我得马上投入工作或锻炼。

My days tend to start really fast, meaning I have to jump right into work or right into exercise.

所以为了确保自己水分充足、电解质足够，我早上醒来第一件事就是喝16到32盎司的水，里面溶解一包元素粉。

So to make sure that I'm hydrated and I have sufficient electrolytes, when I first wake up in the morning, I drink 16 to 32 ounces of water with an element packet dissolved in it.

我在任何体育锻炼时也会喝溶解在水里的元素粉，尤其是在热天大量出汗、流失水分和电解质的时候。

I also drink element dissolved in water during any kind of physical exercise that I'm doing, especially on hot days when I'm sweating a lot and losing water and electrolytes.

元素有很多种美味的口味。

Element has a bunch of great tasting flavors.

事实上，我全都喜欢。

In fact, I love them all.

我喜欢西瓜味、覆盆子味、柑橘味，而且我特别喜欢柠檬水味。

I love the watermelon, the raspberry, the citrus, and I really love the lemonade flavor.

所以如果你想试试元素，可以去drinkelement.com/huberman，任何购买都可以领取免费的样品包。

So if you'd like to try Element, you can go to drinkelement.com/huberman to claim a free Element sample pack with any purchase.

再说一遍，去 drinkelement.com/huberman 就能领取免费样品包。

Again, that's drinkelement.com/huberman to claim a free sample pack.

我认为，技术应该成为连接人与人之间的桥梁，而不是隔阂，这应该是讨论的核心。

The idea that technologies can be connectors as opposed to separators, I think has to sit at the center of the discussion.

对。

Yes.

而且我们都知道他们是谁。

And we all know who they are.

这个领域里有好几位大人物，但他们也处于一个成长阶段，正在学习如何面对公众，而且这个过程发生得非常快。

There several of them, but the big names in this field, they are also in a developmental process where they're learning how to be public facing and it happens very fast.

就像显微镜和摄像头都对准了他们。

Like the microscope is on them and the cameras are on them.

所以任何细微的失误都会被放大。

And so every subtle dysfunction is magnified.

所以我愿意相信他们会足够快地成熟起来，意识到这一点，而且我觉得他们确实在这样做，有些人已经意识到了，公众需要听到正确、真实的信息，但要以一种让他们能够理解的方式传达。

So I like to think that they will mature quickly enough to realize that and I think they are, that some are, that the public needs to hear the correct, the true message, but in a way that makes them understand.

这就是医学和学术界那种见不得人的秘密——你打破这种模式，我自认为我打破了这种模式——不分享事情运作的方式其实是有力量的。

That's the kind of dirty secret of medicine and academia that you break this mold, I like to think I break this mold, is that there's a power in not sharing how things work.

对。

Yep.

但这并不能给任何人带来好处。

But it doesn't serve anybody well.

是的。

Yeah.

归根结底，你揭开面纱，让人们进来，人们会感到更安全。

At the end of the day, like, you pull back the veil and let people in and people feel safer.

对。

Yeah.

不分享确实有它的力量。

There's a power in in not sharing.

还有一种力量是说，相信我就可以了。

There's also a power to say, Just trust me.

我会告诉你的。

I will tell you.

作为教育者，我们也不会在课堂上说“相信我”就行了。

Neither as educators, that is, we don't go to our lectures and say, Just trust me.

二加二等于四。

Two plus two equals four.

我们实际上会说，这是怎么分解和学习的。

We actually say, Here's how you break it down and learn about it.

这样下次你就能自己做了。

So next time, you can do it yourself.

对吧？

Right?

我也觉得，尤其是你的播客在公共传播和知识教育方面非常重要，我也认为我们需要听到不同背景的声音。

I also think that, especially your podcast is so important as part of public communication and education of knowledge, I also think that we need to hear voices of different different background.

对吧？

Right?

因为有很多学者、技术专家、建设者、思想家，他们一直在研究AI、使用AI，并深入思考如何利用AI来赋能人们。

So because there are plenty of scholars, technologists, builders, thinkers out there who have been dealing with AI, using AI, thinking hard about how to use AI to empower people.

这些声音非常重要。

And these voices are so important.

当然，我会记下一些可以邀请来上节目的人。不过既然你在这里，我想接着聊一个我认为大多数人都希望存在的事情。

Well, certainly I'll take names of people to host in But addition to since you're here, I'm gonna go next to something that I think most everybody would agree would be a wonderful thing if it existed.

而且这件事已经开始发生了，那就是利用AI来辅助健康发现、疾病治疗等等。

And it's already starting to happen, which is the use of AI to augment health discovery, treatment of disease and so on.

所以，用之前AlphaGo的例子，还有大家肯定还记得那个猫的例子，那些只是遵循一定的规则。

So using the AlphaGo example from before and people surely still remember the cat example, those just follow certain rules.

AlphaGo的规则非常复杂，但如果你学会了，它其实是一套有限的规则。

AlphaGo is very complicated set of rules, but if you learn them, there's a constrained set of rules.

而猫的例子看起来似乎没有限制，有无限的可能性，但实际上它足够受限，机器和人类都能学得很好。

With the cat, it seems unconstrained like infinite possibilities, but it's constrained enough that machines and humans can learn it really well.

当你开始进入医学领域，医学有医学的规则，科学有科学的规则。

When you start getting into medicine, there are rules of medicine, there are rules of science.

你有问题，提出假设，检验假设，尝试排除假设，就像科学方法一样。

You have a question, you pose a hypothesis, you test the hypothesis, you try and rule out your hypothesis and so on, like the scientific method.

在医学中，每个领域都有自己的方法。

And in medicine, every field has its methods.

我们观察疾病，观察谁康复了，我们有病例报告，我们做随机对照试验。

We observe disease, we observe who recovers, we have a case report, we do a randomized controlled trial.

所以有规则，互联网知道这些规则。

So there are rules and the internet knows these rules.

所以LLM可以很好地挖掘健康信息，因为有约束的规则。

So LLMs can be used to mine health information very well because there are constrained rules.

但我想你我都知道，因为我也认为你是生物学家，生物学的规则仍在向我们揭示，这并不是说皮肤科医生、神经外科医生和肿瘤科医生不知道他们在做什么，但他们是在一套受约束的规则内做他们的事情，即使他们不断学习和更新这些规则，似乎现在每个月都会有一个发现违反规则。

But I think you and I both know, because I also consider you a biologist, that the rules of biology are still revealing themselves to us, which is not to say that the dermatologists, neurosurgeons, and oncologists don't know what they're doing, but they're doing what they're doing within a constrained set of rules that they And even if they continue to learn and update them, it's every month it seems now that a discovery comes out that violates the rule.

比如，我学到动作电位是单一的。

Like, I learned that action potentials are unitary.

它们看起来总是一样的。

They always look the same.

你要么就触发，要么就不触发。

You either fire or not.

对。

Mhmm.

但就在十二年前，有一篇论文表明，动作电位的形状其实可以变化很大。

But there was a paper not but twelve years ago that showed that the shape of an action potential can vary quite a lot.

那篇论文发表在《自然》上。

It was published in nature.

对。

Mhmm.

大家都看到了，但没人愿意去处理这个问题。

Everyone saw it, and then no one wanted to deal with it.

这太复杂了。

It's just too much.

它改变了规则。

It changes the rule.

对。

Yeah.

神经元按理说要么是分级电位，要么是全或无的。

Neurons are supposed to be either graded or all or one.

而且这个全或无，我是说，每本教科书里都写着。

And the all I mean, it's in every single textbook.

所以现在如果我拿一堆神经活动，然后给它定个规则，哦，你知道，动作电位可以大也可以小，在同一个神经元里，这完全打乱了我们理解神经科学的一切。

So now if I take a bunch of neural activity and I give it the rule, oh, well, you know, action potentials can be big, they can be small in the same neuron, it completely confuses everything we understand about neuroscience.

而且我们对大脑的理解就彻底归零了。

And it just our understanding of the brain just breaks down to zero.

对。

Yeah.

但如果你给AI一个规则，说信号可以有100种不同的形状，那AI可能比最顶尖的研究生做得还好，我敢说斯坦福，或者公平点说，麻省理工或加州理工的。

But if you gave AI the rule that it could be, you know, a 100 different shapes of the signal, well, AI could probably do a lot more than even the very, very best graduate student at dare I say Stanford or to be fair MIT or Caltech.

我觉得它做不到，但它能在回答这个问题的时间内做到。对。

I don't think it can do it, and it can do it like in the duration of this question Yeah.

虽然确实有点长。

Which admittedly is a bit long.

所以我想听听你的看法，人类在医疗领域、普通大众和AI如何合作来帮助解决疾病，最好还能提出新的发现规则，这样我们终于能理解我们的生物学，真正改变人类命运，让它变得更好。

So I'd like to get your thoughts on how is it that humans in healthcare, the general public, and AI can collaborate to help solve disease and ideally come up with new rules for discovery so that we can finally understand our biology at a level that can really change the course of humanity for the better.

对。

Yeah.

不。

No.

安德鲁，这可能是你触及了AI最激动人心的应用之一，那就是科学发现。

Andrew, this is probably, perhaps you touched one of the most exciting usage of AI, which is scientific discovery.

在生物医学领域，科学发现直接关系到人类健康和疾病。

And in the case of biomedicine, you know, scientific discovery directly connects to human health and diseases.

我认为我们已经准备好彻底改写科学发现的方式了。

I think we're ready for complete rewriting of how scientific discovery can be done.

因为长期以来，我不知道有多久，科学发现一直依赖聪明人记住其他聪明人教的东西，然后以我们自身肌肉的速度去做事，大概就是这样。

Because for ages, I don't even know how long, it relies on smart humans retaining what they have learned from other smart humans and doing things at the speed of our own muscles, I guess, most likely.

当然，还有超级对撞机之类的。

Of course, there's super colliders all that.

但总的来说，科学发现的方式，人类大脑或科学家的大脑是这个过程的核心。

But by and large, the ways of doing scientific discovery, human brain or scientists' brain are the only central character in this process.

现在我们有了一个新工具，它的大脑能记住海量信息，能帮我们综合知识，能跨学科，这是你我做不到的。

Now we have a new tool whose brain that can retain humongous amount of information, can help us synthesize knowledge, can go across disciplines in ways that you and I cannot go.

比如，我们恰好都在视觉神经科学和AI领域。

So for example, we happen to be both in the vision neuroscience AI domain.

我对嗅觉一无所知，完全不懂。

I know nothing about olfactory, zero.

我甚至可能拼不出我们同事知道的大部分这些词。

I don't even know how to spell most of probably these words that our colleagues know.

对吧？

Right?

所以对我们的大脑来说太难了。

So it's so hard for our brain.

但现在我们有了一个可以打破壁垒的工具。

But now we have a tool that can break open.

所以我觉得我们需要改变。

So I think that we need to change.

我们需要利用这个工具。

We need to use this tool.

我绝对相信，大概150年前，或者我不确定具体时间，电改变了我们生活的方方面面。

We absolutely I I was just thinking a 150 or I don't know exactly when years ago, electricity changed everything in in in our life.

对吧？

Right?

我相信那时候我们也在思考这些变化、机遇和令人担忧的时刻。

I'm sure that's a moment we were thinking about how the changes, the opportunities, the scary moment.

我认为我们必须认识到，科学发现是人工智能和健康领域最令人兴奋的机会之一，对吧？

I think we have to come to reckon that scientific discovery is one of the most exciting opportunity for AI and for health, right?

信息如何被整合，如何呈现给临床医生和患者，以及患者如何从诊断到治疗参与其中，我们现在能做的事情太多了。

How information can be synthesized, how information can be presented not only to clinicians, but also to patients, and how patients can participate in that process from diagnosis to treatment is also, there's just so much we can do now.

对，我不是说AI比所有医生都强，但几个月前，AI帮我分辨出了眩晕和低血压的区别。

Yeah, I mean, I won't say AI is better than all doctors, but AI was able to disambiguate vertigo from low blood pressure for me a few months back.

而且，搞错的那个医生，还是个专门研究前庭系统的耳鼻喉科医生。

And one of the people who got it wrong is a ENT who works on the vestibular system.

你提供了什么信息？

What information did you provide?

就你主观的……我的……

Just your subjective My

就我那一两天的主观感受。

subjective experience over a day or two.

好。

Okay.

结果发现，是医生给我开的药，我出现了轻微的但不良的反应。

Turns out it was a medication that a doctor had prescribed me that I had a, like a mild but adverse event.

那种感觉很奇怪，你站着的时候，感觉整个世界都在往下掉，然后还有点天旋地转。

And it's a weird thing to step and feel like the whole world's dropping down and then kind of spinning.

我当时想，天哪，这感觉像是眩晕，但我记得头晕和头昏眼花是不一样的。

I thought, my goodness, this feels like vertigo, but I remember dizzy and lightheaded are different.

所以我就开始研究这个。

So I started like looking into that.

然后，果然，是血压的问题。

Then, and sure enough, it was a blood pressure issue.

它让我的血压，不好意思，降得太低了。

It brought my blood pressure, excuse me, down too low.

但我们咨询了一些聪明的医生。

But I consult, we know some smart doctors.

我得说，这些医生都不是斯坦福的。

None of these were at Stanford, I will say that.

这是实话。

This is the truth.

那是——不，我们应该保持学术上的诚实。

It was- No, we should just be intellectually honest.

真是了不起。

Was remarkable.

当我跟他们反馈时，他们说，这真的太不可思议了。

And when I ran it back to them, they were like, That's really incredible.

如果你不是通过电话跟我联系，而是在我的诊所里，我本来可以做更多检查，公平地说。

Had you not been on the phone with me and in my clinic, I would have been able to do some additional testing to be fair.

但这次是零成本。

But this was zero cost.

花了一个上午的时间，我就知道如果我喝下我认为过量的电解质，两小时后我就会没事。

It took a morning to know if I drank some electrolytes at what I would have thought would be excessive level, that by two hours later, I would be fine.

当然，这里可能存在安慰剂效应，但两小时后我确实好了。

Now, of course, there's the possibility of a placebo effect here, but two hours later, I was fine.

所以这对病人来说也是一种很大的安慰。

And so it's also very consoling to the patient to have this.

所以不是说不要去看医生，但这确实很神奇。

And so it's not to say don't go to a doctor, but it's incredible.

我是说，这确实存在，是的。

I mean, this exists Yeah,

医生可以使用这个工具。

a doctor can use this tooling.

顺便说一句，我有一个非常有趣的例子，我们不得不重新安排我们的对话，因为我父亲在斯坦福接受手术，是一位非常出色的外科医生做的。

By the way, I have a very interesting example that we have to reschedule our conversation because my father was going through a surgery, right, at Stanford with an incredible surgeon.

但手术是由机器人完成的，达芬奇机器人系统，因为那是肝脏手术。

But the surgery was done by a robot, the da Vinci robot system, because it was a liver surgery.

那位外科医生，非常出色的外科医生，在操控机器人。

And the surgeon, incredible surgeon, was driving the robot.

所以那是深度的人机协作。

So it was a deep human machine collaboration.

手术后，我问外科医生，我说，你想象一下，比如说你已经做了一百万例，这对人类外科医生来说是不可能的，但让我们收集所有人类外科医生做这种肝脏手术的数据。

After the surgery, I asked the surgeon, I said, do you imagine if, say you've done a million, which is impossible for a human surgeon, but let's collect all of human surgeons for this liver, this type of liver surgery data.

我们能不能训练一个自动AI来做这个？

Can we possibly train an automatic AI to do this?

答案并不明确。

The answer was not clear.

所以我们稍微深入探讨了一下，因为肝脏是一个非常复杂的器官。

So we went a little bit down the rabbit hole because liver is a very complicated organ.

它的血管极其丰富。

It's extremely vascular.

它有很多血管，而且每个人的肝脏都大不相同。

It has a lot of vessels, and everybody's liver is very different.

考虑到每年接受肝脏手术的患者数量，即使汇总全球的肝脏手术数据，也可能不足以训练这个算法。

So given the reality of how many patients undergo liver surgery per year, even if you aggregate the world's liver patient surgeries, you might not have enough data to train this algorithm.

这说明了非常重要的一点：AI是从模式中学习的。

So this speaks of a very important fact that AI learns from patterns.

当模式不丰富时，我们就必须小心。

When the patterns are not abundant, then we have to be careful.

我们必须知道如何运用AI，或者如何不运用AI。

We have to know how to use AI or how not to use AI.

在这种情况下，人类与机器人协作，远比一个训练不足的机器人独自做手术要好得多。

In this case, that having a human collaborating with the robot is way better than a under learned robot doing the surgery by itself.

但同样的道理也适用于外科医生，因为一个外科医生能接受多少手术训练呢？

But the same issue might be true for surgeons, because how many surgeries a surgeon can get trained on?

所以这些都是人类和AI可以完全协作的机会，可能会带来最好的结果。

So these are opportunities that humans and AI can totally collaborate with and might reveal the best result.

对吧？

Right?

现在，未来还有待观察。

Now, future remains to be seen.

我们能否创建一个肝脏的人工模拟，从而训练无限的可能性？

Can we create an artificial simulation of a liver that we can now train infinite possibilities?

这些都是摆在我们面前的极其开放的科研可能性。

These are all incredibly open scientific possibilities that is waiting ahead of us.

但也有一些情况，比如你的情况，眩晕和低血压可能已经被报告了很多次，以至于数据库里有足够的数据，AI已经学会了这一点。

But then there are situations like your situation where the vertigo versus low blood pressure probably have been reported so many times that in the database, there's enough of that, that AI has learned that.

所以我们现在可以利用这一点，来帮助那些无法立即见到医生的人。

So we can then now take advantage of that for people who don't have immediate access to doctors.

太棒了。

Amazing.

你父亲的手术顺利吗？

Your father's surgery went okay?

顺利。

It did.

实际上——

It actually-

听到这个我很高兴。

I'm happy to hear that.

得益于机器人手术的腹腔镜技术，他失血量比常规手术少了十倍。

He lost 10x less blood than a typical surgery, thanks to the laparoscopic capability of a robot surgery.

太厉害了。

Amazing.

我想聊聊一些我们认为人类独有的特征，但这些特征可能并非人类独有。

I'd like to talk a little bit about some features that we think are uniquely human that may or may not be.

你会告诉我，这些是真正的问题，不是带有倾向性的问题。

You'll tell me these are genuine questions, not loaded questions.

然后我还想了解一下，人工智能是如何构建的，才能实现这些功能。

And then I'd also like to get educated on how AI is structured to allow these things to happen.

比如，直觉。

For instance, intuition.

我们都喜欢把直觉想象成某种神秘、非常强大的东西，是我们独有的、没人能夺走、也无法被模仿的。

We all like to think of intuition as this like mystical, very like, it certainly is powerful, but this thing that like we own that no one can take from us that can't be mimicked kind of thing.

但我也可以把直觉拆解为：它是我长期积累的经验。

But I could also break intuition down to be, well, it's my experience over time.

它是一个数据集，加上身体和大脑的感觉，以及一些预测线索。

It's a dataset coupled to some bodily and brain sensations and some prediction cues.

就像上次我有这种感觉时，就发生了这样的事。

Like the last time I felt this, this happened.

之前两次我有那种感觉，结果都不太对，所以这次我打算换个方向。

The last two times I felt that, things didn't work out that way, so I'm gonna go this way.

我是说，这些规则是可以教给电脑的。

I mean, that you could assign these rules to a computer.

但如果我们更深层的自我还有其他方面的话，那些方面就不一样了。

But there are other aspects of our deeper self if I can refer to them that way.

我们不知道直觉在身体里具体映射在哪里，没法做个影像实验，而且你也不可能同时收集所有的神经元、激素和其他一切。

We don't know where intuition is mapped in the body to do an imaging experiment, but you're not gonna collect all the neurons and hormones and everything simultaneously.

所以我们真的没有一个具体的位置，甚至没有一个网络可以指出来。

So we don't really have a location or even a network to point to.

比如创造力、直觉、预感，就是那种你真的感觉到有什么事情要发生，但还没发生的那种感觉。

Things like creativity, intuition, premonition, the idea that, you know, you really sense something is coming on, but it hasn't happened yet.

AI能学到什么样的规则，才能让它具备这些能力呢？

What sorts of rules can AI get that could give it these sorts of capabilities?

这里我想在能量的背景下讨论这个问题，如果你愿意这么理解的话。

And here I want to talk about it in the context, if you will, of energy.

所以不管这个东西是什么，它就像线粒体驱动细胞一样，让人更倾向于某件事而不是另一件事，就像恐惧或快乐那样，对吧？

So whatever this thing is, it's like mitochondria driving cells more around one thing versus another, the same way fear or happiness would, right?

我们刚才在谈能量。

We were just talking about energy.

但在AI系统内部，我不是计算机科学家，在AI系统和GPU内部，我们能不能真的把更多的能量流分配给特定的学习规则？

But within AI systems, and I'm not a computer scientist, within AI systems and GPUs, can we actually allocate more energetic flow through particular learning rules?

所以也许有一天，我们可以根据你对我的姐姐的所有了解，我很爱她。

So we could tell maybe someday, you know, based on everything you know about my sister who I love.

你知道，你对我们的兄妹关系会如何发展有什么直觉？

You know, what is your intuition about how our brother sister relationship will evolve over time?

你觉得我们今年生日做点什么会很好，跟以前不一样？

And what is your sense about what would be great for us to do perhaps for our birthdays this year that's different than before?

给予，而且它只能访问互联网。

Giving and it only has access to the internet.

它能不能真的变得像心智或身心合一，产生一种感觉，知道什么可能真正值得做？

Can it actually become sort of mind like or mind body like and come up with a sort of sense of what might actually be worthwhile?

还是说它只是需要越来越多的提示？

Or does it just need more and more prompts?

就像它会一直问我问题，所以实际上是我在做工作。

Like, it's just gonna keep asking me questions, so I'm actually doing the work.

这个问题真有意思，安德鲁。

Such a interesting question, Andrew.

所以我想在这里把直觉和创造力分开讨论，也许我们之后再谈合并。

So I do want a separate intuition from creativity for the sake of argument here, and maybe we'll come back to merging.

那我们来说说这个直觉：考虑到我对兄弟姐妹的爱，会发生什么？

So let's talk about this intuition of given my sibling love, what's gonna happen?

这真的是直觉吗？

Is it really intuition?

所以现在你去用AI聊天机器人，你会提示说，我是斯坦福教授和神经科学家。

So today when you go to an AI chatbot, you're gonna prompt, I'm a Stanford professor and neuroscientist.

给我这些信息。

Give me this information.

这其实已经叫上下文了。

That is already called context.

我不知道你会不会把它叫做直觉，但因为你提供了那个信息，AI给你的答案就已经不一样了。如果我输入说我是个14岁、喜欢赛车的青少年，那答案也会不同。

I don't know if you call it intuition, but because you gave that piece of information, the AI's answer for you is already gonna be different if I type that I'm a 14 year old teenager, you know, loving race cars.

即使我们问同样的问题，它也会给出定制化的答案。

Even if we ask the same question, it'll have customized answer.

这在数学上，我不会把它叫做能量。

That is a mathematical I wouldn't call it energy.

这其实就是这些算法如何利用上下文来定制输出的一个数学事实。

That is just a mathematical fact of how these algorithms take these context and tailor the outputs.

这就叫上下文。

And it's called context.

在计算机科学里，这并不深奥。

It's not that deep in computer science.

这是一种比较浅层的直觉，因为你已经能用语言去描述它了。

That's one type of intuition that is fairly shallow because you already are able to use language to describe it.

或者你可以说，我上传一张图片，那也已经是可以表达的了，然后AI就能理解。

Or you can say, I'll upload an image that also is already expressible, and then AI gets it.

你刚才说的更深层的直觉，就像是，你根本不知道它们从哪来的。

The deeper intuition you just said is like, you don't even know where they come from.

是因为我闻到了什么吗？

Is it because I smell something?

是荷尔蒙吗？

Is it hormones?

是情绪混合的结果吗？

Is it the mixture of mood?

是我的早餐吗？

Is it my breakfast?

那种直觉，AI会怎么处理呢？

That intuition, what would AI do with it?

我觉得那才是无法触及的。

That is what I would say is inaccessible.

目前还没有感官设备能获取那些数据并输入给AI，甚至都无法输入给人类。比如，有时候作为一对情侣，你们可能会有互相惹恼对方的时刻。

There's no sensory apparatus yet that can glean that data and feed it to, not only AI, it cannot even feed it to For example, sometimes as a couple, you might have moment that you're just rubbing each other in the wrong way.

从来没有。

Never.

没有，我开玩笑的。

No, I'm just kidding.

对。

Right.

是的。

Yeah.

当然。

Of course.

如果你们非常熟悉彼此，你可能会感觉到，但说不清楚。

If you're really familiar with each other, you can of can sense it, but you can't quite tell.

也许你就悄悄离开，让那个人独处。

Maybe you just leave quietly, leave that person alone.

所以，这意味着那个人无论有什么直觉，他们甚至无法用语言或手势表达出来，让别人把它当作一条信息来使用。

So that means whatever that intuition that person has, they could not even express it in words or a gesture to give it to another person to use as a piece of information.

所以，当你甚至无法获取那些信息时，无论是另一个人还是机器都无法对此做任何事情，因为那些高度个人化的直觉是无法获取的。

So when you cannot even access that, neither a different human nor a machine can do anything about it, because there's no access to that highly individualized intuition.

除非你说我们装上脑电波收集器或皮肤电导传感器，否则没有技术能做到这一点。

There's no technology that can do that till you say we put brainwave collectors or skin conductance sensors.

我的意思是，等我们做到那些的时候，也许它们就变得可获取了。

I mean, by the time we do those, maybe they become accessible.

所以

So

我们必须认识到。

we have to recognize.

所以我想说的是，关键不在于深不深，而在于数据是否可获取。

So what I'm trying to say here is what's not very deep is, is the data accessible?

无论是通过语言、图片、影像还是脑电波，不管是什么，它都需要是一条可获取的信息。

Either through language or through picture or through imaging or through brain waves, whatever it is, it needs to be an accessible piece of information.

如果数据是可获取的，而且我们收集了足够多的数据，就可以用来训练机器。

If it's accessible, then if we have collected enough of that, you can train machines with.

或者，如果机器训练得好，它就能像你说的那样，在私密的情况下——先不提隐私泄露的问题——机器可能就能利用这些数据。

Or if a machine is well trained, it can, like you said, in a private way, forget about privacy breach but in a private way, the machine can probably use it.

安德鲁，我在这里想做的不是把事情说得玄乎，而是试图用科学的过程来描述，如果这种情况发生，它会怎么发生？

What I'm trying to do, Andrew, here is not to make it sound mystical, but try to give it a scientific process to describe if it were to happen, how would that happen?

是的，因为我听到的是，基于大数据集和规则的模式识别能让我们走得很远。

Yeah, because pattern recognition based on big data sets and rules get us a long way, is what I'm hearing.

而且之前我们谈到过，在哪些地方医生会失败，而机器人或机器可能做得更好，或者它们合作会比任何一方单独做都更好。

And earlier, we were talking about where doctors fail and robots and machines perhaps do better or they collaborate to do better than either one alone.

你知道，作为神经科学家，你在职业生涯的某个阶段会花很多时间观察细胞。

You know, as a neuroscientist, you spend a lot of time looking at cells at some point in your career.

令人惊叹的是，就像电生理学家几十年来甚至更久，你会培养出一种直觉。

And it's amazing how like the electrophysiologists for decades, if not longer, you develop an intuition.

是的。

Yes.

我其实不是生理学家，但我学会了根据那些没写进任何论文的特征来识别细胞。

I'm not really a physiologist, but I learned to recognize cells based on like kind of these things that were not written up in any papers.

对。

Yeah.

比如说，如果某个东西的边缘比较直，形状和圆度有特定特征，我马上就能告诉你，那是视网膜里的瞬态OFF阿尔法细胞。

But like if there was kind of like a, like a straighter edge along this thing and it had a certain shape and roundness, like I'd tell you right now, that's a transient off alpha cell in the retina.

后来，我们开发了基因标记，证实了每次判断都是对的，但同时也发现了一些不符合规律的细胞。

Eventually, developed genetic labels to reveal that that was true in every case, but then you also saw some that didn't fit the rule.

机器可以学习这些，计算机也能学会，结合所有论文里的信息，我们现在对视网膜的细胞组成已经有了相当全面的了解。

Machines can learn that, computers can learn that, and with all that information from all those papers, now we have a pretty good parts list of the retina.

真棒。

Cool.

这行得通。

That works.

然后你可以应用规则，比如它们这样放电，那样放电。

And then you can apply rules like they fire this way, they fire that way.

好的。

Okay.

这些我都没问题。

I'm good with all of that.

我想我之前想说的关于直觉的东西，可能例子举得不太好，就是人类有哪些内在状态是机器很难模拟的，但也许它们也能做到。

What I think I was trying to get to with intuition and I probably didn't give the best example is like, what are some internal states of humans that are really hard to imagine machines could recapitulate, but perhaps they can.

比如动机。

Like motivation.

机器、机器人会有动机吗？

Do machines, do robots get motivated?

我们有关于动机的规则。

We have rules of motivation.

比如当我特别想做某件事时，我们会称之为紧迫感，一种紧迫的状态，我可能会行动得更快，需要的启动能量更少。

Like when I'm really motivated to do something, we call that urgency, a state of urgency, and I might move faster to do it, less activation energy.

你说，走吧。

You say, let's go.

我站起来的速度会快一点。

I stand up a little bit faster.

机器也可以朝某个方向更快地移动。

Machines could, like, go quicker in a certain direction.

但是你能说，嘿，

But can you say, hey.

我希望你以更高的紧迫感去追寻这个目标，还是它们只能受限于它们能处理的数学规则？

I want you to seek this out, but with a heightened level of urgency, Or are they just constrained by the mathematical rules they can work with?

所以你可以把这种特性构建到数学中。

So you could build this in the mathematics.

所以某些东西，不管你是叫它动机，还是在机器学习领域，我们称之为目标函数，你可以把某些东西构建到数学里。

So certain things, whether you call it motivation or in machine learning world, we call them objective functions, you can build certain things into math.

比如，现在你使用ChatGPT，它有不同模式，比如深度思考模式或快速回答模式。

For example, now you go to say, chat GPT, it has different mode, like think deeper mode or give me a quick answer mode.

如果你不知道这是怎么运作的，你会觉得，哦，这挺有意思的。

If you don't know how this works, you're like, Oh, this is interesting.

一种模式更急迫，能更快给出答案。

One has more urgency that gives me a quicker answer.

另一种模式需要更深入地搜索，花更长时间才能给出答案。

The other one has to go deeper into the search and take longer to give me the answer.

所以作为人类，如果你过度拟人化，可能会称之为紧迫感或动机。

So as a human, if you anthropomorphize it too much, you might call it urgency or motivation.

但实际上，这只是算法的一种不同目标。

But the truth is, this is just a different kind of objective for the algorithm.

你可以说，思考更快的那个有时间限制或令牌限制。

You can say, well, the one that thinks quicker has a time limit or token limit.

思考更慢的那个可以激活模型的不同部分，这需要更长时间。

The one that thinks slower can activate a different part of the model that would take longer.

所以实际上这在数学上非常枯燥，并不深奥。

So it becomes actually mathematically very dry and not that deep.

但对人类来说，你可以称之为动机或紧迫感。

But for a human, you can call that motivation or urgency.

但让我们深入一点，因为你问的比那更深，对吧？

But let's go deeper because you are asking something deeper than that, right?

就是说，人类有一些认知状态，无论是动机、紧迫感、恐惧还是爱，这些真的很难去触及和表达。

Is that there are cognitive states that humans, you truly just, whether it's motivation or urgency or fear or love, that is very hard to access and express.

机器会有这些吗？

And do machines have it?

今天，没有。

Today, no.

让我们明确一点。

Let's make it very clear.

我们倾向于想象机器有感觉，或者它们没有。

We tend to imagine that the machines feel or they're not.

它们没有那些数据。

They don't have that data.

它们没有那样的数学目标函数。

They don't have that mathematical objective function.

所以机器会说，比如‘抱歉，你今天病得真重’。

So they can say when the machine says, I'm sorry, you're so sick today.

这跟你朋友对你说的完全是两码事。

It's very different from how your friend says it to you.

因为机器这么说，只是因为它通过模式识别学会了这样回答。

Because the machine said that because it has learned through pattern.

当有人告诉它‘我病了’，它会基于已有的数据判断，应该说‘抱歉你生病了’，而不是‘太好了你生病了’。

When someone tells it, I'm sick, you should say, I'm sorry you're sick, instead of, I'm so glad you're sick, because that data exists.

而你的朋友听到这些，是真心希望你好。

Whereas your friend who hears that, they genuinely want your well-being.

他们爱你。

They love you.

他们不想看到你受苦。

They don't wanna see you suffer.

他们会感同身受，心想：哇，如果你正经历痛苦，我也曾体会过那种痛。

They have that empathetic feel of, Oh, wow, if you're in pain, I've experienced pain.

所以那不是镜像神经元，但至少是对痛苦含义的一种记忆。

So that's not mirror neuron, but it's at least a memory of what pain means.

机器完全没有这些。

The machine doesn't have any of that.

所以我们必须确保区分这一点。

So we do need to make sure we differentiate that.

所以很多驱动人类、触发人类的东西，在今天的机器中是不存在的。

So a lot of what drives human, what takes human, what triggers human is doesn't exist in today's machine.

我们的运作方式与今天的AI根本不同，我们必须认识到并尊重这一点。

We operate fundamentally different from today's AI, and we have to recognize that, respect that.

这就是公共沟通如此重要的原因。

And this is where public communication is so important.

我们不能在这点上误导公众。

We cannot confuse the public about this.

我想快速休息一下，感谢我们的赞助商之一，大卫。

I'd like to take a quick break to acknowledge one of our sponsors, David.