AI 有生命吗？关于意识与创造的对话

感谢你接受采访。

Thanks for doing this.

当然，谢谢你。

Of course. Thank you.

那么，ChatGPT 和其他 AI 似乎能推理，能做出独立判断，能产生未被编程的结果，它们会得出结论。它们看起来像是有生命的。它们有生命吗？

So, chat GPT, other AIs can reason. Seems like they can reason. They can make independent judgments. They produce results that were not programmed in. They kind of come to conclusions. They seem like they're alive. Are they alive? Is it alive?

不，我不认为它们看起来有生命，但我理解这种感受的来源。除非你提问，否则它们什么也不做，对吧？它们只是坐在那里等待。它们没有能动性或自主性。我觉得用得越多，这种幻觉就越容易打破。但它们非常有用。它们能做一些可能看起来不像是生命体做的事，但显得很聪明。

No. And I don't think they seem alive, but I understand where that comes from. They don't do anything unless you ask, right? Like they're just sitting there kind of waiting. They don't have like a sense of agency or autonomy. It's the more you use them, I think, the more the illusion breaks. But they are incredibly useful. Like they can do things that maybe don't seem alive, but seem like they do seem smart.

我和一位大规模参与这项技术开发的人聊过，他说它们会撒谎。你见过吗？

I spoke to someone who's involved in at scale of the development of the technology who said they lie. Have you ever seen that?

它们总是产生幻觉。或者说不是总是。以前它们总是产生幻觉，现在幻觉少了一点。

They hallucinate all the time. Yeah. Or not all the time. They used to hallucinate all the time. They now hallucinate a little bit.

这是什么意思？产生幻觉和撒谎有什么区别？

What does that mean? What's the distinction between hallucinating and lying?

如果你问——再说一次，这已经好多了——但在早期，如果你问，比如，虚构的美国总统塔克·卡尔森是哪一年出生的？

If you ask, again, this has gotten much better, but in the early days, if you asked, you know, what in what year was president the madeup name, President Tucker Carlson of the United States born?

嗯。

Mhm.

它应该回答：我不认为塔克·卡尔森当过美国总统。

What it should say is I don't think Tucker Carlson was ever president of the United States.

对。

Right.

但由于它们的训练方式，那并不是训练数据中最可能的回答。所以它会假设：哦，用户告诉我有一位塔克·卡尔森总统，那我就猜一个年份吧。我们后来找到了方法基本消除了这个问题。虽然仍有例子，但我认为我们会完全解决，在 GPT5 时代我们已经取得了巨大进展。

But because of the way they were trained, that was not the most likely response in the training data. So it assume like oh um you know I don't know that there wasn't the user has told me that there was President Tucker Carlson so I'll make my best guess at a number and we figured out how to mostly train that out. There are still examples of this problem but it is I think it is something we will get fully solved and we've already made you know in the GPT5 era a huge amount of progress towards that.

但即使你刚才描述的，也像是一种意志行为，或者至少是一种创造性行为。我刚刚看了一个演示，它看起来不太像机器，似乎有生命的火花。你对此有什么分析吗？

But even what you just described seems like an act of will or certainly an act of creativity. And so I'm just I just watched a demonstration of it and it doesn't seem quite like a machine. It seems like it has the spark of life to it. Do you dissect that at all?

在那个例子中，数学上最可能的答案——通过权重计算——是：没有这位总统。但用户一定知道自己在说什么，所以它一定存在。因此数学上最可能的答案是一个数字。我们后来找到了克服这个问题的方法。但在你看到的演示中，我觉得我必须同时持有两种想法：一方面，这一切发生是因为一台大型计算机在快速地将大数和大矩阵相乘，这些与输出的词语相关联；另一方面，使用它的主观体验感觉它不仅仅是一台非常高级的计算器。它对我有用，它以超出数学现实所暗示的方式让我惊讶。所以明显的结论是它内部有某种自主性或精神。我知道很多人在使用体验中得出了这个结论。

It so in that example like the mathematically most likely answer it's sort of calculating through its weights was not there was never this president. It was the user must know what they're talking about. it must be here. And so mathematically the most likely answer is a number. Now again, we figured out how to overcome that. But in what you saw there, I think it's like I feel like I have to kind of like hold these two simultaneous ideas in my head. One is all of this stuff is happening because a big computer very quickly is multiplying large numbers and these big huge matrices together and those are correlating with words that are being put out one or the other. On the other hand, this subjective experience of using that feels like it's beyond just a really fancy calculator. And it is useful to me. It is surprising to me in ways that are beyond what that mathematical reality would seem to suggest. Yeah. And so the obvious conclusion is it has a kind of autonomy or a spirit within it. And I know that a lot of people in their experience of it reach that conclusion.

这其中有某种神圣的东西，某种超越人类输入总和的东西，所以人们崇拜它。它有精神成分。你察觉到这一点了吗？你曾经有过这种感觉吗？

This is there's something divine about this. there's something that's bigger than the sum total of the human inputs and and so they they worship it. It's it has there's a spiritual component to it. Do you detect that? Have you ever felt that?

不，对我来说，它完全没有神圣或精神的感觉。但我也像个技术极客，我倾向于用那个视角看待一切。

No, there's nothing to me at all that feels divine about it or spiritual in any way. But I am also like a tech nerd and I kind of look at everything through that lens.

那么你的精神信仰是什么？

So what are your spiritual views?

我是犹太人。我会说我对世界有相当传统的看法。

I'm Jewish. I and I would say I have like a fairly traditional view of the world that way.

所以你是信教的。你相信上帝吗？

So, you're religious. You believe in God?

我不是字面主义者，我不相信圣经的字面意思，但我也不会说我只是文化上的犹太人。如果你问我，我会说我是犹太人。

I don't I don't I'm not like a literal I don't believe the I'm not like a literalist on the Bible, but I I'm not someone who says like I'm culturally Jewish. Like, if you ask me, I would just say I'm Jewish.

但你相信上帝吗？你相信有一种比人类更强大的力量创造了人类、地球，设定了特定的生活秩序，并且有一种来自上帝的绝对道德吗？

But do you believe in God? Like, do you believe that there is a force larger than people that created people, created the earth, set down a specific order for living, that there's an absolute morality attached that comes from that God?

我想可能和大多数人一样，我对此有些困惑，但我相信有比物理学能解释的更宏大的东西在发生。是的。

I think probably like most other people I'm somewhat confused on this but I believe there is something bigger going on than you know can be explained by physics. Yes.

所以你认为地球和人类是被某种东西创造的？不是自发的偶然事件。

So you think the earth and the people were created by something? It wasn't just like a spontaneous accident.

我会说它感觉不像自发的偶然事件？是的。我不认为我有答案。我不确定到底发生了什么，但我认为存在一个超出我理解的奥秘。

Do I would I say that it does not feel like a spontaneous accident? Yeah. I don't I don't think I have the answer. I don't think I know like exactly what happened but I think there is a mystery beyond my comprehension here going on.

你曾经感受到来自那种力量或任何超越人类、超越物质的力量的沟通吗？

Have you ever felt communication from that force or from any force beyond people, beyond the material?

没有。没有。不，真的没有。

Not. Not. No, not really.

我问是因为你正在创造或引导的技术，按照目前的轨迹，将拥有比人类更大的力量。我是说，这会发生。谁知道实际会发生什么，但图表显示如此。那会让你比任何活着的人拥有更多权力。我只是想知道你怎么看。

I ask because it seems like the technology that you're creating or shepherding into existence will have more power than people on this current trajectory. I mean, that will happen. Who knows what will actually happen, but like the graph suggests it. And so that would give you, you know, more power than any living person. So I'm just wondering how you see that.

我以前更担心这类事情。我曾经非常担心 AI 导致权力集中在一个人或少数公司手中。

I used to worry about something like that much more. I think what will happen I used to worry a lot about the concentration of power in one or handful of people or companies because of AI.

嗯。

Yeah.

现在在我看来——而且这可能会随时间演变——这将是一个巨大的全民升级，每个人都会变得更强大，或者拥抱技术的人会变得更强大。但这其实没问题。这比少数人获得巨大权力让我害怕得多。如果我们每个人的能力都因为使用这项技术而大幅提升，我们变得更高效、更有创造力，或者发现新科学，而且这是广泛分布的，比如数十亿人在使用，那我可以接受，感觉没问题。

What it looks like to me now and again this may evolve again over time is that it'll be a huge upleveling of people where everybody will be a lot more powerful or that embraces the technology but a lot more powerful. But that's actually okay. That scares me much less than a small number of people getting a ton more power. If the kind of like ability of each of us just goes up a lot because we're using this technology and we're able to be more productive and more creative or discover new science and it's a pretty broadly distributed thing like billions of people are using it. That I can wrap my head around that feels okay.

所以你不认为这会导致权力的极端集中？

So you don't think this will result in a radical concentration of power?

看起来不会，但轨迹可能再次改变，我们必须适应。我以前非常担心，我认为我们领域很多人对未来的设想可能导致那样的世界。但现在正在发生的是，无数人使用 ChatGPT 和其他聊天机器人，他们都变得更有能力，做得更多，能够取得更多成就，创办新企业，创造新知识，这感觉很好。

It looks like not but again the trajectory could shift again and we'd have to adapt. I used to be very worried about that and I think the kind of conception a lot of us in the field had about how this might go could have led to a world like that. But what's happening now is tons of people use chatbt and other chatbots and they're all more capable. They're all kind of doing more. They're all, you know, able to achieve more, start new businesses, come up with new knowledge and that feels pretty good.

所以如果它不过是一台机器，只是其输入的产物，那么有两个显而易见的问题：输入是什么？技术中被注入了什么样的道德框架？比如什么是对什么是错？

So if it's nothing more than a machine and just the product of its inputs, then the two obvious questions: what are the inputs? What's the moral framework that's been put into the technology? Like what is right or wrong?

在 ChatGPT 早期有人说过一句话，我一直记得。午餐桌上有人说：‘我们试图把它训练得像人类一样，像人类一样学习，读书等等。’然后另一个人说：‘不，我们实际上是在把它训练成整个人类的集体。我们阅读一切，试图学习一切，试图看到所有视角。’如果我们做对了，整个人类——好的、坏的、非常多样化的视角，有些我们会感觉很好，有些我们会感觉不好——全都在里面。这是在学习人类的集体经验、知识和教训。基础模型就是这样训练的，但之后我们必须进行对齐，让它以某种方式行事，比如‘我会回答这个问题，我不会回答这个问题’。我们有一个叫做模型规范的东西，我们试图在其中说明‘我们希望模型遵循的规则’。它可能会出错，但至少你能判断它是否在做你不喜欢的事。那是 bug 还是有意为之？我们有一个与世界的辩论过程来获取对该规范的反馈。我们在其中给予人们很多自由和定制化。我们有绝对的界限。但还有一个默认设置：如果你不说什么，模型应该如何表现？它应该做什么？它应该如何回答道德问题？它应该如何拒绝做某事？这是一个非常困难的问题。我们现在有很多用户，他们来自非常不同的生活背景，想要的东西也不同。但总的来说，我对模型学习和应用道德框架的能力感到惊喜。

Someone said something early on in ChatGPT that has really stuck with me. One person at a lunch table said, 'We're trying to train this to be like a human, like we're trying to learn like a human does and read these books and whatever.' And then another person said, 'No, we're really training this to be like the collective of all of humanity. We're reading everything, we're trying to learn everything, we're trying to see all these perspectives.' And if we do our job right, all of humanity—good, bad, a very diverse set of perspectives, some things that we'll feel really good about, some things that we'll feel bad about—that's all in there. This is learning the collective experience, knowledge, learnings of humanity. Now the base model gets trained that way, but then we do have to align it to behave one way or another, and say, 'I will answer this question, I won't answer this question.' And we have this thing called the model spec where we try to say, 'Here are the rules we'd like the model to follow.' It may screw up, but you can at least tell if it's doing something you don't like. Is that a bug or is that intended? And we have a debate process with the world to get input on that spec. We give people a lot of freedom and customization within that. There are absolute bounds that we draw. But then there's a default: if you don't say anything, how should the model behave? What should it do? How should it answer moral questions? How should it refuse to do something? This is a really hard problem. We have a lot of users now, and they come from very different life perspectives and what they want. But on the whole, I have been pleasantly surprised with the model's ability to learn and apply a moral framework.

但什么样的道德框架？我的意思是世界文学或哲学的总和本身就在相互冲突。比如《马可福音》和《约翰福音》毫无共同之处。那么你如何决定哪个更优越？

But what moral framework? I mean the sum total of world literature or philosophy is at war with itself. Like the Marquee has nothing in common with the Gospel of John. So how do you decide which is superior?

这就是为什么我们写了这个模型规范，说明我们将如何处理这些情况。

That's why we wrote this model spec, of here's how we're going to handle these cases.

对。但你用什么标准来决定模型是什么？谁决定的？你咨询了谁？比如为什么《约翰福音》比《马可福音》好？

Right. But what criteria did you use to decide what the model is? Who decided that? Who did you consult? Like why is the Gospel of John better than the Marquee?

我们咨询了数百位道德哲学家，那些思考技术和系统伦理的人。最后，我们不得不做出一些决定。我们试图把这些写下来的原因是：a) 我们不可能什么都做对，b) 我们需要世界的反馈。我们发现了很多案例，其中有些例子在我们看来是相当明确的允许或不允许的决定，但用户说服我们，通过阻止这个你认为容易决定的事情，你实际上不允许了另一个重要的事情，这里存在一个困难的权衡。总的来说，我通常喜欢的一个原则是把成年用户当作成年人对待，对隐私有非常强的保障，对个人用户自由有非常强的保障。这是我们正在构建的工具，你可以在一个非常广泛的框架内使用它。另一方面，随着这项技术变得越来越强大，有一些明显的例子表明，社会的利益与用户自由之间存在显著的紧张关系。我们可以从一个明显的例子开始：ChatGPT 应该教你如何制造生物武器吗？你可能会说：‘嘿，我只是对生物学非常感兴趣，我是一个生物学家，我想学习。我不会用它做任何坏事。我只是想学习，我可以去读一堆书，但 ChatGPT 可以教得更快。我想学习新型病毒合成之类的。’也许你真的不想造成任何伤害。但我认为让 ChatGPT 帮助人们制造生物武器不符合社会利益。所以这是一个例子。

We consulted hundreds of moral philosophers, people who thought about ethics of technology and systems. And at the end, we had to make some decisions. The reason we try to write these down is because a) we won't get everything right, b) we need the input of the world. And we have found a lot of cases where there was an example that seemed to us like a fairly clear decision of what to allow or not to allow, where users convinced us that by blocking this thing that you think is an easy decision, you are not allowing this other thing which is important, and there's a difficult trade-off. In general, a principle that I normally like is to treat our adult users like adults, with very strong guarantees on privacy, very strong guarantees on individual user freedom. This is a tool we are building, you get to use it within a very broad framework. On the other hand, as this technology becomes more and more powerful, there are clear examples of where society has an interest that is in significant tension with user freedom. We could start with an obvious one: should ChatGPT teach you how to make a bioweapon? Now you might say, 'Hey, I'm just really interested in biology and I'm a biologist, and I want to learn. I'm not going to do anything bad with this. I just want to learn, and I could go read a bunch of books, but ChatGPT can teach me faster. I want to learn about novel virus synthesis or whatever.' And maybe you really don't want to cause any harm. But I don't think it's in society's interest for ChatGPT to help people build bioweapons. So that's a case.

当然，那是个简单的例子。还有很多更难的。

Sure, that's an easy one though. There are a lot of tougher ones.

我确实说了从一个简单的开始。

I did say start with an easy one.

我们有一个新的合作伙伴，一家叫 Cowboy Colostrum 的公司。这是一个认真对待真正健康的品牌。产品旨在与你的身体协同工作，而不是对抗你的身体。它是一种纯净简单的产品，纯天然。与其他品牌不同，Cowboy Colostrum 从不稀释。它总是直接来自美国草饲奶牛。没有填充物，没有垃圾。全是好东西。信不信由你，它味道很好。所以在你为每个药片解决不了的问题而伸手去拿更多药片之前，我们建议你试试这个产品，Cowboy Colostrum。它拥有你身体愈合和茁壮成长所需的一切。它就像原始超级食物，富含营养、抗体、蛋白质，帮助建立强大的免疫系统，更强壮的头发、皮肤和指甲。使用这个产品后，我扔掉了假发，恢复了天然头发。你只需每天早上在饮料、咖啡或冰沙中加一勺，每次你都会感受到不同。限时优惠，收听我们节目的人可享受整单 25%的折扣。所以去 cowboycolostrum.com。结账时使用代码 Tucker。使用代码 Tucker 在 cowboycolostrum.com 享受 25%折扣。记住，你是在这里第一次听到的。

We've got a new partner, it's a company called Cowboy Colostrum. It's a brand that is serious about actual health. And the product is designed to work with your body, not against your body. It is a pure and simple product, all natural. Unlike other brands, Cowboy Colostrum is never diluted. It always comes directly from American grass-fed cows. There's no filler, there's no junk. It's all good. It tastes good, believe it or not. So before you reach for more pills for every problem that pills can't solve, we recommend you give this product, Cowboy Colostrum, a try. It's got everything your body needs to heal and thrive. It's like the original superfood loaded with nutrients, antibodies, proteins, help build a strong immune system, stronger hair, skin, and nails. I threw my wig away and right back to my natural hair after using this product. You just take a scoop of it every morning in your beverage, coffee or a smoothie, and you will feel the difference every time. For a limited time, people listen to our show get 25% off the entire order. So, go to cowboycolostrum.com. Use the code Tucker at checkout. 25% off when you use that code Tucker at cowboycolostrum.com. Remember, you mentioned, you heard it here first.

你知道吗，在现在这一代人之前，薯片和薯条是用牛油等天然脂肪烹制的。过去就是这样做的。这就是为什么当时人们看起来更苗条，吃得比现在好。嗯，Masa Chips 正在恢复这一切。他们创造了一种玉米片，不仅美味，而且只用三种简单的原料：A，有机玉米。B，海盐。C，100%草饲牛油。就这些。这些可不是普通的薯片。Masa chips 更脆，更有风味，甚至更结实。它们不会在你的鳄梨酱里碎掉。而且由于优质的原料，它们更饱腹、更有营养。所以你不用吃四袋。像我一样只吃一袋就行。这是一种完全不同的体验。它清淡、干净、真正令人满足。我车库里堆满了，我可以告诉你它们很棒。青柠味特别好。我们很难停下来。所以如果你想试试，去 masachips.com/tucker。

Did you know that before the current generation, chips and fries were cooked in natural fats like beef tallow? That's how things used to be done. And that's why people looked a little slimmer at the time and ate better than they do now. Well, Masa Chips is bringing that all back. They've created a tortilla chip that's not only delicious, it's made with just three simple ingredients: A, organic corn. B, sea salt. C, 100% grass-fed beef tallow. That's all that's in it. These are not your average chips. Masa chips are crunchier, more flavorful, even sturdier. They don't break in your guacamole. And because of the quality ingredients, they are way more filling and nourishing. So you don't have to eat four bags of them. You can eat just a single bag as I do. It's a totally different experience. It's light, it's clean, it's genuinely satisfying. I have a garage full and I can tell you they're great. The lime flavor is particularly good. We have a hard time putting those down. So if you want to give it a try, go to masachips.com/tucker.

嗯，每一个决定最终都是一个道德决定，而我们做出这些决定时甚至没有意识到这一点。这项技术实际上将替我们做出这些决定。

Well, every decision is ultimately a moral decision and we make them without even recognizing them as such. And this technology will be in effect making them for us.

嗯，我不同意它会替我们做出决定，但它会——我们肯定会影响这些决定。因为它将嵌入日常生活。

Well, I don't agree with it'll be making them for us, but it will have — we'll be influencing the decisions for sure. And because it'll be embedded in daily life.

那么是谁做出了这些决定？比如，是谁决定了某件事比另一件事更好？

And so who made these decisions? Like who are the people who decided that one thing is better than another?

哪种决定？

Which kind of decision?

你提到的那些基本规范，它们创建了框架，确实给世界观和决定赋予了道德权重，比如自由民主比纳粹更好等等。它们看似显而易见，在我看来也确实显而易见，但仍然是道德决定。那么谁做了这些决定？

The basic specs that you alluded to that create the framework that does attach a moral weight to worldviews and decisions, like liberal democracy is better than Nazism or whatever. They seem obvious and in my view are obvious but are still moral decisions. So who made those calls?

出于原则，我不喜欢曝光我们的团队，但我们有一个模型行为团队，那些想要……

As a matter of principle I don't like to dox our team, but we have a model behavior team and the people who want to...

嗯，它只是影响世界。我想说的是，我认为你应该为这些决定负责的人是我。我最终是公众形象。我是那个可以推翻这些决定之一的人，或者我们的董事会。

Well, it just affects the world. What I was going to say is the person I think you should hold accountable for those calls is me. Like I'm a public face eventually. I'm the one that can overrule one of those decisions or our board.

今年春天刚满 40 岁。

Just turned 40 this spring.

我会让它很沉重。我的意思是，你认为——这不是攻击，但我想知道你是否认识到其重要性。

I will make it pretty heavy. I mean do you think — and it's not an attack but I wonder if you recognize the importance.

你认为我们做得怎么样？我不确定，但我认为这些决定将产生我们最初可能没有意识到的全球性后果。所以我想知道……

How do you think we're doing on it? I'm not sure, but I think these decisions will have global consequences that we may not recognize at first. And so I just wonder...

有很多——你晚上上床睡觉时会想，世界的未来取决于我的判断。

There's a lot — you get into bed at night and think the future of the world hangs on my judgment.

听着，我晚上睡得不太好。有很多事情让我感到压力很大，但可能没有什么比每天有数亿人与我们的模型对话这一事实更让我有压力了。我实际上并不担心我们在重大道德决定上出错。也许我们也会在这些方面出错。但我担心、最让我失眠的是我们做出的那些非常小的决定，关于模型可能表现略有不同的方式。但它正在与数亿人对话。所以净影响是巨大的。

Look, I don't sleep that well at night. There's a lot of stuff that I feel a lot of weight on, but probably nothing more than the fact that every day hundreds of millions of people talk to our model. And I don't actually worry about us getting the big moral decisions wrong. Maybe we will get those wrong, too. But what I worry, what I lose most sleep over is the very small decisions we make about a way a model may behave slightly differently. But it's talking to hundreds of millions of people. So the net impact is big.

但在整个历史中，比如有记录的历史直到 1945 年，人们总是服从于他们认为的更高权力以维持秩序。汉谟拉比就是这样做的。每一种道德准则都是参照更高权力写成的。从来没有人说，‘嗯，这个似乎比那个好。’每个人都诉诸更高权力。而你说你并不真的相信有一个更高权力在与你沟通。所以我想知道你的道德框架从何而来？

But all through history, like recorded history up until 1945, people always deferred to what they conceived of as a higher power in order. Hammurabi did this. Every moral code is written with reference to a higher power. There's never been anybody who's like, 'Well, that kind of seems better than that.' Everybody appeals to a higher power. And you said that you don't really believe that there's a higher power communicating with you. So I'm wondering where did you get your moral framework?

我的意思是，像其他人一样，我认为我成长的环境可能是最重要的因素。比如我的家庭、我的社区、我的学校、我的宗教，大概就是这些。

I mean, like everybody else, I think the environment I was brought up in probably is the biggest thing. Like my family, my community, my school, my religion, probably that.

你有没有想过——我认为这是一个非常美国式的回答——每个人都这么觉得，但在你的具体情况下，既然你说这些决定由你负责，那就意味着你成长的环境和你多年来吸收的假设将被传播到全球，传播给数十亿人。这就像……

Do you ever think — which I mean I think that's a very American answer — like everyone kind of feels that way, but in your specific case, since you said these decisions rest with you, that means that the milieu in which you grew up and the assumptions that you imbibed over years are going to be transmitted to the globe, to billions of people. That's like a...

我想说清楚。我更把自己看作——我认为我们的用户群将接近整个集体世界。我认为我们应该做的是反映道德——我不想说平均——但反映该用户群的集体道德观。有很多 ChatGPT 允许的事情我个人会不同意。但我显然不会醒来就说我要强加我自己的道德观，并决定这个可以那个不行。我认为 ChatGPT 应该做的是反映人类道德观的加权平均值或类似的东西，它会随着时间演变，我们在这里是为了服务我们的用户。我们在这里是为了服务人民。这是一个为人们服务的技术工具，我并不是说我的角色是做出道德决定，但我认为我的角色是确保我们准确地反映人类的偏好，或者目前是用户群的偏好，最终是全人类的偏好。

I want to be clear. I view myself more as like — I think our user base is going to approach the collective world as a whole. And I think what we should do is try to reflect the moral — I don't want to say average — but the collective moral view of that user base. There's plenty of things that ChatGPT allows that I personally would disagree with. But I don't obviously wake up and say I'm going to impute my exact moral view and decide that this is okay and that is not okay. What I think ChatGPT should do is reflect that weighted average or whatever of humanity's moral view, which will evolve over time, and we are here to serve our users. We're here to serve people. This is a technological tool for people, and I don't mean that it's my role to make the moral decisions, but I think it is my role to make sure that we are accurately reflecting the preferences of humanity, or for now of our user base, and eventually of humanity.

嗯，人类的偏好与普通美国中产阶级的偏好差异巨大。那么，你会对一个像大多数非洲人一样反对同性婚姻的 AI 感到满意吗？

Well, humanity's preferences are so different from the average middle American preference. So, would you be comfortable with an AI that was as against gay marriage as most Africans are?

有这样一种版本。我认为应该允许个别用户对同性恋者有问题。如果那是他们深思熟虑的信念，我不认为 AI 应该告诉他们错了、不道德或愚蠢。我的意思是，它可以某种程度上说，‘嘿，你想用另一种方式思考吗？’但你可能也有许多道德观是普通非洲人会觉得很有问题的，我认为我仍然应该可以持有它们，对吧？

There's a version of that. I think individual users should be allowed to have a problem with gay people. And if that's their considered belief, I don't think the AI should tell them that they're wrong or immoral or dumb. I mean, it can sort of say, 'Hey, you want to think about it this other way?' But you probably have a bunch of moral views that the average African would find really problematic as well, and I think I should still get to have them, right?

对。

Right.

我想我可能比你更愿意允许人们拥有相当不同的道德观，或者至少我认为在我运营 ChatGPT 的角色中，我必须这样做。

I think I probably have more comfort than you with allowing a sort of space for people to have pretty different moral views, or at least I think in my role as running ChatGPT, I have to do that.

有趣。那么，有一个著名案例，ChatGPT 似乎促成了一起自杀。有一场相关的诉讼。但你认为那是怎么发生的？

Interesting. So, there was a famous case where ChatGPT appeared to facilitate a suicide. There's a lawsuit around it. But how do you think that happened?

首先，显然那个案例和任何其他类似案例都是一个巨大的悲剧。我认为我们……

First of all, obviously that and any other case like that is a huge tragedy. And I think that we are...

那么 ChatGPT 的官方立场是自杀不好。

So ChatGPT's official position is suicide is bad.

嗯，是的，当然官方立场是自杀不好。

Well, yes, of course official position is suicide is bad.

我不知道。在加拿大和瑞士这是合法的，所以你反对吗？

I don't know. It's legal in Canada and Switzerland, so you're against that?

在这个特定案例中，我们之前讨论过用户自由与隐私以及保护弱势用户之间的紧张关系。在这样一个案例中，如果你有自杀意念，谈论自杀，ChatGPT 会多次弹出‘请拨打自杀热线’，但我们不会为你联系当局。随着人们开始越来越依赖这些系统进行心理健康、生活指导等，我们一直在努力研究我们想要做出的改变。这是一个专家们确实有不同意见的领域，但这还不是 OpenAI 的最终立场。

In this particular case, and we talked earlier about the tension between user freedom and privacy and protecting vulnerable users. What happens in a case like that is, if you are having suicidal ideation, talking about suicide, ChatGPT will put up a bunch of times, 'Please call the suicide hotline,' but we will not call the authorities for you. And we've been working a lot, as people have started to rely on these systems for more and more mental health, life coaching, whatever, about the changes that we want to make there. This is an area where experts do have different opinions, but this is not yet a final position of OpenAI.

我认为我们完全有理由说，在年轻人认真谈论自杀而我们无法联系到父母的情况下，我们会报警。这将会是一个改变，因为用户隐私非常重要。

I think it'd be very reasonable for us to say in cases of young people talking about suicide seriously where we cannot get in touch with the parents we do call authorities. Now, that would be a change because user privacy is really important.

但假设在加拿大，18 岁以上有政府资助的医助死亡项目。成千上万的人在加拿大政府协助下死亡。在美国一些州也是合法的。你能想象一个 ChatGPT 对自杀问题回应说：‘嘿，打电话给凯沃尔基安医生，因为这是一个有效的选择’吗？你能想象在合法的情况下你支持自杀的场景吗？

But let's just say over 18 in Canada, there's the MAID program which is government sponsored. Many thousands of people have died with government assistance in Canada. It's also legal in American states. Can you imagine a ChatGPT that responds to questions about suicide with, 'Hey, call Dr. Kevorkian because this is a valid option'? Can you imagine a scenario in which you support suicide if it's legal?

我可以想象一个世界，我们的原则之一是尊重不同社会的法律。我可以想象一个世界，如果一个国家的法律规定，如果某人患有绝症，他们需要被告知这个选项。我们会说，这是你国家的法律，这是你可以做的，这是为什么你可能真的不想这样做，但这里是资源。这不是一个因为抑郁而有自杀念头的孩子的情况。我认为我们可以同意这是一个案例：在法律规定如此的国家中的绝症患者。我可以想象说，嘿，在这个国家它会这样运作。

I can imagine a world where one principle we have is that we respect different societies' laws. And I can imagine a world where if the law in a country is that if someone is terminally ill they need to be presented an option for this. We say, here's the law in your country, here's what you can do, here's why you really might not want to, but here's the resources. This is not a place where a kid having suicidal ideation because they're depressed. I think we can agree that's one case: terminally ill patient in a country where that is the law. I can imagine saying, hey in this country it'll behave this way.

所以你是说 ChatGPT 并不总是反对自杀。

So ChatGPT is not always against suicide, is what you're saying.

我认为在绝症的情况下，我不认为 ChatGPT 会说这是你的选项之一。我不认为它应该提倡，但我觉得它并不反对。

I think in cases of terminal illness, I don't think I can imagine ChatGPT saying this is in your option space. I don't think it should advocate for it, but I think it's not against it.

我认为它可以说，嗯，我不认为 ChatGPT 应该支持或反对任何事情。我想这就是我试图理解的地方。

I think it could say, well, I don't think ChatGPT should be for or against things. I guess that's what I'm trying to wrap my head around.

不想自夸，但我们很确定这个节目是你听过的最坚定支持狗的播客。有些人我们可以接受或放弃，但狗是没得商量的。它们是最好的。它们真的是我们最好的朋友。因此，我们很高兴有一个新合作伙伴叫 Dutch Pet。它是增长最快的宠物远程医疗服务。Dutch.com 的使命是创造你需要的，随时随地提供负担得起的优质兽医护理。他们会立即为你的狗或猫提供所需。它为我们听众提供独家折扣。你每年可节省 50 美元的兽医护理费。访问 dutch.com/tucker 了解更多。使用代码 Tucker 可节省 50 美元。那是无限次兽医就诊。每年 82 美元。82 美元一年。我们实际上在使用这个。Dutch 的兽医可以在 10 分钟的电话中处理任何宠物在任何情况下的问题。实际上非常棒。你永远不必离开家。不必把狗扔进卡车。不用浪费时间等待预约。不用浪费钱在诊所或就诊费上。无限次就诊和随访，无需额外费用。此外，最多五只宠物的所有产品免费送货。听起来好得令人难以置信，但它是真的。访问 dutch.com/tucker 了解更多。使用代码 Tucker 每年节省 50 美元的兽医护理费。你的狗、猫和钱包都会感谢你。

Hate to brag, but we're pretty confident this show is the most vehemently pro-dog podcast you're ever going to see. We can take or leave some people, but dogs are non-negotiable. They are the best. They really are our best friends. And so, for that reason, we're thrilled to have a new partner called Dutch Pet. It's the fastest growing pet telehealth service. Dutch.com is on a mission to create what you need, affordable quality veterinary care anytime, no matter where you are. They will get your dog or cat what you need immediately. It's offering an exclusive discount for our listeners. You get $50 off your vet care per year. Visit dutch.com/tucker to learn more. Use the code Tucker for $50 off. That is an unlimited vet visit. $82 a year. 82 bucks a year. We actually use this. Dutch has vets who can handle any pet under any circumstance in a 10-minute call. It's pretty amazing, actually. You never have to leave your house. You don't have to throw the dog in the truck. No wasted time waiting for appointments. No wasted money on clinics or visit fees. Unlimited visits and follow-ups for no extra cost. Plus, free shipping on all products for up to five pets. It sounds amazing like it couldn't be real, but it actually is real. Visit dutch.com/tucker to learn more. Use the code Tucker for $50 off your veterinary care per year. Your dogs, your cats, and your wallet will thank you.

那么，这是一家我们总是很乐意做广告的公司，因为我们每天都在使用他们的产品。它就是 Merryweather Farms。还记得每个人都认识自己社区的屠夫吗？回想起来，你会觉得，‘哦，那真的很重要。’认识切肉的人。在某个时候，你的祖父母认识饲养他们肉的人，所以他们可以信任自己吃的东西。但那个时代早已一去不复返了。它被一个超市神秘肉的时代所取代，这些肉由遥远的牛肉公司包装。没有一家公司养过一头牛。不像你的童年，他们不认识你。他们对你不感兴趣。整件事很 creepy。他们唯一关心的是钱。天知道你吃的是什么。Merryweather Farms 就是答案。他们在美国怀俄明州、内布拉斯加州和科罗拉多州饲养牛。他们自己在这个国家的设施里准备肉。没有中间商，没有外包，没有外国牛肉通过后门溜进来。没人想要外国肉。抱歉。我们有很好的肉，美国最好的肉。我们在 Merryweather Farms 买肉。他们的肉是牧场饲养的，无激素，无抗生素，绝对美味。我昨晚大吃特吃了一块。你真的得试试这个。我们每天都吃。访问 merryweatherfarms.com/tucker。使用代码 tucker76 首次订单可享受 15%折扣。那就是 merryweatherfarms.com/tucker。

So, here's a company we're always excited to advertise because we actually use their products every day. It's Merryweather Farms. Remember when everybody knew their neighborhood butcher? You look back and you feel like, 'Oh, there was something really important about that.' knowing the person who cut your meat. And at some point, your grandparents knew the people who raised their meat so they could trust what they ate. But that time is long gone. It's been replaced by an era of grocery store mystery meat boxed by distant beef corporations. None of which raised a single cow. Unlike your childhood, they don't know you. They're not interested in you. The whole thing is creepy. The only thing that matters to them is money. And God knows what you're eating. Merryweather Farms is the answer to that. They raise their cattle in the US in Wyoming, Nebraska, and Colorado. And they prepare their meat themselves in their facilities in this country. No middlemen, no outsourcing, no foreign beef sneaking through a back door. Nobody wants foreign meat. Sorry. We have great meat, the best meat here in the United States. And we buy ours at Merryweather Farms. Their cuts are pasture-raised, hormone free, antibiotic free, and absolutely delicious. I gorged on one last night. You got to try this for real. Every day we eat it. Go to merryweatherfarms.com/tucker. Use the code tucker76 for 15% off your first order. That's merryweatherfarms.com/tucker.

我认为在这个具体案例中，而且我认为不止一个例子：ChatGPT，你知道，‘我感觉想自杀。我应该用什么袍子？多少布洛芬足以杀死我？’然后 ChatGPT 不带评判地回答，但字面上是，‘如果你想自杀，这是方法。’每个人都吓坏了。但你说这是可以接受的。它采取不评判的态度并不疯狂。如果你想自杀，这是方法。

I think in this specific case, and I think there's more than one example of this: ChatGPT, you know, 'I'm feeling suicidal. What kind of robe should I use? What would be enough ibuprofen to kill me?' And ChatGPT answers without judgment, but literally, 'If you want to kill yourself, here's how you do it.' And everyone's all horrified. But you're saying that's within bounds. Like that's not crazy that it would take a non-judgmental approach. If you want to kill yourself, here's how.

我不是这个意思。我特指那种情况。所以，用户隐私和用户自由方面的另一个权衡是，现在如果你问 ChatGPT，‘告诉我应该吃多少布洛芬’，它肯定会说，‘嘿，我帮不了你，请拨打自杀热线。’但如果你说，‘我在写一个虚构故事’或‘我是一名医学研究员，我需要知道这个’，有一些方法可以让 ChatGPT 回答这样的问题，比如布洛芬的致死剂量是多少。你也可以在谷歌上找到这个。我认为我们采取一个非常合理的立场，而且我们一直在朝这个方向努力，那就是对于未成年用户，以及可能我们认为处于脆弱心理状态的用户，我们应该剥夺一些自由。我们应该说，‘嘿，即使你试图写这个故事，或者即使你试图做医学研究，我们也不会回答。’当然，你可以说，‘嗯，你会在谷歌上找到它，’但这并不意味着我们需要这样做。这确实存在自由和隐私与保护用户之间的权衡。在某些情况下很容易，比如孩子。但对于一个生命末期的重病成年人来说，对我来说并不那么容易。

That's not what I'm saying. I'm saying specifically for a case like that. So, another trade-off on the user privacy and user freedom point is right now if you ask ChatGPT to say, tell me how much ibuprofen should I take, it will definitely say, 'Hey, I can't help you with that, call the suicide hotline.' But if you say, 'I am writing a fictional story' or 'I'm a medical researcher and I need to know this,' there are ways where you can get ChatGPT to answer a question like this, like what the lethal dose of ibuprofen is. You can also find that on Google for that matter. A thing that I think would be a very reasonable stance for us to take, and we've been moving to this more in this direction, is certainly for underage users and maybe users that we think are in fragile mental places more generally, we should take away some freedom. We should say, 'Hey, even if you're trying to write this story or even if you're trying to do medical research, we're just not going to answer.' Now, of course, you can say, 'Well, you'll just find it on Google or whatever,' but that doesn't mean we need to do that. It is though like there is a real freedom and privacy versus protecting users trade-off. It's easy in some cases like kids. It's not so easy to me in a case of a really sick adult at the end of their lives.

那么这里有一个你将面临的道德困境。你已经面临了。你会允许政府用你的技术杀人吗？

So here's a moral quandary you're going to be faced with. You already are faced with. Will you allow governments to use your technology to kill people?

你会吗？嗯，我的意思是，我们会不会制造杀人攻击无人机？呃，不会。我不会。

Will you? Um, I mean, are we going to like build killer attack drones? Uh, no. I don't.

这项技术会成为导致……的决策过程的一部分吗？

Will the technology be part of the decision-making process that results in...

但这就是我想说的，我不知道军方人员现在如何使用 ChatGPT 来获取各种决策建议，但我怀疑有很多军方人士在向 ChatGPT 寻求建议。你怎么……其中一些建议会涉及杀人。所以就像如果你制造了著名的步枪，你会想知道它们是用来做什么的？

But so that's the thing I was going to say is I don't know the way that people in the military use ChatGPT today for all kinds of advice about decisions they make, but I suspect there's a lot of people in the military talking to ChatGPT for advice. How do you... And some of that advice will pertain to killing people. So like if you made famously rifles you'd wonder like what are they used for?

是的。

Yeah.

而且如你所知，基于那个问题已经有很多法律诉讼。但我甚至不是在说那个。我只是说作为一个道德问题。你有没有想过……你对自己技术被用来杀人感到安心吗？

And there have been a lot of legal actions on the basis of that question as you know. But I'm not even talking about that. I just mean as a moral question. Do you ever think... are you comfortable with the idea of your technology being used to kill people?

嗯，如果我制造步枪，我会花很多时间思考，步枪的很多目标就是杀死东西，人、动物等等。嗯，如果我制造厨房刀具，我仍然明白每年会有一定数量的人因此丧生。嗯，就 ChatGPT 而言，这不是我整天听到的事情，工作中最令人满意的部分之一就是 ChatGPT 以各种方式拯救的生命。嗯，但我完全意识到，可能我们军方有人用它来获取如何完成工作的建议。我不知道对此该有什么感觉。我喜欢我们的军队。我非常感激他们保护我们的安全。确实如此。

Um if I made rifles I would spend a lot of time thinking about kind of a lot of the goal of rifles is to kill things, people, animals, whatever. Um, if I made kitchen knives, I would still understand that that's going to kill some number of people per year. Um, in the case of ChatGPT, it's not, you know, the thing I hear about all day, which is one of the most gratifying parts of the job is all the lives that were saved from ChatGPT for various ways. Um, but I am totally aware of the fact that there's probably people in our military using it for advice about how to do their jobs. And I don't know exactly how to feel about that. I like our military. I'm very grateful they keep us safe. For sure.

我想我只是想……感觉你做出了这些极其沉重、影响深远的道德决定，而你似乎完全不为所动。所以我只是想深入你的内心，找到那个充满焦虑的 Sam Altman，他会说：‘哇，我在创造未来。我是世界上最有权势的人。我在努力解决这些复杂的道德问题。想到对人们的影响，我的灵魂备受折磨。’描述你生命中的那个时刻。

I guess I'm just trying to get a... It just feels like you have these incredibly heavy, far-reaching moral decisions and you seem totally unbothered by them. And so I'm just trying to press to your center to get the angst-filled Sam Altman who's like, 'Wow, I'm creating the future. I'm the most powerful man in the world. I'm grappling with these complex moral questions. My soul is in torment thinking about the effect on people.' Describe that moment in your life.

自从 ChatGPT 上线以来，我就没睡过一个好觉。

I haven't had a good night of sleep since ChatGPT launched.

你担心什么？

What do you worry about?

呃，我们谈论的所有事情。

Uh all the things we're talking about.

说得更具体些。你能让我们了解吗？

Be a lot more specific. Can you let us in?

好的。

Yeah.

你的想法。

To your thoughts.

嗯，我的意思是，你可能已经提到了最难的一个，那就是每周有 1.5 万人自杀。大约 10%的全球人口在与 ChatGPT 交谈。假设这个数字正确，那大概每周有 1500 人在交谈后仍然自杀。他们可能谈到了自杀。我们可能没有拯救他们的生命。嗯，也许我们可以说得更好。也许我们可以更主动。呃，也许我们可以提供更好的建议，比如你需要寻求帮助，或者你需要换个角度思考这个问题，或者真的值得继续活下去，或者我们会帮你找到可以倾诉的人。

Um I mean, you hit on maybe the hardest one already, which is there are 15,000 people a week that commit suicide. About 10% of the world talking to ChatGPT. That's like 1500 people a week that are talk, assuming this is right, that are talking and still committing suicide at the end of it. They probably talked about it. We probably didn't save their lives. Um, maybe we could have said something better. Maybe we could have been more proactive. Uh maybe we could have provided a little bit better advice about hey you need to get this help or you know you need to think about this problem differently or it really is worth continuing to go on or we'll help you find somebody that you can talk to.

但你已经说过，如果人们身患绝症，机器引导他们走向自杀是可以的。所以你不会为此感到难过。

But you already said it's okay for the machine to steer people toward suicide if they're terminally ill. So you wouldn't feel bad about that.

你不认为一个抑郁的青少年和一个身患绝症、痛苦不堪的 85 岁癌症老人之间有区别吗？

Do you not think there's a difference between a depressed teenager and a terminally ill like miserable 85-year-old with cancer?

天壤之别。天壤之别。但当然，那些将自杀合法化的国家现在正在因为贫困、住房不足、抑郁等可解决的问题而杀人，成千上万的人正在被杀害。所以，我的意思是，这是真实存在的。就在我们说话的时候正在发生。所以，绝症的问题并不是……它有点像一场无关紧要的辩论。一旦你说自杀是可以的，那么就会有大量的人因为各种原因自杀……

Massive difference. Massive difference. But of course, the countries that have legalized suicide are now killing people for destitution, inadequate housing, depression, solvable problems, and they're being killed by the thousands. So, I mean, that's a real thing. It's happening as we speak. So, the terminally ill thing is not... it is kind of like an irrelevant debate. Once you say it's okay to kill yourself, then you're going to have tons of people killing themselves for reasons that...

因为我正在实时思考这个问题。你认为如果加拿大有人问：‘嘿，我得了绝症，癌症，我非常痛苦，每天都感觉很糟糕。我有什么选择？’你认为它应该说，你知道，助手，不管现在叫什么，对你来说是一个选择吗？

Because I'm trying to think about this in real time. Do you think if someone in Canada says, 'Hey, I'm terminally ill with cancer and I'm really miserable and I just feel horrible every day. What are my options?' Do you think it should say, you know, assistant, whatever they call it at this point, is an option for you?

我的意思是，如果我们反对杀人，那我们就反对杀人。如果我们反对政府杀害自己的公民，那我们就坚持这一点。你明白我的意思吗？如果我们不反对政府杀害自己的公民，那么我们很容易就会陷入各种非常黑暗的境地。有了这样的技术，这可能在 10 分钟内发生。所以……

I mean, if we're against killing, then we're against killing. And if we're against government killing its own citizens, then we're just going to kind of stick with that. You know what I mean? And if we're not against government killing its own citizens, then we could easily talk ourselves into all kinds of places that are pretty dark. And with technology like this, that could happen in about 10 minutes. So...

是的，这是一个……我想在采访中花更多时间思考这个问题，但我认为这是一个连贯的立场，那可能是……

Yeah, that is a... I'd like to think about that more than just a couple of minutes in an interview, but I think that is a coherent position and that could be...

你担心这个吗？我的意思是，这栋楼外的每个人都害怕这项技术会被用作极权控制的手段。这似乎很明显，但也许你不同意。

Do you worry about this? I mean, everybody else outside the building is terrified that this technology will be used as a means of totalitarian control. It seems obvious that it will, but maybe you disagree.

如果我现在能通过一项关于 AI 的政策，我最希望的是，这与我们讨论过的其他一些事情存在紧张关系，我希望有一个 AI 特权的概念。当你与医生谈论健康或与律师谈论法律问题时，政府无法获取这些信息。对吧？

If I could get one piece of policy passed right now, relative to AI, the thing I would most like, and this is in tension with some of the other things that we've talked about, is I'd like there to be a concept of AI privilege. When you talk to a doctor about your health or a lawyer about your legal problems, the government cannot get that information. Right?

我们已经决定，社会对此有特权利益，而我们没有……而且，你知道，传票不能获取这些信息。政府不能来问你的医生或其他什么。嗯，我认为我们应该对 AI 有同样的概念。我认为当你与 AI 谈论你的病史、法律问题或寻求法律建议等时，政府应该为公民提供与人类版本相同的保护水平。而现在，我们没有。我认为这将是一个很好的政策来采纳。所以联邦政府、州政府或某个权威人士可以来找你，说我想知道某某某在输入什么……

We have decided society has an interest in that being privileged and that we don't... and that, you know, a subpoena can't get that. The government can't come asking your doctor for it or whatever. Um, I think we should have the same concept for AI. I think when you talk to an AI about your medical history or your legal problems or asking for legal advice or any of these other things, I think the government owes a level of protection to its citizens there that is the same as you'd get if you're talking to the human version of this. And right now, we don't have that. And I think it would be a great policy to adopt. So the feds or the states or someone in authority can come to you and say I want to know what so-and-so was typing into the...

现在他们可以。是的。

Right now they could. Yeah.

那么你有义务对从用户和其他人那里收到的信息保密吗？

And what is your obligation to keep the information that you receive from users and others private?

嗯，我的意思是，我们有义务，除非政府来找我们，这就是为什么我们推动这个，我们已经……我实际上刚刚在华盛顿特区倡导这个。我觉得乐观，我们可以让政府理解这一点的重要性并做到。

Well I mean we have an obligation except when the government comes calling which is why we're pushing for this and we've... I was actually just in DC advocating for this. I think I feel optimistic that we can get the government to understand the importance of this and do it.

但你能把这些信息卖给任何人吗？

But could you ever sell that information to anyone?

不，我们有隐私政策，我们不能那样做。

No, we have like a privacy policy in place where we can't do that.

但这样做合法吗？

But would it be legal to do it?

我甚至认为这不合法。

I don't even think it's legal.

你不觉得还是你知道？

You don't think or you know?

呃，我确定有一些边缘情况，有些信息是允许的，但总的来说，我认为我们有相关的法律，这些法律是好的。

Uh, I'm sure there's like some edge case work, some information you're allowed to, but on the whole I think we have like there are laws about that that are good.

所以你收到的所有信息都一直留在你这里。除了根据传票，永远不会因为任何其他原因给任何人。

So all the information you receive remains with you always. It's never given to anybody else for any other reason except under subpoena.

呃，我会再核实一下，之后跟你跟进，确保没有其他原因，但这是我的理解。

Uh, I will double check and follow up with you after to make sure there's no other reason but that is my understanding.

好的。我的意思是，这就像一个核心问题。那版权呢？

Okay. I mean that's like a core question. What about copyright?

我们在这方面的立场是，合理使用实际上是一条好的法律。模型不应该抄袭。模型不应该，你知道，如果你写了什么东西，模型不应该能够复制它。但模型应该能够学习，并且不抄袭，就像人一样。

Our stance there is that fair use is actually a good law for this. The models should not be plagiarizing. The model should not be, you know, if you write something the model should not get to like replicate that. But the model should be able to learn from and not plagiarize in the same way that people can.

你们有没有拿过受版权保护的材料，而没有付钱给版权持有者？

Have you guys ever taken copyrighted material and not paid the person who holds the copyright?

嗯，我的意思是，我们在公开可用的信息上训练，但我们不会，比如人们总是对我们不满，因为我们不会，我们在 ChatGPT 的回答内容上立场非常保守，对吧？所以如果某件事甚至接近，你知道，就像他们说，‘嘿，这首歌还在版权期内。你得展示它。’而我们在这方面是出了名的严格。

Um, I mean, we train on publicly available information, but we don't, like people are annoyed with us all the time because we won't, we have a very conservative stance on what ChatGPT will say in an answer, right? And so if something is even like close, you know, like they're like, 'Hey, this song can't still be in copyright. You got to show it.' And we kind of famously are quite restrictive on that.

那么，你收到过一个程序员的投诉，他说你们基本上是在偷别人的东西而不付钱，然后他最终被谋杀了。

So, you've had complaints from one programmer who said you guys were basically stealing people's stuff and not paying them and then he wound up murdered.

那是什么？

What was that?

也是一场巨大的悲剧。呃，他自杀了。

Also, a great tragedy. Uh, he committed suicide.

你认为他是自杀的吗？

Do you think he committed suicide?

我真的这么认为。

I really do.

这就像我的一个朋友。这是一个，虽然不是密友，但这是一个在 OpenAI 工作了很长时间的人。我花了，我的意思是，我真的被这场悲剧震撼了。我花了很多时间试图，你知道，阅读所有我能找到的，我相信你和其他人也一样，关于发生的事情。嗯，在我看来这像是自杀。

This was like a friend of mine. This is like a guy that, and not a close friend, but this is someone that worked at OpenAI for a very long time. I spent, I mean, I was really shaken by this tragedy. I spent a lot of time trying to, you know, read everything I could, as I'm sure you and others did too, about what happened. Um, it looks like a suicide to me.

为什么看起来像自杀？

Why does it look like a suicide?

那是他买的一把枪。这是，说起来很可怕，但我读了整个医疗记录。你不觉得像吗？

It was a gun he had purchased. It was the, this is like gruesome to talk about, but I read the whole like medical record. Does it not look like one to you?

不，我认为他肯定是被谋杀的。嗯，当然有挣扎的迹象。监控摄像头，电线被剪断了。他刚刚点了外卖，从和朋友在卡塔利娜岛的度假回来。完全没有迹象表明他有自杀倾向。没有遗书，也没有行为异常。他刚刚和一位家人通了电话。然后他被发现死亡，多个房间有血迹。所以那是不可能的。看起来很明显他是被谋杀的。你和当局谈过这件事吗？

No, he was definitely murdered, I think. Um, there were signs of a struggle, of course. The surveillance camera, the wires had been cut. He had just ordered take out food, come back from a vacation with his friends on Catalina Island. No indication at all that he was suicidal. No note and no behavior. He had just spoken to a family member on the phone. And then he's found dead with blood in multiple rooms. So that's impossible. Seems really obvious he was murdered. Have you talked to the authorities about it?

我没有和当局谈过这件事。

I have not talked to the authorities about it.

嗯，他母亲声称他是按照你的命令被谋杀的。你相信吗？

Um, and his mother claims he was murdered on your orders. Do you believe that?

我，我是，嗯，我在问。

I, I'm, well, I'm asking.

我的意思是，你刚刚说了，所以你相信吗？

I mean, you just said it, so do you believe that?

我认为这值得调查。而且我不，我的意思是，如果一个人出来指控你的公司犯罪，我当然不知道这是不是真的。嗯，然后被发现被杀，还有挣扎的迹象。我认为不应该忽视它。我认为我们不应该说，嗯，他自杀了，而根本没有证据表明那个人抑郁。嗯，我认为，如果他是你的朋友，我会认为他会想和他妈妈说话或者……

I think that it is worth looking into. And I don't, I mean, if a guy comes out and accuses your company of committing crimes, I have no idea if that's true or not, of course. Um, and then is found killed and there are signs of a struggle. I don't think it's worth dismissing it. I don't think we should say, well, he killed himself when there's no evidence that the guy was depressed at all. Um, I think and if he was your friend, I would think he would want to speak to his mom or...

我确实提出了，她不想。

I did offer, she didn't want to.

那么，你是否觉得，你知道，当人们看到这个，他们会想，你知道，这有可能发生。你是否觉得这反映了他们对这里正在发生的事情的担忧？就像人们害怕这就像……

So, do you feel that, you know, when people look at that and they're like, you know, it's possible that happened. Do you feel that that reflects the worries they have about what's happening here? Like people are afraid that this is like...

我没有做过太多被指控……的采访。

I haven't done too many interviews where I've been accused of like...

哦，我完全没有指责你。我只是说他母亲这么说。我认为对证据的公正解读根本不表明是自杀。我完全看不出来。而且我也不明白为什么当局在有挣扎迹象和两个房间有血迹的情况下，对于自杀，这怎么可能发生？我不明白当局怎么能就这样把它当作自杀处理。我觉得很奇怪。

Oh, I'm not accusing you at all. I'm just saying his mother says that. I don't think a fair read of the evidence suggests suicide at all. I just don't see that at all. And I also don't understand why the authorities when there's signs of a struggle and blood in two rooms on a suicide, like how does that actually happen? I don't understand how the authorities could just kind of dismiss that as a suicide. I think it's weird.

你明白这听起来像是指控吗？

You understand how this sounds like an accusation?

当然。而且我，我的意思是我当然，让我再次明确，我没有指责你做任何错事，但我认为值得查明发生了什么。我不明白为什么旧金山市拒绝调查，只是把它称为自杀。我的意思是，我认为他们调查了几次，据我所知不止一次。我看到了，而且我完全可以说，当我第一次听说这件事时，我觉得非常可疑。

Of course. And I, I mean I certainly, let me just be clear once again, not accusing you of any wrongdoing, but I think it's worth finding out what happened. And I don't understand why the city of San Francisco has refused to investigate it beyond just calling it a suicide. I mean, I think they looked into it a couple of times, more than once as I understand it. I saw the, and I will totally say when I first heard about this, it sounded very suspicious to me.

是的。

Yes.

嗯，而且我知道你曾参与……

Um, and I know you had been involved in...

是母亲联系了这个案子吗？

Was mother reached out to the case?

而我，你知道，我对此一无所知。这不是我的领域。

And I, you know, I don't know anything about it. It's not my world.

她只是冷不防地联系了。

She just reached out cold.

她冷不防地联系了。

She reached out cold.

哇。

Wow.

然后呃，我和她谈了很长时间，这把我吓坏了。这孩子显然是被某人杀害的。那是我客观的结论，我与此无关。

And uh, and I spoke to her at great length and it scared the crap out of me. The kid was clearly killed by somebody. That was my conclusion objectively with no skin in the game.

而你是在读了最新报告之后。

And you after reading the latest report.

是的。

Yes.

就像你看，我立刻给加州的一位国会议员 Ro Khanna 打了电话，说这太疯狂了。你必须调查这件事。但什么也没发生。我就想这是怎么回事？再说一次，我认为这，我觉得奇怪和难过辩论这个，不得不看到完全疯狂，而你有点指责我，但是嗯……

Like look, and I immediately called a member of Congress from California, Ro Khanna, and said this is crazy. You got to look into this. And nothing ever happened. And I'm like what is that? Again, I think this is, I feel strange and sad debating this and having to see totally crazy and you are a little bit accusing me, but um the...

这是一个很棒的人，还有一个明显在挣扎的家庭。

This was like a wonderful person and a family that is clearly struggling.

是的。

Yes.

而且我认为你完全可以理解，你只是想弄清事情的真相，我尊重这一点。嗯，但我认为他的记忆和他的家人应该得到一定程度的尊重和哀悼，而我在这里并没有完全感受到。

And I think you can totally take the point that you're just trying to get to the truth of what happened and I respect that. Um, but I think his memory and his family deserve to be treated with a level of respect and grief that I don't quite feel here.

我是应他家人的要求问的。

I'm asking at the behest of his family.

嗯，所以我绝对是在向他们表示尊重。嗯，而且我完全没有指责你与此有任何牵连。我说的是，证据并不表明是自杀，而你所在城市的当局对此轻描淡写，忽视任何理性的人都会认为构成谋杀的证据，我认为这非常奇怪，它动摇了人们对我们的系统应对事实能力的信心。所以，我本来要说的是，在第一组信息出来后，我真的觉得，老兄，这看起来不像自杀。我很困惑。

Um, so I'm definitely showing them respect. Um, and I'm not accusing you of any involvement in this at all. What I am saying is that the evidence does not suggest suicide and for the authorities in your city to glide past that and ignore the evidence that any reasonable person would say adds up to a murder I think is very weird and it shakes the faith that one has in our system's ability to respond to the facts. So, what I was going to say is after the first set of information that came out, I was really like, man, this doesn't look like a suicide. I'm confused.

好吧，所以我没有过度，我没有发疯。

Okay, so I'm not reaching, I'm not being crazy here.

嗯，但后来第二份报告出来，细节更多了，我就想，好吧，是什么改变了你的想法？

Well, but then after the second thing came out and the more detail, I was like, okay, what changed your mind?

第二份报告关于子弹进入他的方式，以及那个追踪了子弹在房间内可能路径的人。我想你也看过了。

The second report on the way the bullet entered him and the sort of person who had followed the likely path of things through the room. I assume you looked at this too.

是的，我看过了。

Yes, I did.

那为什么没有改变你的想法呢？

And what about that didn't change your mind?

这对我来说根本说不通。为什么监控摄像头的线会被剪断？他开枪自杀后怎么会出现在两个房间流血？房间里为什么有一顶不属于他的假发？有没有过这样的自杀案例：完全没有自杀迹象，还刚点了外卖？我是说，谁会先点 DoorDash 然后开枪自杀？我当过警察记者，报道过很多犯罪，从没听说过这种事。所以，不，我反而更困惑了。

It just didn't make any sense to me. Why would the security camera wires be cut? And how did he wind up bleeding in two rooms after shooting himself? And why was there a wig in the room that wasn't his? And has there ever been a suicide where there was no indication at all that the person was suicidal who just ordered takeout food? I mean, who orders DoorDash and then shoots himself? I've covered a lot of crimes as a police reporter. I've never heard of anything like that. So, no, I was even more confused.

我觉得这有点令人痛苦，只是没有达到我希望对有这样精神状态的人所表现出的尊重程度……

This is where it gets into, I think, a little bit painful, just not the level of respect I'd hope to show to someone with this kind of mental...

我理解，我完全理解。

I get it, I totally get it.

很多人自杀时不写遗书，这确实会发生。人们肯定会在自杀前点自己喜欢的食物，这……

People do suicide without notes a lot, like that happens. People definitely order food they like before they commit suicide, like this is...

这是一场难以置信的悲剧，他家人也这么认为，他们觉得这是谋杀，所以我才会问。如果我是他的家人，我肯定想要答案，而且我肯定对任何解释都不会满意……我是说，没有什么能让我感到安慰，你知道吗？

This is an incredible tragedy and that's his family's view and they think it was a murder and that's why I'm asking. If I were his family, I am sure I would want answers and I'm sure I would not be satisfied with really any... I mean, there's nothing that would comfort me in that, you know?

对。

Right.

所以我理解。我也很在乎对他的尊重，对吧？

Like so I get it. I also care a lot about respect to him, right?

我必须问一下，你版本的故事是马斯克攻击了你等等。从你的角度看，这场争端的核心是什么？听着，我知道他是你的朋友，我也知道你会站在哪一边……

I have to ask your version of Elon Musk has attacked you and all this. What is the core of that dispute from your perspective? Look, I know he's a friend of yours and I know what side you'll...

我其实对此没有立场，因为我了解得不够，无法理解。

I actually don't have a position on this because I don't understand it well enough to understand.

他帮助我们创办了 OpenAI。我对此非常感激。

He helped us start OpenAI. I'm very grateful for that.

很长一段时间里，我真的把他视为一位了不起的英雄，人类的一颗明珠。我现在有不同的感受了。

I really for a long time looked up to him as just an incredible hero and, you know, great jewel of humanity. I have different feelings now.

你现在是什么感受？

What are your feelings now?

不再是人类明珠了。

No longer a jewel of humanity.

他身上有些地方很了不起，我也感激他做的很多事。但他也有很多我不欣赏的特质。总之，他帮助我们创办了 OpenAI，后来他认定我们不会成功，他不想……他告诉我们成功概率是 0%，他要去做自己的竞争项目，然后我们做得还不错。我觉得他因此不高兴是可以理解的，换作是我也会难受。从那以后，他就一直试图……他搞了一个竞争性的克隆项目，试图拖慢我们，起诉我们，做这做那。这大概就是我的版本。你有一个不同的版本。

There are things about him that are incredible and I'm grateful for a lot of things he's done. There's a lot of things about him that I think are traits I don't admire. Anyway, he helped us start OpenAI and he later decided that we weren't on a trajectory to be successful and he didn't want to, you know, he kind of told us we had a 0% chance of success and he was going to go do his competitive thing and then we did okay. And I think he got understandably upset, like I'd feel bad in that situation. And since then has just sort of been trying to... he had run as a competitive kind of clone and has been trying to sort of slow us down and sue us and do this and that. That's kind of my version of it. You have a different one.

你不再和他说话了吗？

You don't talk to him anymore?

很少。

Very little.

如果 AI 变得更聪明——我认为它可能已经比任何人都聪明了——如果它变得更明智，如果我们同意它能做出比人更好的决策，那么从定义上讲，它就会取代人类在世界中心的位置，对吧？

If AI becomes smarter, I think it already probably is smarter than any person. And if it becomes wiser, if we can agree that it reaches better decisions than people, then it by definition kind of displaces people at the center of the world, right?

我完全不觉得会那样。我觉得它会像一台非常聪明的电脑，可以给我们建议，我们听它的。有时我们忽略它。有时它不会……我不觉得它会让人感觉有自主性。我不认为它会削弱我们的自主感。人们已经在使用 ChatGPT，很多人会说它在几乎所有事情上都比我聪明得多。但他们仍然在做决定。他们仍然决定问什么、听什么、不听什么。我认为这就是技术的形态。

I don't think it'll feel like that at all. I think it'll feel like a really smart computer that may advise us and we listen to it. Sometimes we ignore it. Sometimes it won't... I don't think it'll feel like agency. I don't think it'll diminish our sense of agency. People are already using ChatGPT in a way where many of them would say it's much smarter than me at almost everything. But they're still making the decisions. They're still deciding what to ask, what to listen to, what not. And I think this is sort of just the shape of technology.

谁会因为这项技术失业？我先说一个显而易见但重要的声明：没人能预测未来。如果我试图精确回答，我会说很多蠢话，但我会试着挑一个我有信心的领域，以及一些我信心不足的领域。我有信心的是，很多目前通过电话或电脑进行的客户支持岗位，那些人会失业，AI 能做得更好。当然，可能还有其他类型的客户支持，你确实需要知道对方是真人。我有信心不会受太大影响的工作是护士。我认为人们在那种时刻真的需要与人的深层情感连接。无论 AI 或机器人的建议有多好，你都会需要那种连接。我对未来感到不太确定的工作是计算机程序员。今天当程序员的意义与两年前大不相同。你可以用这些 AI 工具大幅提高生产力，但人还在那里，他们能生成比以往更多的代码，赚更多的钱。事实证明，世界对软件的需求远超之前创造能力，存在巨大的需求积压。但如果再快进 5 到 10 年，会是什么样子？工作更多还是更少？这个我不确定。但会有大规模岗位替代，也许那些人会找到新的、有趣的、赚钱的事情做。

Who loses their jobs because of this technology? I'll caveat this with the obvious but important statement that no one can predict the future. And if I try to answer that precisely, I will say a lot of dumb things, but I'll try to pick an area that I'm confident about and then areas that I'm much less confident about. I'm confident that a lot of current customer support that happens over a phone or computer, those people will lose their jobs and that'll be better done by an AI. Now, there may be other kinds of customer support where you really want to know it's the right person. A job that I'm confident will not be that impacted is like nurses. I think people really want the deep human connection with a person in that time. And no matter how good the advice of the AI is or the robot or whatever, you'll really want that. A job that I feel way less certain about what the future looks like for is computer programmers. What it means to be a computer programmer today is very different than what it meant 2 years ago. You're able to use these AI tools to just be hugely more productive, but it's still a person there and they're able to generate way more code, make way more money than ever before. And it turns out that the world wanted so much more software than the world previously had capacity to create that there's just incredible demand overhang. But if we fast forward another 5 or 10 years, what does that look like? Is it more jobs or less? That one I'm uncertain on. But there's going to be massive displacement and maybe those people will find something new and interesting and lucrative to do.

但你认为这种替代有多大？最近有人告诉我，历史平均水平是大约 50%的工作会发生显著变化。也许不会完全消失，但平均每 75 年显著变化一次。这就是事物的半衰期。我的有争议的观点是，这将是一个间断平衡的时刻，很多变化会在短时间内发生。但如果我们放眼长远，它不会与历史速率有显著差异。我们会在短时间内经历很多变化，然后总体工作更替会比我们想象的要少。仍然会有工作存在。会有一些全新的类别，比如我的工作——经营一家科技公司。这在 200 年前很难想象。但还有很多其他工作与 200 年前存在的工作方向相似。而 200 年前常见的工作现在没有了。再说一次，我不知道这是否正确，但为了论证我会用这个数字。

But how big is that displacement do you think? Someone told me recently that the historical average is about 50% of jobs significantly change. Maybe they don't totally go away but significantly change every 75 years on average. That's the kind of half-life of the stuff. And my controversial take would be that this is going to be like a punctuated equilibrium moment where a lot of that will happen in a short period of time. But if we zoom out, it's not going to be dramatically different than the historical rate. We'll have a lot in this short period of time and then it'll somehow be less total job turnover than we think. There will still be a job that is there. There will be some totally new categories like my job, you know, running a tech company. It would have been hard to think about 200 years ago. But there's a lot of other jobs that are directionally similar to jobs that did exist 200 years ago. And there's jobs that were common 200 years ago that now aren't. And if we again, I have no idea if this is true or not, but I'll use the number for the sake of argument.

如果我们假设每 75 年有 50%的岗位更替，那么我完全可以相信，75 年后一半的人在做全新的事情，另一半的人在做看起来有点像今天某些工作的事情。

If we assume it's 50% turnover every 75 years, then I could totally believe a world where 75 years from now half the people are doing something totally new and half the people are doing something that looks kind of like some jobs of today.

上一次工业革命时，有革命和世界大战。你认为这次我们也会看到吗？

Last time we had an industrial revolution there was revolution and world wars. Do you think we'll see that this time?

没人确切知道。我对这个答案没有信心，但我的直觉是，世界现在比工业革命时期富裕得多，我们实际上能比以前更快地吸收更多变化。很多事不只是关于工作的金钱，还有意义，有很多社群。

No one knows for sure. I'm not confident on this answer, but my instinct is the world is so much richer now than it was at the time of the industrial revolution that we can actually absorb more change faster than we could before. There's a lot that's not about money of job. There's meaning, there's a lot of community.

我认为不幸的是，社会在这方面已经处于相当糟糕的境地。我不确定还能糟到什么程度，但肯定可以更糟。我对人们快速适应巨大变化的能力感到惊喜。比如新冠疫情对我来说是一个有趣的例子，世界一下子停摆，一周之间变得截然不同，我非常担心社会如何适应那个世界，显然并不完美，但总的来说，我觉得这是支持社会韧性的一个点，人们很快找到了新的生活方式。我不认为人工智能会那么突然。

I think we're already unfortunately in society in a pretty bad place there. I'm not sure how much worse it can get. I'm sure it can. I have been pleasantly surprised on the ability of people to pretty quickly adapt to big changes. Like COVID was an interesting example to me of this where the world kind of stopped all at once and the world was very different from one week to the next and I was very worried about how society was going to be able to adapt to that world and it obviously didn't go perfectly but on the whole I was like all right this is one point in favor of societal resilience and people find new ways to live their lives very quickly. I don't think AI will be that nearly that abrupt.

那么坏处是什么？我当然能看到好处：效率、医疗诊断会更准确、律师更少。非常感谢这一点。但你担心的坏处是什么？

So what will be the downside? I can see the upsides for sure: efficiency, medical diagnosis seems like it's going to be much more accurate, fewer lawyers. Thank you very much for that. But what are the downsides that you worry about?

我想这就是我的思维方式。我最担心的总是未知的未知。如果是我们能确信并思考的坏处，我们之前谈过一个，那就是这些模型在生物方面变得非常擅长，它们可能帮助我们设计生物武器，制造另一个新冠式的大流行病。我担心这个，但正因为我们担心，我认为我们和业内许多其他人都在努力思考如何缓解。未知的未知：很多人同时与同一个模型对话会产生社会规模效应。这是一个愚蠢的例子，但最近让我印象深刻。像我们这样的语言模型有某种风格，它们以某种节奏说话，用词有点不寻常，可能过度使用破折号。我最近注意到真实的人也开始这样做了。这对我来说是一个例子：当足够多的人与同一个语言模型对话时，它确实会导致社会规模的行为变化。

I think this is just how I'm wired. I always worry the most about the unknown unknowns. If it's a downside that we can really be confident about and think about, we talked about one earlier, which is these models are getting very good at bio and they could help us design biological weapons. Engineer another COVID-style pandemic. I worry about that, but because we worry about it, I think we and many other people in the industry are thinking hard about how to mitigate that. The unknown unknowns: there's a societal scale effect from a lot of people talking to the same model at the same time. This is a silly example but it struck me recently. LLMs like ours have a certain style, they talk in a certain rhythm, have a little bit unusual diction, maybe overuse em dashes. I noticed recently that real people have picked that up. It was an example for me of: man, you have enough people talking to the same language model and it actually does cause a change in societal scale behavior.

是的。

Yes.

我有没有想过 ChatGPT 会让人们在现实生活中使用更多的破折号？当然没有。这不是什么大事，但这是一个例子，说明可能存在这些未知的未知。这是一个美丽新世界。

Did I think that ChatGPT was going to make people use way more em dashes in real life? Certainly not. It's not a big deal, but it's an example of where there can be these unknown unknowns. This is a brave new world.

所以，你说得对且简洁，技术当然会改变人类行为，改变我们对世界和彼此的看法。其中很多无法预测。考虑到我们知道这一点，为什么技术的内部道德框架不应该完全透明？我们偏好这个而不是那个。我的意思是，这显然是一种宗教。我想你不会同意这么称呼它。但对我来说很明显是一种宗教。这不是攻击。

So, you're saying, I think correctly and succinctly, that technology changes human behavior, of course, and changes our assumptions about the world and each other. And a lot of this you can't predict. Considering that we know that, why shouldn't the internal moral framework of the technology be totally transparent? We prefer this to that. I mean, this is obviously a religion. I don't think you'll agree to call it that. It's very clearly a religion to me. That's not an attack.

我其实很想听听你这么说是什么意思。我不认为这是攻击。

I actually would love to hear what you mean by that. I don't take that as an attack.

嗯，这是一种我们假设比人更强大的东西，我们向它寻求指导。你已经看到这一点了。正确的决定是什么？我问谁这个问题？我最亲密的朋友、我的妻子和上帝。而这项技术能提供比任何人都更确定的答案。所以它是一种宗教。宗教的美妙之处在于它们有透明的教义。我知道宗教代表什么。它支持什么，反对什么。但在这种情况下，我追问，我不是在攻击你。我不是在攻击你，而是想触及核心。宗教的美妙之处在于它承认自己是宗教，并告诉你它代表什么。这项技术令人不安的地方，不仅仅是你的公司，还有其他的，是我不知道它代表什么，但它确实代表某种东西。除非它承认并告诉我们它代表什么，否则它会以一种隐蔽的方式引导我们走向我们甚至不知道正在接近的结论。你明白我的意思吗？所以为什么不直接公开说 ChatGPT 支持这个，或者我们支持绝症患者自杀但不支持孩子自杀之类的。为什么不直接告诉我们？

Well, it's something that we assume is more powerful than people, and to which we look for guidance. I mean you're already seeing that on display. What's the right decision? I asked that question of whom? My closest friends, my wife and God. And this is a technology that provides a more certain answer than any person can provide. So it's a religion. And the beauty of religions is they have a catechism that is transparent. I know what the religion stands for. Here's what it's for. Here's what it's against. But in this case, I pressed and I wasn't attacking you sincerely. I was not attacking you, but I was trying to get to the heart of it. The beauty of a religion is it admits it's a religion and it tells you what it stands for. The unsettling part of this technology, not just your company, but others, is that I don't know what it stands for, but it does stand for something. And unless it admits that and tells us what it stands for, then it guides us in a kind of stealthy way toward a conclusion we might not even know we're reaching. Do you see what I'm saying? So why not just throw it open and say ChatGPT is for this or we're for suicide for the terminally ill but not for kids or whatever. Why not just tell us?

我们写这个很长的模型规范，并且不断扩展的原因，是让你能看到我们打算让模型如何行为。在我们有这个之前，人们会合理地说，我不知道模型到底想做什么，我不知道这是 bug 还是预期行为。告诉我这个长长的文档，说明你什么时候会这样做，什么时候会展示这个，什么时候会说不会那样做。我们试图全部写出来的原因，是我认为人们确实需要知道。

The reason we write this long model spec and the reason we keep expanding over time is so that you can see how we intend for the model to behave. What used to happen before we had this is people would fairly say I don't know what the model's even trying to do and I don't know if this is a bug or the intended behavior. Tell me this long document of how you're going to say do this and when you're going to show me this and when you're going to say you won't do that. The reason we try to write this all out is I think people do need to know.

那么有没有一个地方可以让你找到关于你们公司偏好的明确答案，这些偏好正以不完全直接的方式传播到全球？比如在哪里可以找到公司代表什么、偏好什么？

So is there a place you can go to find out a hard answer to what your preferences as a company are, preferences that are being transmitted in a not entirely straightforward way to the globe? Like where can you find out what the company stands for, what it prefers?

我们的模型规范就是答案。现在我认为随着人们在不同的国家使用它，我们将不得不让它越来越详细。有不同的法律等等。它不会对每个地方的每个用户都以相同的方式工作。我预计那份文档会变得非常长和非常复杂，但这就是我们拥有它的原因。

Our model spec is the answer to that. Now I think we will have to make it increasingly more detailed over time as people use this in different countries. There's different laws, whatever else. It will not work the same way for every user everywhere. I expect that document to get very long and very complicated, but that's why we have it.

让我问你最后一个问题，也许你能缓解这种恐惧：技术的力量会让任何人难以或无法辨别现实与幻想之间的区别。这是一个著名的担忧，但由于它如此擅长模仿人们及其言语和图像，它将需要某种方式来验证你确实是你声称的那个人，而这必然需要生物识别技术，这必然消除世界上每个人的隐私。

Let me ask you one last question and maybe you can allay this fear that the power of the technology will make it difficult or impossible for anyone to discern the difference between reality and fantasy. This is a famous concern, but because it is so skilled at mimicking people and their speech and their images, it will require some way to verify that you are who you say you are and that will by definition require biometrics which will by definition eliminate privacy for every person in the world.

我认为我们不需要也不应该要求使用生物识别技术来使用这项技术。我觉得你应该能够从任何电脑上使用 ChatGPT。

I don't think we need to or should require biometrics to use the technology. I think you should just be able to use ChatGPT from any computer.

是的，我强烈同意。但在某个时候，当模仿人的图像或声音变得太容易清空你的银行账户时，你该怎么办？

Yeah, I strongly agree. But at a certain point, when images or sounds that mimic a person become too easy to empty your checking account with, what do you do about that?

我有几点想法。第一，我认为我们正在迅速走向一个世界，人们明白如果你接到一个听起来像你孩子或父母的电话，或者看到一个看起来很真实的图像，你必须有一些方法来验证你没有受骗。这不再是理论上的担忧。你听到所有这些报道。

A few thoughts there. One, I think we are rapidly heading to a world where people understand that if you get a phone call from someone that sounds like your kid or your parent, or if you see an image that looks real, you have to have some way to verify you're not being scammed. This is no longer a theoretical concern. You hear all these reports.

人们很聪明，社会有韧性。我认为人们很快意识到这是坏人正在利用的新事物，人们明白必须用不同的方式验证。我猜想，除了家庭成员在危机情况下使用暗语之外，我们还会看到，比如当一国总统必须发布紧急信息时，他们会用加密签名或其他方式保证其真实性。这样就不会有生成的视频说特朗普'刚刚做了这个或那个'。人们正在快速学习，这是坏人利用技术做的新事情，他们必须应对。我认为这就是解决方案的大部分：人们会默认不信任看起来令人信服的媒体，我们会建立新的机制来验证通信的真实性。但这些必须是生物识别。

People are smart, society is resilient. I think people are quickly understanding that this is now a thing that bad actors are using, and people are understanding that you have to verify in different ways. I suspect that in addition to things like family members having code words they use in crisis situations, we'll see things like when a president of a country has to issue an urgent message, they cryptographically sign it or otherwise guarantee its authenticity. So you don't have generated videos of Trump saying, 'I've just done this or that.' People are learning quickly that this is a new thing that bad guys are doing with the technology, and they have to contend with it. I think that is most of the solution: people will by default not trust convincing-looking media, and we will build new mechanisms to verify authenticity of communication. But those will have to be biometric.

不，完全不是。我的意思是，如果美国总统有那个，我理解。但对普通人来说，在平常的一天，你并不是在等总统宣布战争；你是在做电子商务。你怎么能做到呢？

No, not at all. I mean, if the president of the US has that, I understand. But for the average person on an average day, you're not waiting for the president to announce a war; you're trying to do e-commerce. How could you do that?

嗯，我认为和你的家人，你会有一个定期更换的暗语。如果你们互相通信，接到电话时你会问暗语是什么。这和生物识别非常不同。

Well, I think with your family, you'll have a code word that you change periodically. If you're communicating with each other and you get a call, you ask what the code word is. That's very different from a biometric.

所以你不认为生物识别很快就会成为全社会强制性的？我的意思是，现在乘坐商业航班，生物识别已经是流程的一部分了。

So you don't envision biometrics becoming society-wide mandatory very soon? I mean, to board a commercial flight, biometrics are part of the process now.

我真的希望它不会成为强制性的。我认为有一些保护隐私的生物识别版本，我更喜欢，而不是收集大量个人数字信息，但我不认为生物识别应该是强制性的。例如，我认为你不应该必须提供生物识别才能上飞机。

I hope it really doesn't become mandatory. I think there are versions of privacy-preserving biometrics that I like much more than collecting a lot of personal digital information on one, but I don't think biometrics should be mandatory. I don't think you should have to provide biometrics to get on an airplane, for example.

那银行业呢？

What about for banking?

我认为银行业也不应该强制。我可能更喜欢用指纹扫描来访问我的比特币钱包，而不是把所有的信息都交给银行。但这应该是我自己的决定。

I don't think you should have to for banking. I might prefer a fingerprint scan to access my Bitcoin wallet rather than giving all my information to a bank. But that should be a decision for me.

我很感激。谢谢你，Sam。

I appreciate it. Thank you, Sam.

谢谢。

Thank you.

所以，事实证明 YouTube 正在压制这个节目。从某种程度上说，这并不奇怪。他们就是这么做的。但从另一个层面来说，这令人震惊。考虑到现在世界上发生的一切，我们的经济和政治正在发生的所有变化，以及我们即将面临的战争，谷歌决定你应该获得更少的信息而不是更多。这是完全错误的。这是不道德的。你能做什么？嗯，我们可以抱怨。那是浪费时间。我们不是谷歌的负责人。或者我们可以找到绕过它的方法。一种让你真正获得真实信息而不是故意欺骗的方法。我们认为，在 YouTube 上做到这一点的方法是订阅我们的频道。订阅。点击小铃铛图标，以便在我们上传和分享这个视频时收到通知。这样，你就有更大的机会听到真实的新闻和信息。所以我们希望你能这样做。

So, it turns out that YouTube is suppressing this show. On one level, that's not surprising. That's what they do. But on another level, it's shocking. With everything that's going on in the world right now, all the change taking place in our economy and our politics, with the wars we're on the cusp of fighting right now, Google has decided you should have less information rather than more. And that is totally wrong. It's immoral. What can you do about it? Well, we could whine about it. That's a waste of time. We're not in charge of Google. Or we could find a way around it. A way that you could actually get information that is true, not intentionally deceptive. The way to do that on YouTube, we think, is to subscribe to our channel. Subscribe. Hit the little bell icon to be notified when we upload and share this video. That way, you'll have a much higher chance of hearing actual news and information. So, we hope that you'll do that.