---
title: "OpenAI 总裁 Greg Brockman：网络安全的「物理规律」正在改变，AI 将重塑药物研发"
title_original: "The Changing Physics of Cybersecurity and Drug Development with OpenAI President Greg Brockman"
guest: "Greg Brockman"
guest_role: "OpenAI 联合创始人、总裁"
host: ["Sarah Guo", "Elad Gil"]
show: "No Priors"
date: 2026-10-15
duration: "39:22"
language: en
source: "https://www.youtube.com/watch?v=AwuTTOJGEi0"
source_platform: "YouTube"
page: "https://ai-daily-cn.pages.dev/watch/np-gdb/"
tags: ["研几", "研几AI", "访谈"]
---

# OpenAI 总裁 Greg Brockman：网络安全的「物理规律」正在改变，AI 将重塑药物研发

嘉宾 **Greg Brockman** · OpenAI 联合创始人、总裁 · 主持 Sarah Guo、Elad Gil · No Priors · 2026-10-15 · 时长 39:22 · [YouTube 原片](https://www.youtube.com/watch?v=AwuTTOJGEi0) · [研几 · AI 访谈页](https://ai-daily-cn.pages.dev/watch/np-gdb/)

## 核心总结

OpenAI 总裁 Greg Brockman 在 No Priors 访谈中称，AI 已能解决约 400 个重要开放数学问题，证明其可生成真正的新知识，并主张安全支出应远超软件支出。

- **数学突破证明 AI 能产出新知识**：OpenAI 公布了约 400 个重要开放数学问题，如今已被解决；Brockman 称这标志着 AI 能带来新知识已“无法否认”，但强调重点不只是答案，而是人类借助这些工具能取得哪些此前无法想象的成就，以及数学界正逐一研究这些成果中的洞见。

- **药物发现是首要应用领域，但受数据与监管瓶颈制约**：Brockman 认为药物发现是最清晰、最应被彻底变革的领域，理由是神经网络没有理由在生物学上不如自然语言；但他指出生物学数据瓶颈严重，监管申报（如撰写临床试验材料）约需一年，湿实验室与模拟创新将变得极其重要，并称短期内影响“非常可能”。

- **Dots 是始终在线的主动式智能体**：Brockman 描述 Dots 为主动式智能体，使用 Astra 级别模型、拥有自己的云计算机，会追踪他未完成的工作、指出日历冲突、整合收件箱与 Slack 信息，每天早晨给出待办清单，且经常注意到他从未要求它关注的事情。

- **无 UI 与生成式 UI 并行**：Brockman 称“最好的 UI 就是没有 UI”，AI 可通过 Slack、iMessage 等界面交互；同时 OpenAI 已引入智能 UI，AI 可即时生成生成式 UI，例如为组装自行车这类一次性需求实时生成定制应用。

- **网络安全处于“防御者窗口”，安全支出应远超软件支出**：Brockman 称前沿 AI 能力尚可帮助防御者，但广泛扩散的网络能力正在上升，韩国已出现由不同模型驱动的备受关注的银行黑客攻击；OpenAI 通过 Daybreak 项目向防御者提供模型访问，做出十亿美元承诺帮助一线防御者，并曾抽调约 25% 生产工程师暂停项目专做安全、用巅峰模型探查基础设施直至饱和。他主张安全支出应“远远超过软件支出”，因为软件正变得极其廉价，可靠、安全、可信赖的软件才是价值所在。

- **人类必须坐在驾驶座上**：Brockman 重申 OpenAI 使命是确保 AGI 惠及全人类，人类应牢牢掌握方向盘、掌控未来，技术应放大人类能动性；他以 Steve Jobs 在 iMac 上装把手作比喻，称人类未来可能更像 CEO 或董事会成员，但根本问题始终是“这是你想要的吗、这让你的生活变得更好吗”。

## 金句

- 「我们公布了近 400 个非常重要、如今已被解决的开放问题。」 —— Greg Brockman · [5:32](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=332s)
  > we published about almost 400 very important open problems that are now resolved.
- 「我认为最好的 UI 就是没有 UI。我认为 AI 正朝这个方向走。」 —— Greg Brockman · [23:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1419s)
  > I think that the best UI is no UI. And I think in AI, that is where we're going.
- 「我们让大约 25% 的生产工程师停下手中的所有项目，说：你们现在只做安全工作。」 —— Greg Brockman · [31:03](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1863s)
  > we took about 25% of our production engineers and said, all of your projects are on hold. You are just securing now.
- 「我感觉网络支出、安全支出实际上应该远远超过软件支出。」 —— Greg Brockman · [34:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2057s)
  > I have a feeling that cyber spend, security spend should actually vastly outstrip software spend.
- 「有个故事讲 Steve Jobs 为什么给 iMac 装上把手，对吧？这样人们就能感觉到自己可以控制这台机器：你随时都能把它拎起来扔出窗外。我认为 AI 也需要同样的精神。」 —— Greg Brockman · [37:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2259s)
  > there's the story of why Steve Jobs put the handle on the iMac, right? So that that way that people could feel that they could control the machine, right? That you can always pick it up and throw it out the window. And I think that the same spirit needs to apply to AI.

## 精华片段

0. [5:31–7:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=331s) AI 已解开近 400 个重要数学难题
1. [23:36–25:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1416s) 最好的 UI 就是没有 UI
2. [30:46–32:22](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1846s) 抽调 25% 工程师专做安全
3. [34:17–35:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2057s) 安全投入应远超软件本身
4. [37:35–38:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2255s) 像 iMac 的把手：人要能掌控 AI

## 文字稿（中英对照）

### [0:00](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=0s) 开场
*Cold open*

[0:00](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=0s) **Greg Brockman** 确保 AGI 惠及全人类。我们认为人类应当牢牢掌握方向盘，应当掌控未来，而这项技术应当放大人类的能动性。当我审视我们将要身处的世界时，软件在某种程度上正成为准入门槛，它变得如此廉价。

> To ensure AGI benefits all of humanity. We think that humans should be firmly in the driver's seat, should have control over the future, and this technology should amplify human agency. When I look at the world we're going to be in, where software in some ways is becoming like the barrier to entry, it's becoming so cheap.

[0:14](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=14s) **Greg Brockman** 你想要的不仅仅是软件，对吧？你想要的是可靠、安全、值得信赖的软件。因此，网络安全方面的支出实际上应当远远超过软件支出。我认为人们还没有真正内化网络安全相关的物理规律将会发生多大的变化。

> You don't just want software, right? You want software that is reliable, that is secure, that you can believe in. And so cybersecurity spend should actually vastly outstrip software spend. I think that people have not yet really internalized how much the physics around cybersecurity are going to change.

[0:29](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=29s) **Greg Brockman** 我认为我们有机会最终迎来一个安全得多的世界。

> I think there's an opportunity to end up with a much more secure world.

### [0:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=39s) Greg Brockman 介绍
*Greg Brockman Introduction*

[0:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=39s) **Sarah Guo** 嗨，听众朋友们，欢迎回到《No Priors》。今天，Elad 和我与 OpenAI 联合创始人兼总裁 Greg Brockman 进行了对话。我们讨论了 OpenAI 最新的数学成果、Astra，以及为什么 Greg 认为 AI 如今能够生成真正全新的知识。我们聊到了药物发现、Dots，以及他关于安全支出为何应当远远超过软件支出的论点——这是一个相当大胆的主张。

> Hi listeners, welcome back to "No Priors." Today, Elad and I talk with Greg Brockman, the co-founder and president of OpenAI. We discuss OpenAI's latest math results, Astra, and why Greg believes AI can now generate genuinely new knowledge. We get into drug discovery, Dots, and his case for why security spending should actually vastly outstrip software spending, which is quite a claim.

[1:02](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=62s) **Sarah Guo** 以下是我们与 Greg Brockman 的对话。Greg，非常感谢你接受这次访谈。

> Here's our conversation with Greg Brockman. Greg, thanks so much for doing this.

[1:06](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=66s) **Elad Gil** 感谢你加入我们。

> Thanks for joining us.

[1:07](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=67s) **Greg Brockman** 感谢你们的邀请。

> Thank you for having me.

### [1:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=68s) AI 在数学上的突破
*AI and Advancements in Mathematics*

[1:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=68s) **Elad Gil** 在 AI 领域，每一周都感觉像过了一年。而本周 OpenAI 发生了很多事情。你们向超过十亿人推出了 GPT-6，这相当了不起。在数学方面，你们也有一些非常令人兴奋的消息。

> So every week in AI feels like a year. And a lot's happened this week for OpenAI. You rolled out GPT-6 to over a billion people, which is pretty amazing. There's some really exciting news that you have on the math side.

[1:22](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=82s) **Elad Gil** 毫无疑问，内部还有各种各样的事情在发生。无论是 OpenAI 内部还是更广泛的 AI 世界，当下正在发生的事情中，最让你兴奋的是什么？

> There's all sorts of things, undoubtedly, that are happening internally. What are you most excited about that's happening right now, both at OpenAI and in the broader AI world?

[1:29](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=89s) **Greg Brockman** 你看，我认为在 AI 领域，我对 OpenAI 的运作方式有这样一个信念，就是那句循环往复的话：它永远不会变得更容易，你只会跑得更快。我认为实际情况正是如此，活动如此之多，每当你以为自己已经处于影响力的顶峰、事情就是这样了，你就会意识到其实还有另一个层次是可能的。

> Look, I think that in AI, like I have this belief about how OpenAI operates, that there's this cycling quote of it never gets easier, you just go faster. And I think that's very much what happens, is that there's just so much activity and that every time you think you're at kind of peak sort of impact and how things are happening, you realize that there's actually another level that's possible.

[1:50](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=110s) **Greg Brockman** 而且这不仅仅关乎某一家公司。不仅仅关乎某一项技术。它真正关乎的是我们作为一个世界、一个社会共同携手来采用这项技术，最大限度地发挥它的作用，改善每个人的生活。

> And it's not just about any one company. It's not just about any one technology. It's really about we as a world and society are all coming together to adopt this technology, get the most out of it, to improve everyone's lives.

[2:00](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=120s) **Greg Brockman** 所以我认为，对我来说，本周的数学成果代表着一个真正的里程碑，对吧？这是一个真正的时刻，

> And so I think to me, the math results this week represent a real milestone, right? It's a real moment where

[2:07](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=127s) **Greg Brockman** 如果你想一想，过去一千年里人类生活质量提升的根本驱动力是技术，对吧？是科学进步。而我认为我们现在有能力大幅加速科学发现。一直以来都有这样一个问题：AI 是否真的能够带来新知识？

> the if you think about like the fundamental driver of human quality of life improving over the past thousand years is technology, right? It's scientific advancement. And I think that we now have the ability to massively accelerate scientific discovery. And there's always been this question of is AI actually able to come with new knowledge?

[2:30](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=150s) **Greg Brockman** 而到了现在这个阶段，这已经无法否认了，对吧？答案绝对是肯定的，是可能的。当然，还有更多要做。我认为我们还没有提出全新的理论，但我们能够解决这些有百年历史的数学难题。

> And at this point, it's just impossible to deny, right? It is definitely yes, possible. Now, there's still more to go. I think we're not yet coming up with whole brand new theories, but we are able to resolve these 100-year-old math problems.

[2:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=164s) **Greg Brockman** 而对我来说，非常令人兴奋的一点是，这不仅仅是机器直接给你一个答案，对吧？真正重要的是，配备了这些工具的人类能够取得哪些此前无法想象的成就？

> And to me, the thing that's very exciting is it's not just about the sort of just the machine just sort of gives you an answer, right? It's really what can a human equipped with these tools achieve that was unthinkable before?

[2:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=178s) **Greg Brockman** 我认为这几乎就像是人类去做任何想做的事情的杠杆率正在大幅、大幅提升，对吧？现在，如果你有一个想解决的数学问题，你就能解决它。

> And I think it's almost like the leverage of humans to do whatever you want to do, it's going way, way up, right? It's now that you can, if you have a math problem you want to go solve, you can solve it.

[3:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=188s) **Greg Brockman** 但把同样的能力应用到生物学上会怎样？如果你是一个小企业主，想要把自己的生意经营好，又会怎样？那种针对特定领域调优的智能水平，将意味着这些几乎难以想象的、与你的目标和愿望相一致的结果，现在将触手可及，人人都能获得。

> But what about taking that same capability, applying it to biology? What about if you're a small business owner and you want to be able to operate your business well? Like that level of intelligence that's tuned for that domain, it's just going to mean that almost this sort of unthinkable results that are aligned with your goals, with your desires, is now going to be within the reach of everyone.

[3:31](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=211s) **Greg Brockman** 而这对我来说，实在太令人兴奋了。

> And that, to me, is so exciting.

[3:32](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=212s) **Elad Gil** 进步的速度也相当惊人，因为如果我看一下，过去模型在算术上都有问题，对吧？在早期模型里，我记得人们抱怨，哦，为什么它连基础数学都做不了？而现在你们已经在解决这些高深的定理了。

> The rate of progress has been pretty amazing too, because if I look at it, you used to have problems with arithmetic, right? In early models, I remember people complaining, oh, why can't it do basic math? And now you're solving these advanced theorems.

[3:43](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=223s) **Elad Gil** 随着这些东西不断进步，并开始真正影响数学、物理、材料等领域以及各种应用，如果这些模型变得如此聪明，是否需要在如何有效地向人们传达这一点上加以关注？

> As these things continue to advance and this starts to really impact areas of math, physics, materials, et cetera, and various applications, is there some need if these models get so smart to focus on how to communicate that effectively to people?

[3:57](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=237s) **Elad Gil** 或者，当你做出一个非常高级的定理时，你如何看待实际的沟通层面？因为如果你看数学的任何领域，只有极少数人才能真正评判一个人产出的定理。

> Or how do you think about the actual communication side of this as you do a very advanced theorem? Because if you look at any field of math, only a very small number of people can ever really adjudicate a theorem that's produced by a person.

[4:07](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=247s) **Greg Brockman** 是的。

> Yes.

[4:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=248s) **Elad Gil** 对吧？所以你就有了这种专业化分工。

> Right? And so you kind of have this specialization that happens.

[4:10](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=250s) **Greg Brockman** 嗯，顺便说一句，这一直是——我出身于数学背景。我曾以为自己会成为一名数学家。我深受 Galois 或 Gauss 这样的数学家的启发，对吧？

> Well, and this, by the way, is something that has always been like I came from a math background. Like I thought I was going to be a mathematician. And I was very inspired by mathematicians like Galois or Gauss, right?

[4:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=261s) **Greg Brockman** 这些人研究的是 100 年、400 年的时间尺度，对吧？我认为数学关乎理解宇宙的结构，这其中有着某种深刻的美。所以它不仅仅是——而且一直都有这些问题：数学是被发明的还是被发现的？

> These people who were working on like a 100, 400-year time horizon, right? And I think there's something deeply beautiful about mathematics being about sort of understanding the structure of the universe. And so it's not just about and there's always these questions about is math invented or discovered?

[4:40](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=280s) **Greg Brockman** 在我看来，数学的美和意义，在某种意义上并不是客观的。它存在于观者的眼中。它是能打动我们的东西。其中有部分是像物理定律一样的东西，关乎简单性、压缩理论、复杂性理论等等，这些也许更偏向物理学，但其中也有重要的属于人的一面。

> And that in my mind, it's something that the beauty and the meaning of mathematics, it's not something that is objective in some sense. It's in the eye of the beholder. It's something that appeals to us. And there is some amount that's like kind of physical laws and it's about simplicity and compression theory, complexity theory, and all these things like that are maybe are more physics, but there's something that is also importantly human about it.

[5:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=305s) **Greg Brockman** 所以对我来说，这不仅仅是——而且我认为很多关于数学的讨论不仅仅是说这里有一个证明，没人看得懂，太好了，我们解决了问题。真正重要的是从中得出了哪些洞见？

> And so to me, it's not just about and I think a lot of the conversation about mathematics is not just about here's a proof, no one understands it, great, we resolved the problem. It's really about what are the insights that are derived?

[5:16](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=316s) **Greg Brockman** 而我认为我们将会看到——而且我认为已经有令人鼓舞的生命迹象——这些 AI 所做的工作中蕴含着美妙的洞见。我认为我们现在看到的是，

> And the thing that I think we will see, but I think that there's promising signs of life on, is that there's beautiful insights within the work that these AIs have done. And I think what we're seeing right now is that

[5:32](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=332s) **Greg Brockman** 我们公布了大约近 400 个非常重要的开放问题，如今它们已被解决。数学界正在逐一研究所有这些成果，寻找洞见。如果你读一读人们发布的一些内容，这些特定问题的领域专家，他们说，这就像是——有一些几乎非常出人意料的运用思路的方式，或者令人惊叹地——等等，他们为什么在这里用了那个结果？

> we published about almost 400 very important open problems that are now resolved. And the mathematical community is going through all of that to look for insights. And if you read some of the things that people have posted, domain experts in these particular problems, and they say that it's like that there's these almost like sort of very surprising ways of applying ideas or incredibly like wait, why do they use that result over here?

[5:59](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=359s) **Greg Brockman** 而且，产出的并不是那些难以理解的定理或证明——顺便说一句，数学史上确实有难以理解的证明。四色定理就是一个完美的例子，就像是，好吧，我们只是让机器检查了一大堆情况。

> And that what's coming out is not these incomprehensible theorem or proofs, which by the way, there is a history of incomprehensible proofs in mathematics. Like four-color theorem is a perfect example of this that is it's like, OK, like we just had the machine check a bunch of cases.

[6:16](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=376s) **Greg Brockman** 那里面没有美。所以我认为我们进入了机器发现的另一个阶段。

> Like there's no beauty in that. And so I think we're into a different phase of machine discovery.

[6:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=381s) **Sarah Guo** 我有些乐观地认为，随着模型也获得更好的心智理论，也就是弄清楚我们人类知道什么、以及如何让我们跟随它们走上通往新理解之路的能力，也会被铺平，对吧？

> I have some optimism that like as models also get to better theory of mind, like the ability to figure out like what we as humans know and how to get us to follow them on the path to new understanding will also be paved, right?

[6:34](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=394s) **Sarah Guo** 所以，举例来说，我是个数学人，但在学校的某个时候，我停止了学数学。后来我用了 Math Academy。再后来我用 ChatGPT 来帮我学习新的数学。所以我就觉得，知识水位上涨的预期——你可以期望人们理解的东西——也会随之而来。

> And so like for example, like I'm a math person, but at some point in school, I stopped learning math. And then I did Math Academy. And then I did like ChatGPT to help me learn new math. And so I just feel like the expectation for the rising tide of what knowledge like you can expect people to understand will also be there.

### [6:52](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=412s) AI 沟通与监督
*AI Communication and Oversight*

[6:52](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=412s) **Greg Brockman** 我认为这是一个非常重要的点。这也是我期望、希望并从这项技术中开始看到的、也许被低估的好处之一：帮助人们知道更多、理解更多、更好地理解彼此、更好地理解世界。

> I think that this is a really important point. And one of maybe the underappreciated benefits that I expect from and hope for and we're starting to see from this technology is helping people know more, understand more, understand each other better, understand the world better.

[7:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=425s) **Greg Brockman** 而且我认为有一些技术机制可以作为基础——你有一个能够去解决这些问题、看到各领域之间所有联系的 AI，但它究竟如何把它传达给你？你如何理解它？

> And I think that there are technical mechanisms that can underlie the you have an AI that's able to go and solve these problems, see all the connections between domains, but how does it actually communicate it to you? How do you understand it?

[7:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=438s) **Greg Brockman** 实际上，这甚至与 AI 安全和对齐有关——让 AI 能够被我们理解它在做什么。所以有一些我们从 2017 年就开始讨论的技术，比如辩论和迭代放大，它们就像是同一枚硬币的两面，你真正要思考的是：如果你有 AI 在解决人类难以核查的问题。

> And actually, there's even connections to AI safety and alignment on having AI that we're able to understand what they're doing. And so there are techniques that we've been talking about since 2017, such as debate and iterative amplification, which are kind of two sides of a similar coin, where you really think about if you have AIs that are solving problems that are just too hard for humans to check.

[7:42](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=462s) **Greg Brockman** 甚至，举例来说，总结一本书就是一个很难核查的任务，对吧？除非你读完了整本书，否则真的很难知道这个总结好不好。所以如果一个 AI 直接给你一本书的总结并说，这就是总结，那它到底是一个好总结还是不好的总结？

> And even, for example, summarizing a book is an example of a task that's really hard to check, right? It's like it's really hard to know is the summary any good unless you've read the whole book. And so if an AI just gives you a summary of a book and says, this is the summary, is it a good summary or not a good summary?

[7:55](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=475s) **Greg Brockman** 这相当困难。除非你有另一个 AI 试图证明这个摘要是一个糟糕的摘要。那么，如果那个 AI 回来说，嘿，比如我在这本书里找到了这一段，它显然包含一个非常重要的情节点，而摘要并没有体现这一点，你不需要读完整本书就能说，哦对，这确实是一个很有力的论据。

> It's pretty hard. Unless you have another AI who is trying to prove that the summary is a bad summary. And so then if that AI comes back and says, hey, like here's this like paragraph in the book that I found, it's got a clearly a very important plot point, and that does not represent the summary, you don't have to read the whole book to be like, oh, yeah, that's actually a very good argument.

[8:14](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=494s) **Greg Brockman** 因此你意识到，有一些训练 AI 的方法实际上非常符合让人类能够进行监督、能够坐在驾驶座上的理念，这是 OpenAI 视角下未来重要的基本价值和基本事项，但也非常符合你所说的：能否让 AI 的成果成为让你学到真正为人类量身定制的东西，而不仅仅是为了某种抽象知识而存在的知识。

> And so you realize that there are ways to train the AI that are actually very aligned with having humans be able to have oversight, to be able to be in the driver's seat, which is a fundamental value, a fundamental thing that is important for the future from OpenAI's perspective, but also are very much aligned with what you're saying of can you make the results of an AI be something that means you learn something that is really tuned for the human rather than just sort of knowledge for some abstract knowledge sake.

### [8:42](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=522s) AI 克服生物学数据瓶颈
*AI for Overcoming Biology Data Bottlenecks*

[8:42](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=522s) **Sarah Guo** 对大多数人来说，数学理论的进步仍然有些抽象，对吧？所以会有那种“哇，这确实正在今天发生”的感觉。然后还有 AI 用于科学，比如有一种说法或信念认为它能在未来若干年内解决所有疾病，对吧？

> For most people, advancements in math theory are still somewhat abstract, right? So there's like, wow, this is absolutely happening today. And then also AI for science, like there's a claim or a belief that it could solve all disease in some number of future years, right?

[8:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=538s) **Sarah Guo** 你如何看待我们在这之间可能看到的进步，比如一步一步地？

> How do you look for like one step in terms of what we might see in advancements in between?

[9:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=545s) **Greg Brockman** 嗯，我认为我们已经在这项技术的迭代中经历了数年，看到了预期形态的轮廓。我觉得这有点像当你拥有一个在某个维度上有能力的 AI，并且如果你理解它为什么在那个维度上有能力，那么所有具有类似属性的人类其他领域，你也应该期待类似的结果。

> Well, I think that we've had a number of years in iteration of this technology to see the shape of what to expect. And I think that it's kind of like when you have an AI that is capable in a certain dimension, and if you understand the like why it got capable in that dimension, then all other fields of human endeavor that have similar properties, you should also expect similar results.

[9:29](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=569s) **Greg Brockman** 所以在我看来，为什么数学是我们成功实现了真正变革、取得了如此惊人成果的领域？部分原因在于我们的 AI 现在达到了某种能力水平和原始智能。

> And so in my mind, like why is mathematics something that we've managed to really transform in the way that we have had such incredible results? And that part of it is that our AI is now at a certain level of capability and raw intelligence.

[9:43](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=583s) **Greg Brockman** 部分原因在于，你可以认为数学是可检查的，但这已经不太正确了，对吧？比如我们确实是从非常可检查的问题开始的，比如一个算术问题，有一个数值答案。

> Part of it is that you can think of it as that, well, mathematics is checkable, but that's not quite true anymore, right? Like we really started with very checkable problems of here's an arithmetic problem. There's like some numeric answer.

[9:55](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=595s) **Greg Brockman** 所以你可以训练一个 AI 来检查它是否答对了，因为我们很容易看出 AI 提出的答案是否与实际答案相同。

> And so you have an AI that you can train to just like check if it got the answer right because we can very much just see is the answer that the AI is proposing the same as the actual answer.

[10:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=605s) **Greg Brockman** 但对于证明来说，情况非常不同。证明非常难以检查。它是否真的准确？形式化证明，好吧，是的，那是可检查的。但自然语言证明，也就是我们在这里真正产生的东西，要困难得多。

> But for a proof, very different. Proofs are very hard to check. Is this actually accurate or not? Formalized proof, OK, yes, that is checkable. But natural language proof, which is what we're really producing here, much, much more difficult.

[10:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=618s) **Greg Brockman** 但我认为真正的核心洞察其实在于数据，对吧？你可以在数学中获得大量数据。你可以在理论物理中获得大量数据。你能在生物学中获得大量数据吗？

> But the thing that actually I think is the core insight is really about data, right? It's that you can get lots of data in mathematics. You can get lots of data in theoretical physics. Can you get lots of data in biology?

[10:28](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=628s) **Greg Brockman** 这显然是一个数据瓶颈非常严重、非常困难的领域。现在，幸运的是，我们也一直在研究许多技术，以便能够用更少的数据工作。如果你想想这意味着什么，就像，好吧，我们现在有了聪明的模型，如果某个领域有大量数据，这就是深度学习的基本物理规律，对吧？

> That's obviously something very data bottlenecked, very difficult. Now, fortunately, we also have been working on lots of techniques to be able to work with less data. And if you think about what that means, it's like, OK, we now have smart models that if there's something where there's a lot of data, and this is the fundamental physics of deep learning, right?

[10:46](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=646s) **Greg Brockman** 算力、数据、算法等于能力等于结果。我们能够在这些困难的领域做到这一点，而人们已经花费了数百年时间，我们最聪明的人都在试图解决这些问题。现在 AI 可以解决它。

> Compute data, algorithms equals capability equals results. We're able to do it in these hard domains where people have spent hundreds of years of our smartest people all trying to solve the problem. And now the AI can solve it.

[10:59](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=659s) **Greg Brockman** 太棒了。当你把这一点转移到生物学时，显然我们想要的是药物发现，所有这些都是生物学及其理解的根本性变化，以改善生活质量，这就是我们想要应用的。

> Amazing. When you transfer that over to biology, and clearly what we want is drug discovery, all of these things that are like fundamental changes in biology and our understanding of it to improve quality of life, that's what we want to apply.

[11:13](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=673s) **Greg Brockman** 问题将是，而且尚未解决：我们有数据吗？我们有能力获取数据吗？所以我认为将会发生的是，我们将采用这些——我们已经这样做了——我们将这些模型、这些相同的技术应用到生物领域。

> The question will be and is not yet resolved, do we have the data? Do we have the ability to get the data? And so I think that what will happen is that we will take these and we're already doing this, where we take these models, these same techniques, and apply them in the biological domain.

[11:26](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=686s) **Greg Brockman** 有些领域会很快取得进展，对吧？我们会看到过去极其痛苦的事情现在变得极其容易，出现大规模加速。但还会有其他瓶颈。问题将是，好吧，我们如何缓解这些瓶颈？

> And there are some areas that will fall quickly, right? That we'll just see the massive acceleration of things that used to be incredibly painful, now incredibly easy. But there will be other bottlenecks. And there'll be a question of, well, how do we alleviate those bottlenecks?

[11:38](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=698s) **Greg Brockman** 顺便说一句，其中一些甚至不是真正关于深层生物学理论的。例如，监管申报，巨大的瓶颈，对吧？撰写整个临床试验案例等需要大约一年时间。

> And some of these, by the way, are not even really about deep biological theory. For example, regulatory filings, huge bottleneck, right? Takes like a year to write up your whole sort of case for clinical trial and all these things.

[11:50](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=710s) **Greg Brockman** 而用 AI 写这些东西要容易得多。所以我认为我们将不得不审视药物发现过程的每一步。有些地方根本数据就是不存在。

> And it's just like, you can write those things with an AI much more easily. And so I think that we're going to have to look at every step of the drug discovery process. And there are going to be places where the fundamental data just does not exist.

[12:04](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=724s) **Greg Brockman** 因此，湿实验室的创新、如何实际进行模拟的创新，所有这些都将变得极其重要。所以我认为我们会看到这些突破的到来，但重要的是要知道，有些东西与数学共享，即关于科学发现，而有些东西在数据来源方面非常不同。

> And so then innovations on wet labs, innovation on how you actually do simulations, all of this will become incredibly important. So I think that we will see these breakthroughs coming, but it's important to know there are things that are shared with mathematics that is about scientific discovery and things that are very different in terms of where the data comes from.

### [12:23](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=743s) 彻底改变药物发现
*Revolutionizing Drug Discovery*

[12:23](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=743s) **Elad Gil** 你提到了生物学，有没有具体的——我认为这是一个很好的例子，很多早期研究，AlphaFold 之类的东西，真正专注于设计分子或理解某物如何相互结合。

> Are there specific you mentioned biology, and I think that's a great example where a lot of the early research, AlphaFold and things like that were really focused on designing a molecule or understanding how something binds to each other.

[12:33](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=753s) **Elad Gil** 正如你所说，开发一种药物需要 15 年。花费 15 亿美元。大约 1500 万美元用于临床前、小鼠研究、分子，然后 14 亿美元以上用于临床试验。所以你的观点是，这也是两三年 versus 另外 12 年。

> And to your point, it takes 15 years to develop a drug. It's $1.5 billion. $15 million or so tends to be preclinical, mouse studies, the molecule, and then $1.4 billion plus tends to be the clinical trial. And so your point is and it's also two, three years versus another 12 years.

[12:50](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=770s) **Elad Gil** 所以你的观点非常正确，这些模型的早期重点很多是“这东西能否结合”，而不是下游还要担心哪些其他事情才能真正将其转化为对人类事业有生产力的东西？

> And so your point is really well taken in terms of a lot of the early emphasis of these models has been can the thing bind versus what are all the other things you have to worry about downstream to actually turn it into a productive thing for human endeavors?

[12:59](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=779s) **Elad Gil** 在 OpenAI，你们目前如何对不同科学领域进行排序，就你们关注的重点、或你们想要生成更多数据或更多结果或更多能力让其他人受益的地方而言？

> How do you at OpenAI currently stack rank the different areas of scientific endeavor in terms of where you're focused or where you want to generate more data or more results or more capabilities that others can benefit from?

[13:11](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=791s) **Greg Brockman** 看，我认为药物发现是最清晰、最明显的领域，如果你要选一个想要彻底变革的领域。而且那些真正能提升每个人生活质量的地方，纯粹重要，也许是所有人长期以来一直在思考的 AI 应用之首，我们显然必须去做。

> Look, I think that drug discovery is the most clear and obvious area if you get to pick one that you want to fully revolutionize. And just the places where you're going to really uplift everyone's quality of life that is just like so purely important and maybe the top one application of AI that everyone's been kind of thinking about for a long time, we've clearly got to do it.

[13:36](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=816s) **Greg Brockman** 正如你所说，整个流程太长了。我认为我们会发现人类在生物学上非常糟糕，在药物发现上非常糟糕。如果你想想，有一个特性：对于人类的自然语言，我们非常擅长。

> And to your point, like there's this whole pipeline, it's just so long. And I think we're going to find that humans are incredibly bad at biology, incredibly bad at drug discovery. And if you think about it, there's this like property where for humans' natural language, we're very well suited to it.

[13:53](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=833s) **Greg Brockman** 这显然非常人性化，非常符合我们与世界互动的方式。而生物学则完全不同。但对于神经网络来说，我并不认为有任何理由相信它在自然语言上应该比在生物学上更好。

> Like it's something that very clearly it's very human. It's like very well tuned to how we have interacted with the world. Whereas biology, it's totally foreign. But for a neural net, I don't really think there's any reason to believe it should be better at natural language than it should be better compared to biology.

[14:10](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=850s) **Greg Brockman** 所以，我们在自然语言中看到神经网络取得如此惊人的结果，从第一性原理出发，你应该能够真正破解生物学的语言。所以对我来说，这是第一位的。

> And so the fact we're seeing such incredible results with neural nets in natural language, you're just like from first principles, you should be able to really be able to crack the language of biology. And so to me, that is number one.

[14:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=861s) **Greg Brockman** 从第一性原理来看，这应该可行，就我们所看到的而言。我们在许多行动上已经非常接近了。

> It's like first principles that should work in terms of what we're seeing. We're so close in so many actions.

[14:24](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=864s) **Elad Gil** 是的，这非常令人兴奋。

> Yeah, that's very exciting.

[14:26](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=866s) **Greg Brockman** 是的。而且我认为你可以大规模增加并行性。我认为你可以增加——我认为你可以减少许多步骤的持续时间。顺便说一句，我认为药物发现有很多维度，例如 NF1 疾病。

> Yeah. And I think you can massively increase the parallelism. I think you can increase I think you can decrease so many of these steps in terms of their duration. And I think, by the way, that there are many dimensions to drug discovery, for example, NF1 diseases.

[14:41](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=881s) **Greg Brockman** 甚至有很多人，治愈他们特定疾病的药物已经存在。只是它存在于文献中的某个地方，有一篇论文。如果你能找到那篇与你的疾病相关的论文，你就能治愈那个人。

> There's even like lots of people who the drug to cure their particular condition exists. It's just that it's just out there somewhere in the literature that there's a paper. And if you can find that one paper that's relevant to your conditions, then you'll be able to cure the person.

[14:54](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=894s) **Greg Brockman** 而且有很多先例。我认为我们可以大规模扩展这类操作。所以我认为短期影响实际上非常可能。但材料科学，另一个例子。我认为还有一个长尾的其他领域，我们应该看到人类能够实现的事情发生真正的变化。

> And there's plenty of precedent for it. I think we can massively scale up those kinds of operations. So I think that impact in the short run is actually very possible. But material science, another example. And I think there's a long tail of other domains where we should just see real changes in kind of what is possible for humanity to accomplish.

### [15:12](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=912s) 改变法规以加速进展
*Changing Regulations to Accelerate Progress*

[15:12](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=912s) **Sarah Guo** 我们想想为什么一种药物可能需要 15 年才能上市。部分原因是你正在进行 successively 更昂贵的试验，这些试验有不同的安全考虑，最终导致大规模人体试验。

> It seems just we think about why a drug might take 15 years to go to market. Part of it is you're doing like successively more expensive trials that have a different set of safety considerations resulting in large-scale human trials.

[15:27](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=927s) **Sarah Guo** 而今天的这些试验是由法规规定的，对吧？有充分的理由。如果你相信技术会起作用，并且部分数据可以获取或生成，那么限制因素之一似乎将是说服监管机构我们应该改变期望

> And then these trials today are they're ordained by regulation, right? For good reason. It feels like one of the limiting factors if you believe the technology will work and parts of the data can be acquired or generated is going to be convincing regulators that we should change the expectations

[15:47](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=947s) **Sarah Guo** 关于哪些药物被接受以及步骤是什么。

> around what drugs are accepted and what the steps are.

[15:51](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=951s) **Greg Brockman** 我认为这里有一个时刻，而且要达到这一点还有更多工作要做。但如果你真的能证明药物发现的物理学已经改变，那么这当然应该渗透到系统的每个部分。

> I think there is a moment here where and there's more work to be done to reach this point. But if you can really prove that the physics of drug discovery have changed, then of course it should trickle through to every part of the system.

[16:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=965s) **Greg Brockman** 如果你想一想，我的意思是，哪怕只是举一个非常不同的例子。想想自动驾驶汽车，对吧？我们的道路运营方式、安全性以及所有这些事情，很大程度上都是为人类驾驶员调校的。

> And if you think about I mean, even just take a very different example. Think about self-driving cars, right? We have so much of our how roads operate and safety and all these things that are really tuned for human drivers.

[16:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=978s) **Greg Brockman** 如果你达到这样一个点，自动驾驶汽车比人类驾驶安全10倍或100倍，对吧？你已经到了这样一个点，AI实际上能够在一个领域中交付结果，而这些结果毫无疑问就是你想要的东西，那么你当然应该问，好，我们该如何重新思考整个交通系统？

> And if you have if you reach a point where self-driving cars are 10 times or 100 times safer than human ones, right? That you've reached that point where AI is actually able to deliver the results in a domain that just are so unambiguously the thing that you want, of course you should then say like, OK, how do we rethink the whole transportation system?

[16:37](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=997s) **Greg Brockman** 也许你应该制定法规，规定只允许在某些区域使用自动驾驶汽车，比如有很多潜在的变化，会改变你如何看待一个系统究竟试图产出什么，而这些也需要反过来渗透到监管中。

> Maybe you should have regulation that says that you only are allowed to have self-driving cars in certain areas, like that there are a lot of potential sort of changes to the way that you think about what is even trying to come out of a system that needs to trickle back into regulation as well.

[16:52](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1012s) **Greg Brockman** 当然，监管在全球并不统一，对吧？在临床试验方面，澳大利亚和美国的法规就非常不同。而且你已经看到很多人把试验转移到监管环境更合理的地方，更符合事情的发展方式、他们想要推进的方向，以及人们认为对他们有效的东西。

> And of course, regulation is not uniform across the world, right? That there's very different regulations in Australia versus the US when it comes to clinical trials. And that you're already seeing a lot of people moving their trials to places where the regulatory environment makes more sense for how things are, how they want to pursue, and what people think is efficacious for them.

[17:13](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1033s) **Greg Brockman** 所以我认为，事情究竟应该如何改变，很大程度上取决于我们实际能证明什么，对吧？是不是真的到了这样一个点：你能从神经网络获得比实际运行临床试验更好的结果？

> And so I think that the exactly how things should change, I think, will depend a lot on what we actually can prove, right? Is it the case that you actually reached the point that you can get better results from a neural net than from actually running a clinical trial?

[17:27](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1047s) **Greg Brockman** 比如我们还没到那一步。

> Like we're not there.

[17:29](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1049s) **Sarah Guo** 是的。

> Yes.

[17:29](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1049s) **Greg Brockman** 但你可以想象我们会达到那一步。

> But you could imagine we will get there.

[17:30](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1050s) **Sarah Guo** 但如果它们在有效性或毒性等方面更具预测性，那么显然你应该改变它。

> But if they were more predictive on efficacy or tox or something, then obviously you should change it.

[17:35](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1055s) **Greg Brockman** 当然。我认为，在一些领域，社会应该并且我希望会大力投入，因为可能性已经发生了根本性变化。然后让我们很快地围绕它重新调整方向。

> Of course. And I think that of areas of areas where I think society should and I hope will be very sort of leaned into the what is possible has fundamentally changed. And let's very quickly then reorient around it.

[17:53](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1073s) **Greg Brockman** 再说一次，你必须去做。你必须真正思考所有后果，对吧？以及所有影响。而且，如果你在考虑长期研究、人，以及类似这样的问题：这个神经网络在预测模拟方面，对于长期结果实际上是否准确？

> And again, you have to do it. You have to really think through all the consequences, right? And all of the implications. And that one of the if you're thinking about long-term studies, people and like question of like, is this neural net actually in terms of what's predicting the simulation actually accurate for long-term outcomes?

[18:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1088s) **Greg Brockman** 要做到至少没有一个长时间刻度来对应这一点，是相当困难的。但我认为，我们会看到80%的可能性很快实现。

> It's pretty hard to not have at least had one long time tick to correspond to that. But I think that I think we will see like 80% of what is possible come to fruition very quickly.

### [18:24](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1104s) 主动式智能体 Dots
*Dots the Proactive Agent*

[18:24](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1104s) **Sarah Guo** 距离Astra发布和Dev Day已经过去一段时间了。就人们内部或外部如何使用一些新产品或能力而言，有什么让你感到惊讶的吗？

> You're a little ways out from the Astra release and dev day now. Is there anything in terms of how people internally or externally are using some of the new products or capabilities that surprises you?

[18:38](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1118s) **Greg Brockman** 嗯，我认为Dots确实改变了许多人的工作方式。而Dots所代表的东西是……

> Well, I think Dots has really changed how many of us work. And the thing that Dots represents is.

[18:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1124s) **Sarah Guo** 你能简要描述一下Dots吗？

> Can you describe Dots briefly?

[18:46](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1126s) **Greg Brockman** 可以。Dots是一个主动式智能体。它始终在线，始终工作。你停止工作，它仍继续工作。我的Dots会做的事情，对吧？我有一个Dot在留意我本该做的工作，比如我本该发送的消息、我说过要做但还没完成的事情。

> Yeah. So Dots is a proactive agent. It's always on, always working. You stop working, it keeps working. And the kind of thing that my Dots does, right? So I have a Dot that's looking out for work that I was supposed to do, like messages I was supposed to send, things that I said I would do that I haven't gotten to yet.

[19:04](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1144s) **Greg Brockman** 我每天早上醒来，都会有一份清单，上面是我当时想，哦对，这是我的待办事项。而且它经常注意到我甚至从未要求它注意的事情。如果我日历上的日程安排有冲突，它会指出来。

> And I always wake up in the morning. There's like a list of things that I was like, oh, yeah, this is my to-do list. And that it often notices things that I never even asked it to notice. If there's like a conflict on my calendar in terms of scheduling, it'll point that out.

[19:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1158s) **Greg Brockman** 比如它会说，嘿，你实际上需要提前离开这个会议，才能赶上另一件事。所以它在语境上非常敏锐。它知道我一天在做什么。它理解我们正在做的项目。

> If it's like, hey, you actually need to leave this meeting earlier in order to get to the other thing. And so it's like so contextually aware. It knows what I'm doing with my day. It understands the projects we're working on.

[19:28](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1168s) **Greg Brockman** 它理解，或者说它知道我的收件箱里有什么，知道我的Slack里有什么，并且能够跨这些进行整合。它知道我一直问你的问题。所以它知道哪些领域是我最关心的。

> It understands sort of it knows what's in my inbox, knows what's in my Slack, and it's able to integrate across these. And it knows the questions I've been asking you about. So it knows what areas are most top of mind for me.

[19:38](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1178s) **Greg Brockman** 而且我认为，从某些方面说，Dots是什么，它改变的只是形态，而不是能力，对吧？它就像是一个Astra级别的模型。它有自己的云计算机。但在工作场所环境中，这种形态变化加上这种智能水平，是游戏规则的改变者。

> And I think that having again, some ways what Dots is, it's just changed to the form factor, not the capability, right? It's like it's an Astra-level model. It's got its own cloud computer. But that the change in that form factor in a workplace setting with this level of intelligence, it's a game changer.

[19:57](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1197s) **Elad Gil** 你有多个Dots在运行，还是只有一个在覆盖你所有的后台任务之类的？

> And do you have multiple Dots running or you just have one that's sort of covering all your background tasks or anything?

[20:01](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1201s) **Greg Brockman** 我有一个。我有一个。它的图标是一只小猫头鹰。你可以让它更改图标。它只是在我的工作生活中处理各种事情。

> I've got one. I've got one. It's a little owl is the icon. You can ask it to change the icon. And it's I just sort of doing everything across my work life.

### [20:13](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1213s) 打通工作与个人场景的智能体
*Unifying Work and Consumer Agents*

[20:13](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1213s) **Sarah Guo** OpenAI显然有研究努力，然后还有消费者业务、企业业务和开发者业务，也就是今天的平台业务。你认为，比如说帮助你在工作的助手，比如Dots，或者我主要就是这样理解它的，和像Instinct Muses那样的消费者助手之间，会出现分离吗？

> OpenAI obviously has research efforts and then a consumer and an enterprise and developer business, platform business today. Do you think that there's going to be a separation between, let's say, assistants that help you at work like Dots or just how I primarily think of it and then like consumer assistants like the Instinct Muses of the world?

[20:32](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1232s) **Greg Brockman** 我认为这些东西会融合，毫无疑问。而且我认为在不同环境中有一些非常不同的压力。所以对于大众市场消费产品来说，真正弄清楚如何扩展规模，以及如何免费或以极低成本提供东西，非常重要。

> There is some I think these things are going to converge, no question. And I think that there are some pressures that are very different in different environments. So for a mass market consumer product, really figuring out how to scale and to provide things for free or at very low cost is very important.

[20:49](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1249s) **Greg Brockman** 而在工作场所，如果你是生产力场景，如果你是高管、个人贡献者，无论你的角色是什么，那么说“峰值智能”就容易得多。而且实际上，只要你能从支付更多中获得回报，投入更多算力就是值得的。

> Whereas in the workplace and if you're productivity, if you're the executive, individual contributor, whatever your role is, then it is much easier to say peak intelligence. And really, as long as you're getting returns to paying more, it's worthwhile to throw more compute at it.

[21:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1268s) **Greg Brockman** 所以我认为，主要区别确实在于成本。我们在ChatGPT上已经看到了这一点，对吧？对于ChatGPT，我们投入了大量精力来打造最大规模的免费层级，并且把记忆等功能带给所有免费用户是一项了不起的努力，是真正巨大的工程投入。

> And so I think that that is the main distinction is really around cost. And we've seen this in ChatGPT, right? That ChatGPT, we put in a lot of effort to have the largest scale, free tier, and bringing things like memory to all of our free users was an incredible endeavor, like a real huge engineering effort.

[21:23](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1283s) **Greg Brockman** 而为极少数人以高得多的成本实现同样的功能，则非常容易。所以我认为，我们在个人助手工作空间中看到的是，将会有这些不同的市场细分，从根本上会消耗不同数量的算力。

> And it's just very easy to implement the same functionality at much higher cost for a very small number of people. And so I think that what we're seeing in the personal assistant workspace is that there are going to be these different market segments where different amounts of compute are going to be expended fundamentally.

[21:41](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1301s) **Greg Brockman** 所以有时候，当算力水平不同时，就意味着产品看起来会有些不同，对吧？可能一个产品中嵌入了更多功能，而另一个没有。当然，使用场景也不同。

> And so sometimes when there's different levels of compute, it then means the product can look a little different, right? It can be that there's just more functionality that's embedded in one versus the other. And there's, of course, different use cases.

[21:51](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1311s) **Greg Brockman** 如果你想想支付、旅行之类的事情，那更多是面向消费者的。但我不认为有任何根本区别，对吧？它真的就只是一个在做事情的AI。只是碰巧在某个方向上有一点轻微的压力。

> If you think about things like payments and travel and things like that are much more for the consumer. But I don't think there's any fundamental difference, right? It's like really just like an AI that's doing things. And it just like happens to be like slight sort of pressure in one way or another.

[22:07](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1327s) **Greg Brockman** 所以OpenAI的优势之一，就是我们ChatGPT上有12亿消费者。我们还有领先的编程和企业业务。没有其他公司同时拥有这两者。所以我们的许多战略是，让我们把这些统一起来，这样我们就不会在两条战线上竞争。

> And so one of the strengths of OpenAI is the fact that we have 1.2 billion consumers on ChatGPT. And we have also a leading coding and enterprise business. And no one else has both in-house. And so that a lot of our strategies to say, let's unify these so we're not sort of competing on two fronts.

[22:24](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1344s) **Greg Brockman** 实际上是在一条战线上竞争。而且我认为，这也是AI的大方向，即模糊消费者和企业之间的区别。我们将迎来企业家的复兴。

> It's really competing on one front. And I think that this is also the general direction of travel of AI is to blur the distinction between what is consumer, what is enterprise. We're going to have this renaissance of entrepreneurs.

[22:36](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1356s) **Greg Brockman** 比如我们已经看到了。我的意思是，我想你正在非常具体地看到这一点。我认识一些人辞掉工作去创业，他们说，如果我没有一个我知道可以依赖的AI，我绝不会这么做。

> Like we're already seeing it. I mean, I imagine you're seeing this very concretely. I know people who are quitting their job to go start companies who say, I never would have done it if I didn't have an AI that I knew I could rely on.

[22:47](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1367s) **Greg Brockman** 它可以处理掉许多正常的进入壁垒。所以我认为，这些区别已经存在了这么久，我认为它们将会改变。

> It could take care of so much of the normal barriers to entry. And so I think that we're going to see just like the distinctions have held for so long, I think they're going to change.

### [22:56](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1376s) UI 将如何演变
*How UI Will Evolve*

[22:56](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1376s) **Elad Gil** 传统上区分企业和消费者世界的一件事，是你如何看待UI。显然，你提到了成本和以某种保真度为某人交付东西的底层能力。

> So one of the things that traditionally has separated the enterprise and consumer world is how you think about UI. Obviously, there's your point on cost and underlying ability to deliver something for somebody at a certain level of fidelity.

[23:07](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1387s) **Elad Gil** 但另一部分就是用户界面。而且通常在企业环境中，用户界面或你访问它的方式，与消费者环境非常不同。正如你所说，你正朝着统一这些层次的方向努力。

> But the other part of it is just the user interface. And often the user interface in an enterprise setting or how you access it is very different from the consumer setting. To your point, you're looking towards unifying these layers.

[23:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1397s) **Elad Gil** 你也一直在思考动态UI之类的问题。这一直是人们很早就谈论的事情之一：如何动态生成对用户有用的UI？

> You've also been thinking quite a bit about like dynamic UIs. And that's always been kind of one of the things people have talked about quite early is how do you just dynamically generate a UI that's useful for a user?

[23:27](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1407s) **Elad Gil** 或者在某些情况下，这些东西会退到后台。它只是你聊天的一部分。你认为UI将如何为这些不同用例演变？或者当你处理这两个细分市场时，你如何看待这个问题？

> Or in some cases, these things kind of move into the background. It's just part of your chat. How do you think UI will evolve for these different use cases? Or how do you think about that as a problem as you sort of address both of these segments?

[23:36](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1416s) **Greg Brockman** 首先，我认为最好的 UI 就是没有 UI。我认为 AI 正朝这个方向走。它就是一个你可以与之交谈的东西，你可以直接开口跟它说。

> So at a place so first of all, I think that the best UI is no UI. And I think in AI, that is where we're going. It's just one of the things you can talk to. And you can speak to it verbally.

[23:48](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1428s) **Greg Brockman** 你可以通过 Slack、iMessage，或者任何你能调用它的界面与它交流。而这正是 Dots 现在就能做到的事，就是字面意义上的今天。但个人场景和工作场所场景之间存在一些根本性的差异。

> You can talk to it over Slack, iMessage, any of the surface you can call it. And that's literally what you can do with Dots, like literally today. But there are some fundamental differences between a personal setting and a workplace setting.

[24:02](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1442s) **Greg Brockman** 在工作场所中，你有管理员，对吧？你有公司的安全团队。你只是有更多的利益相关方。而且它不只是一个为你服务的 AI，它代表你，你拥有它，你也拥有它的结果。

> And a workplace, you have an admin, right? You have kind of your company security team. You have just more stakeholders. And it's not just sort of a AI that's for you, that's representing you, that you own and you own its outcomes.

[24:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1461s) **Greg Brockman** 所以我认为这就是我看到的 enterprise 和家庭使用场景之间的主要区别。当然，这也会带来产品上的差异。有时候它们是非常小的事情。比如，当我们发布一个新功能时，企业希望控制它何时被引入。

> And so I think that that is the main distinction I see between the enterprise and the home use case. And of course, this leads to product differences as well. And sometimes they're very small things. Like for example, when we release a new feature that enterprises want control over when it's introduced.

[24:35](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1475s) **Greg Brockman** 因此我们有管理员控制和开关。所以你会遇到这样一种体验：由于它处于工作场所中，你可用的功能反而更少。而这确实会在某种程度上改变……嗯，以某种方式向你呈现这一点。

> And so that we have admin controls and toggles. And so you can have an experience where there's just less functionality available to you as a result of the fact that it is in the workplace. And that does then change the well, in some way, to surface that to you.

[24:51](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1491s) **Greg Brockman** 这就回到了整个要点——没有 UI，对吧？所以问题是，功能缺失意味着什么？你又该如何呈现它，并让人们知道情况就是如此？

> And this comes back to what the whole point was that there was no UI, right? So it's like, what does it mean for there to be functionality that's missing? And how do you even surface it and let people know that that's the case?

[24:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1498s) **Greg Brockman** 所以这是一个持续的挑战。我认为它不会神奇地消失。我认为你需要拥抱这种复杂性。但对我来说，从根本上说，如果你仔细想想，这既疯狂又美丽。

> So that's an ongoing challenge. And I think that that will not magically go away. I think that you need to lean into that complexity. But yeah, to me, fundamentally, it's kind of crazy and beautiful if you think about it.

[25:09](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1509s) **Greg Brockman** ChatGPT，那个文本框，对吧？它就是从这里开始的。它就是一个文本框。而我们正在走向的，是甚至连文本框都没有，对吧？就像，那就是我们曾被承诺的 AI。

> The ChatGPT, the text box, right? That's where it started. It was a text box. And what we're heading to, there isn't even a text box, right? It's just like, that is the AI we were promised.

[25:20](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1520s) **Sarah Guo** 当我思考什么能让一个没有 UI 的产品更好时，这得由你来说。但我认为包括：我能在这上面花多少算力？模型本身有多好？还有记忆系统。

> When I think about what might make a product that doesn't have a UI better, there's like you tell me. But I think there's like, how much compute can I spend on it? How good is the model itself? There's a memory system.

[25:34](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1534s) **Sarah Guo** 还有，我不知道，关于行为的决策。模型在“应该多努力思考某件事”上的校准现在有多好？

> There's like, I don't know, decisions about behavior. How good are models that being calibrated on how hard they should think about something now?

[25:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1544s) **Greg Brockman** 所以这是我们正在变得非常擅长的事情。随着 GPT-6 向所有人发布，我们现在有能力让 AI 在要深入思考一个问题时，先给你一个快速答案，然后继续思考，再给你更完整的答案。

> So that's something we're getting very good at. And with the release of GPT-6 to everyone, that we have the ability now for the AI, if it's going to go think hard about a problem, it actually can give you a quick answer and then keep thinking and then give you the more fulsome answer.

[25:59](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1559s) **Greg Brockman** 另外，关于没有 UI 这一点，以及你之前说的，我们还引入了智能 UI。所以现在 AI 可以即时生成生成式 UI。因此我认为，即使没有固定的 UI，没有按钮，没有需要学习的菜单，拥有一个能真正把结果打造得在视觉上完全符合你想要的 AI，也是很有价值的。

> And also to the point of no UI and to the thing you were saying earlier, too, is that we introduced intelligent UI. So now the AI can do generative UI on the fly. And so I think that even though there's no fixed UI, there's no buttons, there's no menus to learn, having an AI that can actually craft the result to be visually exactly what you want.

[26:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1579s) **Greg Brockman** 比如，即使你想想订机票这样的事情，你在那里想要一种复杂的 UI，对吧？这样你就能看到所有不同的选项。而对于你可能问的任何问题，比如我该怎么组装一辆自行车之类的，你可能想要一些非常定制的应用，通常人们会花一个月去写这种应用。

> Like if you think about even something like flight booking, you want kind of a complex UI there, right? So you can see all the different options. And that for any question you might ask, like how do I assemble a bicycle or something, you might want some very sort of bespoke app that normally a person would have gone off and written for a month.

[26:35](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1595s) **Greg Brockman** 当然，你永远不会雇人为你写一个一次性应用来帮你组装自行车。但让 AI 在你需要的那一刻实时为你做这件事，则完全没问题。

> And of course, you're never going to hire someone to write a one-off app for you to help assemble a bike. But totally fine for an AI to go do that for you in real time in the moment you need it.

[26:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1604s) **Greg Brockman** 所以我认为，关于这些，有一点是：我们现在处于一个 AI 越来越多地为你塑造自身形态的世界，对吧？用户界面正在变成你当下所需要的那件东西。

> And so I think that there is something about these like we're now in a world where the AI is molding itself much more to you, right? That the user interface is becoming exactly the thing you need at the moment you need it.

[26:57](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1617s) **Greg Brockman** 这是我们大量关注的重点。我们正在从我们思考产品的每一个方面来做这件事。

> And that is a lot of our focus. And we're doing that across every aspect of how we think about the product.

### [27:02](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1622s) AI 时代的网络安全
*Cybersecurity in the AI Age*

[27:02](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1622s) **Elad Gil** 最近围绕 AI 的讨论中最大的转变之一是关于网络安全的。你如何构建能够抵御 AI 驱动的网络攻击方式的安全系统？所以我有点好奇你如何看待这一点，你如何构建防护栏，或者你如何应对网络安全的各个不同方面。

> One of the biggest shifts in the discourse around AI recently has been around cyber. And how do you build secure systems that are resistant to AI-driven cyber approaches? And so I'm just a little bit curious how you're thinking about that, how you're building guardrails or how you're approaching the different aspects of cybersecurity.

[27:18](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1638s) **Greg Brockman** 嗯，首先，我确实认为网络安全的物理规律正在改变。我们正处于我们所说的“防御者窗口”。我对这个窗口的理解是，我们拥有巅峰的前沿 AI 能力，可以帮助防御者保护自己。

> Well, first of all, I do think the physics of cybersecurity are changing. And we're in what we call the defender's window. And the way that I think about this window is that we have peak frontier AI capabilities that can help defenders secure themselves.

[27:33](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1653s) **Greg Brockman** 这些能力尚未广泛扩散。但广泛扩散的能力正在逐步提升。你可以从世界各地供应商的每一次模型发布中看到这一点。所有人都能获得的网络能力正在上升。

> They're not broadly diffused yet. But the broadly diffused capabilities are ratcheting. And you can see this from each model release to model release from providers across the world. The capabilities on cyber that are available to everyone are going up.

[27:47](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1667s) **Greg Brockman** 而且你开始看到一些影响的领先边缘，对吧？比如，你开始看到在韩国，有一些备受关注的银行黑客攻击是由不同的模型驱动的。而我认为至关重要的一点是，防御者需要利用这段时间。

> And you're starting to see the leading edges of some of the impacts, right? You're starting to see, for example, in South Korea, that there were some high-profile hacks of banks that were driven by different models. And I think that the picture that is so critical is that defenders need to use this time.

[28:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1685s) **Greg Brockman** 所以我们一直在非常努力地工作，以便能够通过我们的 Daybreak 项目负责任地向防御者提供访问权限，让他们能够使用我们的巅峰智能，或使用非常强大的模型来发现并修复他们系统中的漏洞。

> And so we have been working very hard to be able to responsibly give access through our Daybreak program to defenders to be able to use our peak intelligence or to use very capable models to find and fix vulnerabilities in their systems.

[28:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1699s) **Greg Brockman** 而有一件非常重要的事情需要认识到：这些缺陷一直存在，对吧？从某种意义上说，每一行人类编写的代码，你都得问，这安全吗？开发者是否以正确的方式受过训练？然后你就会开始意识到，我们构建系统的方式从根本上就是不安全的。

> And one thing that is very important to realize is these flaws have existed, right? It's every line of human-written code in some ways that you have to ask, is this secure? Was the developer trained in the right way? And you just start to realize that the way in which we built systems were kind of fundamentally insecure.

[28:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1719s) **Greg Brockman** 而我们一直都知道这一点，对吧？从来没有哪个 CISO 拥有过充足的资源。

> And we always knew it, right? There's never been a CISO who's been over-resourced.

[28:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1724s) **Elad Gil** 是啊，对吧？

> Yeah, right?

[28:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1724s) **Sarah Guo** 我是说，你可以看看 XKCD 那幅漫画，讲的是全世界的基础设施都依赖某个家伙在地下室里写的一个开源项目，对吧？或者 Heartbleed，对吧？

> I mean, you can look at XKCD in the comic of like all the world's infrastructure reliant on like one open source project written by one dude in his basement, right? Or Heartbleed, right?

[28:54](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1734s) **Greg Brockman** 没错。

> Exactly.

[28:54](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1734s) **Sarah Guo** 那早于所有这些 AI 问题。

> Predated like all these AI issues.

[28:56](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1736s) **Greg Brockman** 没错。所以我认为，对我来说，而且我认为非常重要的一点是，要认识到这些其实是人的问题，对吧？这些是我们选择构建互联网的方式。安全是被事后附加上的，对吧？

> Exactly. And so I think that to me, and I think very importantly, it's very important to realize that these are really human issues, right? That these are the way that we chose to build the internet. It was security got kind of bolted on, right?

[29:08](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1748s) **Greg Brockman** 而且我们有过备受关注的事件，对吧？你想想早期的蠕虫和病毒。然后人们开始意识到，嗯，我们需要病毒扫描软件，对吧？我们需要对 HTTP 请求进行加密，应该用 HTTPS。所有这些机制，我认为都是我们走了较慢的路线，以便事后添加安全性。

> And we had high-profile incidents, right? You think about early worms and viruses. And then people started to realize, well, we need virus scanning software, right? And we need to have encryption for requests over HTTP, should be over HTTPS. All of these mechanisms were ones where I think that we kind of took the slower route in order to add security on afterwards.

[29:34](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1774s) **Greg Brockman** 就像互联网确实是为韧性而建的，而不一定是为了安全。而我认为我们现在将看到的是，很多这样的债务，很多这样的技术债务，将会浮出水面。

> Like the internet's really been built for resilience, not necessarily for security. And I think that what we're going to see now is a moment where a lot of that debt, a lot of that tech debt is going to come to the surface.

[29:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1784s) **Greg Brockman** 是时候让我们偿还它了。而作为一家公司，我们确实在努力帮助应对这一时刻。例如，我们做出了十亿美元的承诺，帮助一线防御者获得我们的模型，并将其用于防御。

> It's time for us to pay it down. And that as a company, we are really leaning into trying to help with this moment. So for example, we have a billion-dollar commitment that we've made to help frontline defenders get access to our models and utilize them for defense.

[30:00](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1800s) **Greg Brockman** 我们正在研究并试图找出各种不同的漏洞。然后思考我们如何帮助世界做出响应？我们如何协调事件响应？我们如何与政府合作？我们如何与社区合作，以便尽可能好地落地这些工作，并帮助人们度过难关？

> That we're studying and trying to find different vulnerabilities out there. And then thinking about how do we help the world respond? How do we coordinate incident response? How do we work with government? How do we work with the community in order to land these as well as possible and to help people navigate through?

[30:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1817s) **Greg Brockman** 但我认为人们还没有真正内化网络安全的物理规律将会发生多大的变化。我认为我们最终会——我认为有机会最终拥有一个安全得多的世界，对吧？

> But I think that people have not yet really internalized how much the physics around cybersecurity are going to change. I think we will end up I think there's an opportunity to end up with a much more secure world, right?

[30:27](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1827s) **Greg Brockman** 你想想这样一个事实：AI 让你能够以以前完全不可想象的方式偿还技术债务。你能够做诸如软件形式化验证这样的事情。这些模型显然非常擅长证明事情。

> You think about the fact that AI lets you pay down your tech debt in a way that was just previously unthinkable. The fact that you could do things like formal verification of software. Like these models are obviously great at proving things.

[30:40](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1840s) **Greg Brockman** 为什么不证明软件的安全属性呢？这项技术在某种程度上已经为人们准备好了。只是它要复杂得多。

> Why not prove security properties of software? That's something that the technology has kind of been there for people. But it's just so much more complex.

[30:46](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1846s) **Sarah Guo** 痛苦、小众、昂贵。

> Painful, niche, expensive.

[30:47](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1847s) **Greg Brockman** 没错。没有人会为任何重要的事情去做它。所以我认为所有这些都可以改变。而我认为我们现在正在走向的世界——我们已经在 OpenAI 内部这样做了——是每当有新模型发布，你都需要应用它来保护自己，找出你的漏洞，然后饱和。

> Exactly. No one does it for anything significant. And so I think all of that can change. And I think that the world we're moving to right now, and we've done this internally within OpenAI, is every time a new model drops, you need to apply it to secure yourself, find your vulnerabilities, and then saturate.

[31:02](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1862s) **Greg Brockman** 例如，我们抽调了大约 25% 的生产工程师，说，你们的所有项目都暂停。你们现在只做安全工作。然后拿出我们当时的巅峰智能模型，让它探查我们所有的基础设施，给一堆东西打分，构建了一些工具，以便随着时间推移可以轻松地再做一次。

> And so for example, we took about 25% of our production engineers and said, all of your projects are on hold. You are just securing now. And took our peak intelligence model at the time, had it poke at all of our infrastructure, rate a bunch of things, built some tools so that it's easy to do it again over time.

[31:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1879s) **Greg Brockman** 我们发现了许多问题。我们修复了它。而在某个时候，我们饱和了这个模型能够发现的问题。这是一个非常好的故事，对吧？就像，我认为需要认识到的是，我们将会有一个下一个模型。

> And we found a number of issues. We fixed it. And at some point, we saturated the issues that this model was capable of finding. And that's a very good story, right? It's like, I think that the thing to realize is that we're going to have a next model.

[31:32](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1892s) **Greg Brockman** 我们将不得不重复这个过程。而这将会继续。但在某个时候，再次，我认为如果你想想，正在编写的代码，我们现在实际上有那么多是 AI 辅助的。我们实际上可以训练这些模型。

> We're going to have to repeat that process. And that that will continue. But at some point, again, I think if you think about, well, the code that's being written, we're actually so much of it is now AI-assisted. We can actually train these models.

[31:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1904s) **Greg Brockman** 而我们正在训练这些模型从一开始就编写安全的代码。所以我认为，通过以正确的方式使用和应用这些 AI 模型，我们可以拥有从根本上更安全的系统。

> And we are training these models to write secure code from the beginning. And so I think we can have sort of fundamentally more secure systems through the use and application of these AI models in the right way.

[31:54](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1914s) **Elad Gil** 所以听起来基本上会有一段时间，也许我们会看到更多这类网络攻击、更多事件等等，因为人类或人们正在逐步赶上，让他们的代码库变得越来越安全，使用更多 AI 生成的代码，而这些代码预设为更安全。

> So it sounds like basically there's going to be a period of time where maybe we're going to see more of these cyber attacks, more incidents, et cetera, as humanity or as people sort of catch up in terms of their code bases getting more and more secure, using more AI-generated code that's sort of preset to be more secure.

[32:10](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1930s) **Elad Gil** 因此，随着时间的推移，我们可能最终会经历一段过渡期，在这段时期里攻防都会更多，然后防御会赶上。我的意思是，这大概就是我在整个行业里一直听到的共识。

> And so we'll probably end up with some transitionary period over time where we'll have more offense and defense and then defense will catch up. I mean, that's kind of the consensus I've been hearing in the industry in general.

[32:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1939s) **Elad Gil** 而且至少这似乎是最有可能的前景。

> And it seems like the most likely forward case, at least.

[32:22](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1942s) **Greg Brockman** 是的，你看，我对我们如何度过这个时刻非常乐观。我认为这确实需要协调一致的应对，对吧？它确实需要防御者内化网络安全的基本规律正在如何变化，确实存在这样一个窗口期，而这个窗口期不会永远持续下去，对吧？

> Yeah, look, I have a lot of optimism about how we can navigate this moment. I think it does require a coordinated response, right? It really requires defenders to internalize how the physics of cybersecurity are changing, that there is this window where that window is not going to last forever, right?

[32:42](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1962s) **Greg Brockman** 而且你应该这样想：我们正在转向这样一个世界，在这个世界里，我们组织的安全性既与他们已经做对的基本功有关——这算是一种一次性投资——也与他们用来防御自己的网络能力水平有关。

> And that you should think about we're shifting to this world where the security of our organization is related to both the fundamentals that they've gotten right, which is kind of a one-time investment, but also the level of cyber capability they've used to defend themselves.

[32:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=1978s) **Greg Brockman** 随着我们转向那个世界，我认为我们可以处在一个防御者能够超越攻击者的世界里。但这需要大量的改变，对吧？这需要我们在思考每个组织如何运作的方式上发生某种改变，也需要我们共同思考关键基础设施，并为那些迄今尚未拥有这些资源的人提供资源。

> And as we shift to that world, I think that we can be in a world where defenders are able to outstrip attackers. But it is going to take a lot of change, right? It's going to take a sort of change in terms of how we think about every organization operating and how we collectively think about critical infrastructure and resourcing those that have sort of not had those resources to date.

### [33:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2001s) 安全投资的转变
*Shift in Security Investment*

[33:21](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2001s) **Sarah Guo** 我觉得你的例子，我的意思是，你对这个的描述就像是，比如说，互联网范围的技术债务。然后你在内部对此的回应非常好。因为我会想，好吧，我参与过一些公司，在那里技术债务非常难以处理。

> I think your example, I mean, your description of this is like, let's say, like internet-wide tech debt. And then your response to it internally is a really good one. Because I'm like, OK, I've been part of companies where tech debt is very painful to address.

[33:34](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2014s) **Sarah Guo** 而要说，我们要拿出 25% 的工程师，或者天哪，我们要冻结功能开发一个季度，直到我们修好一些确实需要修好的东西。这感觉还不像是世界已经决定采取的状态。

> And to say, we're going to take 25% of the engineers or God forbid, we're going to do a feature freeze for a quarter until we fix some stuff that really needs to be fixed. That doesn't feel like the state that the world has decided on quite yet.

[33:50](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2030s) **Sarah Guo** 当你思考整个世界从安全投资的角度处于什么位置时，传统上，人们认为大概是 IT 预算的低个位数百分比。那么，在这个时刻，你认为应该是多少？

> When you think about where the entire world is from a security investment perspective, traditionally, it's thought of as like, oh, maybe it's like low single-digit percent of the IT budget. Like, what do you think it should be at this moment of time?

[34:04](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2044s) **Sarah Guo** 还有，是不是你从网络工具、网络、终端、传统企业安全角度投资的所有其他东西都变得不那么相关了？还是我们只需要花两倍的钱？

> And like, is it all of this other stuff that you were investing in from a cyber tooling perspective, network, endpoint, traditional corporate security is less relevant? Or we just need to go spend twice as much?

[34:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2057s) **Greg Brockman** 我有一种感觉，网络支出、安全支出实际上应该远远超过软件支出。我真的这么认为。

> I have a feeling that cyber spend, security spend should actually vastly outstrip software spend. I really think that.

[34:25](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2065s) **Sarah Guo** 哇，这话说得够大胆。

> Well, that's a statement.

[34:26](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2066s) **Greg Brockman** 是的，我真的相信。我可能不对。但当我展望我们将要进入的世界时，软件在某种程度上正在成为准入门槛。正是如此。

> Yeah, I really believe it. And I may not be right. But when I look at the world we're going to be in, where software in some ways is becoming like the barrier to entry. Exactly.

[34:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2079s) **Sarah Guo** 近乎免费。

> Free-ish.

[34:39](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2079s) **Greg Brockman** 正是如此。它变得如此便宜。

> Exactly. It's becoming so cheap.

[34:40](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2080s) **Sarah Guo** 是的。

> Yeah.

[34:40](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2080s) **Greg Brockman** 所以当然，确保你不只是想要软件，对吧？你想要的是可靠的、安全的、你可以信赖的软件。所以对我来说，那种软件的韧性就是那种。

> And so of course, making sure you don't just want software, right? You want software that is reliable, that is secure, that you can believe in. And so to me, the resilience of that software is the sort of.

[34:53](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2093s) **Sarah Guo** 那就是价值所在。

> That's the value.

[34:54](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2094s) **Greg Brockman** 正是如此。那将变得更加重要。所以我认为我们拥有这个时刻。我们拥有这个机会，你知道，就像你说的那个 XKCD 漫画，底部有一个小小的方块，画着某个开源维护者。

> That's exactly. That's going to become even more important. And so I think that we have this moment. We have this opportunity to, you know, to your point of the XKCD with the little block at the bottom of some open source maintainer.

[35:03](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2103s) **Sarah Guo** 是的。

> Yeah.

[35:03](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2103s) **Greg Brockman** 就像，那从来都说不通。我们之所以有那个世界，是因为某些基本规律，对吧？也就是写软件和写可靠软件有多难。而且每个人都有这条供应链。

> Like, that never made any sense. And the reason we had that world was because of some of this physics, right? Of kind of how hard it is to write software and to write reliable software. And that everyone has this supply chain.

[35:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2117s) **Greg Brockman** 而审计自己供应链的公司少之又少，对吧？我们已经看到过供应链漏洞，有人上传了一个恶意软件包，然后一些公司中招。所有这些都需要改变。而用人类软件开发者来改变是完全不可能的，因为人太少了，对吧？

> And so few companies audit their supply chain, right? If there's we've seen this with supply chain vulnerabilities where someone uploads a malicious package and a number of companies get hit. All of that needs to change. And it was totally impossible to change with human software developers because there's so few, right?

[35:30](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2130s) **Greg Brockman** 拥有这种专长的人太少了，这涉及非常冷门的东西，比如了解不同的软件包和不同的安全技术。

> There's so few people who have this expertise in this very obscure thing of like sort of knowing different packages and like different security techniques.

[35:38](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2138s) **Elad Gil** 现在像 Socket 和其他公司这样的第三方软件会做这件事，举个例子，它们会扫描你的开源供应链，试图发现漏洞并为你识别出来。或者还有人为因素，比如某个维护者突然接管了一个项目，然后安装了某种东西，相对于之后使用那个软件包的人来说是恶意的。

> Third-party software now like Socket and others will do this as an example where they'll scan through your open source supply chain and try and find vulnerabilities and identify them for you. Or there's the human element of it where a maintainer would suddenly take over a project and then install something that would be malicious relative to people then using that packet.

[35:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2158s) **Elad Gil** 所以看到第三方生态系统也在发展，以支持 OpenAI 和其他公司在这个市场上所做的努力，这真的很有趣。

> So it's really interesting to see that shift in terms of the third-party ecosystem as well developing to support the efforts that OpenAI and others are doing in terms of this market.

[36:06](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2166s) **Greg Brockman** 正是如此。而且我认为这非常重要。所以我们正在与社区一起做这件事。我们正在与 CrowdStrike 这样的公司合作。我认为这同样是协调一致的应对。但我们确实把它视为 OpenAI 引领的核心优先事项，对吧？

> Exactly. And I think it's so important. So we're doing this with the community. We're partnering with companies like CrowdStrike. And I think that it is, again, a coordinated response. But we really view it as a core priority for OpenAI to lead, right?

[36:17](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2177s) **Greg Brockman** 对我们来说真正因为拥有前沿模型。我们拥有巅峰智能模型。当然，在网络安全领域，我们认为向善的变革潜力是存在的。但这意味着我们作为一家公司，我们一直在经历一场安全转型，对吧？

> For us to really because we have the frontier model. We have the peak intelligence model. Certainly in cybersecurity, we think that there is transformative potential for the good possible. But that means that we as a company and we've been going through a security transition, right?

[36:31](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2191s) **Greg Brockman** 在过去几个月里，我们谈到了我们开发流程的安全性、安全性和对齐。但我们也在为整个世界思考这一点，即如何真正帮助世界变得更根本安全，并帮助人们度过整个时期。

> That over the past couple of months, we've talked about safety security alignment of our development process. But we're also thinking about that for the world of how to really help the world become more fundamentally secure and help people through this whole period.

### [36:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2204s) OpenAI 以人为本的理念
*OpenAI’s Human-Centric Ethos*

[36:44](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2204s) **Elad Gil** 关于 OpenAI 的一点是，它一直非常亲人类，非常支持人类能动性和人类进步。让我们用 AI 来让每个人变得更好。我认为 OpenAI 最初的 premise 是，让我们拥有某种东西，能够在创造可以非常广泛和普遍地以积极方式使用的 AI 方面，与其他公司形成制衡。

> One of the things about OpenAI is that it's always been very pro-human and very pro-human agency, human progress. Let's use AI to better everyone. I think the original premise of OpenAI was let's have something that can counterbalance other companies in terms of creating AI that can be used very broadly and universally in a positive way.

[37:01](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2221s) **Elad Gil** 这种精神在 OpenAI 延续下来了吗？或者你在 AI 社区的背景下如何看待这一点？

> Has that ethos continued at OpenAI? Or how do you think about that in the context of the AI community?

[37:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2225s) **Greg Brockman** 我认为这种精神是我们使命的核心，对吧？确保 AGI 造福全人类。而这里的“人类”非常关键。我们认为人类应该牢牢坐在驾驶座上，应该掌控未来。

> I think that ethos is very core to our mission, right? To ensure AGI benefits all of humanity. And the humanity there is very key. We think that humans should be firmly in the driver's seat, should have control over the future.

[37:16](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2236s) **Greg Brockman** 而且这项技术应该放大人类的能动性。

> And that this technology should amplify human agency.

[37:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2239s) **Sarah Guo** 即使模型在能力上超越人类。

> Even as models surpass humans in capability.

[37:22](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2242s) **Greg Brockman** 而且我认为这绝对至关重要，对吧？我认为，要记住，这些模型是非常参差不齐的，对吧？有些地方显然这些 AI 比今天任何人都更有能力，对吧？甚至 GPT-4 知道的也比任何人都多得多。

> And I think this is absolutely critical, right? That I think that the picture of and keep in mind, these models are very jagged, right? That there are places where obviously these AIs are more capable than anyone today, right? Even GPT-4 knows far more than any person.

[37:36](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2256s) **Greg Brockman** 但非常重要的一点是，我想起了，比如，有个故事讲Steve Jobs为什么在 iMac 上装了一个把手，对吧？这样

> But the thing that is so important and I'm reminded, for example, of the there's the story of why Steve Jobs put the handle on the iMac, right? So that that way

[37:45](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2265s) **Greg Brockman** 人们就能感觉到他们可以控制这台机器，对吧？你总能把它拿起来扔出窗外。我认为同样的精神需要适用于 AI。人类坐在驾驶座上的精神，人类判断、人类价值观。

> that people could feel that they could control the machine, right? That you can always pick it up and throw it out the window. And I think that the same spirit needs to apply to AI. The fact of the spirit of humans are in the driver's seat, that these human judgment, human values.

[38:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2285s) **Greg Brockman** 也许你应该思考未来人类在某些领域和某些维度上将如何运作，不管是什么，我们会变得更像 CEO 或董事会成员。但重要的是我们要有那种程度的能动性。

> And maybe you should think of how humans will operate in certain fields and certain dimensions in the future, whatever it is, is we become more like CEOs or like board members. But it's important that we have that level of agency.

[38:19](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2299s) **Greg Brockman** 而且我认为我们将建造的机器就像，你知道，如果你是 CEO，你会有领域专家。你会有团队，他们会在如此多的事情上拥有知识，甚至可能在许多不同的维度上比你更有想法。

> And that I think that we're going to build machines just like, you know, if you're a CEO, you're going to have domain experts. You're going to have teams that are going to have knowledge across so many things and maybe even better ideas than you on many different axes.

[38:34](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2314s) **Greg Brockman** 但根本上，问题是，你开心吗？这是你想要的吗？这是让你的生活变得更好的东西吗？对我来说，这是我们所有人都应该共同思考的，这不仅仅是关于 AI。这是关于技术，对吧？技术是为了提升人类文明。

> But fundamentally, the question is, are you happy? Is this what you wanted? Is this what makes your life better? And that to me is something that we should all collectively it's not just about AI. It's about technology, right? Technology is to lift up human civilization.

[38:50](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2330s) **Greg Brockman** 它是改善和生活质量的最根本驱动力。而这正是我们应该从 AI 中寻求的。这就是 OpenAI 的信念。

> It's the single most fundamental driver of improvement and quality of life. And that is what we should be seeking from AI. That is the belief of OpenAI.

### [38:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2338s) 结语
*Conclusion*

[38:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2338s) **Sarah Guo** 非常感谢你来做这件事，Greg。

> Thank you so much for doing this, Greg.

[38:58](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2338s) **Elad Gil** 是的，感谢你抽出时间。

> Yeah, thanks for your time.

[39:01](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2341s) **Greg Brockman** 谢谢你们两位。非常感谢。

> Thank you both. Appreciate it.

[39:05](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2345s) **Sarah Guo** 在 Twitter 上找到我们：No Priors Pod。如果你想看到我们的面孔，请订阅我们的 YouTube 频道。在 Apple Podcasts、Spotify 或任何你收听的地方关注节目。这样你每周都能收到新一集。

> Find us on Twitter at No Priors Pod. Subscribe to our YouTube channel if you want to see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week.

[39:16](https://www.youtube.com/watch?v=AwuTTOJGEi0&t=2356s) **Sarah Guo** 并注册电子邮件或在 no-priors.com 上找到每集的文字记录。

> And sign up for emails or find transcripts for every episode at no-priors.com.

---

*来源：[研几 · AI](https://ai-daily-cn.pages.dev/watch/np-gdb/) · 原节目 [YouTube](https://www.youtube.com/watch?v=AwuTTOJGEi0)。文字稿：节目方发布的文字稿（podhood），时间点按音频语音识别逐段对齐（与 YouTube 原片时间轴一致）；章节标题取自 YouTube 章节，起点按实际对话校准；逐段译文 · 机器翻译（DeepSeek）+ 人工抽检校订*