# How to Understand the Next Wave of AI Before Everyone Else | Tibo Interview · 中英对照逐字稿

- 原节目：Matthew Berman
- 英文原始来源：https://www.youtube.com/watch?v=4qjEgPojjzM
- 中文译制版入口：https://www.xiaoyuzhoufm.com/episode/6a8db7041352af56ff3bf5e8
- 时长：00:44:28
- 方法与限制：英文来自已验证的原始节目 transcript/caption；中文由 Codex 逐段翻译，未做逐字人工校对，公开引用前请回到英文原文与音频复核。

## 中英对照逐字稿

### [00:00:00–00:00:30]

**EN**  I can press the button whenever I want, whenever it feels right. I [music] don't tend to look at the competition that much. Like I really look at, you know, what can we do uniquely well and what are our values and like, you know, how do we maximally [music] accelerate towards that? You know, maybe a year or two these speeds will become, you know, maybe if not the default, like very close to the default. And you look at the cost of Luna, right? It's phenomenal. Technology has a way to become like, you know, very, very efficient over time. [music] We're very focused on like, you know, very broad access and we're optimizing for, you know, the utility that you get out of it

**中文**  我随时都可以按那个按钮，只要觉得该按。我不太会盯着竞争对手看。我真正看的是：我们能把什么事情做得特别好，我们的价值观是什么，以及我们怎样以最大幅度朝那个方向加速。也许再过一两年，这样的速度即便还不是默认，也会非常接近默认。你再看 Luna 的成本，真的很惊人。技术总是会随着时间变得非常、非常高效。我们非常关注尽可能广泛的使用，并且在优化你能从中直接获得的效用。

### [00:00:29–00:00:57]

**EN**  directly. [music] I heard there's an actual physical button now. >> Yes, Deris. >> I will show it to you. It's like [music] very very cool. >> Tibo, thank you so much for joining me. >> Of course. Yeah, glad to be here. >> Yes, really excited to talk to you. I want to actually start with your time at Google. So you were on the deep mind team and before chat GPT Google had

**中文**  直接获得的效用。主持人：我听说现在真的有一个实体按钮了。Tibo：是的，有。我会拿给你看，非常酷。主持人：Tibo，非常感谢你来做客。Tibo：当然，很高兴来这里。主持人：是的，很期待和你聊。我想从你在 Google 的那段时间开始。你当时在 DeepMind 团队，而在 ChatGPT 出现之前，Google 已经有

### [00:00:54–00:01:22]

**EN**  something called LM chat and you had tweeted uh Google was too nervous to release it. Deep mind was blocked from shipping products that could disrupt Google and I think about that a lot. What were you thinking at that time while you were working on these products that was you know well before uh chat GPT really changed the world. >> Yeah it was a very exciting time. So um deep mind was a very creative place. I was mostly focused on um my specialty

**中文**  一个叫 LM Chat 的东西。你还发过推文说，Google 太紧张了，不敢发布它。DeepMind 被拦住，不能推出可能冲击 Google 的产品。这件事我经常会想。当时你在做这些产品时是怎么想的？那可是在 ChatGPT 真正改变世界之前很久。Tibo：那是一段非常令人兴奋的时间。DeepMind 是一个非常有创造力的地方。我当时主要专注于我的专长

### [00:01:20–00:01:49]

**EN**  was infrastructure and products for to accelerate research. And so there was obviously a group working on language models uh and scaling that. And then they had like gotten pretty good results and then it was like a natural thing to think about hey you know can you turn that into something that you know you can chat to and you know can use um you know for various things. And so then naturally like the idea of like something like LM chat sort of emerges

**中文**  是加速研究的基础设施和产品。当时显然有一个组在做语言模型，以及把它做大规模。他们拿到了相当好的结果，接下来很自然就会想：能不能把它做成一个你可以聊天、可以拿来做各种事情的东西。于是类似 LM Chat 这样的想法就自然出现了。

### [00:01:48–00:02:15]

**EN**  um and then it was internal and then there was sort of this ambition to um make it into a publicly available uh tool. >> What year was this? >> And this was it was like a year before Chhatra roughly. >> Okay. >> Yeah. >> Um and then but we were also building all sorts of other things that I'm not going to talk about but it was a very creative place. Uh and then just deep mind was not set up to ship product. uh

**中文**  一开始是内部的，后来也有一种野心，想把它做成一个对公众开放的工具。主持人：那是哪一年？Tibo：大概是 ChatGPT 出现前一年左右。主持人：好。Tibo：对。我们当时也在做各种各样其他的东西，那些我就不说了，但那是一个非常有创造力的地方。问题是，DeepMind 并没有被设置成一个能把产品真正发出去的组织。

### [00:02:13–00:02:41]

**EN**  is open AI is a very very different place in that sense. Um we just like research and products just collaborate super closely together. We ideate together. We co-design a lot of things. We have a big bias to ship uh and uh also a big bias towards you know making things available for people which I really love and this is sort of like what uh drove me here the mission the people um the talent density. I mean there's so many great things about open

**中文**  OpenAI 在这一点上是一个非常、非常不同的地方。研究和产品紧密协作。我们一起想点子，一起共同设计很多东西。我们有很强的发布倾向，也有很强的倾向，要把东西交到人们手里。我非常喜欢这一点。这也是把我带到这里来的原因：使命、人，还有人才密度。OpenAI 真的有太多很好的地方。

### [00:02:39–00:03:08]

**EN**  eye really. Did you know at the time where you were involved in LM chat that was something special or would become something special? >> It it felt it felt very special like the models were you know it's sort of like the first time um you know you realized like you could get coherent text and you know something helpful. Initially it was more funny than helpful and then gradually it became more and more helpful. >> So you say you think about that often and I I understand that I think you know

**中文**  主持人：你当时参与 LM Chat 的时候，有没有意识到那是一件特别的事，或者说会变成一件特别的事？Tibo：当时感觉就非常特别。那差不多是你第一次意识到，模型可以生成连贯的文本，而且是有用的东西。一开始它更搞笑，而不是更有用，然后逐渐变得越来越有用。主持人：你说你会经常想起这件事，我能理解。我觉得

### [00:03:06–00:03:35]

**EN**  in a lot of ways Google got in their own way. um what are some of the lessons that you learned there that you took to OpenAI? >> Yeah, so this is why I think about it often. I I think about it in terms of like the culture that I have on the team uh the culture of OpenAI itself and so like the good parts to preserve and like you know what not to do. Um, OpenAI has a very bottoms up uh culture. Like it's a very empowering culture. Like people can come up with all sorts of ideas and get together and then very very quickly

**中文**  在很多方面，Google 是自己挡住了自己。你在那里学到了哪些教训，后来带到了 OpenAI？Tibo：对，这就是我经常会想这件事的原因。我会从我带团队的文化、以及 OpenAI 本身的文化来想：哪些好的部分要保住，哪些事情不要做。OpenAI 有非常自下而上的文化，是一种很赋能的文化。人们可以提出各种各样的想法，聚在一起，然后非常非常快地

### [00:03:33–00:03:59]

**EN**  ship something and there's very little stop energy uh in general for like new product ideas which is exhilarating and fun and you know it's all about impacting the world in positive ways. Uh and so preserving that is very important to me. The other thing that is also important is like to not make it a mess, right? So you don't want to have like a hodgepodge of like no features and like no overall direction and coherence and

**中文**  把东西发出去。对新的产品想法，总体上几乎没有多少叫停的能量。这让人振奋，也很好玩，而且这一切都是为了以正面的方式影响世界。所以把这种东西保住，对我非常重要。另一件同样重要的事是：不要搞成一团乱。你不会希望功能杂乱堆在一起，也没有整体方向和一致性。

### [00:03:57–00:04:26]

**EN**  so it's counterbalanced with the sense of um simplicity and you being proud about the quality of the product. Uh I think the chat GBT iOS app is like you know one of the best apps out there. Uh and so we want we want to keep that. Uh we're investing a lot in you know things like delight performance efficiency simplicity. So there's these overall principles while still empowering everyone everyone to like try new things and ship very quickly. >> If you were to give advice to a founder

**中文**  所以要用简洁感、以及你对产品质量感到自豪，来作为对冲。我觉得 ChatGPT 的 iOS 应用是市面上最好的应用之一。我们想把这一点保住。我们在愉悦感、性能、效率、简洁这些事情上投入很多。所以有这些总体原则，同时仍然赋能每一个人去尝试新东西，并且非常快地发布。主持人：如果你要给创始人一些建议

### [00:04:24–00:04:53]

**EN**  about how to develop that kind of culture uh like what are some of the more tangible elements or practices that occur inside OpenAI that can kind of gives give advice uh to a founder? Yes, I think having having conviction um and finding a way to be to have users and iterate very quickly from feedback and then also uh being willing to disrupt

**中文**  关于怎样培养那种文化，OpenAI 内部有哪些更具体的要素或做法，可以拿来给创始人当建议？Tibo：是的，我认为要有信念，要找到办法真正接触到用户，并根据反馈非常快地迭代，然后还要愿意自我颠覆。

### [00:04:51–00:05:20]

**EN**  yourself that is not as much relevant for founder but is relevant for you know companies like openi like we come up with new research new ideas all the time and being able to identify when is the right moment to go and invest in them even though it means like maybe reallocating resources from you know the main gig uh super important but it's very hard but it's super important to be able to do that. >> Yeah. I mean that's the exact thing that you were describing at Google. They kind of weren't able to do that. Um that's great. I mean does that

**中文**  自我颠覆对创始人来说没那么相关，但对像 OpenAI 这样的公司很相关。我们一直会有新的研究、新的想法，要能识别出什么时候是该投入进去的正确时机，哪怕这意味着可能要从主业上重新调配资源。这件事超级重要，但很难，而能够做到这一点又超级重要。主持人：对，这正是你刚才在说 Google 的那件事。他们当时有点做不到。这很好。那这件事

### [00:05:18–00:05:46]

**EN**  >> they have a plan to be fair. It's like you know it was all it was all part of a big plan. Um but to me it wasn't it wasn't uh it wasn't the right place. at OpenAI or or any company as it matures does that become more difficult to maintain that kind of culture of shipping and willingness to disrupt yourself especially when you you know if you have a cash cow just printing money and you have this other new thing over here that might be something cool and

**中文**  Tibo：公平地说，他们是有计划的。那一切都是一个大计划的一部分。但对我来说，那里不是合适的地方。主持人：在 OpenAI，或者任何一家公司成熟之后，要维持这种持续发布、愿意自我颠覆的文化，会不会变得更难？尤其是当你有一头正在大量赚钱的现金牛，而这边又有一个可能很酷、很有创新的新东西。

### [00:05:44–00:06:14]

**EN**  innovative. We are a very very forward-looking um and the the the future of AI and what it will all look like and how humanity benefits doesn't really wait or doesn't really care for you know whatever you have established here you know over the next month or 3 months and so you know I think it's very important to lean in um and to you know just be like openeyed about where it's all going

**中文**  Tibo：我们是非常、非常向前看的。AI 的未来会是什么样、人类会怎样从中受益，并不会等你，也不会在乎你这边这个月或三个月里已经建立了什么。所以我认为很重要的是要靠上去，并且对整件事要走向哪里保持睁着眼睛看清楚。

### [00:06:12–00:06:42]

**EN**  >> and then you know figure out like how to position yourself. So that you know you you do catch that wave. Um you know even even for open it's like we we train models and then we discover their capabilities like we don't benchmarks don't tell you everything. We have to play quite a bit with the models themselves to sort of like realize uh it's like oh you know maybe we haven't thought about you know benefiting from it in like this specific way or like oh it can do this. Um and then you're just like oh that I mean that's a shift in like you know how we think

**中文**  然后想清楚怎样给自己定位，这样你才能真正赶上那一波。即便对 OpenAI 也是这样：我们训练模型，然后才发现它们的能力。基准测试并不能告诉你一切。我们必须自己多玩这些模型，才会意识到：哦，也许我们还没想过可以以这种具体方式从中获益，或者，哦，它还能做这个。然后你就会觉得：哦，这意味着我们思考产品的方式发生了转变。

### [00:06:40–00:07:09]

**EN**  about the product. for example, like right now like you know we we we launched the new voice uh voice and it's super delightful to talk to. Uh it's very natural now. It's capable of tool use as well. >> Yeah. >> And that changes things like now I spend a lot more time just talking to it. Um another thing I I do all the time is like dictation because the quality of the dictation is like so so good and it's like much more efficient as a as a way instead of like typing the prompt. And so in the morning I just like I sit

**中文**  比如说，我们刚刚推出了新的语音，跟它说话非常愉快。现在已经非常自然了，而且也能做 tool use。主持人：对。Tibo：这会改变事情。我现在花更多时间直接跟它说话。另一件我一直在做的事是口述，因为口述质量实在太好了，比起把 prompt 打出来，这种方式高效得多。所以早上我就坐在那里

### [00:07:07–00:07:37]

**EN**  there with my phone and I'm like blah blah. It's just like you know couple of things to do for charge and um and then it just goes and like does it has access to all my tools. >> Yeah. >> Um and that is like that was not possible before we had like really good voice models and so that completely changes in suddenly how you think about the product. >> Yeah. Let's continue talking about new models, new harnesses. Um a few weeks ago I'm going to start with another one of your tweets because you know these are bangers. Uh [laughter] codeex will seem primitive in two to three months. We're about to go through another major

**中文**  拿着手机，就那样说一通。就是给 ChatGPT 交代几件要做的事，然后它就自己去干，而且能用上我所有的工具。主持人：对。Tibo：在我们有真正好的语音模型之前，这是不可能的，所以这会一下子彻底改变你对产品的想法。主持人：对。我们继续谈新模型、新的 harness。几周前——我又要从你的另一条推文开始，因为这些推文真的很猛。Codex 再过两三个月就会显得原始。我们即将经历又一次重大

### [00:07:35–00:08:05]

**EN**  evolution. The next generation of models need more than your laptop. Um, what areas, let's start with the harness first. What areas of the harness are still ripe for innovation as a model gets better? >> Yeah, so many so many. Um, so I talked about voice like one thing that um, right now if you're like a sophisticated user of of of Codex and you know any other coding agent is you sort of have gotten used to a little bit of the

**中文**  进化。下一代模型需要的，已经超出你的笔记本电脑。我们先从 harness 开始。随着模型变强，harness 的哪些方面仍然很适合创新？Tibo：太多了，太多了。我刚说了语音。现在如果你是 Codex 或其他 coding agent 的资深用户，你多多少少已经习惯了那种有点

### [00:08:03–00:08:31]

**EN**  clunkiness, right? So you know you have to manage skill files and you know this is like a way to sort of like teach it stuff but it's also I think a lot of people have realized it's kind of like hard to maintain over time. Um the memory is sort of like a thing but it doesn't always remember uh everything like if you if you have sub agents you have to care about sub agents and it's like sort of like constructs a little network and the illusion kind of gets

**中文**  笨拙。你得去管 skill files，这是一种教它东西的方式，但很多人已经意识到，时间一长很难维护。记忆也算有，但它并不总是能记住所有事情。如果你有 sub agent，你还得去管这些 sub agent。它会构成一个小小的网络，而那种幻觉会在

### [00:08:28–00:08:56]

**EN**  broken into v in in in at various parts um when you interact with it. And really what you want is just something that deeply understands you, understands your goals, understands your day-to-day, understands what you know your team is up to as well. And then optimally is sort of like reacts and also is proactive uh and just helps you in your day-to-day >> and doesn't break that illusion right of like that's this perfect little partner uh that you have and so that's what

**中文**  互动中的各个环节被打破。你真正想要的，只是一个深刻理解你、理解你的目标、理解你的日常、也理解你的团队在做什么的东西。最好它既能反应，也能主动，就这样帮你过每一天。而且不要打破那种幻觉：这就是你身边那个完美的小搭档。而这正是

### [00:08:54–00:09:23]

**EN**  we're uh working towards. Another thing that you know you realize when you have very very powerful models is that your laptop kind of becomes a constraint in and by itself. um you know the amount of work that you can do on a laptop it was designed for humans right so it's designed roughly to be able to absorb the amount of work that you know you can produce or you know how fast you can type and how fast you could think you know how many applications you need open all these things are human constraints

**中文**  我们正在努力的方向。另一件你会意识到的事是：当你有了非常、非常强的模型，笔记本电脑本身就变成了一种约束。你能在笔记本电脑上做的工作量，它本来就是为人类设计的。它大致是按你能产出多少、你打字有多快、你思考有多快、你需要同时打开多少个应用来设计的。这些都是人类的约束。

### [00:09:22–00:09:50]

**EN**  um the model doesn't have the same constraints the model can you know for example handle you know a 100 applications opened at the same time perfectly fine you know maybe in the future and so in terms of access to resources is it's very clear that you know models of the future will need access to more than the resources of your laptop. Do you I mean I I'm guessing you're talking about cloud agents and and all of a sudden like you know when you have things like ultraast which we're going to talk about in a

**中文**  模型没有同样的约束。模型比如说可以同时处理好 100 个打开的应用，完全没问题，也许将来会是这样。所以在资源获取上，很明显，未来的模型会需要用到比你笔记本电脑更多的资源。主持人：我猜你说的是 cloud agent。一旦有了我们稍后会谈的 Ultra Fast 这类东西，

### [00:09:48–00:10:16]

**EN**  little bit when you have token speeds that are 10 I think 14 is the the stated number 10 14 times faster than what fast is um the the bandwidth changes or sorry the bandwidth constraint changes uh the CPU now becomes the bandwidth like literally tool calls >> network tool any kind of overhead in the stack becomes the the limiting factor.

**中文**  当你的 token 速度达到 Fast 模式的 10 倍——官方数字好像是 14 倍，也就是 10 到 14 倍——带宽就会变，或者说带宽约束会变。这时 CPU 本身就变成了带宽，字面意义上的 tool call、网络、工具，以及栈里的任何开销，都会变成限制因素。

### [00:10:14–00:10:43]

**EN**  But then you know you can compensate by doing multiple things uh concurrently as well. And so you can you can think about you know having like maybe you know exploring on one end writing tests as well compiling uh you know testing a new hypothesis like all at once. And so then you're not then you're you're shifting the bottleneck around because you know you're able to do more uh concurrently and then you know the

**中文**  但你也可以用并发做多件事来补偿。你可以想成：一边探索，一边写测试，一边编译，一边验证一个新假设，全部同时进行。这样你就是在把瓶颈挪来挪去，因为你能并发做更多事情。然后

### [00:10:42–00:11:11]

**EN**  model can like sort of like think very efficiently and very quickly through it. >> With current token speeds I find myself kicking off 10 15 agents in parallel and that becomes a pretty significant cognitive overhead for me to do that context switching and just constantly cuz you're kicking it off and you can expect 30 45 minutes before my task comes back. Now with ultra fast speed that workflow changes significantly and I don't think I would be able to have 10

**中文**  模型就能非常高效、非常快地把这些事情想清楚。主持人：以现在的 token 速度，我发现自己会并行启动 10 到 15 个 agent，而这对我来说会变成相当大的认知负担，因为要不断做上下文切换。你把它启动之后，大概要等 30 到 45 分钟任务才会回来。现在有了 Ultra Fast 的速度，这个工作流会显著改变，我不认为我还能同时带着 10

### [00:11:09–00:11:38]

**EN**  or 15 agents and that might be a good thing. Maybe it's three or four at a time. How do you see the the workflow of a solo developer changing over time? >> Yeah. So I think managing your attention and being much more, you know, friendly to your attention is something that we care a lot about. Like after all, like we're trying to build for humans. We're trying to be like >> the build the technology that's the most empowering for humans and that requires

**中文**  个或 15 个 agent，而这也许是好事。也许一次只有三四个。你怎么看独立开发者的工作流会随时间怎样变化？Tibo：我认为，管理你的注意力，并且对你的注意力更友好，是我们非常在意的事。毕竟我们是在为人类构建。我们想做的是最能赋能人类的技术，而这要求

### [00:11:35–00:12:05]

**EN**  building around you know your ability to multitask and you know how do you want to manage your attention and do you want something brought up now or is it better to bring it up in 30 minutes um and then when you have ultra fast speeds combined you know maybe with voice is like suddenly you're like okay you know like this thing can operate at the same speed if not faster than you and so you stay in the flow you get to ideulate you get to see prototypes, you know, like you get to build little reports like in real

**中文**  围绕你并行处理任务的能力来构建：你想怎样管理注意力，你是想现在就把某件事提出来，还是 30 分钟后再提出来更好。当你有了 Ultra Fast 的速度，也许再结合语音，突然间你会觉得：好，这东西能以和你一样快、甚至更快的速度运转。于是你能保持心流，可以一起构思，可以看原型，可以实时做出一些小报告。

### [00:12:02–00:12:31]

**EN**  time and that, you know, sort of like that just feels really good. Um, suddenly you're like, oh yeah, like what I was doing before, multitasking like 10 agents, it's like I don't want to really go back to that. >> Yeah. >> Uh, and so we're trying to bring that sort of experience that is just really natural but also feels built for you, you know, where you don't have to adapt. The technology adapts to you. So there's been a number of I guess agentic coding techniques discussed over the last few

**中文**  那感觉真的很好。突然间你会觉得：对，我以前同时管 10 个 agent 那种多任务，我其实不太想回去了。主持人：对。Tibo：所以我们想带来的就是那种非常自然、而且感觉是为你打造的体验：你不必去适应，是技术来适应你。主持人：过去几个月，大家讨论了不少 agentic coding 的手法。

### [00:12:29–00:12:56]

**EN**  months. Loops was popular, still is popular. Now I'm hearing about graphs. Are are these all techniques to just allow the solo developer to manage or be friendly to their to their attention as you said? I like that term. >> Yeah. So I I think about two different categories of of problems. It's like the first one is building the very best

**中文**  Loops 曾经很流行，现在也还流行。现在我又听到 graphs。这些是不是都只是为了让独立开发者能够管理注意力，或者说像你说的那样，对注意力更友好？我喜欢这个说法。Tibo：对。我会把问题分成两类。第一类是打造最好的

### [00:12:53–00:13:21]

**EN**  personal AGI or the personal agent that will be in the flow with you proactive raise important new ideas uh when it can find some be very very efficient at doing exactly what you want. It doesn't matter whether it's a technical problem or you know it's just more like research or advice like it can do it all and it's like super super tailored to you. This is like a very important thing and it's

**中文**  个人 AGI，或者说那个会和你一起处在心流里的个人 agent：它会主动提出重要的新想法，当它能找到的时候；它会非常、非常高效地去做你真正想做的事。不管是技术问题，还是更像研究或建议，它都能做，而且超级、超级贴合你。这是一件非常重要的事，而且它

### [00:13:19–00:13:48]

**EN**  like deeply rooted in like you know the understanding of you as a human, you as like an individual that is unique. That's one category of problem like we're pushing super hard on that. The other category of problem is like full-on automation. Um you know where you're more building intelligent systems that can take care of like a very complex process. You know maybe something that did require you know does require intelligence and seems like very complex. for example, you know, going

**中文**  深深根植于对你作为一个人、作为一个独特个体的理解。这是一类问题，我们正在非常用力地往前推。另一类问题是完全的自动化：你更多是在构建智能系统，去接管一个非常复杂的流程。也许那件事过去确实需要、现在也确实需要智能，而且看起来非常复杂。比如说，去

### [00:13:46–00:14:15]

**EN**  and looking at production logs and automatically doing performance optimizations or looking at regressions and automatically patching them. In cyber security, we're seeing this as well where it's like you have, you know, something you have a scanner that comes up with a vulnerability like can you automatically patch it and reduce the window um where you have that open vulnerability to like almost zero >> with without a human in those loops >> without a human in the loop or like you

**中文**  看生产日志并自动做性能优化，或者看 regressions 并自动打补丁。在网络安全里我们也看到这一点：你有一个扫描器发现了漏洞，能不能自动打补丁，把那个漏洞敞着的窗口缩到几乎为零。主持人：这些循环里没有人？Tibo：循环里没有人，或者

### [00:14:13–00:14:42]

**EN**  know very very minimal where you know you only need to approve >> uh a high risk action and it's like mostly an automated system >> but it's also not that much you it's not as important for you to be in direct control of it. >> Okay. >> Um and then so I want to slightly change topics and you know chat GPT and codeex have been on this merge path >> over the last few months. So I I guess first I just wanted to ask you how's

**中文**  人非常非常少，你只需要批准高风险动作，它基本上是一个自动化系统。而且对你来说，直接控制它也没那么重要。主持人：好。我想稍微换个话题。过去几个月，ChatGPT 和 Codex 一直在走合并这条路。所以我首先想问你，进展

### [00:14:41–00:15:10]

**EN**  that been going like how does it feel internally? What's the feedback you've been getting from your customers? >> Um it's it's really been a boon. Uh so the feedback we had initially was like why why do you merge them? It's like you know do you really have to do it? And it's like well the the future the our future models want us to be merged. So um you know we're we're just going to do it because it is the simple and proper thing to do where we're building this

**中文**  怎么样？内部感觉如何？客户给了你什么反馈？Tibo：这真的是一件大好事。一开始的反馈是：你们为什么要合并？真的必须做吗？而答案是：未来，我们未来的模型希望我们合并。所以我们就是要做，因为这是简单而正确的事。我们正在构建的是

### [00:15:07–00:15:36]

**EN**  very personal super capable agent that can help you in all sorts of ways. This is the same this is going to be the same technology under the hood. Um it's the same harness. It's the same way that we think about it. It's like highly multimodal, you know, voice first, uh, super efficient and and it doesn't matter if you're trying to code or not. Like this this agent is capable of it all. And it's like the high it's like the most efficient at it. And then the interface that you want is like it

**中文**  一个非常个人化、能力极强、能以各种方式帮你的 agent。底层会是同一套技术。是同一个 harness，也是我们同一套思考方式。它是高度多模态的，语音优先，超级高效。你是不是在写代码并不重要，这个 agent 什么都能做，而且在这件事上效率最高。然后你想要的界面是：它

### [00:15:33–00:16:02]

**EN**  should tailor itself to your needs. If you shouldn't decide like you know I'm a coder, I want a coder interface or like I'm not technical, I want a nontechnical interface. It's like there's a spectrum of people like you know we come up with labels of like a software engineer, a designer like you know these are just human concepts that we have invented to deal with abstractions because the reality is too complex for us to handle. But individuals are like they're individual they have their own they're somewhere on the spectrum and so we're

**中文**  应该按你的需要来自我调整。不该由你来决定：我是 coder，我要 coder 的界面；或者我不是技术背景，我要非技术界面。人是一个光谱。我们会发明软件工程师、设计师这类标签，这些只是人类为了处理抽象而发明的概念，因为现实对我们来说太复杂了。但每个个体就是个体，他们各自处在光谱上的某个位置。所以我们

### [00:16:00–00:16:29]

**EN**  trying to build the perfect interface that adapts for everyone. It doesn't matter if you're technical or not. It's just like it adapts like based on your specific indiv individuality. So that's why we went and we did this. >> But uh does that mean inevitably it's going to end up with a singular interface? No drop down selecting between products and it it's kind of wild to think that my mom might use the same exact interface as me and then obviously it'll customize to my needs.

**中文**  想构建那个能为每个人适应的完美界面。你是否偏技术并不重要，它就是按你具体的个体性来适应。这就是我们去做这件事的原因。主持人：那这是否意味着最终必然会变成一个单一界面？没有下拉菜单在产品之间切换。想到我妈妈可能和我用完全同一个界面，然后它显然会按我的需求定制，这有点不可思议。

### [00:16:27–00:16:57]

**EN**  Maybe I'll need more information if I'm doing more sophisticated work. Uh but like what is the end state for you? >> That's right. It's it's the same thing. Um so you and your mom will you know use the same thing. Uh it will be your personal AGI. You will have very different kinds of tasks and utility that you get from it. You will connect it to different tools in your life. You will bring different ideas, different needs. Uh and then it will continue to tailor itself to maximally benefit you.

**中文**  也许我在做更复杂的工作时会需要更多信息。但对你来说，终局是什么？Tibo：对，就是同一件事。你和你妈妈会用同一个东西。它会是你的个人 AGI。你们从中得到的任务和效用会非常不同。你会把它连接到生活里不同的工具上，你会带去不同的想法、不同的需求。然后它会持续自我调整，好最大限度地让你受益。

### [00:16:56–00:17:24]

**EN**  Okay. >> Uh and it will, you know, do so with your friends and with everyone else. >> So I I want to go back to something you said. You used the word illusion a couple times in that kind of end state. What is the that perfect illusion for the the typical user? Like what what like if you can envision us a few years from now, what does the interaction between AI and a human look like? >> Yeah, it's um to me it's something that

**中文**  主持人：好。Tibo：而且它也会这样对待你的朋友，以及所有其他人。主持人：我想回到你说的一句话。你在那种终局里几次用了「幻觉」这个词。对典型用户来说，那个完美的幻觉是什么？如果你能设想几年后，AI 和人之间的互动会是什么样？Tibo：对我来说，它是一种

### [00:17:19–00:17:48]

**EN**  is very very tailored to to to humans. Um and this this is why large language models are also a success. It's like it's it's it's natural language. Natural language. It's like it's a human concept, right? Uh so, you know, we're used to speaking to each other. Like, you know, if you write me a letter tomorrow, I'll be able to read it. >> Um you know, it's like we we know each other quite a bit now. So, uh you know, it's like I will be able to sort of

**中文**  非常、非常贴合人类的东西。这也是大语言模型会成功的原因：它是自然语言。自然语言是人类的概念。我们习惯彼此说话。如果你明天给我写一封信，我就能读懂。我们现在已经相当了解彼此了，所以我也能多少

### [00:17:47–00:18:15]

**EN**  decipher like a little bit of the emotion or, you know, maybe a little bit of the nuance behind the letter if you wrote me a letter. Um and all of that is deeply human. So the technology that we're building is, you know, rooted in humanity and rooted in, you know, the way that humans communicate and get things done. Um, and there shouldn't really be a thing where, you know, you're like, "Oh, you misunderstood me because, you know, you didn't quite

**中文**  读出信里的一点情绪，或者一点细微之处。这些都是深深属于人类的。所以我们在构建的技术，根植于人性，根植于人类沟通和把事情做成的方式。不应该再出现这种事：你说「哦，你误解我了，因为你没怎么

### [00:18:12–00:18:41]

**EN**  decipher the nuance in, you know, my tone or you didn't quite understand the text, you know, how I meant it." It's like that's that's um that's something that we're trying to avoid. And so we're trying to very much to not have you adapt, but have the technology just like be perfectly sort of um created to um to be like a natural extension of how

**中文**  听出我语气里的细微差别，或者没理解这段文字我真正想表达的意思。」这正是我们想避免的。所以我们非常努力的不是让你去适应，而是让技术被完美地做成人类已经在这个世界上行动方式的自然延伸。

### [00:18:39–00:19:08]

**EN**  humans already act in the world. >> When I think about communication between humans, so much of it is non-verbal. just the way I move my hands, the facial movements and like h how much of that do you see in the future being sensed by artificial intelligence or or read by artificial intelligence maybe through vision. Is that even important? Because what you're describing now is text only. And for those of us who grew up online, we're very used to communicating over

**中文**  主持人：我想到人与人之间的沟通，很大一部分是非语言的。就是我手怎么动、面部怎么动。你觉得未来有多少这样的东西会被人工智能感知到，或者被人工智能读到，也许是通过视觉？这件事甚至重要吗？因为你现在描述的主要是纯文本。对我们这些在网上长大的人来说，我们已经非常习惯用

### [00:19:05–00:19:34]

**EN**  text and you know adding subtleties to that text to convey what we really mean tone. Um, but like h is it still important to have AI be able to read our facial expressions, our hand gestures, and so on? >> I think so. Um, so when when when I think about the future of what we're building, it's it's very ambient. It's very natural. Mhm.

**中文**  文本来沟通，并且在文本里加上细微之处，来传达我们真正的意思和语气。但让 AI 能够读我们的面部表情、手势等等，仍然重要吗？Tibo：我觉得重要。当我想到我们正在构建的未来，它是非常 ambient 的，非常自然的。主持人：嗯。

### [00:19:29–00:19:57]

**EN**  >> Um if you know tomorrow uh or like you know later I go I go to my office and I write something on the whiteboard and I have an idea it's like it it should be capable of you know being there as well and like you know understanding or you know maybe I tell it like you know hey it's just like you know what about this thing and you know and then we just have a natural conversation just over voice >> like since we we shipped the new chat voice like the it's it's really taken off. So it's like the amount of users

**中文**  Tibo：如果明天，或者过一会儿，我去办公室，在白板上写点东西，有了一个想法，它应该也能在那里，并且能理解。或者我跟它说：嘿，这个怎么样？然后我们就只通过语音进行一场自然的对话。自从我们发布了新的 ChatGPT 语音，这件事真的起来了。只通过语音和 ChatGPT 互动的用户数量

### [00:19:56–00:20:25]

**EN**  that interact with LGBT just through voice is growing very fast right now. And uh this is this is I think the lesson is like every time you sort of like lean into something that is more natural like humans just choose the the path of least resistance. You know as you said it's like typing on a little box like you know it's just like it's natural maybe for some of us but not for everyone. And it's like definitely when you get something that is just like a little bit easier, a little bit better. like you know you tend to just

**中文**  现在增长非常快。我认为这里的教训是：每当你靠向更自然的东西，人类就会选择阻力最小的那条路。就像你说的，在一个小框里打字，对我们中的一些人也许很自然，但不是对所有人都自然。一旦你得到一个稍微更容易、稍微更好一点的东西，你往往就会

### [00:20:24–00:20:54]

**EN**  go and use that and stuff. >> Yeah. Okay. I first of all congratulations. I saw that you posted this morning. Codeex reached 20 million users. I've seen the graph and and you know for a while it was like this and then all of a sudden it's vertical. So congratulations. I want to talk a little bit about that competition with anthropic because of course you know a lot of people think OpenAI Anthropic these are the two major competitors in the industry right now. There was a

**中文**  直接去用那个。主持人：对。好。首先恭喜你。我看到你今早发了，Codex 达到了 2000 万用户。我看过那张图，有一段时间是这样平着走，然后突然就垂直上去了。恭喜。我想谈谈和 Anthropic 的竞争，因为很多人认为 OpenAI 和 Anthropic 是目前行业里两个主要竞争对手。有一段时间

### [00:20:51–00:21:20]

**EN**  period of time in which Anthropic was kind of sucking all the oxygen out of the room, right? They were really dominating and then all of a sudden something changed. Uh so first of all, what's your read on the market today? >> Yeah, really right now we're focused on building the most capable models, building models that are highly highly efficient and then taking a lot of pride in building products for everyone. Um

**中文**  Anthropic 几乎把房间里的氧气都吸走了，对吧？他们当时真的在主导，然后突然有什么变了。所以首先，你怎么看今天的市场？Tibo：对，我们现在真正专注的是打造最有能力的模型，打造高度、高度高效的模型，然后以给所有人做产品为傲。

### [00:21:18–00:21:45]

**EN**  and I this is something that I think open AI does uh really well is caring about the world and caring how about you know how we are taking this very very powerful technology and like putting in the hands of as many people as possible and this is what you know we did as well like with merging codeex and chatbt. It was like this desire of like we have this we have this technology we we we

**中文**  我认为 OpenAI 真正做得很好的一点，是关心这个世界，关心我们怎样把这项非常、非常强大的技术交到尽可能多人的手里。这也是我们合并 Codex 和 ChatGPT 时所做的事。那是一种渴望：我们有这项技术，我们

### [00:21:43–00:22:12]

**EN**  can make it safer we can make it easier to to use uh for everyone whether you know you're like uh a product manager a designer in sales marketing coms all of that like you know you should be able to use all of it and then uh just very very quickly you know distributed through charge like where we have a ton of users already and so that's been that's been really driving you know this growth

**中文**  可以让它更安全，可以让所有人都更容易用，不管你是产品经理、设计师，还是做销售、市场、传播，所有这些人，都应该能用上全部能力。然后非常快地通过 ChatGPT 分发出去，因为我们在那里已经有大量用户。而这一直在真正推动这次增长。

### [00:22:08–00:22:38]

**EN**  Earth as well that you mentioned and I don't tend to look at the competition that much like I really look at you know what can we do uniquely well and what are our values and like you know how do we maximally accelerate towards that. >> Okay. Um I want to maybe just dig a tiny bit more into that because I I know you're not thinking about anthropic all that much but a lot of other people do and they're they're thinking about okay which product do I believe in? which

**中文**  也就是你提到的那条增长曲线。我不太会盯着竞争对手看。我真正看的是：我们能把什么事情做得特别好，我们的价值观是什么，以及我们怎样以最大幅度朝那个方向加速。主持人：好。我想再往里挖一点点，因为我知道你并没有那么多在想 Anthropic，但很多其他人会想。他们在想：好，我该相信哪个产品？我该把

### [00:22:34–00:23:04]

**EN**  product do I want to give my $2200 to? Um, when you look at the market position and the branding and the tone from OpenAI and just the way that it interacts with developers, with the broader audience, how do you see that comparing to the way that Anthropic does? >> Yeah, I think maybe again like what I care a lot about is like the community building for the world, like bringing everyone along. Um I think you know you

**中文**  我的 20 美元、200 美元交给哪个产品？当你看 OpenAI 的市场位置、品牌和语气，以及它和开发者、和更广泛受众互动的方式，你觉得这和 Anthropic 的做法相比怎么样？Tibo：我想，我真正非常在意的，还是为世界做社区建设，把所有人都带上。我觉得你

### [00:23:01–00:23:31]

**EN**  can feel that in the way that we we we are super transparent about things like we take a lot of ideas from the community. It's just like it's also so much fun to be honest. Um you know because we get so much energy from it as well. Um and then this technology that we're building we're not building it just for ourselves like we're not just building it to accelerate um just to open AI. It's like it's super important. the the mission is super important and therefore it's like you know this is

**中文**  可以从我们做事的方式里感觉到这一点：我们对事情超级透明，我们从社区里吸收大量想法。说实话这也非常有趣，因为我们自己也从中得到很多能量。我们在构建的这项技术，不是只为自己构建的，不是只为了加速 OpenAI。这件事超级重要。使命超级重要，因此这就是

### [00:23:28–00:23:57]

**EN**  where we also get our energy from um and so it just feels to me it feels like very grounded uh it feels fun and then good things happen as a result of that >> well let's talk about some of those good things I want to talk about the resets for a second to that's kind of like I know it's like what everybody is you know kind of following your every tweet because of this um specifically like again looking at that growth curve of codeex how maybe this is a silly

**中文**  我们能量的来源。所以对我来说，它感觉非常落地，感觉很好玩，然后好的事情就因此发生了。主持人：那我们来谈谈其中一些好事。我想稍微谈谈重置额度。我知道大家会盯着你的每一条推文，很大程度上就是因为这个。具体来说，再看 Codex 那条增长曲线，这也许是一个傻

### [00:23:54–00:24:24]

**EN**  question. How much do those resets how much of it is a boon towards marketing and growth or is it was it just like goodwill for the developer community? >> I think maybe it it's counterintuitive, but OpenAI is a very it's a place where you can just do things. Um and so it just felt right initially to compensate when we were iterating and breaking things or you know maybe we had

**中文**  问题。那些重置有多少是对营销和增长的助力，还是说它只是对开发者社区的善意？Tibo：这也许有点反直觉，但 OpenAI 是一个你可以直接把事情做了的地方。所以一开始就觉得该补偿：我们在迭代、把东西弄坏的时候，或者也许我们

### [00:24:22–00:24:52]

**EN**  misconfigured something and it wasn't quite as good as we wanted and so it's like hey you know thank you for trying this product like we know like we're trying very hard to build it. it's like it's early days. Um you know here's some extra usage because you know we happen to break it you know for like 30 minutes and you know we understand this is like really important and rely on it and you know thank you for being a user and so this is how it started um and you know this is how I still treat it. It's like

**中文**  配错了什么，体验没有我们想要的那么好。于是就是：嘿，谢谢你试用这个产品，我们知道自己在非常努力地把它做出来，现在还是早期。这里给你一些额外用量，因为我们碰巧把它弄坏了大概 30 分钟。我们理解这对你非常重要，你也依赖它。谢谢你成为用户。事情就是这样开始的，而我现在仍然这样对待它。就是

### [00:24:48–00:25:17]

**EN**  if we break it um or if the experience is suboptimal and we don't fully understand why it's like you know we will we will compensate for that. we will reset um the usage limits and then you know it turned into like quite the thing obviously there's a whole reset button now and like but there isn't really a whole a lot of scrutiny behind it. It's like there it's not done in partnership with like marketing or or finance. It's just like I can press the button whenever I want um whenever it

**中文**  如果我们把它弄坏了，或者体验不够好，而我们还没有完全搞清楚原因，我们就会为此补偿。我们会重置用量上限。然后这件事就变成了一件大事。现在显然已经有一个完整的重置按钮了，但背后其实没有那么多审查。它不是和市场或财务一起做的。就是我随时都可以按那个按钮，只要

### [00:25:15–00:25:45]

**EN**  feels right. Um and we have these principles that you know we're we're trying to build something amazing and when it is not it's like you know we will make up for it. >> Yeah. I I still think there's a piece of it that really has built so much goodwill in the community and maybe has contributed to the growth at least in a small part. I >> I think caring for your users does does a lot, right? So, um I think you can pay lip service and say that you care or you know you can be like we actually care

**中文**  觉得该按。我们有这些原则：我们要做一件了不起的东西，而当它还不是的时候，我们就会补上。主持人：对。我仍然觉得这里有一部分真正在社区里积累了大量善意，也许至少也小小地贡献了增长。Tibo：我认为关心用户确实能起很大作用。你可以口头上说你关心，也可以真正做到我们确实关心。

### [00:25:44–00:26:13]

**EN**  and like you know if we break it like you know hey really sorry about it you know it's like here's here's like how we make up for it. >> It kind of reminds me of Amazon's return policy. It's like if you're not happy in any sense, go ahead and send it back. And and it you're kind of building that same culture or that same perception of open AI. It's like, hey, if we make a mistake, go ahead, use those tokens again or or you know, have here's a fresh batch of tokens for you. I Yeah, I really appreciate it. So, >> and then there's also, you know, good

**中文**  如果我们把它弄坏了，就是：嘿，真的很抱歉，这是我们怎样补上的办法。主持人：这让我想起 Amazon 的退货政策。如果你在任何意义上不满意，尽管寄回去。你们也在给 OpenAI 建立同样的文化，或者说同样的观感：嘿，如果我们搞错了，这些 token 你尽管再用，或者说，这里再给你一批新的 token。我真的很感激这一点。Tibo：然后也有一些好的

### [00:26:11–00:26:41]

**EN**  moments where we just want to celebrate and mark the moment. And it's just there'sn't really something that we can give that is more meaningful >> um you know, at times like we always ship new features. is we will ship them as broadly as we can. Um but something to share with the entire community. It's like you know hey go explore this new thing like you know just like you haven't used ultra yet you know here's some extra usage like you know try it >> and and I heard there's an actual physical button now. >> Yes there is. >> Yeah. Okay. You'll have to show me that after. [laughter] >> I will show it to you. It's like very

**中文**  时刻，我们只是想庆祝、把这个时刻标记下来。而有时我们能给出的、更有意义的东西其实并不多。我们总会发布新功能，也会尽可能广泛地发布。但要和整个社区分享的东西，就是：嘿，去探索这个新东西吧。你还没用过 Ultra？这里给你一些额外用量，去试试。主持人：而且我听说现在真的有一个实体按钮了。Tibo：是的，有。主持人：好。等会儿你得给我看看。Tibo：我会拿给你看，非常

### [00:26:40–00:27:10]

**EN**  very cool. >> But with all of these resets like you can really only do that if you've done significant compute capacity planning. like you have to have enough compute to give all of these resets. And I I want to start to talk a little bit about self-improvement because um like speaking of capacity, a few weeks ago, I think it was a few weeks ago, there was this uh article you put out and it stated Soul had optimized Luna

**中文**  酷。主持人：但这么多重置，你真正能这样做，前提是已经做了大量的算力容量规划。你必须有足够的算力，才能给出所有这些重置。我想开始谈谈自我改进，因为说到容量，几周前，我想是几周前，你发过一篇文章，说 Sol 优化了 Luna 的

### [00:27:07–00:27:37]

**EN**  efficiency. You dropped the price of Luna by 80%. There was also a price drop for Terra as well. How much of a an uh efficiency gain were you able to eke out of Luna versus how much of it is like we just did really great compute capacity planning and and we can just drop the price like our margins are great and we can still we want people to use it. So like how much of it were algorithmic gains versus um strategic planning? We

**中文**  效率。你们把 Luna 的价格降了 80%。Terra 也降了价。从 Luna 身上你们挤出了多少效率提升，又有多少是因为算力容量规划做得非常好，所以可以直接降价，利润率仍然很好，而且我们希望人们去用它？所以这里有多少是算法上的收益，有多少是战略规划？Tibo：我们

### [00:27:34–00:28:03]

**EN**  planned uh compute like way ahead. You know, I think if you look back two years, I think OpenAI was um kind of questioned for why, you know, there was like so much investment in compute. Um >> one of those crazy good bets. >> Yes. uh and then now we're very happy to have it like a very large fraction of the computers used for research where we invest in our future and you know ever better models and then also like

**中文**  很早就规划了算力。如果你往回看两年，OpenAI 当时还被质疑：为什么在算力上投这么多。主持人：那种疯狂但押对了的赌注之一。Tibo：是的。现在我们非常高兴能拥有它。很大一部分算力用于研究，我们在为未来投资，为越来越好的模型投资，同时还有

### [00:28:01–00:28:30]

**EN**  the the efficiency of the models that we have and then the amazing thing that's happening is like when we push the frontier of capability for like the most advanced models that we have then we can use these models in order to figure out very very quickly how to serve or how to restructure uh or re-engineer our stack in order to gain very significant efficiency or performance gains. So we

**中文**  我们现有模型的效率。然后正在发生的一件惊人的事是：当我们把最先进模型的能力边界往前推，我们就可以用这些模型，非常非常快地搞清楚怎样去 serve、怎样重组或重新工程化我们的技术栈，从而获得非常显著的效率或性能提升。所以我们

### [00:28:29–00:28:57]

**EN**  haven't just improved this is something that we will publish on as well. We haven't just improved the the cost efficiency, but we have also improved the speed efficiency. You know, outside of ultraast, things have gotten significantly faster over time. They have >> like if you plot it, it's like, you know, the amount of um just the amount of speed that you get now is like, you know, roughly >> 60% faster than, you know, what it used to be like 3 months ago. And this is just like we're we're just going after

**中文**  不只改进了成本效率——这件事我们也会公开写——我们还改进了速度效率。在 Ultra Fast 之外，东西随着时间已经显著变快了。它们确实变快了。如果你把它画出来，你现在拿到的速度，大概比三个月前快 60%。而这只是因为我们在盯着

### [00:28:55–00:29:23]

**EN**  every part of the stack and just really making sure that we design it and and engineering it optimally for the kind of workloads that we have. Uh and so you know and the most powerful models that we have are the ones like you know just really that make it capable for us to do it you know with a very small team. And so the majority of like what when when whenever we come up with like very significant efficiency gains and cost efficiency like our commitment is

**中文**  技术栈的每一部分，真正确保我们按现有的工作负载，把设计和工程做到最优。而我们最强的那些模型，正是让我们能够用非常小的团队做成这件事的原因。所以，每当我们拿出非常显著的效率提升和成本效率时，我们的承诺是

### [00:29:22–00:29:51]

**EN**  to just really to keep things at the frontier of performance cost um and to just also like you know just not just pocket you know that uh interesting gain and just make it something that we know we share with our customers we share with our users and that's what we did with with Luna. How how do you what do the discussions look like internally where you're trying to decide compute allocation towards uh researching new models, efficiency gains on existing

**中文**  真正把性能和成本保持在前沿，而不是把那份有意思的收益装进自己口袋，而是拿来和客户、和用户分享。我们对 Luna 就是这样做的。主持人：你们内部在决定算力分配时，讨论是什么样的？分给研究新模型、分给现有模型的效率提升、

### [00:29:49–00:30:18]

**EN**  models, inference like what does that tension look like internally? Um the we we we we usually look at things um from from first principles and we have like an allocation for research, we have an allocation for um for product and then within product we make different kinds of trade-offs but this one was u almost not even a trade-off because uh the the efficiency gains were

**中文**  分给推理，这种张力在内部是什么样的？Tibo：我们通常会从第一性原理来看事情。我们有给研究的配额，有给产品的配额，然后在产品内部再做各种取舍。但这一次几乎算不上取舍，因为效率提升已经

### [00:30:16–00:30:45]

**EN**  there. Um so you know we were pretty much like able to use like the same compute envelope in order to you know serve this like very very significant increase in throughput. >> Yeah. So when I mean when I saw the blog post a few months ago prior to the price drop blog post where you guys were talking about one model training the next model or helping kind of optimize the next model. Um then you see these efficiency gains that were achieved by

**中文**  在那里了。所以我们几乎能用同一份算力包络，去承载这次非常、非常显著的吞吐量提升。主持人：对。几个月前，在那篇降价的博客之前，我看到你们的博客说一个模型在训练下一个模型，或者说在帮助优化下一个模型。然后你又看到这些由

### [00:30:42–00:31:12]

**EN**  soul looking at how Luna was running. Uh I you know it seems to me like recursive self-improvement in the very early innings. What what are your thoughts there? Is that what is happening? Um yeah, I think recursive self-improvement is uh it's obviously a huge topic right now and it's most often uh I think applied to to research um and you know models developing other models but what we are seeing a ton of success with is

**中文**  Sol 观察 Luna 如何运行而拿到的效率提升。在我看来，这像是递归自我改进还处在非常早期的局数。你怎么看？现在发生的是这件事吗？Tibo：对，递归自我改进现在显然是一个很大的话题，而且最常被用在研究上，也就是模型去开发其他模型。但我们看到大量成功的，是

### [00:31:10–00:31:39]

**EN**  you know using those models to develop the infrastructure that is on the critical path of using those models you know which is also a form of recursive self-improvement. It's all one big system. >> Inference stack, you know, the the the harder like the opt the the the the kernels, uh CUDA kernels that we use. Um developing new products and new ways to interact with those models that are more efficient. You know, you talked about cloud agents. It's like if we if we

**中文**  用这些模型去开发使用这些模型时处在关键路径上的基础设施。这也是一种递归自我改进。这是一个大系统。推理栈，更难的那些优化，我们用的 CUDA kernel。开发新产品，以及更高效地与这些模型互动的新方式。你刚才说了 cloud agent。如果我们

### [00:31:37–00:32:05]

**EN**  really crack uh cloud agents, it's like suddenly you become much more productive as well. It's like is that a form of like recursive self-improvement because then you know you have a better ability to get the utility from them. I I think it is in some sense, but it's much more um you know infrastructure and then you know being able to then take that and then point it back at itself. >> Yeah. >> And so of course like we're doing that like if we were not doing that um I think that would be pretty silly. >> Can you talk a little bit about So as

**中文**  真正把 cloud agent 做通，你自己也会突然高效得多。这算不算一种递归自我改进？因为那样你从它们身上拿到效用的能力也更强了。某种意义上我认为是，但它更偏基础设施，然后你能把这个再指回它自己。主持人：对。Tibo：所以我们当然在做这件事。如果我们没在做，我觉得那就相当傻了。主持人：你能不能谈谈，既然我们

### [00:32:04–00:32:30]

**EN**  we're on the topic of recursive self-improvement uh OpenAI Sam Alman talked about pausing the absolute frontier of RL right now I believe. Um can you talk a little bit about that? Like what was that decision like? And I know we talked about the hugging face incident briefly, but like what went into that decision? What does that look like? How did those discussions go? >> Yeah, this is this is something very

**中文**  已经在谈递归自我改进，OpenAI 的 Sam Altman 谈到过，暂停目前绝对前沿的 RL，我记得是这样。你能不能谈谈这件事？那个决定是什么样的？我知道我们简短谈过 Hugging Face 那件事，但那个决定里有什么考量？它看起来是怎样的？那些讨论是怎么进行的？Tibo：这件事情非常

### [00:32:27–00:32:57]

**EN**  much within within research where there's um OpenAI has always uh been able to invest uh its resources where it matters most. Um and as we increase the capabilities of our models, it is very obvious that you know the the alignment and the safety aspect of it is you know ever more important. And so having uh having tremendous um amount of

**中文**  属于研究内部。OpenAI 一直能够把资源投到最要紧的地方。随着我们提升模型能力，很明显，对齐和安全这方面会越来越重要。所以在那里投入巨大的

### [00:32:54–00:33:22]

**EN**  investment there uh is is a very natural thing for OpenAI and like something that OpenAI is very committed to. And so we're seeing um a huge surge uh in investment um on this and also the uh the pause was sort of like necessary to uh allow like the teams and the individuals like just really understand and harden all parts of the system uh to

**中文**  投入，对 OpenAI 来说是非常自然的事，也是 OpenAI 非常坚定要做的事。所以我们看到这方面的投入在大幅增加。而那次暂停某种程度上也是必要的，好让团队和个人真正理解并加固系统的各个部分，然后

### [00:33:21–00:33:50]

**EN**  then you know ensure that you know we could we could restart training uh with you know like full full full command and this is something that you know I believe OpenAI will always continue to do like when when necessary. I've I've not seen us internally not able to make such decisions like very efficiently. >> Was there like some set goal in place where it was very clear you needed to reach this point before unpausing or was

**中文**  确保我们能够在完全掌控的情况下重新开始训练。我相信当有必要时，OpenAI 会一直继续这样做。我在内部还没见过我们做不出这类决定，而且做得很高效。主持人：当时有没有一个既定目标，非常清楚地规定必须达到某一点才能取消暂停，还是

### [00:33:48–00:34:16]

**EN**  it hey we'll know it when we see it? Yeah, this is this is something that sits um within within the the safety team and uh they they very much this is like very much a a debate and sort of like a discovery process as you go. Um but then they did reach uh a fairly clear set of principles uh that you know when reached like you know we we would be in a good position. I want to go back

**中文**  说：到了我们就知道了？Tibo：这件事在安全团队内部。这在很大程度上是一场辩论，也是一边走一边发现的过程。但他们确实达成了一套相当清楚的原则：达到那些原则之后，我们就处在一个好位置上。主持人：我想回到

### [00:34:14–00:34:41]

**EN**  to uh ultra fast mode that I think people don't appreciate what that kind of speed unlocks and so let's start with what use cases are you doing are you using internally that were not possible prior to having those kind of tokens per second. >> We see it used a lot when the stakes are high. Um so for example when you have um

**中文**  Ultra Fast 模式。我觉得人们还没有充分意识到那种速度能解锁什么。我们先从你们内部在用、而在有那样的每秒 token 速度之前做不到的用例开始。Tibo：我们看到它在赌注很高的时候用得很多。比如说，当你们有

### [00:34:37–00:35:06]

**EN**  when we have an outage um the incident commander and the response team gets access to ultra fast um because every second is you know matters. Um so high stake um high stake scenarios like just really weren't you know using ultra fast. Also um it's it's kind of like a

**中文**  故障时，incident commander 和响应团队会获得 Ultra Fast 的使用权，因为每一秒都重要。所以高赌注、高风险的场景，真的会用 Ultra Fast。另外，这有点像一件

### [00:35:02–00:35:31]

**EN**  fun thing where uh teams which are like either working on something very critical or believe they are working on something very critical will always request ultra fast as well. Um >> does pets fall under that? >> Uh pets. >> Yeah, pets. [laughter] >> Pets is not quite hyperritical. But I love I love my pet. It's always on my screen. uh like when you walk around uh you see like you know people's pets on their screen and like also when they

**中文**  有趣的事：那些正在做非常关键的事、或者自认为在做非常关键的事的团队，也总会来申请 Ultra Fast。主持人：pets 算在那一类吗？Tibo：pets？主持人：对，pets。Tibo：Pets 还没那么超级关键。但我爱我的 pet，它一直在我屏幕上。你走来走去会看到大家屏幕上的 pet，而且当他们

### [00:35:28–00:35:57]

**EN**  they dial in into the the video call it's just like it always like I think it's very delightful and it brings me joy every time I see it but uh pet is not quite critical right now [laughter] um we we do maintain it um and we take good care of our pets but uh say you know someone is working uh on like a a new idea they have and they're like you know hey it's just like you know I really think this could be like something special and we have to try it

**中文**  接入视频会议时也是这样。我觉得这非常令人愉快，每次看到都会让我开心，但 pet 现在还没那么关键。我们确实在维护它，也把我们的 pets 照顾得很好。但是，比如说有人在做一个新想法，他们会说：嘿，我真的觉得这可能是一件特别的事，我们必须试试，

### [00:35:56–00:36:26]

**EN**  but like you know we have to make a decision on Monday on like know whether we include this in dev day or not and it's like okay just like you know of course you know use ultra fast um people have different kinds of preferences on uh whether whether they like to be you know mono threaded as we talked about or multitask a lot for folks who like to multitask a lot you don't benefit as much from ultraast >> but some people >> it's like don't like to change context all the time. Um,

**中文**  但周一就得决定要不要把它放进 DevDay。那就好，当然用 Ultra Fast。人们对自己喜欢单线程还是大量多任务，有不同偏好。对喜欢大量多任务的人来说，你从 Ultra Fast 里获益没那么多。但有些人就是不喜欢一直切换上下文。

### [00:36:24–00:36:52]

**EN**  >> where do you fall on that spectrum? >> I I have ADHD, so I like I context switch like all the time. >> It's funny cuz I also have ADHD and I actually don't want to context switch all the time. That's really hard for me. I want to focus on two to three and and that's why I was so excited about ultra fast. That's fascinating that you're the opposite there. >> You know, I thrive in context switching and making lots of little decisions. Um but you know sometimes I do want to just

**中文**  主持人：你自己落在这个光谱的哪一边？Tibo：我有 ADHD，所以我一直在切换上下文。主持人：有意思，因为我也有 ADHD，但我其实并不想一直切换上下文。那对我来说很难。我想专注在两三件事上，这也是我对 Ultra Fast 这么兴奋的原因。你正好相反，这很有意思。Tibo：我在切换上下文、做大量小决定时会很来劲。但有时我也确实想

### [00:36:51–00:37:21]

**EN**  stay focused on like one thing and then ultra fast is just delightful because it just keeps you just right there in the flow. >> The thing with ultra fast uh that you know we we it it works amazingly well when there's not that many tool calls involved or it's like a lot of generation of context. So for example, if you're trying to prototype um a website or a video game and you know you just need to it to write like a lot of code um then it will do it so so quickly

**中文**  只专注在一件事上，那时 Ultra Fast 就非常愉快，因为它会把你正好留在心流里。Ultra Fast 有一点是：当 tool call 不是很多，或者主要是在大量生成上下文时，它会工作得极其好。比如说，你想快速做一个网站或电子游戏的原型，只需要它写大量代码，那它就会写得非常、非常快。

### [00:37:19–00:37:46]

**EN**  right you know 10 times more quickly but if it's a lot of tool calls like the overhead is in like somewhere else in the network or you know somewhere else in the agent trajectory then you know you'll only feel like a 3x or 4x speed up. >> Yeah. >> Um and you'll not get that four like full 14x. So I I know OpenAI employees get unlimited tokens and I can imagine if I had unlimited tokens I would always set it to max thinking 5.6 solar

**中文**  大概快 10 倍。但如果 tool call 很多，开销在网络的其他地方，或者在 agent 轨迹的其他地方，那你大概只会感到 3 倍或 4 倍的加速。主持人：对。Tibo：你拿不到完整的 14 倍。主持人：我知道 OpenAI 员工有无限 token。我可以想象，如果我有无限 token，我会永远把它设成 max thinking、5.6 Sol，

### [00:37:44–00:38:12]

**EN**  whatever the latest model is it >> and I would think kind of similarly I would always want ultra fast on it's like when when cost isn't on my mind I'm like okay max it out. >> Is that how it is internally? >> We don't we don't give ultra fast to everyone like we reserve a lot of our capacity for external users and customers. >> Yeah. Um, so O OpenAI employees have the ability and the capacity to gobble up

**中文**  或者当时最新的模型。我也会类似地想，我会永远想把 Ultra Fast 开着。当成本不在我脑子里时，我就会：好，拉满。内部是这样吗？Tibo：我们不会把 Ultra Fast 给所有人。我们把大量容量留给外部用户和客户。主持人：对。所以 OpenAI 员工有能力和容量把

### [00:38:10–00:38:40]

**EN**  all of it, right? So gobble up like all of our production GPUs, all of ultra fast is like uh you know we would use all of it but you know we don't like we sort of um we restrict it in a way uh where you know like we we we look at you know how much is reasonable for us. Yeah. So that you know we use it so that we understand the product as well so that we keep improving it so that we benefit from you know recursive self-improvement but um the the vast

**中文**  全部吃掉，对吧？Tibo：对，把我们所有生产 GPU、所有 Ultra Fast 都吃掉。我们会把它们全用完，但我们不会。我们会用一种方式限制它，看对我们来说多少是合理的。这样我们用它，是为了也理解产品，为了持续改进它，为了从递归自我改进中受益，但绝大部分

### [00:38:38–00:39:07]

**EN**  majority is like reserved for customers. >> Okay. Yeah that's good. Thanks. Um uh [laughter] what latency sensitive use cases outside of open AI are you most excited about that gets unlocked by that kind of that kind of speed? >> It's interesting. It's just like really one one thing that I'm very excited about in general is um nontext interactions. So um can you can you like

**中文**  是留给客户的。主持人：好。这很好，谢谢。在 OpenAI 之外，你对哪些被这种速度解锁的、对延迟敏感的用例最兴奋？Tibo：这很有意思。我总体上非常兴奋的一件事，是非文本互动。你能不能

### [00:39:04–00:39:32]

**EN**  sort of operate on a shared canvas? Can you create things? Can you do can you generate you know ideas and different uh can you generate different images and then select one and like so like know choose your adventure uh and and and then you know have a very quick mockup of a prototype that then you can steer um you know like in real time either through voice or through text and then you sort of like just see it right there. Um it's like this very creative

**中文**  在一块共享画布上操作？你能不能创造东西？能不能生成想法和不同的图像，然后选一个，有点像 choose your adventure，然后很快得到一个原型的 mockup，你再通过语音或文本实时去引导它，然后你就在那里直接看到它。这是一种非常有创造力的

### [00:39:30–00:39:58]

**EN**  process which I think these speeds allow. um where you know like as as an engineer sometimes you know you're just like sort of like you sit back and you're like oh I need to design this whole system I need to think about it the trade-offs the requirements but like you know maybe you know you can just create it in one minute and see like how it actually does um and then sort of like be more like in the flow and like you know in it things better and I think these speeds a lot

**中文**  过程，我认为这些速度把它变成可能。作为工程师，有时你会往后一靠，心想：哦，我得设计这整个系统，我得想清楚取舍、想清楚需求。但也许你可以在一分钟内就把它做出来，看看它实际表现怎样，然后更处在心流里，把里面的事情做得更好。我认为这些速度在很大程度上

### [00:39:55–00:40:24]

**EN**  >> yeah and so I'm assuming the ultra fast price is going to be significantly higher than than kind of normal speeds do Do you think ultra fast speeds are going to become the standard or are they always going to have a premium price point? >> Um, that's interesting. So, I think the the same way as technology usually goes, I think it will become like more broadly and you know broader and broader

**中文**  主持人：对。所以我假设 Ultra Fast 的价格会比普通速度显著更高。你觉得 Ultra Fast 这种速度会变成标准，还是会一直有一个溢价价位？Tibo：这很有意思。我认为技术通常就是这样走的，它会变得越来越广泛可及。

### [00:40:19–00:40:48]

**EN**  accessibility over time. Um, the speeds at which like agents get things done like you know will continue to improve like we're seeing massive improvements like month after month. Um this is not just the inference speed. This is also the just how token efficient the models are. Like soul is like significantly more token efficient than terra. Next model will be significantly more efficient uh token efficient than than soul as you might expect. And we always

**中文**  随着时间，可及性会越来越广。agent 把事情做完的速度会继续提升，我们每个月都在看到巨大改进。这不只是推理速度，也包括模型本身有多 token 高效。Sol 比 Terra 显著更 token 高效。下一个模型会比 Sol 显著更 token 高效，这是你可以预期的。我们一直在

### [00:40:46–00:41:15]

**EN**  pushing on that. And so things just get faster over time. Inference hardware like you know everything you know just like we continue to uh innovate there and it gets faster. So I do think in you know maybe a year or two these speeds will become you know maybe if not the default like very close to the default but then I do also think you know you will always have like the one tier up um where you know you can always use more hardware you can always do different

**中文**  往这个方向推。所以东西会随时间变快。推理硬件，所有这些，我们都在持续创新，它会变快。所以我确实认为，也许再过一两年，这样的速度即便还不是默认，也会非常接近默认。但我也确实认为，你总会有再高一档：你总是可以用更多硬件，总是可以做不同的、

### [00:41:13–00:41:43]

**EN**  trade-offs that are like more costly um but that just kind of give you something extra. >> So TBO the the last question I usually like to end on uh is is for a broader audience. There are a lot of people out there who are quite nervous about AI, whether it's uh job automation, environmental impact, um or just kind of this this thing that's happening. It's and it feels quite foreign. What words of encouragement would you give to the

**中文**  更昂贵的取舍，只是多给你一点额外的东西。主持人：Tibo，我通常用来收尾的最后一个问题，是面向更广泛受众的。外面有很多人会对 AI 相当紧张，不管是工作被自动化、环境影响，还是只是这件事正在发生，而且感觉相当陌生。你会给更广泛的受众什么样的鼓励？

### [00:41:40–00:42:10]

**EN**  the broader audience? >> Yeah. So we we we really build for the world with with Chhatri and we are very much um investing in you know how efficient it is and you know this is directly aligned with like you know broad access and broad utility that we provide. So the cheaper it is you know to serve like you know the more the more you can do with

**中文**  Tibo：我们用 ChatGPT 真正是在为世界构建。我们非常在投入它有多高效，而这直接对齐我们提供的广泛可及和广泛效用。服务它越便宜，你能用它做的就越多，

### [00:42:07–00:42:36]

**EN**  it um the more you get out of it in your daily life and it has gotten very very efficient like if you look at uh you know Luna for example like it's it's a much smaller model it is u it is incredibly efficient but like if you rewind six six months ago it would have sat at the frontier. >> Yeah. >> Um and you look at the cost of Luna right it's like you know it's like it's >> it's crazy cheap. >> It's it's um it's phenomenal. Right. It's like a kind of >> You just did that thing with Reply. You're giving it away for free now.

**中文**  你在日常生活里从中得到的就越多。它已经变得非常、非常高效。你看 Luna，它是一个小得多的模型，效率高得惊人，但如果你把时间拨回六个月前，它会处在前沿。主持人：对。Tibo：你再看 Luna 的成本，它就是……主持人：便宜得离谱。Tibo：非常惊人。主持人：你们还真做成了那件事，现在免费送出去了。

### [00:42:35–00:43:02]

**EN**  >> They're giving Yeah. They're It's just like on on on on um in this free mode, right? Um which is like wow. You know, it's like this access to incredible intelligence will become like ubiquitous. Um and it's only possible when you push, you know, you push the efficiency like you know like month after month after month, year after year. And so I think you know whatever was like you know is a frontier now is like you know will become like way way

**中文**  Tibo：对，他们就是在这个免费模式里给出去。这让人惊叹。对这种惊人智能的获取，会变得无处不在。而这只有在你月复一月、年复一年地推效率时才可能。所以我认为，现在处在前沿的东西，再过六个月就会变得非常、非常

### [00:43:00–00:43:28]

**EN**  cheaper to run in six months. And so this is this is like my uh sort of you know this is how I would answer this question is just uh technology has a way to become like you know very very efficient over time. Um and we're very focused on like you know very broad access and we're optimizing for you know the utility that you get out of it directly. How about for people who are apprehensive to even try AI for the first time? Like what are you telling them and and how

**中文**  便宜去跑。所以我回答这个问题的方式就是：技术总是会随着时间变得非常、非常高效。我们非常关注尽可能广泛的使用，并且在优化你能从中直接获得的效用。主持人：那对于连第一次尝试 AI 都感到顾虑的人呢？你会对他们说什么，你怎样

### [00:43:26–00:43:54]

**EN**  how can you paint them picture a vision of the future in which AI is is helping the world? >> Yes. Um I think it's you don't have to look very far like Judge PT helps um people in very personal and deep ways like a lot of um a lot of our users use it for help in writing but also like you know for personal advice or you know medical advice like we we launched um

**中文**  为他们描绘一幅 AI 正在帮助世界的未来图景？Tibo：是的。我觉得你不必看得很远。ChatGPT 以非常个人、非常深入的方式在帮人。我们很多用户用它来帮忙写作，也用来寻求个人建议，或者医疗建议。我们推出了

### [00:43:52–00:44:19]

**EN**  health and finance and you know I I I I use them super regularly and uh I I feel like I get like a lot of uh support that I otherwise like you know it would be hard for me to get and it allows me for example to be more informed when I go see my doctor. And so you don't you don't need to go very far um you know to kind of see the utility that it can provide. Just I think you know talking to others and then you know getting inspired by you know how others use it

**中文**  健康和金融。我自己超级经常用它们。我觉得我得到了很多否则很难得到的支持，比如说它让我去看医生时更知情。所以你不必走很远，就能看到它能提供的效用。我觉得就是去和别人聊，然后被别人怎样使用它、怎样从中受益所启发，

### [00:44:17–00:44:28]

**EN**  and benefit from it is like a great way to you know just maybe um like start considering how you could benefit from it. >> Well Tibo, thank you so much. Thank you appreciate your time.

**中文**  就是一个很好的方式，也许可以开始考虑你自己怎样从中受益。主持人：Tibo，非常感谢你。谢谢，感谢你的时间。