AGI终局之战！山姆·奥特曼斯坦福专访：预言AI三波浪潮，聊
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:04 | Good afternoon, everyone, and welcome to another View from the Top. | 下午好,各位,欢迎来到另一景点 |
| 0:09 | It's my pleasure today to introduce our guest, Sam Altman, who is a co-founder and | 我今天很高兴介绍我们的客人 萨姆·阿尔特曼 他是一个共同创始人 |
| 0:16 | the CEO of OpenAI, and one of the most prominent voices right now, | OpenAI的首席执行官, 和最突出的声音之一 现在, |
| 0:22 | shaping how the world thinks about artificial intelligence. | 塑造世界对人工智能的看法。 |
| 0:27 | As many of you know, Sam began his career as an entrepreneur, | 你们中很多人都知道 山姆开始他的企业家生涯 |
| 0:30 | co-founding a company called Looped, before becoming an investor and | 在成为投资人之前 合作创办了一家叫Loopered的公司 |
| 0:35 | an entrepreneur. | 一个企业家。 |
| 0:37 | He went on to serve as president of Y Combinator, | 他继续担任Y组合总裁, |
| 0:40 | where he worked closely with many early stage founders and | 他与许多早期创始人密切合作 |
| 0:44 | helped shape the next generation of startups. | 帮助塑造下一代初创企业. |
展开字幕全文（714 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | Good afternoon, everyone, and welcome to another View from the Top. | 下午好,各位,欢迎来到另一景点 |
| 2 | It's my pleasure today to introduce our guest, Sam Altman, who is a co-founder and | 我今天很高兴介绍我们的客人 萨姆·阿尔特曼 他是一个共同创始人 |
| 3 | the CEO of OpenAI, and one of the most prominent voices right now, | OpenAI的首席执行官, 和最突出的声音之一 现在, |
| 4 | shaping how the world thinks about artificial intelligence. | 塑造世界对人工智能的看法。 |
| 5 | As many of you know, Sam began his career as an entrepreneur, | 你们中很多人都知道 山姆开始他的企业家生涯 |
| 6 | co-founding a company called Looped, before becoming an investor and | 在成为投资人之前 合作创办了一家叫Loopered的公司 |
| 7 | an entrepreneur. | 一个企业家。 |
| 8 | He went on to serve as president of Y Combinator, | 他继续担任Y组合总裁, |
| 9 | where he worked closely with many early stage founders and | 他与许多早期创始人密切合作 |
| 10 | helped shape the next generation of startups. | 帮助塑造下一代初创企业. |
| 11 | OpenAI has become one of the most closely watched companies in the world, | OpenAI已经成为世界上最密切的监控公司之一, |
| 12 | and Sam's views on where this technology is headed continue to influence | 山姆对科技方向的看法继续影响着 |
| 13 | conversations across business, policy, and culture. | 跨越商业、政策和文化的对话。 |
| 14 | Today, Sam is going to share his perspective, | 今天,山姆将分享他的观点, |
| 15 | the share of the perspective he has built at the center of one of the most | 他所构建的观点在其中最核心 |
| 16 | consequential technological shifts of our time. | 我们时代技术的转变 |
| 17 | We could not be more happy to have him with us today. | 今天他和我们在一起,我们非常高兴。 |
| 18 | Please join me in welcoming Sam Altman. | 请和我一起欢迎萨姆·阿尔特曼 |
| 19 | Hi, Sam. | 嗨,萨姆。 |
| 20 | Hello. | 哈罗 |
| 21 | Welcome to the GSB. | 欢迎来到GSB. |
| 22 | Thank you for having me. | 谢谢你邀请我 |
| 23 | We are all so excited to have you here. | 我们都很高兴你能来 |
| 24 | One of the things I've been doing to get ready is I've been talking to an AI | 为了准备,我一直在跟一个AI说话 |
| 25 | version of you, filled by Codex, of course. | 当然是用Codex填的 |
| 26 | And it's been very helpful, answered most of my questions. | 这非常有帮助,回答了我的大部分问题。 |
| 27 | But I have to say, I still don't think anything beats an in-person conversation, | 但我不得不说,我仍然认为 没有什么比当面谈话更棒的 |
| 28 | so thank you. | 谢谢你 |
| 29 | Excited to be here. | 很高兴能来 |
| 30 | Sam, many in this audience know you left Stanford to build your first startup. | 山姆,很多观众都知道你离开斯坦福 来建立你的第一个启动。 |
| 31 | But fewer probably know that you almost did an internship in investment banking instead. | 但很少有人知道你几乎在投资银行做了实习 |
| 32 | I did, that's true. | 是的,没错。 |
| 33 | Could you take us back to that decision and your mindset at the time? | 你能带我们回到那个决定 和你当时的心态吗? |
| 34 | I never wanted to be an investment banker, but | 我从来没有想成为投资银行家,但 |
| 35 | it was what all the cool kids wanted to do, and I fell victim to peer pressure. | 这是所有酷孩子想做的事 我成为同伴压力的受害者 |
| 36 | No, if any of you want to be investment bankers or are more power to you, | 不,如果你们谁想成为投资银行家 或对你们更有利 |
| 37 | it was just not my organic thing. | 它只是不是我的有机的东西。 |
| 38 | I was like a nerd, but I got to Stanford, and all of the smartest kids were | 我像个书呆子 但我去了斯坦福 所有最聪明的孩子都是 |
| 39 | trying to get to be investment bankers, and that was the cool thing. | 试图成为投资银行家, 这是很酷的事情。 |
| 40 | And I was like, well, I want to be a smart kid, and | 我想成为一个聪明的孩子, |
| 41 | maybe I should want to go into investment banking too. | 也许我也应该去投资银行 |
| 42 | So for no reason other than peer pressure, literally, | 所以除了同龄人的压力之外 没有任何其他原因 |
| 43 | I did the whole thing and got an internship and felt super cool. | 我做了整件事,得到了一个实习 觉得超级酷。 |
| 44 | At the time, they would fly you out to New York, and there was this interview, | 当时,他们会把你飞到纽约, 有一次采访, |
| 45 | and then they had this super nice dinner, and it was like this very, | 然后,他们有 超级漂亮的晚餐, 它就像这个非常, |
| 46 | it's like a 19 or 18, whatever, it was very impressive. | 这就像一个19或18, 不管怎样,它非常令人印象深刻。 |
| 47 | I'd never experienced anything like that. | 我从来没有经历过这样的事 |
| 48 | And that made me even more think, this must be the right path for me. | 这使我更深思熟虑 这对我来说一定是一条正确的道路 |
| 49 | And then I had this increasing feeling of dread, and | 然后,我越来越感到恐惧, |
| 50 | I just didn't think I was gonna like it. | 我只是没想到我会喜欢它。 |
| 51 | Meanwhile, I had been building this sort of, not even a startup at the time, but | 与此同时,我一直在建造这种, 甚至不是一个启动在当时,但 |
| 52 | this project with some friends. | 这个项目与一些朋友。 |
| 53 | And then Y Combinator was like, we will give you money for the summer to live on, | 然后Y Combinator说, 我们会给你钱 夏天生活, |
| 54 | and you can try to do a startup. | 你可以尝试做一个启动。 |
| 55 | And I was like, okay, this is like a sign from the universe, | 我当时想,好吧, 这就像一个来自宇宙的征兆, |
| 56 | I should try to go do that. | 我该去试试 |
| 57 | And you joined the very first batch of Y Combinator, and | 你加入了第一批Y联合体 |
| 58 | eventually become president of YC. | 最终成为共青团主席. |
| 59 | What did that experience teach you going from running a startup to | 是什么让你从启动到 |
| 60 | running the world's most famous startup accelerator? | 运行世界上最著名的启动加速器? |
| 61 | I mean, one of the best things for me about, so I was sad to drop out of school | 我的意思是,我最好的事情之一, 所以我难过的辍学 |
| 62 | cuz I loved school, and I was also worried I just was not gonna keep learning stuff. | 因为我喜欢学校,我还担心 我只是不会继续学习的东西。 |
| 63 | And one of the coolest things for me about doing a startup is it was like, and | 对我来说最酷的事情之一 做一个启动是,它喜欢, |
| 64 | still is, it's the highest rate of learning environment I had ever been in. | 这还是我所经历的最高学习环境。 |
| 65 | I think if you're like, this is maybe the wrong room to say this in. | 我想如果你喜欢,这可能是 说这个错误的房间。 |
| 66 | But a thing that someone said to me, I think it was maybe Paul Graham, | 但有人对我说,我觉得可能是保罗·格雷厄姆 |
| 67 | maybe someone like that said to me around then is, | 也许有人是这样对我说, |
| 68 | you will get an entire MBA in one month of doing a startup. | 一个月后你就会得到整个MBA |
| 69 | And I don't have an MBA, so I can't say whether or not that's true. | 我没有MBA,所以我不能说这是否是真的. |
| 70 | But I did learn a lot very quickly about business, and product, and | 但我对商业和产品学得很快 |
| 71 | customers, and sales, and all of this other stuff that has been very important | 顾客,销售,以及其他所有 非常重要的东西 |
| 72 | the rest of my career that I think I just wouldn't have been able to learn in school. | 余生,我想我只是 无法在学校学习。 |
| 73 | It was one of these things that you can just learn by doing, and | 这是其中之一 你可以只是学习 通过做,和 |
| 74 | I don't, maybe there's no other good way to learn it. | 我没有,也许没有其他好方法去学习它。 |
| 75 | So the rate of learning for that first period of a startup was unbelievable. | 初创阶段的学习速度令人难以置信。 |
| 76 | I definitely have never learned more in any other period of my life. | 我这辈子绝对没有学到更多的东西 |
| 77 | Running YC was its own kind of learning. | 管理共青团是它自己的学习方式。 |
| 78 | That's almost like a government job. | 这就像一个政府的工作。 |
| 79 | You don't have the power over these companies, but | 你对这些公司没有权力,但是 |
| 80 | you're trying to build a healthy community and make sure everyone thrives. | 你试图建立一个健康的社区, 确保每个人都繁荣。 |
| 81 | So that was another interesting and very different learning experience. | 所以,这是另一个有趣的和非常不同的学习经历。 |
| 82 | And I like everything that comes before, and you'll all find this out in your own | 我喜欢以前的一切 你们自己会发现的 |
| 83 | lives, you build on everything no matter how unrelated they seem. | 生活,你建立在一切之上 不管它们看起来多么无关紧要。 |
| 84 | So I learned a lot of things about running a startup that were helpful for YC. | 因此,我学到了很多关于运行一个对共青团有帮助的启动器的事情. |
| 85 | But it was a very different job. | 但它是一个非常不同的工作。 |
| 86 | How did you go about learning all of those skills that were so | 你怎么学会这些技巧的? |
| 87 | different than running a startup? | 与运行启动不同吗 ? |
| 88 | Do you feel like you learned them on the job, or you had mentors? | 你觉得你是在工作上学会的 还是有导师的? |
| 89 | How did you actually go about that? | 到底怎么样? |
| 90 | I mean, definitely had great mentors. | 我的意思是,肯定有伟大的导师。 |
| 91 | But there's no, I think of a job that has been done many times before. | 但是,没有,我想一个以前做过很多次的工作。 |
| 92 | You can maybe learn it in a classroom, maybe. | 你可以在教室里学 |
| 93 | You can learn a lot of skills that are helpful and | 你可以学习很多有用的技能 |
| 94 | a lot of sort of knowledge that's helpful. | 很多知识是有用的。 |
| 95 | But I think things like any specific startup or running Y Combinator or | 但我认为事情 像任何具体的启动 或运行Y Combinator或 |
| 96 | any other kind of like not that many people have done it, jobs, | 任何其它类型的 不喜欢,许多人已经做到这一点,工作, |
| 97 | you're just gonna learn as you go. | 你只是要去学习 当你去。 |
| 98 | And I am a big believer that people spend too much time trying to prepare to do | 我非常相信 人们花太多时间准备去做 |
| 99 | the thing, than just doing the thing and learning as they go. | 事情,不只是做事情 和学习他们去。 |
| 100 | So I didn't, you're gonna make a lot of mistakes, you're gonna learn very quickly. | 所以我没有 你会犯很多错误 你会很快学的 |
| 101 | As long as you really approach it with a spirit of, I think people call it | 只要你真正接近它的精神, 我认为人们称它为 |
| 102 | deliberate practice, where you are running a lot of experiments and then actually | 深思熟虑的练习, 在那里你正在运行 很多实验,然后实际上 |
| 103 | reflecting on the learnings, positive or negative, and trying to update your model. | 思考学习,无论是正面还是负面 并试图更新你的模型。 |
| 104 | And not just sort of like random flailing. | 也不仅仅是随机挥舞 |
| 105 | I think the rate of learning is super high and people almost always surprise | 我觉得学习的速度超高 人们几乎总是惊讶 |
| 106 | themselves on if you just like jump in and do the thing, | 如果你喜欢跳进去做这件事 |
| 107 | how quickly you can get good at it. | 有多快才能学好 |
| 108 | YC gave you a front row seat to all kinds of ambitious ideas. | 共青团给予你各种雄心勃勃的想法前排席位. |
| 109 | Yet in 2015, when you co-founded OpenAI, most people told you artificial general | 然而在2015年,当你们共同创建OpenAI时,大多数人告诉你人工将军 |
| 110 | intelligence was impossible, at least for the foreseeable future. | 至少在可预见的将来,情报是不可能的。 |
| 111 | What do you think you and your team saw then that other people were missing? | 你觉得你和你的团队看到其他人失踪了吗? |
| 112 | Deep learning was working. | 深入学习是有效的。 |
| 113 | So I think it, I actually, when I was a student at Stanford, | 所以我想,实际上, 当我是斯坦福的学生, |
| 114 | I was, I worked in the AI lab and I like was planning to work on AI for my career. | 我在人工智能实验室工作 我喜欢为我的事业计划人工智能 |
| 115 | And I remember very memorably in 2005, | 我记得在2005年 |
| 116 | a Stanford CS professor telling me that the one thing we knew for sure, | 斯坦福CS教授告诉我 我们所知道的一件事 |
| 117 | there were a lot of things about AI, it was probably gonna take another century at | AI有很多东西 可能要再过一个世纪 |
| 118 | least, but the one thing we knew for sure is deep learning was not gonna work. | 至少,但我们肯定知道 深层学习是行不通的。 |
| 119 | And, you know, it's like, okay, I believe you, you're like, you know, impressive. | 而且,你知道,这就像,好吧, 我相信你, 你就像,你知道,令人印象深刻。 |
| 120 | And, and then in 2012, there was the famous AlexNet result. | 之后在2012年出现了著名的AlexNet结果. |
| 121 | And then by 2014, we had seen that deep learning seemed | 然后到2014年,我们看到 深层学习似乎 |
| 122 | to get better with scale in all these different ways. | 以所有这些不同的方式,通过规模提高。 |
| 123 | And the world didn't accept this at all. | 世界完全不接受这一点。 |
| 124 | It's, it's actually hard to overstate how much the world was just like, | 很难夸大世界的大小 |
| 125 | there is no way this works. | 这不可能 |
| 126 | You know, there, there are still a few, like, people online that you can find that | 你知道,那里,仍然有一些人,喜欢, 网上的人,你可以找到 |
| 127 | will spend all their time talking about how deep learning is not gonna work and | 将花全部时间 谈论如何深入的学习 是要去工作 |
| 128 | it's a parlor trick and this and that. | 这是一个客厅的诡计 和这个和那个。 |
| 129 | At this point, that's sort of the radical view. | 在这一点上,这就是激进的观点。 |
| 130 | At the time, the radical view was there is no way this is going to get | 当时,激进的观点是,这是不可能得到的 |
| 131 | to any serious level of intelligence. | 任何严肃的情报 |
| 132 | This is a ridiculous thing. | 这太荒谬了 |
| 133 | But I had kind of learned, and this has been, to the degree there is, | 但我已经学到了一些东西, 并且,在程度上, |
| 134 | has been a central sort of insight of my career, | 一直是我事业的核心见解 |
| 135 | that there are surprising effects to scale. | 有惊人的影响 规模。 |
| 136 | And that when you see a beautifully scaling exponential of some sort, | 当你看到一个美丽的指数 某种类型的, |
| 137 | you are supposed to keep going. | 你应该继续走 |
| 138 | And something about human intuition does not do a good job with that. | 关于人类直觉的东西 做不了什么好事。 |
| 139 | And in all of these different things, the scaling returns of individual companies, | 在所有这些不同的事物中, 单个公司的回报率, |
| 140 | networks like Y Combinator, neural networks, lots of other examples I could give, | 网络,如Y联合体,神经网络, 很多其他例子我可以给出, |
| 141 | but we'll skip in the interest of time. | 但为了时间起见我们会跳过 |
| 142 | I just always found this, and it is something that I have, like, | 我只是总是找到这个, 这是我有的东西,喜欢, |
| 143 | developed a high personal conviction to continue to bet on. | 我个人很有信心继续赌注 |
| 144 | And so when we saw that this was working and we saw that it got better with scale, | 所以,当我们看到这是工作 和我们看到它得到了更好的 规模, |
| 145 | we didn't know at the time how predictably better it got with scale. | 我们当时还不知道 规模比这要好得多 |
| 146 | That came a few years later. | 这是几年后的事 |
| 147 | It seemed like we should keep pushing on it. | 似乎我们应该继续推进 |
| 148 | And, you know, maybe it would break at some point. | 而且,你知道,也许它会打破 在某些时候。 |
| 149 | We weren't sure, like, there were various times where we almost lost conviction | 我们不确定,像,有 不同的时间,我们几乎失去了信念 |
| 150 | and thought we were hitting some sort of wall. | 还以为我们撞到墙了 |
| 151 | But we just kept going and it was so non-obvious to any, pretty much, | 但我们只是继续走, 它是如此不明显 任何,几乎, |
| 152 | anyone that was not inside the walls of OpenAI at that time didn't seem to believe it. | 当时不在OpenAI墙内的人似乎不相信. |
| 153 | That's not literally true. | 这不是真的 |
| 154 | They're obviously people who did. | 他们显然就是这么做的 |
| 155 | But it was, like, it was a very unpopular opinion. | 但它是一个非常不受欢迎的意见。 |
| 156 | OpenAI starts as a non-profit research lab with no intention of becoming a product company. | OpenAI最初是一个非营利研究实验室,无意成为一家产品公司. |
| 157 | And today, over a billion people use ChatGPT every week. | 今天,超过十亿人每周使用ChatGPT。 |
| 158 | How has your own leadership had to evolve as OpenAI has gained tremendous influence? | 随着OpenAI获得巨大影响力,你们自己的领导地位如何演变? |
| 159 | Um, like many other people in the sort of many other, like, tech people | 嗯,像许多其他人 那种其他,喜欢,科技的人 |
| 160 | in the mid-2010s of Silicon Valley, I got enamored with the idea of research labs. | 在硅谷的2010年代中期 我被研究实验室的想法迷住了 |
| 161 | This is a time when, like, everybody was talking about how Silicon Valley had, | 这时候,像, 每个人都在谈论 硅谷是如何, |
| 162 | you know, lost the plot on research. | 你知道,失去了研究的情节。 |
| 163 | And we needed to get, you know, back to Xerox PARC or Bell Labs | 我们需要,你知道,回到 Xerox PARC 或贝尔实验室 |
| 164 | or whatever your favorite example was. | 或是你最喜欢的榜样 |
| 165 | And it was, like, kind of a cool thing to talk about and do. | 它一样,是一种很酷的事情 谈论和做。 |
| 166 | And I thought, I always thought research was awesome. | 我想,我总是认为 研究是真棒。 |
| 167 | I always kind of wanted to run a research lab. | 我一直想经营一个研究实验室 |
| 168 | And I was always interested in AI. | 我总是对AI感兴趣 |
| 169 | So this was kind of supposed to be my retirement project. | 所以这是我的退休计划。 |
| 170 | And for the first, like, five years, I mean, truly the coolest job I could imagine. | 在第一个五年里, 我的意思是,真正最酷的工作 我能想象。 |
| 171 | It was like there was a small set of us. | 好像我们有一小撮人 |
| 172 | It was a research lab. | 那是个研究实验室 |
| 173 | We had not, we were right about many things, | 我们没有,我们是对的很多事情, |
| 174 | but we were wrong about the way that we thought AI was going to go once developed. | 但是我们错误地认为 AI会发展起来 |
| 175 | And so it truly hadn't occurred to us that the most likely path was that we were going | 因此我们真的没有想到 最可能的道路是,我们要去 |
| 176 | to have to build a product company. | 必须建立一个产品公司。 |
| 177 | We mostly thought that we were going to, like, develop this unbelievable superintelligence. | 我们大多认为,我们将发展出 不可思议的超级情报。 |
| 178 | And I realize this sounds ridiculous now. | 我意识到这听起来很荒谬 |
| 179 | And that we wouldn't then have to do all of the hard work of figuring out how to get it | 那样我们就不会费尽心思 想办法得到它了 |
| 180 | usefully into people's hands and for society to benefit. | 让社会受益, |
| 181 | It was like there was that meme from a long time ago of, like, do the first thing, | 就像很久以前的我一样 做第一件事 |
| 182 | you know, do the second thing, question mark, question mark, question mark, question mark, | 做第二件事 问号 问号 问号 |
| 183 | profit about how, like, startups were going to work. | 初创企业如何运作是有好处的。 |
| 184 | Ours was, like, do research, build superintelligence, | 我们的研究 建立超智能 |
| 185 | many question marks, people are benefit and happy. | 许多问题标记,人们是受益和幸福的。 |
| 186 | And the technology went in a very different path than we originally thought it was going to. | 技术走在了与我们原先想象的完全不同的道路上。 |
| 187 | I think it's, like, a much better one. | 我觉得这比以前好多了 |
| 188 | But it means we did have to productize this. | 但它意味着我们不得不生产这个。 |
| 189 | And it also means, and this is why I think it's one reason I think it's a good direction, | 这也意味着, 这就是为什么我认为这是原因之一 我认为这是一个好的方向, |
| 190 | is that this is much more like a normal economic force than we imagined at the time. | 这比我们当时想象的更像正常的经济力量。 |
| 191 | This is like electricity or transistors or some other major scientific discovery that enables | 这就像电力或晶体管 或其他重大科学发现 |
| 192 | many other companies, technologies, scientific discoveries to happen. | 许多其他公司,技术,科学的发现将会发生. |
| 193 | And we have to put infrastructure, which is building products like ChatDBT, | 我们必须把基础设施, 它正在制造产品 像ChatDBT, |
| 194 | it's also building, like, an amount of compute bigger than any project | 也造了一个比任何工程都大的计算器 |
| 195 | humanity has ever taken on in history into people's hands. | 人类在历史上曾经落入人们的手中。 |
| 196 | So it became clear by, I would say, 2020, late 2020, | 所以到2020年,我说,2020年末, |
| 197 | early 2021 that we were going to have to build a big company. | 2021年初,我们要建个大公司。 |
| 198 | Didn't think it would happen so quickly. | 没想到会这么快发生 |
| 199 | And then when we launched ChatDBT, which was almost four years ago, | 然后,当我们推出ChatDBT, 这是近四年前, |
| 200 | then overnight everyone was like, okay, we're going to be a big company very fast. | 然后一夜之间,每个人都喜欢, 好吧,我们将是一个大公司 非常快。 |
| 201 | Very big company you are. | 你是一个大公司 |
| 202 | I'm curious to know what sorts of decisions are hardest for you to make today? | 我想知道你今天最难做出什么样的决定? |
| 203 | I think that would take the rest of our time. | 我认为这需要我们剩下的时间。 |
| 204 | I'd be happy to talk about it, but the number one one that comes to mind is | 我很乐意谈论它, 但头一个想到的是 |
| 205 | people call it different things, pausing, pacing, intelligence, speed limits. | 人们称之为不同的东西, 穷困, 速度,智能,速度限制。 |
| 206 | But like fundamentally, if you believe, which I very much do, | 但基本上,如果你相信, 我非常相信, |
| 207 | that people always should be in control of the future and we should not either intentionally | 人民应该永远掌握未来 我们不应该故意 |
| 208 | handover irreversibly decision-making power to AI or accidentally lose control. | 不可逆地将决策权移交给AI或意外失去控制. |
| 209 | And also you believe that power needs to be pretty distributed in the future, | 而且你相信 权力需要相当的分布 在未来, |
| 210 | both of which are saying a different version of like we want the kind of democratic | 都说我们想要民主 |
| 211 | pluralistic ideals that have gotten us this far to continue and we don't want to see | 多元理想让我们得以延续 我们不想看到 |
| 212 | too much power concentration either in an AI with IQ 10 billion or in, you know, | 太多的动力集中 要么在一个AI 与100亿IQ 或者,你知道, |
| 213 | a small number of people that have AI and can do whatever they want. | 少数有AI的人,可以做他们想做的事. |
| 214 | Everybody else can't. | 每个人都不能。 |
| 215 | If you believe both of those, then I think you get to very difficult questions about | 如果你相信这两个,那么我认为你 得到非常困难的问题 关于 |
| 216 | how much smarter than the smartest people can we let an AI system get? | 比最聪明的人还要聪明 我们能让人工智能系统得到多少? |
| 217 | How quickly and how much do we need to enhance people's ability to be smarter or to understand | 我们需要多快多快才能提高人们的聪明或理解能力 |
| 218 | these systems? | 这些系统? |
| 219 | And I think this is solvable. | 我觉得这可以解决 |
| 220 | It just takes time. | 这只是需要时间。 |
| 221 | And then if you believe that there's like a limited number of ways that I think the | 如果你相信有 有限的方法 我认为 |
| 222 | world can come together and agree on how we're going to say what those limits are. | 世界可以走到一起 商定我们如何说出这些限制是什么。 |
| 223 | How we're going to handle them. | 我们怎么对付他们? |
| 224 | And these are very tricky stuff because you have extremely strong opinions on all sides. | 这些都是非常棘手的事情 因为你在各方面都有非常强烈的看法。 |
| 225 | You know, you have people saying like any regulation is bad, which I think is wrong. | 你知道,你有人说 像任何规章是坏的, 我认为是错误的。 |
| 226 | People saying that we have to like absolutely stop all progress right now, which I think | 人们说,我们必须喜欢 绝对停止所有的进展 现在,我认为 |
| 227 | is absolutely wrong. | 绝对是错误的。 |
| 228 | And like in many other of society's highest stakes debates, trying to like be the pragmatic | 和许多其他社会最高利益的辩论一样, 试图喜欢务实 |
| 229 | centrists in the middle of this is a very difficult place to be in. | 中间的中间派是一个非常困难的地方。 |
| 230 | How has your own thinking about being a pragmatic centrist and AI safety evolved over the past | 过去你对做一个务实的中间派 和AI安全的想法是如何演变的 |
| 231 | six months as you've seen capabilities accelerate? | 6个月,你看到的能力加速? |
| 232 | You know, I still believe in a lot of our earliest tenets about how we think AI should | 你知道,我仍然相信很多 我们最早的信条 关于我们认为AI应该如何 |
| 233 | get deployed in society. | 在社会上部署。 |
| 234 | I still believe in iterative deployment. | 我仍然相信反复部署。 |
| 235 | I still believe that society has got to contend with this technology as we develop it, | 我仍然相信当我们开发这种技术时,社会必须与之抗衡, |
| 236 | even though it's, you know, difficult and an incredible rate of change. | 虽然,你知道,这是困难 和令人难以置信的变化速度。 |
| 237 | And I still believe that people are incredibly resilient and adaptable and honestly don't | 我仍然相信,人们是令人难以置信的 弹性和适应性,诚实地不 |
| 238 | care that much about the work of machines. | 这么关心机器的工作 |
| 239 | They're just so focused on people that the future is going to look, even if the technological | 他们把注意力集中在人们身上 以至于未来会展望未来 即使技术 |
| 240 | landscape of the future looks very different, that this sort of societal shape of the future | 这种未来的社会形态 |
| 241 | is going to be much more similar today than people think. | 今天比人们想象的要相似得多 |
| 242 | All that said, watching the rate of progress the last six months has tested the strength | 所有这一切,看过去六个月的进展速度 测试了实力 |
| 243 | of my convictions on some of those things. | 我对一些事情的信念 |
| 244 | It is, you know, in some narrow senses, like certainly mathematical reasoning, as far as | 这在某种狭义上是 就像数学推理 |
| 245 | I'm concerned, we have superintelligence now and I didn't think it was going to happen | 我很担心,我们有超级情报了 我没想到会这样 |
| 246 | this fast. | 这么快。 |
| 247 | And if I play out another year or two of a similar rate of progress, which I think without | 如果我再玩一两年 类似的进步率, 我认为没有 |
| 248 | any changes to the current landscape in the world and sort of the alignment at the level | 世界目前状况的任何变化,以及级别上的调整 |
| 249 | of companies and governments and societies we will have, then it seems to me like that's | 公司、政府和社会,我们将拥有, 那么在我看来,这是 |
| 250 | not going to work to make no changes to how we do things now. | 不想改变我们现在的行事方式 |
| 251 | Like the systems are just getting so capable so quickly that if we want to hold on to those | 就像系统变得这么快 如果我们想坚持下去 |
| 252 | two things I mentioned a little bit ago, we're going to have to make some changes to how | 我刚才提到的两件事, 我们必须做一些改变 |
| 253 | we think about pacing progress from here. | 我们想着从这里取得进步。 |
| 254 | People have very strong opinions about this. | 人们对此有非常强烈的看法. |
| 255 | I'm curious, like, how does that debate play out inside of OpenAI today? | 我很好奇,比如,今天在OpenAI内部 辩论会如何进行? |
| 256 | I'd say the more people see and feel how good the models are getting, the less spread there | 我想说,人们越是看到和感觉到模特儿越来越好, 在那里的传播就越少 |
| 257 | is in the debate and the more people are like, okay, we're going to have to think about | 我们不得不考虑 |
| 258 | the strategy for the future and what it takes to make the safety guarantees we want to make | 未来的战略 以及做出我们想要做出的安全保证所需的一切 |
| 259 | somewhat differently than we did in the past. | 和过去不一样 |
| 260 | I realized that a lot of the debate inside of OpenAI, a lot of the difference in opinion | 我意识到OpenAI内部的很多争论 意见上的分歧 |
| 261 | inside of OpenAI, and I think it's also a good learning for the difference in opinion | 在OpenAI内部, 我认为这也是一个很好的学习 在意见分歧 |
| 262 | society, is there was a lot of debate about how good the models were going to get, how | 社会,是否有很多争论 关于模型会有多好,如何 |
| 263 | quickly, that got reframed as a debate about what the right strategy was. | 很快,这被重新描述为一场关于什么是正确的战略的辩论。 |
| 264 | Once that debate is answered and it's like, okay, the models are about to get extraordinary, | 一旦辩论得到答案,就像,好吧,模型 即将变得非凡, |
| 265 | then there's less space on the what should we do about it questions. | 那么我们该怎么处理这个问题的空间就更少了。 |
| 266 | Still some. | 还有些 |
| 267 | You recently spoke at the UN about AI safety and you emphasize the importance of the government | 你最近在联合国谈到AI的安全,你强调政府的重要性 |
| 268 | when it comes to regulating AI. | 在管制AI方面。 |
| 269 | Traditionally, policy and technological change move at pretty different paces. | 传统上,政策和技术变革的速度相当不同。 |
| 270 | What do you think are some ways that AI labs and the government can better work together? | 你认为AI实验室和政府能更好地合作有什么方法? |
| 271 | I'm actually pretty happy with how it's trending. | 我其实很满意它的趋势。 |
| 272 | Recently, I didn't think most people in most governments took it seriously enough. | 最近,我觉得大多数政府的人 都不够认真。 |
| 273 | Last few months, it's changed and now I think the conversation is being, | 几个月前,事情变了 现在我觉得谈话是... |
| 274 | we'll see if something actually happens. | 我们看看是否真的发生了什么 |
| 275 | But I didn't think anything good was on a trajectory likely to happen and now I think probably yes. | 但我不认为什么好的 是在可能发生的轨迹 现在我想可能是的。 |
| 276 | There's also this question about AI safety at an individual or organizational level. | 还有一个关于个人或组织层面AI安全的问题. |
| 277 | What can people or business leaders do today to better prepare themselves for AI's potential risks? | 今天人们或商界领袖可以做些什么来更好地准备AI的潜在风险? |
| 278 | There's so many levels of that. | 有这么多的级别。 |
| 279 | I think the one, just so that I don't go through a laundry list of 30 things, | 我想那个,只是为了不让我看30件东西的洗衣单 |
| 280 | the cybersecurity, I don't even know a dramatic enough word to use, | 网络安全,我甚至不知道 一个戏剧性足够的词使用, |
| 281 | but the problem coming towards us I think is still something that we have not, | 但问题降临到我们身上 我认为我们还是没有 |
| 282 | as organizational leaders or countries or individuals, taken seriously. | 作为组织领导人或国家或个人,认真对待。 |
| 283 | There was like a moment earlier this year where it was like the mythos freak out of | 今年早些时候的某一刻 就像是神话般的惊慌 |
| 284 | whatever people called it and I think things did get somewhat better. | 不管人们怎么称呼它 我认为事情确实有所改善 |
| 285 | Like, people definitely patched some bugs, but it was another one of these cases where | 就像,人们肯定 补补了一些虫子, 但这是另一个案例 |
| 286 | people didn't really intuit the exponential. | 人并没有真正在意指数。 |
| 287 | So it's like, okay, there are these models, they can find a bunch of security vulnerabilities, | 所以它就像,好吧,有这些模型, 他们可以找到一堆安全弱点, |
| 288 | those are probably all of the security vulnerabilities that exist | 这些可能是所有存在的安全弱点 |
| 289 | or all the security vulnerabilities that models are ever going to find. | 或所有可能发现的安全漏洞 |
| 290 | And I don't think most people sort of said, okay, the models are going to get much smarter, | 我不认为大多数人会说, 好吧,模型会变得更聪明, |
| 291 | they can probably find bugs, like crazy bugs, you know, bugs that are outside of the | 他们可能可以找到虫子, 像疯狂的虫子, 你知道,虫子在外边 |
| 292 | normal threat model of what we think a hacker can do, you know, hardware level, insane stuff, whatever. | 正常的威胁模型 我们认为黑客可以做的, 你知道,硬件水平, 疯狂的东西,不管什么。 |
| 293 | And so people fixed some bugs and, you know, security professionals still freaking out, | 所以人们修了些窃听器 保安人员还是很紧张 |
| 294 | CISOs still freaking out, but largely speaking, I think most organizational leaders have somewhat | CISO仍然惊慌失措,但基本上,我认为大多数组织领导人都有些 |
| 295 | moved on and the models have kept going. | 继续前进,模特们继续前进 |
| 296 | And I no longer think we're going to be able to | 而且我认为我们不能再 |
| 297 | just patch all the bugs, I think we're going to have to completely switch the defense model to | 只要补补所有的bugs, 我认为我们必须 完全切换防御模型 |
| 298 | using the same technology for sort of a new kind of defense shield. | 用同样的技术来做新型防御盾牌 |
| 299 | And I think it's a huge safety and security issue. | 我认为这是一个巨大的安全保障问题。 |
| 300 | And I am somewhat astonished at how relatively little attention it's getting. | 我有点惊讶的是,它得到的注意很少。 |
| 301 | And to bring it back to the people in this room, you know, as business school students at Stanford, | 为了把它带回这间屋子里 作为斯坦福的商学院学生 |
| 302 | we're right by the people building the future of AI, but it's not always clear where we fit in. | 我们是正确的人 建设AI的未来, 但并不总是清楚 我们在哪里适合。 |
| 303 | So what advice would you give us on how we can shape what comes next? | 那么,你给我们什么建议 如何塑造接下来会发生什么? |
| 304 | Um, my honest belief is that it's not, it's not going to be as different as quickly as people | 嗯,我诚实的信念是,它不是, 它不会像人们一样快 |
| 305 | think. | 以为. |
| 306 | Like it is totally true that huge numbers of tasks that are done today by people will be | 就像今天人们完成的大量任务一样 |
| 307 | done better, cheaper, faster, you know, completely differently by AI. | 做得更好,更便宜,更快,你知道的, 完全不一样的AI。 |
| 308 | And it's also true that wide | 这也是事实 |
| 309 | swaths of current jobs will just, the way they're done today, at least, will be entirely done | 现今的一线工作, 将只是, 他们今天的做法,至少, 将完全完成 |
| 310 | by AI. | 由大赦国际编写。 |
| 311 | It's actually been interesting to watch over the last 10 days, the reaction of the | 过去十天里,人们一直很感兴趣, |
| 312 | mathematical community. | 数学界. |
| 313 | You know, the day zero reaction was like, this is fake. | 零反应的那天,这是假的 |
| 314 | The day one | 有一天,我开始... |
| 315 | reaction was like, we're cooked. | 反应就像,我们是熟的。 |
| 316 | And our field is over. | 我们的战地结束了 |
| 317 | And you know, we're telling our grad | 你知道,我们正在告诉我们的研究生 |
| 318 | students to switch majors. | 学生换专业 |
| 319 | The day three reaction was maybe like, it's horrible at opening. | 三天的反应可能像, 它的可怕打开。 |
| 320 | I did | 没错 |
| 321 | this and no one should be allowed to, you know, if this is the price of understanding more | 如果这是理解的代价 |
| 322 | mathematics, we don't want it. | 数学,我们不要它。 |
| 323 | But then by kind of maybe day five, it shifted a lot. | 但也许从第五天开始,它转移了很多。 |
| 324 | And I started | 我开始 |
| 325 | seeing, I started hearing from mathematicians who are like, yes, there is one kind of math job that | 我从数学家那里听来 有一种数学工作 |
| 326 | is over. | 结束了。 |
| 327 | But mostly what we did was not solve millennium prize problems. | 但我们所做的大多不是解决千年奖问题。 |
| 328 | Most of us never did | 我们多数人从来没有 |
| 329 | that. | 那个 |
| 330 | Um, most of what we did was figure out how to make sense of knowledge, how to communicate it, | 嗯,我们所做的大多数 是想出 如何理解知识, 如何沟通它, |
| 331 | how to like present it in a human intelligible way. | 如何用人类能理解的方式呈现它。 |
| 332 | And we've now seen a lot of people say, | 现在我们看到很多人说, |
| 333 | actually, I'm going to have a different, but hopefully better kind of career as a mathematician | 实际上,我将会有一个不同的, 但希望更好的职业 一个数学家 |
| 334 | where I'm going to, I get that AI is going to like drive mathematical discovery from here to | 我想,我得到AI 会喜欢驱动数学发现 从这里到 |
| 335 | a significant extent. | 相当大程度上。 |
| 336 | But in terms of figuring out what to work on, what it means, how to present it, | 但是,在想出什么工作, 它意味着什么,如何呈现它, |
| 337 | how to explain it, how to like write a great proof, how to do all bizarre things, you know, | 如何解释,如何喜欢写一个伟大的证据, 如何做所有奇怪的事情,你知道, |
| 338 | I think there's going to be a whole new set, a whole new set of stuff. | 我想会有一套全新的东西 |
| 339 | And I, | 还有我 |
| 340 | that was the fastest that I think we've all gotten to watch a field go through this cycle. | 我认为这是最快的,我们每个人都 得到了看一个田地 经过这个循环。 |
| 341 | But versions of it happened more slowly in other fields. | 但是,其版本在其他领域发生得比较慢. |
| 342 | You know, I'm sure many of you watch what | 你知道,我敢肯定,你们很多人 看什么 |
| 343 | happened the last two years of computer programmers saying our job is over to actually like, looks like | 发生了过去两年的计算机程序员 说我们的工作已经结束 实际上喜欢,看起来像 |
| 344 | we're going to keep paying ever higher salaries to people that can get computers to do things. | 我们将不断支付更高的工资 给那些可以让计算机做事情的人。 |
| 345 | It's just the day to day of the job is going to be different. | 只是日干日干将不同. |
| 346 | Consultants went through a period | 顾问经过一段时间 |
| 347 | where they were like, this is totally over. | 他们喜欢的地方, 这完全结束了。 |
| 348 | And then they found out that actually like, | 然后他们发现,实际上, |
| 349 | no matter how good the AI models are at making like nice looking PowerPoint slides, | 不管AI模型有多好 制作像好看的PowerPoint幻灯片, |
| 350 | which they are good at, people like really want to interact with another person. | 他们擅长, 人喜欢真的想与另一个人互动。 |
| 351 | And actually, | 事实上, |
| 352 | almost no one wants the AI consultant to show up and tell them what to do. | 几乎没有人希望AI顾问出现,告诉他们该怎么做. |
| 353 | And you get the point, | 你明白这一点, |
| 354 | this is going to keep happening. | 这会继续发生。 |
| 355 | So I don't, I don't think there's going to be like a job | 所以,我不,我不认为 有会像一个工作 |
| 356 | apocalypse in the way that many people talk about. | 许多人谈论的启示录。 |
| 357 | There will be huge amounts of short term | 会有很长的短期时间 |
| 358 | change over in terms of the tasks people do and some kinds of jobs. | 改变人们的任务 和一些类型的工作。 |
| 359 | But if you are like heading | 但如果你喜欢去 |
| 360 | into the world right now to go work on something, I think you kind of, you want to be starting your | 现在进入世界 去工作的东西, 我认为你有点, 你想开始你的 |
| 361 | career at a time of like the highest degree of change. | 在一个变化程度最高的时期 |
| 362 | And you want like the most open playing | 你想玩得最开朗 |
| 363 | field. | 字段。 |
| 364 | I think people with like potential and drive and ideas and sort of a new way of looking | 我认为人们有潜力 驱动力 想法 和一种新的外观 |
| 365 | at things and new skills do relatively the best when the ground is shaking the most. | 当地面最摇晃的时候 新的技能相对来说是最好的 |
| 366 | My like chronic disappointment in MBA students is that they always want to just like go do the thing | 我对MBA学生的 长期失望是 他们总是喜欢去做这件事 |
| 367 | that was cool like five years ago and not do the new thing. | 这就像五年前的酷 而不是做新的事情。 |
| 368 | And I think this is a great time to | 我觉得现在是时候 |
| 369 | go do the new thing. | 去做新东西。 |
| 370 | So go do the new thing, even if AI is the cool thing right now. | 所以,去做新的事情, 即使AI是现在的很酷的事情。 |
| 371 | You, yeah, I mean I don't, I don't think this is like, I don't think this is like a good time to | 你,是的,我的意思是我不, 我不认为这是 像,我不认为这是一个好时机 |
| 372 | like start another copy of OpenAI. | 就像开始OpenAI的另一本。 |
| 373 | But I think there are many companies that can be built on top | 但我认为有很多公司可以在上面建设 |
| 374 | of technology that OpenAI has put out into the world that will be like much, I hope, bigger and | OpenAI向世界投放的技术 我希望它能大得多 |
| 375 | 比OpenAI更重要。 | |
| 376 | I want to shift now to AI's impact on who we are as humans and future | 我现在想转而谈谈AI对我们人类和未来的影响 |
| 377 | potential. | 潜力。 |
| 378 | AI makes it possible for people to build almost anything without prior experience. | AI使得人们有可能在没有经验的情况下建造几乎所有的东西. |
| 379 | In this world, what should people still learn to do for themselves even if AI can do it for them? | 在这个世界里,即使AI能为他们做,人们还能学着为自己做什么呢? |
| 380 | An example of a thing that someone promised me was important to learn to do for myself, | 一个有人答应我的事情的例子是 学会为我自己做 |
| 381 | even though there were computers, is my second grade teacher was absolutely convinced I had to | 虽然有电脑,但我的二年级老师绝对相信 我必须 |
| 382 | learn to write cursive. | 学会写咒语 |
| 383 | I was like the last kind of generation of students that had to learn to | 我就像最后的一代学生 不得不学习 |
| 384 | write cursive. | 写咒语。 |
| 385 | And there was a whole, I was like, I am never going to use this again. | 我本来想,我再也不会用这个了 |
| 386 | I am so sure. | 我这么确定。 |
| 387 | I'm sure this is, I don't even sign my name in cursive. | 我敢肯定,这是, 我甚至不签我的名字 在诅咒。 |
| 388 | I was like, I am just never, I couldn't | 我当时想,我只是从来没有,我不能 |
| 389 | write cursive today probably if I had to. | 今天写咒语,也许如果我不得不。 |
| 390 | And I was like, if you're making me learn this because | 我当时想,如果你让我 学这个,因为 |
| 391 | you know, I have like bad handwriting and I need to like learn fine motor skills or you know, | 你知道,我有像糟糕的笔迹 我需要学习良好的运动技能 或者你知道, |
| 392 | however a seven-year-old would articulate that, I can believe it. | 无论七岁的孩子会说出这一点,我相信。 |
| 393 | But if you think I'm actually | 但如果你认为我是真的 |
| 394 | need to learn cursive for the sake of cursive, then I don't believe you. | 为了咒语需要学习咒语,那么我不相信你。 |
| 395 | And my teacher was like, | 我的老师说, |
| 396 | nope, you're gonna need this as like a functioning member of society. | 不,你需要这个 像一个正常运转的社会成员。 |
| 397 | You have to be able to write | 你必须能写 |
| 398 | cursive. | 诅咒。 |
| 399 | Okay. | 摆 |
| 400 | And then I had another teacher a couple years later who was like, you have to | 几年后,我又有个老师,你必须 |
| 401 | learn to write. | 学会写作。 |
| 402 | And I was like, I have no desire to be a writer. | 我本来不想当作家的 |
| 403 | And that teacher was like, well, | 那个老师说 |
| 404 | that's fine. | 没关系 |
| 405 | But writing is mostly an important tool to learn how to think. | 但写作大多是学习如何思考的重要工具. |
| 406 | And whether or not | 无论是否 |
| 407 | I was like, I'm gonna be a mathematician. | 我当时想,我要成为一个数学家。 |
| 408 | And the teacher was like, well, you still have to learn | 老师说,嗯,你还是要学 |
| 409 | how to think. | 怎样思考。 |
| 410 | And you still have to learn how to, you know, structure ideas. | 你还需要学会如何,你知道的, 结构的想法。 |
| 411 | And even if you never | 即使你从不 |
| 412 | write as a grownup, for anyone else to read, it is an important skill to learn how to think. | 写作是成年人,任何其他人阅读,学习如何思考是一个重要的技能. |
| 413 | And my takeaway is that there was an important learning from both of those. | 和我的外卖是,有一个重要的学习 从这两个。 |
| 414 | Still learning skills that help you think is incredibly valuable. | 仍然学习技能 帮助你 认为是令人难以置信的价值。 |
| 415 | Learning skills that are | 学习技能 |
| 416 | like clearly going to be in the realm of machines and not something that people value are clearly | 像是在机器领域 而不是人们所珍视的东西 |
| 417 | not valuable. | 没有价值。 |
| 418 | And those are like kind of obvious, right? | 这就像一种明显的,对不对? |
| 419 | So I think like most people would agree | 所以我认为大多数人会同意 |
| 420 | directionally that we should still learn these kind of general things. | 从方向上讲,我们仍应该学习这些一般的东西。 |
| 421 | We should not learn things | 我们不应该学东西 |
| 422 | that AI is going to do really well. | AI会做得很好。 |
| 423 | But then there's a question of, we develop our general | 但有一个问题,我们发展我们的将军 |
| 424 | skills. | 技术。 |
| 425 | We do our exercise as people. | 我们做我们作为人的活动。 |
| 426 | Like what remains in the domain of people in the future? | 比如说未来在人们的领域内还剩下什么? |
| 427 | Number one, I think obvious but worth stating, human values have inherent values because they | 第一,我觉得很明显,但值得指出的是,人类的价值观有内在的价值,因为它们 |
| 428 | are human. | 是人类 |
| 429 | We really do not seem to care, and I think this is great that we don't, about what | 我们似乎并不关心, 我认为这是伟大的 我们不,关于什么 |
| 430 | machines do for their own purposes. | 机器是为了自己的目的 |
| 431 | And I think no matter how good AI gets, this will continue to | 我觉得不管AI有多好, 这一切会继续下去 |
| 432 | be the case. | 如果是这样的话。 |
| 433 | In some sense the history of science is this like history of successive disappointments | 从某种意义上说,科学的历史就像 连续失望的历史 |
| 434 | to the main character energy of humanity. | 给人类的主角能量 |
| 435 | You know, we used to think that like the sun rotated | 你知道,我们曾经认为 像太阳旋转 |
| 436 | around the earth, and that we were like really at the center of that, and then we like realized | 环绕着地球,我们像 真正处于中心, 然后我们喜欢意识到 |
| 437 | there was this whole solar system, and then there was this whole galaxy, and this goes on and on, | 有整个太阳系, 然后有整个银河系, 并不断的, |
| 438 | you get to this like multiverse, and like there's really like you have to squint, and you're why | 感觉就像你必须精采地看 |
| 439 | people are the center of it. | 人们是其中的中心。 |
| 440 | But we feel like we are very much at the center of what matters, | 但是,我们觉得我们非常 处于重要的中心, |
| 441 | and we're at the center of our own stories for sure. | 并且我们绝对是自己故事的中心 |
| 442 | And I think this is great, and I think this | 我认为这是伟大的,我认为 |
| 443 | will keep going. | 继续前进。 |
| 444 | And I think that no matter how far AI goes from a capability perspective, as long as | 我认为,无论AI从能力角度来说有多远,只要 |
| 445 | we get the safety right, as long as we don't have one of these like loss of control incidents, human | 只要我们没有像失去控制事件 人类 |
| 446 | values will matter to all of us because there is a person behind them. | 价值观对我们所有人都很重要,因为后面有一个人。 |
| 447 | Work will matter because a | 工作很重要 因为 |
| 448 | person does it, and we're just always going to care that way. | 人这样做, 我们只是总是会 关心的方式。 |
| 449 | So figuring out how to be there, | 所以想着怎么去那儿 |
| 450 | how to be useful to other people, how to like create stuff of central importance to other people, | 如何对其他人有用, 如何喜欢创造一些对其他人至关重要的东西, |
| 451 | that's going to go on forever. | 这会永远继续下去。 |
| 452 | And I think learning how to do that is a surprisingly | 我觉得学习如何做到这一点是令人惊讶的 |
| 453 | learnable skill. | 学习技能。 |
| 454 | And let's say we get the distribution and development of AI right. | 让我们假设我们得到了AI的传播和发展右。 |
| 455 | What do you think is a problem facing the future today that people still dramatically underestimate | 你认为今天面对的 问题是 人们仍然极度低估 |
| 456 | our ability to solve? | 我们的解决能力? |
| 457 | Well, it's like a funny time to ask that because my conversation with | 嗯,这就像一个有趣的时间 问,因为我与 |
| 458 | people the last few days, even the people who are skeptics are like, you know what? | 过去几天的人, 即使是怀疑的人 都一样,你知道吗? |
| 459 | At this point, | 此时此刻 |
| 460 | 我了解,我记得 I get it. | |
| 461 | Like AI is going to solve every piece of knowledge. | 就像AI会解决每一个知识 |
| 462 | And I don't quite agree with that, | 我不太同意这一点, |
| 463 | but I directionally kind of feel that way. | 但我有这种感觉 |
| 464 | You know, like what it means to have a great life | 你知道,像它意味着 拥有一个伟大的生活 |
| 465 | and enable everybody to have the choice, freedom, flexibility to still make mistakes. | 并且让每个人都有选择、自由、灵活 来犯错误 |
| 466 | But | 不过 |
| 467 | you know, when people look back and zoom out and say, yeah, I had like a great life, | 你知道,当人们回头看,放大 说,是的,我有一个伟大的生活, |
| 468 | I think that will still be hard. | 我觉得这还是很难的 |
| 469 | I think that will always be hard in some sense. | 我认为这在某种意义上总是很困难。 |
| 470 | Maybe it's important that's hard. | 也许这很重要 这很难。 |
| 471 | I am hopeful that as these next few years roll on, | 我希望随着这几年的展开 |
| 472 | we push much more towards what feels like a new renaissance than a new industrial revolution. | 我们推动的更像是新的复兴 而不是新的工业革命 |
| 473 | But I think a lot of the energy in AI has been towards the industrial revolution. | 但我认为AI中的许多能量都是为了工业革命. |
| 474 | We're going | 我们要走了 |
| 475 | to automate everything. | 把一切自动化 |
| 476 | You know, we're just going to like perfectly take humans out of the equation | 你知道,我们只是会喜欢 完美地把人类从方程式 |
| 477 | and make the economy run faster and better. | 让经济运行得更快、更好 |
| 478 | And we're going to like give everybody some | 我们会喜欢给大家一些 |
| 479 | money and AI will decide what's important. | 钱和AI会决定什么是重要的。 |
| 480 | And people will have, you know, be this like | 人们会有,你知道, 像这样 |
| 481 | just cog running more smoothly in a machine. | 只要在机器里更顺利地运行 |
| 482 | And the machine is going to just be this incredible | 机器会变得不可思议 |
| 483 | malloc thing. | 杂货店的事。 |
| 484 | And I hope we get that one right. | 我希望我们把那个弄对了 |
| 485 | What do you think is the biggest thing standing in the way between this renaissance version of | 你觉得什么是最大的东西 挡住了这个复兴版本 |
| 486 | the future versus this industrial revolution style version of the future? | 未来与这个 工业革命风格版本的未来? |
| 487 | I actually, I think people want the renaissance version. | 实际上,我想人们想要复兴版。 |
| 488 | I think people want the creativity, | 我认为人们想要创造力 |
| 489 | the ability to explore stuff, the ability to discover knowledge, the ability to sort of | 探索事物的能力 发现知识的能力 排序的能力 |
| 490 | have this like wide open space of new freedom and directions and exploration. | 拥有如此广阔的空间 新的自由 方向和探索。 |
| 491 | I think a lot of the pressures of society and the economy are so | 我认为许多 社会和经济的压力是如此 |
| 492 | just positioned to like ruthless efficiency that probably there will have to be, you know, some | 只是定位为 喜欢无情的效率 可能有,你知道,一些 |
| 493 | force against that from the companies or the governments or people as a whole. | 公司或政府或全体人民的武力。 |
| 494 | We're going to turn it over to our audience questions now. | 我们现在要把它交给观众提问 |
| 495 | Okay. | 摆 |
| 496 | Hey Sam, my name is David Lee. | 嘿,山姆,我的名字是大卫李。 |
| 497 | I'm part of the MBA class of 2027 and I'm originally from | 我是2027年MBA班的一员 我最初来自 |
| 498 | the Bay Area. | 湾区 |
| 499 | My question for you is OpenAI had the foresight to invest heavily in compute in 2024 | 我的问题是OpenAI有远见 在2024年大量投资计算 |
| 500 | and 2025. | 和2025年的报告。 |
| 501 | And I think that investment is paying off with Astra and the speed of chat GPT. | 我认为投资与Astra 和聊天的GPT的速度是相抵的。 |
| 502 | What infrastructure are you all betting on now to further the mission of OpenAI? | 你们现在打赌什么基础设施来推进OpenAI的任务? |
| 503 | Um, robots. | 机器人们 |
| 504 | So I think people talk a lot about recursive self-improvement for | 所以我觉得人们经常谈论 递归自我改善 |
| 505 | AI模型. | |
| 506 | There is another loop, which is if you have robots that can build more robots | 还有另一个循环,那就是如果你有机器人 可以制造更多的机器人 |
| 507 | and then data centers that can effectively build more data centers, including like operating | 然后是能够有效建立更多数据中心的数据中心,包括运行中的数据中心 |
| 508 | chip fab facilities and such, you can like make a lot more compute all the way down to like | 芯片泡沫设施等等,你可以喜欢 做更多的计算 一路下来喜欢 |
| 509 | building more power plants. | 建造更多的发电厂。 |
| 510 | Um, but I think this is like, this is as big of a deal or almost as a | 嗯,但我觉得这就像, 这很重要 或几乎像一个 |
| 511 | good deal, let's say as the kind of potential recursive loop in AI research, uh, it gets much | 不错,比如说,作为AI研究中的潜在循环, 嗯,它得到了很多 |
| 512 | less talk. | 少说话。 |
| 513 | So that's one. | 所以这是一个。 |
| 514 | Another is we made our own chip, um, called jalapeno and we have another | 另一个是我们自己制造的芯片,嗯,叫做 jalapeno,我们有另一个 |
| 515 | one coming after it that are huge wins from a sort of intelligence per watt perspective relative to | 从每瓦的某种智能角度来看, |
| 516 | anything else out in the market. | 市场里的其他东西 |
| 517 | And we will continue to push there. | 我们将继续推进。 |
| 518 | And as the models get | 随着模特们的到来 |
| 519 | smarter, they'll be able to design, help us design smarter and smarter chips. | 更聪明,他们可以设计 帮助我们设计更聪明的芯片 |
| 520 | So, uh, that's like | 所以,呃,这就像 |
| 521 | another kind of one of these loops. | 另一种这样的循环。 |
| 522 | Um, we're investing a lot in the, the whole supply chain, | 嗯,我们投资了很多 在,整个供应链, |
| 523 | like fabs power, all the equipment and data centers. | 像电源, 所有的设备和数据中心。 |
| 524 | Uh, so we'll do a lot of that, | 呃,所以我们会做很多, |
| 525 | but those are probably the three big areas. | 但这些可能是三大领域。 |
| 526 | Hi, Sam. | 嗨,萨姆。 |
| 527 | I'm Sruti MBA class of 27 from India. | 我是来自印度的Sruti MBA级27人. |
| 528 | On your 30th birthday, you wrote a list of life lessons called the days are long, but the decades | 在你30岁生日的时候 你写了一份生活课列表 叫做"日子" |
| 529 | are short. | 简称很短。 |
| 530 | And now that you're 41 and a father, I was curious, what would you change about that | 现在,你41岁了,我很好奇,你会有什么变化? |
| 531 | list? | 列表吗 ? |
| 532 | Are there any that you would rewrite? | 你有要改写的吗? |
| 533 | I remember that I wrote that. | 我记得是我写的 |
| 534 | I don't remember | 我不记得了 |
| 535 | everything in there, but I could do a very short version right now. | 里面的一切, 但我现在可以做一个非常简短的版本。 |
| 536 | I think it was pretty long. | 我觉得很长 |
| 537 | It | 这个 |
| 538 | would be like, have a family, like build something that is important to you, you know, like spend | 就像一个家庭, 像建造一些对你很重要的东西, 你知道,像花费 |
| 539 | time having fun and doing the things that you find fulfilling, stay healthy. | 时间有乐趣和做 你发现满足的东西, 保持健康。 |
| 540 | It will be very short. | 它会很短。 |
| 541 | Uh, I, I kind of think this is a case where you should just like listen to old people | 呃,我,我觉得这是一个案件 你应该只是喜欢听老人 |
| 542 | when they look like they were all right. | 当他们看起来像 他们没事。 |
| 543 | Hello, Sam. | 你好,萨姆。 |
| 544 | Thank you for the time. | 谢谢你抽空来 |
| 545 | My name is Jericho. | 我叫杰里科,你好吗? |
| 546 | I'm from the class of 20, 28. | 我从20,28级。 |
| 547 | I'm from Peru and I have a question on what is a strong belief you | 我来自秘鲁,我有一个问题 关于什么是你的坚定信念 |
| 548 | held back in 2020 about how am AI will unfold that has come to be wrong and how it changed how you | 在2020年被搁置 关于AI将如何展开 已经变得错误 以及它是如何改变你 |
| 549 | I didn't think that we were going to be able to get to models that were, I guess the better | 我没想到我们能找到模特儿 我觉得这样更好 |
| 550 | version of saying is I thought the alignment problem was going to be harder. | 我觉得校对问题会更严重 |
| 551 | It may still | 可能还是 |
| 552 | turn out to be very hard from now. | 从现在开始 很难过 |
| 553 | Uh, in fact, I suspect it will get much harder. | 呃,事实上,我怀疑 它会变得更难。 |
| 554 | And I also think | 我也觉得 |
| 555 | it is really incredibly important from where we go from here. | 我们从这里出发,这是非常重要的 |
| 556 | But the fact that we got to models | 但事实上,我们得到了模型 |
| 557 | that are superhuman in some important, very superhuman in some important ways, | 在一些重要和非常超人的方式, |
| 558 | and alignment has so far been easier than expected. | 迄今为止,调整工作比预期容易。 |
| 559 | I would have said one of those two was wrong with | 我会说,其中之一是错的 |
| 560 | like a high degree of certainty. | 像高度的确定性。 |
| 561 | Now, the thing I'm worried about is that some people look at | 现在,我担心的是,有些人看着 |
| 562 | where we've gotten and say for N plus one, it's going to keep going and alignment is going to stay | 在那里,我们得到了 并说N+1, 它会继续 和调整会保持 |
| 563 | easy. | 轻点 |
| 564 | And I think it's actually not at all obvious to me that that's the case. | 我认为事实并非如此。 |
| 565 | And I also think that | 我也认为 |
| 566 | potential cost of alignment being wrong has gone up enormously. | 调整错误的潜在成本大幅上升。 |
| 567 | I increasingly believe, you know, | 我越来越相信,你知道, |
| 568 | for a while we can sandbox these systems and I think we can make some safety guarantees based | 暂时我们可以把这些系统装进沙盒 我想我们可以做出一些安全保障 |
| 569 | off the sandbox. | 离开沙盒。 |
| 570 | And then when that fails, because the models get so smart, we can't like | 当它失败了,因为模型变得如此聪明,我们不能喜欢 |
| 571 | really understand what a sandbox would even mean. | 真正理解一个沙盒会意味着什么。 |
| 572 | Then I think we'll have a monitorability that'll | 然后,我想我们将有一个可监测性 |
| 573 | work for a while. | 工作一段时间。 |
| 574 | But eventually, and the reason that I was talking about like that speed limit | 但最终,我所说的 速度限制的原因 |
| 575 | earlier, nothing but alignment I think actually will work. | 之前,除了调整外,别无他法 我认为实际上会奏效。 |
| 576 | I don't think any and alignment that | 我觉得没有调整 |
| 577 | like we're smart enough to understand and can really say like with confidence, it's okay to | 就像我们足够聪明 理解和可以真正说 像自信,没关系 |
| 578 | run this model and people will stay in control of the future. | 运行这个模式 人们会控制未来 |
| 579 | And I worry that some AI efforts | 我担心一些AI的努力 |
| 580 | around the world are so confident in alignment given how easy it's been so far that they're | 全世界都对调整充满信心 因为他们已经变得如此容易 |
| 581 | too willing to run ahead. | 太愿意继续前进。 |
| 582 | Hey Sam, thanks for being here. | 嘿,山姆,谢谢你在这里。 |
| 583 | I'm Josh Island, MBA class of 27. | 我是乔希岛,MBA级27人. |
| 584 | You said recently on the Decoded podcast that you think that the world should accept some bad | 你最近在解码播客上说 你觉得世界应该接受一些坏消息 |
| 585 | things happening in exchange for the benefits of AI. | 事情的发生 交换AI的好处。 |
| 586 | As you think about the world that your son's | 当你想起你儿子的世界时 |
| 587 | going to be growing up in, what are the bad things you think we should be okay with specifically | 长大后,什么是坏事 你觉得我们应该 具体地说, |
| 588 | versus where you draw the line? | 和你画的线对吗? |
| 589 | Yeah, I wish I had phrased that answer differently. | 是啊,我希望我用不同的措辞 回答。 |
| 590 | I like | 我喜欢 |
| 591 | the way I meant it seemed like so obvious in my head that it was non-controversial, | 我的意思是,它看起来 在我的脑海里是如此明显 它是无争议的, |
| 592 | but I see why it didn't come across that way. | 但我明白为什么它没那样出现 |
| 593 | I like you know people are going to have open | 我喜欢你知道人们会敞开心扉 |
| 594 | source models that are going to output a bunch of hate speech and troll internet forms and that are | 源代码模型,将输出 一堆仇恨言论和巨魔的互联网形式 |
| 595 | going to like hack into some systems and that are like you know try to commit fraud in some | 喜欢黑入一些系统 就像你知道 尝试欺诈一些 |
| 596 | sorts of ways on behalf of their user. | 代表用户的方式 |
| 597 | Like there will be a new generation of phone scams for example | 比如说会有新一代的电话诈骗 |
| 598 | much worse than the previous generation. | 比上一代还糟糕 |
| 599 | And I think there are people who would say if that's | 我认为有人会说,如果 |
| 600 | going to happen let's ban all open source models. | 我们禁止所有开源模型 |
| 601 | And the point I was trying to make in that podcast | 而我想在播客里说 |
| 602 | but I think got clipped out was banning all open source models would be an extremely high cost to | 但我认为被剪掉是禁止所有开源模型 将是一个极高的成本 |
| 603 | innovation and liberty in society and we shouldn't do it. | 社会的创新和自由 我们不应该这样做。 |
| 604 | But if you are going to have open source | 但是如果你要开源的话 |
| 605 | models and if you're going to let people run these locally on their computer, no matter how good you | 模型,如果你要让 当地人运行这些 在他们的计算机,不管你有多好 |
| 606 | are at like global agent monitoring, there's going to be some stuff that you wish someone hadn't | 就像全球特工监控 会有你希望有人没有的东西 |
| 607 | done. | 结束了。 |
| 608 | And I wouldn't take the trade of saying no open source models whatsoever. | 我不会接受说没有开源模型的交易 |
| 609 | Now there is some | 现在有一些 |
| 610 | threshold and this is where I think society needs a debate above which the models can do sufficient | 我觉得社会需要辩论, |
| 611 | harm that we don't give people the liberty to use them however they want. | 伤害我们不给人们自由 使用他们想要的。 |
| 612 | And obvious examples are | 显而易见的例子是 |
| 613 | no one should get to take down the internet, no one should be able to like cause physical harm to | 任何人都不能破坏互联网, 任何人都不能喜欢造成身体伤害 |
| 614 | another person, no one should be able to you know like yeah make a new kind of like bio terror attack | 别人,你不应该知道 像耶,做一种新型的生物恐怖袭击 |
| 615 | that they put out into the world. | 他们向世界投放。 |
| 616 | One of my learnings from that thing is if you tell people | 我从这件事上学到的一件事 就是如果你告诉别人 |
| 617 | to accept any amount of bad things happening, they'll just say I don't want anything bad happening | 他们只会说我不想发生什么坏事 |
| 618 | at all and they will lose the subtlety of like liberty is important too and what really matters | 他们会失去像自由一样的微妙性 也很重要 真正重要的是 |
| 619 | is where we draw the line. | 就是我们画线的地方 |
| 620 | Like if we go back to the example of like scams on phones you could | 碞钩и钩筿杠 scam |
| 621 | just take away all phones and no one would ever have a phone scam and we decide that's not worth it. | 只要拿走所有的电话 没有人会有一个电话骗局 我们决定这不值得。 |
| 622 | So that's the point I was trying to make. | 这就是我想说的 |
| 623 | I don't think we should take big risks because of society | 我觉得我们不该因为社会大冒险 |
| 624 | and in fact that's why I think AI is different than any technology that has come before it is | 事实上,这就是为什么我认为AI 与任何技术是不同的 |
| 625 | it is a a really new step in the amount of harm one person can cause and there are a bunch of | 这是一个真正的新步骤 伤害的程度 一个人可以造成和有一堆 |
| 626 | uncomfortable consequences that are going to follow from that about how society has to adapt | 社会必须如何适应 |
| 627 | and some freedoms we do have to give up but I don't think it means we like shouldn't get to | 有些自由我们不得不放弃 但我不认为这意味着我们不应该 |
| 628 | like you know run neural networks on our computers. | 就像你知道我们电脑上的神经网络 |
| 629 | Hi Sam I'm Shiraz MBA Class of 2028. 夏拉兹MBA班(英语:Hi Sam I'm Shiraz MBA Class of 2028) 夏拉兹MBA班(英语:Hi Sam I'm Shiraz MBA Class of 2028) 夏拉兹MBA班(英语:Hi Sam I'm Shiraz MBA Class of 2028) 夏拉兹MBA班(英语:Hi Sam I'm Shiraz MBA) 夏拉兹MBA班(英语: | |
| 630 | Thanks for being here with us. | 谢谢你陪我们 |
| 631 | So I'm coming in as a software developer so one of my questions was | 所以我来做软件开发员 所以我的一个问题就是 |
| 632 | as model capabilities get better over time one of the problems we're seeing is that | 由于模型能力随着时间推移而改善 我们所看到的一个问题是 |
| 633 | software emotes are increasingly being swallowed by better and better | 软件模具日益被更好和更好的吞噬 |
| 634 | AI models so for the next generation of you know budding entrepreneurs what advice would you have | AI模型,这样下一代 你认识新兴的企业家 你会有什么建议 |
| 635 | on what kind of motes within software or generally for people that are software developers you know | 对于软件开发者来说 软件里面的是什么种类的调子 |
| 636 | what kind of technologies or what kind of software can we can we kind of work on to build like a | 什么样的技术 什么样的软件 我们可以工作 建设像一个 |
| 637 | startup or you know an entrepreneurial venture in the future? | 创业还是你认识未来的创业企业? |
| 638 | To a first approximation I think | 我想第一个近似点 |
| 639 | there haven't been that many software motes in history there's other kind of motes there's like | 历史上没有过很多软件沼泽 还有别的沼泽 |
| 640 | network effects like Facebook or something there are scale effects like Google there are like data | Facebook之类的网络效应 谷歌之类的规模效应 数据之类的 |
| 641 | motes like a lot of enterprise software company has but software itself has never been like a | 就像很多企业软件公司一样 但软件本身从来没有像一个 |
| 642 | super strong mode it's you use it to get a temporary advantage to build up one of the traditional | 超强模式,就是用它来获得一个暂时的优势来建立传统模式之一 |
| 643 | like motes of a business and I kind of think that's just going to keep going the same way. | 就像一个生意的沼泽 我觉得那只是会 保持同样的方式。 |
| 644 | I'm Adam I'm an MBA too I'm also a chief of staff at the OpenAI Foundation | 我是亚当 我是MBA 也是OpenAI基金会的幕僚长 |
| 645 | just throwing that out there so I've kind of got a bit of a related or follow-up question to this | 刚刚扔出去 所以我有点 关联或后续的问题 |
| 646 | one as AI becomes more powerful there's a growing tension between making the technology broadly | 随着人工智能的增强 技术的普及 矛盾越来越大 |
| 647 | accessible to all of humanity with while also ensuring that it's developed and deployed safely | 供全人类使用,同时确保安全开发和部署 |
| 648 | so I wonder how do you think about the trade-off between open weight and closed weight models | 所以我想知道你如何看待 开放重量和封闭重量模型之间的权衡 |
| 649 | {\fn黑体\fs22\bord1\shad0\3aHBE\4aH00\fscx67\fscy66\2cHFFFFFF\3cH808080}什么才能让你觉得舒服 {\fn黑体\fs22\bord1\shad0\3aHBE\4aH00\fscx67\fscy66\2cHFFFFFF\3cH808080}对最有能力的人来说 | |
| 650 | models? | 模特儿吗? |
| 651 | Well I think open weight models are very important for society to have up to some fairly | 我认为开放体重模型对社会来说非常重要 能够公平掌握一些 |
| 652 | high level of capability and the question of exactly how much damage will society accept | 能力水平高的问题,以及社会会接受多少损失的问题。 |
| 653 | for the benefits of open weight models is something that is going to get very heavily debated in the | 开放重量模型的好处 将会引起激烈争论 |
| 654 | couple of years. | 两年。 |
| 655 | It's like also a moving target because the more that society builds up resilience | 这就像一个感人的目标 因为社会越多 建立复原力 |
| 656 | like OpenAI Foundation is I think hopefully going to do a lot of work to help prevent against future | 像OpenAI一样 基金会是,我认为 希望会做很多工作 帮助防止未来 |
| 657 | pandemics and then maybe the world can have more capable biomodels because we can respond better | 传染病,然后也许世界可以有 更有能力的生物模型,因为我们可以更好的反应 |
| 658 | than we could today so it's not a static question it'll like evolve as the world builds up more | 所以它不是一个静态的问题 它会喜欢进化 随着世界的建设 |
| 659 | resilience with AI but I don't think we could responsibly open source our current best internal | 人工智能的适应力,但我觉得我们无法负责任地打开 我们目前最好的内部源 |
| 660 | model anytime soon. | 任何时间的模型。 |
| 661 | I'm trying to think what would have to happen to the world like | 我试着想世界会怎样 |
| 662 | the total cyber defense model of the way the internet works would have to change. | 互联网工作方式的网络防御模式必须改变。 |
| 663 | We'd have | 我们本来可以 |
| 664 | to build up this bio layer and a long list of other stuff too. | 建立这个生物层 和一长串其他的东西。 |
| 665 | Now the reason that this is an | 现在,这是原因 |
| 666 | interesting and live question is how far ahead does anyone really believe that model is from | 有趣且活生生的问题就是 未来有多远 谁真的相信模型来自 |
| 667 | someone who will open source a model of that level? | 一个会打开该级别模型的人? |
| 668 | Not super and so I think this like | 不是超级,所以我觉得这样 |
| 669 | whatever this rolling advantage is of closed source closed weight models to help the world build the | 无论这种滚动优势是什么 闭源闭塞重量模型 帮助世界建设 |
| 670 | resilience and defend against the open models that I think the world does deserve is super important | 抗御世界所应得的开放模式, |
| 671 | иぃ笵 Open诀穦 | |
| 672 | like part of why I think it's cool that we have now the biggest foundation in the world is we | 就像为什么我觉得这很酷 我们现在最大的基础是我们 |
| 673 | can try to sprint at helping the world build a defense shield. | 可以尝试冲刺帮助世界建立防御盾牌. |
| 674 | My name is Serim first of all | 我的名字是塞里姆 首先 |
| 675 | thank you for being with us today. | 谢谢你今天和我们在一起。 |
| 676 | Last year when you joined David Perls how I write podcasts you | 去年你加入David Perls时 我如何写播客 |
| 677 | shared the analog note taking system you have where you carry a notebook with you all day | 共享模拟音符 使用系统,您拥有一个笔记本 |
| 678 | write down your thoughts like are there any areas in your life both personal and like professional | 写下你的思想 就像你生活中的任何领域 无论是个人还是职业 |
| 679 | where you try to stay more like low tech and more analog to preserve some skills you have | 你尽量保持低科技和模拟 保存你的一些技能 |
| 680 | as a person and as a professional and like what would be your suggestions for us to follow the | 作为一个人,作为一个专业人员,喜欢什么 你会建议我们遵循 |
| 681 | same path? | 同样的道路? |
| 682 | Like the note pad I think I'm just wrong there and I probably should switch to | 就像纸板,我觉得我错了 我也许应该换个 |
| 683 | a computer but it's like an old force of habit I think the analog stuff that matters is like all | 电脑,但就像旧的习惯力 我认为模拟的东西很重要 |
| 684 | of the stuff with is speaking of like old people are always right it's like spend a lot of time | 里面的东西是说,像老人总是对的 它就像花了很多时间 |
| 685 | with other people like have meals together like go out into nature a lot that you know exercise | 和别人一起吃饭 喜欢去自然界 很多你知道的锻炼 |
| 686 | there's like a long list of analog stuff that has worked for a long time but I think it's all the | 有一长串类似的东西 已经工作了很长时间,但我想这是所有 |
| 687 | obvious stuff and trying to be like oh you know someone writes with a note pad like they're | 明显的东西 并试图像 哦 你知道有人写 纸条像他们 |
| 688 | just wrong that one you can probably just stop. | 只是错了,你也许可以停止。 |
| 689 | Sam I have one final question for you today | 山姆 我今天还有最后一个问题要问你 |
| 690 | at the GSB we talk about changing lives changing organizations and changing the world | 改变生活 改变组织 改变世界 |
| 691 | as you look at the world today what do you think future leaders need to focus on to drive change? | 当你看着今天的世界,你认为未来的领导人需要专注于什么来推动变革? |
| 692 | I mean my very obvious answer is AI like if we if we pick the the easiest one of those maybe | 我的意思是,我非常明显的答案是AI 就像如果我们选择最简单的一个 |
| 693 | are the most tractable when you talk about like changing organizations the degree to which | 当你谈论像改变组织的程度时, 是最能吸引人的 |
| 694 | I think it's possible to design a different kind of organization now with a lot of AI relative to | 我认为,现在可以设计一个不同的组织, 有很多AI相对 |
| 695 | each person in it and I think it's just like an underexplored sort of area of study | 里面每个人,我觉得这就像一个 探索不足的领域 |
| 696 | like what it means to set up the best possible organization today I think it's like very very | 就像今天建立最好的组织意味着什么 我想它就像非常 |
| 697 | different my kind of like this is pure speculation I've never studied to know if this is true or | 我从未研究过这些是不是真的 |
| 698 | not but my sense is that the way like most organizations are designed came from copying | 但我觉得大多数组织设计的方式 来自于复制 |
| 699 | the military and it was like a way that we set up this very hierarchical command and control thing | 军队和它就像一种方法 我们设置了这个非常等级 指挥和控制的事情 |
| 700 | and to scale beyond a certain point that was the one example we had of something that could do it | 超越某一点,这是我们唯一能做的 |
| 701 | organizations seem to have like some of that and then over the last few decades | 组织似乎也喜欢一些,然后在过去几十年里 |
| 702 | the tech industry and other industries too have innovated on that and like somewhat improved it | 科技产业和其他产业也对此进行了创新,并有所改进。 |
| 703 | still inspired by that same basic principle and one of the things that's been extremely interesting | 仍然受到同样的基本原则的启发 以及非常有趣的事情之一 |
| 704 | to me about watching these AI systems is if you look at I think it was 10,000 the 10,000 agents | 对我来说,看这些人工智能系统是 如果你看看,我认为是10,000个特工 |
| 705 | that solve Navier-Stokes they work together in a way that is unlike any organization I could | 解决了纳维耶-斯托克斯 他们合作的方式 不像任何组织 我能够 |
| 706 | imagine there was like no leader there were no credit squabbles there was like perfect alignment | 想象没有领袖 没有信用争吵 有像完美对齐 |
| 707 | there was no bickering everyone's just trying to like solve this goal and that's there's like | 没有争吵 每个人都想解决这个目标 |
| 708 | beautiful flaws in humans that are going to mean we never do that people are going to work in you | 人类的美丽缺陷 意味着我们永远不会做 人们会为你工作 |
| 709 | know many of the same ways that they've worked in the past and I wouldn't like want to just be that | 他们以前工作过的许多方法 我也不想这样 |
| 710 | kind of like you know emotionless world obviously but I think there is still something to learn from | 那种像你了解无情感的世界 显然, 但我认为还有 从学习的东西 |
| 711 | that and as we have watched inside of OpenAI teams that really embrace AI and be much smaller | 正如我们所看到的 OpenAI 团队 真正拥抱AI 并且要小得多 |
| 712 | change the way that they work I'm sure you all will be the people that go figure out | 改变他们的工作方式 我相信你们都会是想出办法的人 |
| 713 | how to design new organizations and I think the world will really benefit from it | 如何设计新的组织 我认为世界会真正从中获益 |
| 714 | Sam this has been a pleasure thank you please join me in thanking Sam Altman | 山姆,我很高兴,谢谢你 和我一起感谢山姆・阿尔特曼 |
该视频共有字幕 714 条。解锁更多字幕为会员功能，请移动到 价格