【中英字幕】26.11.12黄仁勋最新专访 【双字&原声】
共 1 集
剧集目录
| 集 | 标题 | 时长 |
|---|---|---|
| P1 | 【中英字幕】26.11.12黄仁勋最新专访 【双字&原声】 | 40:33 |
P1 · 【中英字幕】26.11.12黄仁勋最新专访 p01 【双字&原声】 (P1)
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:01 | NVIDIA is really in its biggest crisis in its three-decade history of operating. | NVIDIA确实处于30年运作史上最大的危机之中. |
| 0:05 | But the episode has transformed the meaning of our time with Huang. | 但剧集改变了我们与黄 Huang的时代意义. |
| 0:08 | We're probably the only AI company in the world that works with every AI company in | 我们可能是世界上唯一一家 与每个AI公司合作的AI公司 |
| 0:12 | the world. | 这个世界。 |
| 0:12 | So at what point does that dependence on one single company become a systemic risk? | 那么,对单一公司的依赖在何时成为系统性风险? |
| 0:17 | Well, they don't depend on us, they choose us. | 他们不依赖我们 他们选择我们 |
| 0:20 | What began as a window into the massive influence of a man at the peak now looks more like a | 刚开始 进入一个男人在顶峰的巨大影响力的窗口 现在看起来更像一个 |
| 0:24 | portrait of concentrated power. | 集中力量的肖像 |
| 0:26 | It's sent an entire industry into crisis and raises tough questions about how much that | 这让整个行业陷入危机, 并提出了一个棘手的问题 关于这个多少钱 |
| 0:31 | industry should depend on one person and their company. | 工业应该依赖一个人和他们的公司。 |
展开字幕全文（699 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | NVIDIA is really in its biggest crisis in its three-decade history of operating. | NVIDIA确实处于30年运作史上最大的危机之中. |
| 2 | But the episode has transformed the meaning of our time with Huang. | 但剧集改变了我们与黄 Huang的时代意义. |
| 3 | We're probably the only AI company in the world that works with every AI company in | 我们可能是世界上唯一一家 与每个AI公司合作的AI公司 |
| 4 | the world. | 这个世界。 |
| 5 | So at what point does that dependence on one single company become a systemic risk? | 那么,对单一公司的依赖在何时成为系统性风险? |
| 6 | Well, they don't depend on us, they choose us. | 他们不依赖我们 他们选择我们 |
| 7 | What began as a window into the massive influence of a man at the peak now looks more like a | 刚开始 进入一个男人在顶峰的巨大影响力的窗口 现在看起来更像一个 |
| 8 | portrait of concentrated power. | 集中力量的肖像 |
| 9 | It's sent an entire industry into crisis and raises tough questions about how much that | 这让整个行业陷入危机, 并提出了一个棘手的问题 关于这个多少钱 |
| 10 | industry should depend on one person and their company. | 工业应该依赖一个人和他们的公司。 |
| 11 | How do you go to all these cities? | 你怎么去这些城市? |
| 12 | How do you keep up? | 你怎么跟上来? |
| 13 | Like, what are your travel must-haves? | 喜欢,什么是你的旅行必须拥有? |
| 14 | I just packed my bag. | 我刚打包了 |
| 15 | I was up at midnight and I've been working since. | 我午夜就起来工作了 |
| 16 | And here you are. | 你在这里。 |
| 17 | Without dragging this out too long, I would actually like to welcome a surprise guest, | 不想拖太久,我其实想欢迎一位惊喜嘉宾 |
| 18 | Mr. Jensen Huang. | 黄延生先生 |
| 19 | On day one of his European tour, Huang hyped up AI agents. | 在他欧洲巡演的第一天,黄hy对AI的特工们大叫. |
| 20 | This is really the beginning now of AI becoming useful. | 这确实是AI变得有用的开始. |
| 21 | We're seeing agents emerging in just about every single industry. | 我们看到每个行业都有代理商出现 |
| 22 | And raffled off two of NVIDIA's new souped-up AI computers that retail for thousands of | 并捣毁了NVIDIA的两台 新的人工智能电脑 用于零售数千台 |
| 23 | dollars each. | 每个美元。 |
| 24 | Leaving the stage, he was instantly mopped by fans. | 离开舞台,他立刻被粉丝们迷倒. |
| 25 | Can I get the love for you? | 我能得到你的爱吗? |
| 26 | It's my second time to meet you. | 这是我第二次见到你。 |
| 27 | Can I take a picture? | 我能拍照吗? |
| 28 | Just a picture. | 只是一个图片。 |
| 29 | So this is Jensen. | 这是詹森。 |
| 30 | Everyone's trying to get a picture with Jensen. | 大家都想和Jensen拍个照 |
| 31 | Clearly it's very contagious. | 很明显是传染性的 |
| 32 | Excuse us, ladies. | 对不起,女士们。 |
| 33 | We're heading this way. | 我们往这边走 |
| 34 | Excuse us. | 对不起 |
| 35 | Huang was here to drumbeat his vision for a global AI ecosystem, announcing investments | 黄先生是来打鼓的 他的愿景 全球AI生态系统,宣布投资 |
| 36 | and partnerships at all levels of the technology. | 和各级技术伙伴关系。 |
| 37 | France is jockeying to become Europe's AI center. | 法国正在骑马成为欧洲的AI中心. |
| 38 | Noteworthy technical talent has come out of the country, powering major companies like | 著名的技术人才已经从国内涌现出来,为大公司提供动力 |
| 39 | 总部设在巴黎的Mistral和位于纽约的Hugging Face,NVIDIA同意以12.9美元收购. | |
| 40 | billion. | 10亿 |
| 41 | Jensen likes to say AI is like a five-layer cake. | Jensen喜欢说AI就像一个五层的蛋糕. |
| 42 | From the bottom up, energy, chips, infrastructure, the models, and the apps on top. | 自下而上,能量,芯片,基础设施,模型,以及顶部的应用. |
| 43 | Now, NVIDIA started out squarely in layer two, making the chips. | 现在,NVIDIA在第二层开始, 制造芯片。 |
| 44 | But that has since changed, and now NVIDIA is investing up and down all the layers of | 但是,这已经改变了, 现在NVIDIA正在投资 所有层次 |
| 45 | the cake, in power, in data centers, in the AI labs, in startups, which means no matter | 蛋糕,在权力, 在数据中心, 在AI实验室, 在初创企业, |
| 46 | how you slice it, more and more of the AI industry runs through NVIDIA. | 你如何切开它, 越来越多的AI产业运行 通过NVIDIA。 |
| 47 | You know what? | 你知道吗? |
| 48 | Since Emily's going to be filming, I want to put my whiskey down. | 既然艾米莉要拍了 我想把威士忌放下来 |
| 49 | In the last year alone, we have invested $8 billion in European startups. | 仅去年一年,我们就向欧洲创业企业投资80亿美元。 |
| 50 | And in the next week, I'm going to announce another two more billion dollars. | 在接下来的一周里,我还要再宣布另外两亿美金. |
| 51 | There's been a lot of talk that AI was going to end humanity. | 有不少说法认为AI会终结人类. |
| 52 | You're saying it's a zero-person chance. | 你说这是零人机会 |
| 53 | It's just an engineering problem, and you got it solved. | 这只是个工程问题 你解决了 |
| 54 | That's what you're saying. | 汝所说法. |
| 55 | No. | 没有 |
| 56 | What I say is that it is in our control. | 我说的是,这是在我们控制之下。 |
| 57 | Between now and the end of humanity, we're going to have a lot of work to do. | 从现在到人类的终结,我们还有许多工作要做. |
| 58 | Advancing capability, advancing safety are not in conflict. | 推进能力,推进安全不发生冲突. |
| 59 | They are, in fact, one. | 事实上,它们就是其中之一。 |
| 60 | OK, you guys, I want to thank all of you for coming. | 好了,各位,我要感谢你们所有人前来。 |
| 61 | This is the beginning of AI in Europe. | 这是AI在欧洲的开始. |
| 62 | Let's go build it together, OK? | 让我们一起建,好吗? |
| 63 | Thank you. | 谢谢 |
| 64 | Cheers. | 欢呼声。 |
| 65 | This way, sir. | 这边请,先生 |
| 66 | So, there's more stuff happening later tonight. | 所以,还有更多的事情 今晚晚些时候发生。 |
| 67 | He's going to be meeting with government officials, other CEOs, other important people in the | 他将会见政府官员、其他CEO、其他重要人物 |
| 68 | tech industry in France, and next stop, London. | 法国的科技产业 下一站伦敦 |
| 69 | Estimates that the EU will need to spend at least $3 trillion by 2035 just to be truly | 估计到2035年欧盟至少需要花费3万亿美元, |
| 70 | self-sufficient in technology. | 技术自给自足。 |
| 71 | Just got off the train in London, and I'm off to meet Jensen. | 刚从伦敦的火车上下车 我去找詹森 |
| 72 | Twenty-seven years ago, NVIDIA went public, and they put me in this hotel. | 27年前,NVIDIA公开了 他们把我关在这酒店里 |
| 73 | Oh, really? | 是吗? |
| 74 | Yes. | 对 |
| 75 | It was the first time I ever had a butler. | 这是我第一次有管家。 |
| 76 | It was 2000, and you'll be blown away $300 million. | 那是2000年,你会被吹走3亿美元。 |
| 77 | Things have changed. | 事情变了 |
| 78 | $300 million. | 3亿美元。 |
| 79 | That was a lot back then, though. | 那时候是很多 |
| 80 | No, no, no, it was fine. | 不,不,不,这是罚款。 |
| 81 | I was just happy to make it. | 我很高兴能来 |
| 82 | You were such an unlikely winner back then. | 那时候你是这么不可能的赢家 |
| 83 | Yeah, I was just happy to stay in business. | 是啊,我只是很高兴 留在商业。 |
| 84 | You didn't know what I was doing. | 你不知道我在干吗 |
| 85 | Didn't know how to write a business plan. | 不知道怎么写商业计划 |
| 86 | And now? | 现在呢? |
| 87 | No, still don't know how to write a business plan. | 不,还是不知道 如何写一个商业计划。 |
| 88 | What makes this trip a success? | 是什么使这次旅行成功? |
| 89 | Is it, like, government commitments, selling chips, data centers? | 像是政府承诺、销售芯片、数据中心吗? |
| 90 | You know, I don't have an endgame. | 你知道,我没有结束游戏。 |
| 91 | I don't have a roadmap. | 我没有地图 |
| 92 | You know, I don't have a plan. | 你知道,我没有计划。 |
| 93 | I'm not trying to get anywhere. | 我可没想走 |
| 94 | You know, there's no completion. | 你知道,没有完成。 |
| 95 | I'm kind of one of those people that the journey is the destination. | 我是那种那种人 那些旅程是目的地。 |
| 96 | Huang's bio might make you think otherwise. | 黄先生的生物资料可能让你不这么想 |
| 97 | The highlight reel goes something like, | 亮点卷轴会像... |
| 98 | In 1963, | 1963年时. |
| 99 | moved to the US and earned a national ranking in table tennis, | 搬到美国 并获得全国排名 在桌球, |
| 100 | famously loved his job washing dishes at Denny's, | 他喜欢在丹尼家洗碗 |
| 101 | skipped a couple grades, | 跳过几个年级 |
| 102 | started college at 16 at Oregon State, | 在俄勒冈州16岁开始上大学 |
| 103 | studied electrical engineering, | 学习电气工程, |
| 104 | became a chip designer at competitor to BAMD. | 成为BAMD竞争对手的芯片设计师. |
| 105 | Then in 1993, | 后来在1993年 |
| 106 | NVIDIA诞生之初谦卑. | |
| 107 | For $200, I bought 15%, I think it was 20% of NVIDIA. | 200美元,我买了15%, 我想是20%的NVIDIA。 |
| 108 | Huang and his two business partners wanted to create chips, | 老黄和两个生意伙伴想创造筹码 |
| 109 | 特别是GPU, | |
| 110 | that would make for better video game graphics. | 这样可以改进电子游戏图形。 |
| 111 | But it was a rocky start. | 但它是一个摇滚的开始。 |
| 112 | 2006年,NVIDIA揭幕CUDA, | |
| 113 | software that allowed its chips to do more than run video games, | 软件允许它的芯片做更多的 运行电子游戏, |
| 114 | from weather forecasting to gene sequencing. | 从天气预报到基因测序 |
| 115 | AI researchers soon discovered those same chips | AI研究人员很快发现了同样的芯片 |
| 116 | could be used to train their models. | 可以用来训练他们的模型。 |
| 117 | When word got back to Huang, he went all in, | 传回黄帝时,他全进去了 |
| 118 | betting big on tech built for AI, | 以人工智能技术为赌注 |
| 119 | before most of the industry caught on. | 之前,大多数行业 抓住。 |
| 120 | NVIDIA powered the early breakthroughs that brought AI mainstream, | NVIDIA推动早期的突破, 带来了AI的主流, |
| 121 | and it's been one of the most lucrative bets in corporate history. | 这是公司历史上最赚钱的赌注之一 |
| 122 | Today, you'll find NVIDIA at the core of most major tech companies. | 今天,你会发现NVIDIA是大多数主要技术公司的核心. |
| 123 | OpenAI, Meta, Google, Anthropic. (原始内容存档于2018-09-29). | |
| 124 | What happens when your views differ from the president's? | 当你的观点与总统不同时会怎样? |
| 125 | How do you walk that line? | 你怎么走这条线? |
| 126 | We don't have to agree on anything. | 我们不必就任何事情达成一致。 |
| 127 | It's not my job to agree or disagree. | 同意或不同意不是我的职责. |
| 128 | It's my job to inform him. | 我的职责是通知他 |
| 129 | The decisions that we make impact not just the American citizens, | 我们所做的决定不仅影响美国公民 |
| 130 | the prosperity of our future, it impacts the world. | 我们未来的繁荣 影响着世界 |
| 131 | And so, if ever I have an opportunity to help inform him | 所以,如果我有机会帮助他 |
| 132 | so that he could make a better decision, | 以便他做一个更好的决定, |
| 133 | whatever the decision is, whether it serves me, | 不管决定是什么 不管它对我有用 |
| 134 | whether it's counter my interests, it's not my concern. | 不管这是否违背我的利益, 这不关我的事。 |
| 135 | It's only my concern that... | 我只关心... |
| 136 | The main point here is that nobody wants to seed competition | 这里的要点是,没有人想要种子竞争 |
| 137 | when it comes to China. | 当於中国. |
| 138 | In late September, | 9月下旬, |
| 139 | Hwang was among a group of tech titans called to the White House for a summit | 黄某1是一群科技巨头 被召到白宫参加峰会 |
| 140 | where lunch was served with a side of super intelligence. | 在那里,午餐得到了 一边的超级智能。 |
| 141 | We're using the word super because the other word is a fake word. | 我们使用超级词是因为另一个词是一个假词. |
| 142 | Hwang stood alongside the president as he rebranded AI as super intelligence | 黄先生站在总统旁边 他把AI重新命名为超级情报 |
| 143 | and lauded a new White House accord that puts the tech industry in charge | 并称赞新的白宫协议 让科技行业负责 |
| 144 | of voluntarily policing itself as it races to advance AI. | 自愿维持治安本身 因为它竞相推进AI。 |
| 145 | You've spent the last three months trying to demystify AI. | 你最近三个月都在试图解密AI |
| 146 | You've said it's not a new species, it's just a bunch of code. | 你说这不是一个新物种 只是一堆密码 |
| 147 | Isn't calling it super the opposite of that? | 说它超乎想象吗? |
| 148 | Actually, I think calling it super is the opposite of artificial. | 事实上,我认为称它为超级 与人工相反。 |
| 149 | And super intelligence is the new word. | 而超级智能是新词. |
| 150 | Yeah, I do think it's better than artificial. | 是的,我觉得这比人造的好 |
| 151 | And I think President Trump's point is that artificial just says fake. | 我认为川普总统的观点是 人造的只是说假 |
| 152 | Well, everything is fake. | 嗯,一切都是假的。 。 。 |
| 153 | We built it. | 我们造的 |
| 154 | And so I think at some level, this is fake. | 所以我觉得从某种程度上来说 这是假的 |
| 155 | We're in a fake structure. | 我们在一个假结构。 |
| 156 | It was built by humans, but we don't say that. | 它是由人类建造的, 但我们不说。 |
| 157 | You know, it's a fake horse, otherwise known as a car. | 你知道,这是一匹假马,否则被称为汽车. |
| 158 | But we don't say that. | 但我们不说。 |
| 159 | Is it super intelligence? | 是超级情报吗? |
| 160 | Absolutely. | 当然 |
| 161 | I mean, it does a better job researching than I do. | 我的意思是,它做一个更好的工作 研究比我。 |
| 162 | It does a better job writing than I do. | 写作比我好 |
| 163 | It's not super at everything that we do, but it's super at many things that we do. | 我们做的每件事都比不上超级 但很多事情都比不上超级 |
| 164 | That day at the White House, one critic said the pledge he made them sign | 当天在白宫,一位评论家说,他让他们签的誓言 |
| 165 | was great for the cameras, but utterly meaningless. | 对摄像头来说很好 但完全毫无意义 |
| 166 | The people building the technology are effectively helping | 技术的建设 正在有效地帮助 |
| 167 | to write the rules that govern it. | 写规则来管理它。 |
| 168 | Is that appropriate? | 这样合适吗? |
| 169 | I think it's very appropriate. | 我觉得这很合适 |
| 170 | And the reason for that is this, before we expect governments | 原因是,在我们期待政府之前, |
| 171 | and third parties to regulate us, we should hold ourselves | 和第三方管理我们, 我们应该坚持自己 |
| 172 | to the highest possible responsibilities, standards, accountability. | 最高责任、标准、问责制。 |
| 173 | The structure that was described is very similar to the way that safety systems | 所描述的结构与安全系统的方式非常相似 |
| 174 | are internally audited and externally audited by many industries. | 由许多行业进行内部审计和外部审计。 |
| 175 | Even if we would have had a government regulation | 就算我们有政府条例 |
| 176 | or international government regulation, | 或国际政府条例, |
| 177 | who would do the auditing if we didn't have the ecosystem of companies | 如果我们没有公司生态系统 谁会进行审计 |
| 178 | that have the capability to do so? | 是否有能力这样做? |
| 179 | And so we have to go and create that industry. | 因此,我们必须去 创建这个行业。 |
| 180 | We have to go find startups. | 我们得去找新手 |
| 181 | We have to encourage countries and | 我们必须鼓励各国和 |
| 182 | companies to get involved. | 公司参与。 |
| 183 | And so we've got a lot of developing to do. | 因此,我们有很多工作要做。 |
| 184 | This call for development is a familiar part of Huang's message to the world | 这个发展呼吁是黄帝向世界传达的信息中一个熟悉的部分 |
| 185 | that all countries need to build and control their own AI. | 所有国家都需要建立和控制自己的AI。 |
| 186 | He's popularizing a term for this, sovereign AI. | 他在宣传这个术语 主权AI |
| 187 | The realization that control of data is power has compelled other countries | 由于认识到控制数据是权力,其他国家不得不这样做。 |
| 188 | to quickly develop their own models and infrastructure. | 迅速发展自己的模式和基础设施。 |
| 189 | Establishing an independent AI ecosystem entirely from scratch would take decades. | 完全从零开始建立一个独立的人工智能生态系统将需要几十年的时间。 |
| 190 | So Huang's pitch is each country can simply build their own capabilities | 所以黄先生的投手是每个国家 都能建立自己的能力 |
| 191 | on Nvidia, which also means the more sovereign AI, | 在Nvidia, 也意味着更主权的AI, |
| 192 | the more demand for Nvidia chips. | 对Nvidia芯片的需求越大 |
| 193 | You're traveling across Europe, promoting this idea of sovereign AI. | 你正在穿越欧洲,宣传这个主权AI的想法. |
| 194 | Is it still sovereign if it's built on American chips? | 如果它是在美国芯片上建造的,它仍然拥有主权吗? |
| 195 | Yes. | 对 |
| 196 | Sovereignty basically means that you have to control your own experience. | 主权基本上意味着你必须控制自己的经验。 |
| 197 | And in the case of AI, that means control your own data, your data, | 就人工智能而言 这意味着控制自己的数据 你的数据 |
| 198 | your life experience, your company experience, | 你的生活经验,你的公司经验, |
| 199 | your country's culture, values, the institutional knowledge | 文化、价值观、机构知识 |
| 200 | of your people, the flywheel of creating more data, more experience, | 创造更多数据 更多经验的飞轮 |
| 201 | more intelligence that belongs to you. | 更多的情报属于你。 |
| 202 | Sovereignty means that you have the ability to determine your own destiny. | 主权意味着你有能力决定自己的命运. |
| 203 | You might need teamwork. | 你可能需要团队精神 |
| 204 | You might need collaboration, might partnership and support from other people. | 你可能需要合作, 可能伙伴关系和他人的支持。 |
| 205 | You could, of course, stand on the shoulder of giants. | 你当然可以站在巨人的肩上 |
| 206 | You could, you know, start with somebody else's technology, open models, for example. | 你可以,你知道,从别人的技术开始, 比如开放模型。 |
| 207 | But then you could adapt it and fine tune it. | 但那样你就可以调整它 并调整它。 |
| 208 | So I think in every single way, open models, Nvidia's technology, | 所以我想在每一个方面,开放的模型, 恩维迪亚的技术, |
| 209 | their own ecosystem, their own data flywheel, their own agency, | 他们自己的生态系统, 他们自己的数据飞轮, 他们自己的机构, |
| 210 | their own aspirations and hopes, that gives them sovereignty. | 他们自己的愿望和希望,这给了他们主权。 |
| 211 | I understand we're in the home of humanoids. | 我知道我们在人类的家中 |
| 212 | Oh, I see a human right there. | 哦,我看到一个人在那里。 |
| 213 | And there's a couple right in front of me. | 还有一对夫妇 就在我面前。 |
| 214 | In London, Huang bounced around between events, | 在伦敦,黄飞鸿在事件之间反弹 |
| 215 | handing out computers, taking few, if any, breaks. | 发送计算机, 占用很少, 如果有中断。 |
| 216 | All I've had today is a quarter of a half of a sandwich. | 我今天只吃了四分之一半个三明治 |
| 217 | And going full Oprah with startup capital. | 和去满奥普拉与启动资本。 |
| 218 | I'm investing in you. | 我在投资你。 |
| 219 | While hammering on his message that AI can be safe and controllable. | 一边敲打着他的信息,说AI可以是安全的,可以控制的. |
| 220 | And so if you're asking an autonomous system to make you an omelet, | 所以如果你要求一个自主系统 让你做个煎蛋卷 |
| 221 | you don't have to give it access to a chainsaw, OK, or a shotgun. | 你不必让它接触电锯,OK,或猎枪。 |
| 222 | AI news was coming thick and fast as Huang made the rounds. | AI的消息随着黄飞鸿的到来而变得浓密迅速. |
| 223 | At one event, he gave a shout out to London based data center | 有一次,他向伦敦的数据中心大喊大叫 |
| 224 | 操作员 Nscale, 由 Nvidia 支持 。 | |
| 225 | I don't know, is Josh here? | 我不知道,乔希在吗? |
| 226 | Nscale在吗? | |
| 227 | You could be the most valuable UK startup. | 你可能是英国最有价值的创业者 |
| 228 | Created in the last 10 years. | 创建于近十年. |
| 229 | But just a few days later, investors got a case of the AI build out | 但就在几天后,投资人 得到了一个案件AI建立 |
| 230 | 和质疑Nscale的估价。 | |
| 231 | Microsoft pulled out of a deal with the company, citing concerns | 微软以担心为由与公司解除交易 |
| 232 | about the cost of electricity in the UK. | 英国的电费 |
| 233 | You know, UK is really the Goldilocks moment. | 你知道,英国真的是金锁的时刻。 |
| 234 | You have every aspect necessary for to take advantage of this new | 你拥有利用这个新的 |
| 235 | industrial revolution. | 工业革命。 |
| 236 | And you just got to activate it. | 你只需要启动它。 |
| 237 | There's a strange general behavior in the UK. | 在英国有一种奇怪的一般行为. |
| 238 | If there was a million ways to win, you find the one reason not to do it. | 如果有百万种方法可以赢, 你找到一个理由不这样做。 |
| 239 | And I think you overthink things. | 我觉得你想得太过分了 |
| 240 | As more countries develop more powerful AI, does that become harder | 随着更多的国家发展出更强大的AI,这是否变得更加困难? |
| 241 | to police and align on safety? | 警方和保安部门? |
| 242 | Do we need something like a UN for AI? | 我们需要一个像联合国一样的AI吗? |
| 243 | Maybe. | 也许 |
| 244 | First of all, dialogue needs to happen. | 首先,需要进行对话。 |
| 245 | You know, as you know, with any movement, it takes two people to | 你知道,你知道,任何运动, 它需要两个人 |
| 246 | start anything. | 开始什么。 |
| 247 | And so I think every country wants to be part of this | 所以我认为每个国家都想参与其中 |
| 248 | dialogue. | 对话框。 |
| 249 | How could any country not be part of the dialogue of | 任何国家如何不参加对话? |
| 250 | intelligence? | 情报? |
| 251 | Despite Huang's call to action, international tensions over AI | 尽管黄帝呼吁采取行动,但国际间对AI的紧张关系 |
| 252 | development have been rising. | 发展一直在增加。 |
| 253 | He's been pushing the White House for favorable regulation and playing | 他一直在推动白宫 争取有利的监管和演奏 |
| 254 | both sides of the AI arms race. | AI军备竞赛的双方. |
| 255 | Since AI advancement is limited by how good the chips are, Nvidia's | 由于AI的推进受到芯片有多好的限制,Nvidia的 |
| 256 | hardware is a crown jewel for the United States, technology that some | 硬件是美国的皇冠珠宝 有些技术 |
| 257 | leaders say should be safeguarded as a unique advantage. | 领袖们说,应该把保护作为一个独特的优势。 |
| 258 | But Huang has been among the loudest voices in Silicon Valley, arguing | 但黄帝是硅谷里最吵的声音之一 |
| 259 | that limiting China's access to U.S. | 限制中国进入美国 |
| 260 | technology will only make domestic capabilities stronger in the long run. | 从长远来看,技术只会加强国内能力。 |
| 261 | The U.S. approved selling older Nvidia chips to China, but there are | 美国批准向中国出售更古老的Nvidia芯片,但有 |
| 262 | companies like DeepSeek that are prioritizing building on Chinese | 像DeepSeek这样的公司 以中文为优先 |
| 263 | hardware. | 硬件。 |
| 264 | Is the promise of that market slipping away? | 市场的承诺会消失吗? |
| 265 | At all? | 一点都没有? |
| 266 | Or shrinking at all? | 还是完全缩小? |
| 267 | Well, it's completely blocked from us today. | 嗯,它今天完全 封锁了我们。 |
| 268 | I did my best to inform the policymakers in the United States of | 我尽力告知美国决策者 |
| 269 | the importance of engaging the Chinese market and making sure that we | 参与中国市场和确保我们 |
| 270 | serve that market for as long as we can. | 尽我们所能为市场服务 |
| 271 | I think we made a mistake and this mistake is hard to unwind. | 我认为我们犯了一个错误,这个错误很难解开。 |
| 272 | But I did my best. | 但我尽力了 |
| 273 | That's all I could do. | 这就是我能做的。 |
| 274 | There are more Chinese AI researchers than any other nation in the world, | 中国AI的研究人员比世界上任何其他国家都多 |
| 275 | and many of them working at the labs. | 很多人在实验室工作 |
| 276 | I've been very clear that even though although we compete against China, | 我很清楚 虽然我们和中国竞争 |
| 277 | they're not our enemies. | 他们不是我们的敌人。 |
| 278 | If we care deeply about the safety of the world with respect to artificial | 如果我们深深地关心世界的安全 与人工 |
| 279 | intelligence, then of course we must open dialogues. | 当然,我们必须开放对话。 |
| 280 | We must work with them in every possible way. | 我们必须以一切可能的方式与他们合作。 |
| 281 | In a memo to employees, Huang wrote, I owe my family and all of you an | 在一份给员工的备忘录中,黄先生写道,我欠我的家人和你们一个 |
| 282 | apology. | 对不起。 |
| 283 | I was unfaithful to my wife. | 我对我妻子不忠 |
| 284 | My actions hurt. | 我的行为伤害。 |
| 285 | To calm anxiety about rogue AI, NVIDIA released an open source safety | 为了平息对流氓AI的焦虑,NVIDIA发布了开源安全 |
| 286 | platform of its own called Open Shell, designed to govern AI agents, even | 它自己的平台叫做Open Shell,旨在管理AI代理,甚至 |
| 287 | shut them down if necessary. | 必要时关闭它们。 |
| 288 | But this announcement suggests the current safety issues require an | 但这一宣布表明,当前的安全问题需要 |
| 289 | engineering solution rather than a regulatory one. | 工程解决方案而不是监管解决方案。 |
| 290 | AI is deeply unpopular and people are scared. | 大赦国际非常不受欢迎,人们很害怕。 |
| 291 | Sam Altman, Dario Amadei, Elon Musk all say we should slow down. | 萨姆·阿尔特曼,达里奥·阿玛迪,埃隆·穆斯克都说我们应该慢下来. |
| 292 | You have been very consistent in your optimism. | 你一直很乐观 |
| 293 | What do you see that they don't? | 你看到什么他们没有? |
| 294 | You don't advance capability, leaving safety behind. | 你不提高能力, 留下安全。 |
| 295 | You don't advance safety, leaving capability behind. | 你不推进安全, 留下能力。 |
| 296 | So let me give you an example. | 让我举个例子。 |
| 297 | When AI first came out, the first criticism of the original chat GBT was | 当AI首次出道时,最初对聊天GBT的批评是: |
| 298 | of course hallucination. | 当然是幻觉 |
| 299 | It was just spewing nonsense. | 只是胡说八道 |
| 300 | And so in order for us to make these chatbots grounded in truth, we have to | 所以我们要让这些聊天员 扎根于真理,我们必须 |
| 301 | invent something like retrieval augmented generation. | 发明了像检索 增强生成。 |
| 302 | We have to expand the context window. | 我们必须扩大上下文窗口。 |
| 303 | We connect it to tools so that it could do a search before it answers the | 我们把它和工具连接起来 这样它就可以在回答之前 进行搜索 |
| 304 | question, knows when to do a search, because maybe the information requires | 问,知道何时进行搜索, 因为也许信息需要 |
| 305 | contextual information or contemporary information. | 背景资料或当代信息。 |
| 306 | And so it has to go on and do a search. | 所以它必须继续搜索 |
| 307 | And all of that is part of its conditioning before it generates the answer. | 所有这一切都是它产生答案之前 的一部分条件。 |
| 308 | It also reasons afterwards. | 这也是后来的原因。 |
| 309 | It reflects on the answer it gave you. | 它反映了它给你的答案。 |
| 310 | And you go, you know, is this the best answer I can provide? | 而你去,你知道, 这是最好的答案,我可以提供? |
| 311 | Or it might even generate two or three different versions of the answer and | 或者它甚至可以生成两个或三个不同的版本 答案和 |
| 312 | reflects on which one's the best one. | 想想哪一个是最好的 |
| 313 | All of that require technology. | 所有这些都需要技术。 |
| 314 | So safety and capability to me is one idea. | 所以对我来说安全和能力是一个想法。 |
| 315 | Microsoft is calling for an AI emergency break and says AI is less | 微软要求AI紧急休息 并说AI更少 |
| 316 | regulated than washing machines. | 比洗衣机受管制。 |
| 317 | Anthropic is saying their models could evade shutdown in a matter of | Anthropic说他们的模型可以避免关闭 在一个问题中 |
| 318 | months. | 几个月。 |
| 319 | What would you have to see that might convince you that the industry should | 你会看到什么 可能说服你 行业应该 |
| 320 | slow down or stop several things? | 慢下来还是停止一些事情? |
| 321 | One, if we're developing agenting systems or autonomous systems, the | 第一,如果我们开发代理系统 或自主系统, |
| 322 | agents have to be in rigorous containment systems and those agents | 毒剂必须处于严格的封存系统之中,而且这些毒剂 |
| 323 | must be monitored, just job one. | 必须被监视,只是工作之一。 |
| 324 | If you're not able to do that, you should not test it. | 若无此法不应验. |
| 325 | If you're not able to do that, you should not deploy it. | 若不能达者,不应部署. |
| 326 | Well, these models that went rogue are unreleased models. | 这些无赖的模型是无法释放的 |
| 327 | So is that reason that they should shut down, shut the labs down? | 所以他们应该关闭实验室吗? |
| 328 | No, what they should do is that they should contain and monitor those | 不,他们应该做的是 他们应该控制和监督这些 |
| 329 | experiments. | 实验 |
| 330 | But they say they can't contain them. | 但他们说不能控制他们 |
| 331 | No, this is just made up nonsense. | 不,这只是胡说八道。 |
| 332 | I cannot be proven. | 我无法证明 |
| 333 | You can't prove that to me. | 你无法证明这一点给我。 |
| 334 | Yeah, you can absolutely contain, you can absolutely monitor. | 是的,你可以绝对控制, 你可以绝对监视。 |
| 335 | You know, until you do, until you try to do it, don't tell me you can't do it. | 你知道,直到你这样做, 直到你试图做到这一点, 不要告诉我你不能做到这一点。 |
| 336 | And so I'm just not going to listen to that. | 所以我不听这个 |
| 337 | So are you saying that you don't believe what they're saying? | 你是说你不相信他们在说什么? |
| 338 | I don't know all that they're saying. | 我不知道他们说什么。 |
| 339 | If you're developing agenting systems and you can't monitor it, you | 如果你正在开发代理系统 并且你无法监视它,你 |
| 340 | should not do so. | 不应该这样做。 |
| 341 | That's just the bottom line. | 这只是底线。 |
| 342 | Even the biggest AI optimists see risks. | 即使最大的AI乐观主义者也看到了风险. |
| 343 | They also see solutions. | 他们也看到了解决办法。 |
| 344 | What is the biggest risk you see? | 你看到的最大风险是什么? |
| 345 | Developing agentic systems without containment, developing agentic systems | 发展无约束的代理系统,发展代理系统 |
| 346 | without monitoring, it is extremely high risk. | 没有监测,风险就极高。 |
| 347 | You should not do that. | 你不应该这样做。 |
| 348 | And until you have the ability to contain and monitor, you should not run | 在你有能力控制和监测之前 你不应该跑 |
| 349 | large scale experiments. | 大规模实验。 |
| 350 | It just sounds like that's already happening. | 这听起来就像已经发生了。 |
| 351 | Then you just, you got to stop. | 然后,你必须停止。 |
| 352 | That's really simple. | 很简单 |
| 353 | So that would be the message to Anthropic, to OpenAI, where we've seen | 所以这就是给Anthropic 和OpenAI的讯息 在那里我们已经看到了 |
| 354 | these. | 这些 |
| 355 | You should contain your technology, contain your experiments. | 你应该控制你的技术 控制你的实验 |
| 356 | You should monitor your experiments. | 你应该监视你的实验 |
| 357 | You should give these agents, when you deploy them, minimum rights. | 你应该给这些特工, 当你部署他们, 最低限度的权利。 |
| 358 | They should be treated with zero trust. | 应当以零信任的态度对待他们。 |
| 359 | They should be monitored all the time. | 他们应该一直受到监视。 |
| 360 | If one of the labs was doing something that was too dangerous, that you felt | 如果其中一个实验室在做一些 太危险的事,你觉得 |
| 361 | was too dangerous, would Nvidia stop selling to them? | 太危险了,Nvidia会停止出售给他们吗? |
| 362 | Well, it's a hypothetical, but of course, yeah, it's a hypothetical. | 嗯,这是一个假设, 但当然,是的,这是一个假设。 |
| 363 | But of course, we don't run other people's companies and there's a | 但当然,我们不经营别人的公司 而且有一个 |
| 364 | limit to the judgment by which we apply to other companies' use of our | 限制我们对其他公司使用 |
| 365 | technology, our technology's general purpose. | 技术,我们技术的总目的。 |
| 366 | But if we know for certain that an application of our technology is | 但如果我们确信 我们技术的应用是 |
| 367 | bringing harm to people, of course. | 当然是给人带来伤害 |
| 368 | It's a Thursday. | 这是一个星期四。 |
| 369 | Xites of distress are emerging in leveraged AI computing firms. | 以杠杆方式利用AI计算公司正在出现苦难。 |
| 370 | Credit jitters sink chip stocks on a six-day losing streak, the longest | 信用紧张的芯片股票在6天的亏损中沉没,是最长的 |
| 371 | run of the year so far. | 目前为止的一年中 |
| 372 | And now we're in Stockholm. | 现在我们在斯德哥尔摩 |
| 373 | This is a city known for its canals and its islands, but it's also a | 这是一个以运河和岛屿而闻名的城市,但它也是一个 |
| 374 | thriving tech hub. | 科技中心蓬勃发展 |
| 375 | AI, of course, is a bigger and bigger part of that, and much of it | 当然,AI是其中更大和更大的部分, 并且大部分 |
| 376 | Nvidia正在提供电力。 | |
| 377 | Our guest of honor needs little introduction. | 我们的贵宾不需要什么介绍 |
| 378 | Welcome, welcome to Stockholm. | 欢迎光临斯德哥尔摩 |
| 379 | Hwang头条为瑞典Nvidia支持的主权AI踢球 | |
| 380 | supercomputer, attended by tech luminaries like Spotify co-founder | 超级计算机,由Spotify公司联合创始人等技术名流参加 |
| 381 | Daniel Ek. | 丹尼尔·艾克. |
| 382 | We have a tremendous educational system that produces incredible | 我们有一个巨大的教育制度 产生不可思议的, |
| 383 | talent across many different disciplines and industries. | 跨越许多不同学科和行业的人才。 |
| 384 | I don't think that there's a problem of potential. | 我不认为有潜力的问题。 |
| 385 | I think what we do need is to speed things up and we need to be more | 我认为我们需要的是加快速度 我们需要更多 |
| 386 | ambitious. | 雄心勃勃。 |
| 387 | Meanwhile, more tough AI news was coming in. | 同时,更严酷的AI消息即将传来. |
| 388 | 数据中心公司CoreWeave 在Nvidia的支援下 正在挨打 | |
| 389 | in the market over concerns about its debt load. | 在市场上 担心它的债务负担。 |
| 390 | A seemingly unfazed Hwang told reporters CoreWeave's strength is | 一个看似虚弱的黄帝告诉记者 科里韦的实力是 |
| 391 | improving every day. | 每天都在进步 |
| 392 | But other AI stocks were selling off, too, as markets seemed to be | 但其他的AI股票也在卖掉 就像市场一样 |
| 393 | questioning the economics of the AI boom. | 质疑AI繁荣的经济学. |
| 394 | More than 100 data center projects have been blocked or delayed in the | 超过100个数据中心项目在 |
| 395 | U.S. just so far this year. | 美国今年刚刚结束 |
| 396 | There's pushback in Europe, too. | 欧洲也有后推. |
| 397 | You have AI IPOs getting delayed, getting scrapped. | 你有AIIPO 被拖延, 被废掉。 |
| 398 | Are we seeing cracks in the AI boom? | 我们在AI爆炸中看到裂缝吗? |
| 399 | And could you be left holding the bag? | 你能把包放一边吗? |
| 400 | It's too fundamental. | 这太根本了 |
| 401 | I think this is the future of computing. | 我认为这就是计算的未来。 |
| 402 | We're at the fundamental level. | 我们处于基本层面。 |
| 403 | Computing will never go back. | 电子计算永远也回不去了 |
| 404 | We used to have computers that the software is written by hand and the | 我们曾经有电脑 软件是用手写的 |
| 405 | computer is waiting for us to use it. | 电脑在等我们使用 |
| 406 | In the future, and where we are now, computers are able to learn | 未来,现在,计算机可以学习 |
| 407 | environments and problems that are too complicated for any human to | 环境和问题过于复杂,任何人类都无法 |
| 408 | program. | 程序。 |
| 409 | The size of it, the way that computers are used instead of a tool, it's not | 它的大小,计算机的使用方式 而不是工具,它不是 |
| 410 | an infrastructure, that's never going to go back. | 基础设施,这是永远不会回头。 |
| 411 | The capabilities of AI, now that we know what it can do, that's never | AI的能力,现在我们知道它能做什么,那永远不会 |
| 412 | going to go back. | 准备回去 |
| 413 | I think in many ways, it's fundamental, irrespective of near-term challenges. | 我认为在许多方面,这是根本性的,不管近期的挑战如何. |
| 414 | So I think the narrative of the industry is really out of control and | 所以我觉得这个行业的叙述 已经失控了 |
| 415 | ought to be stopped, but that's just an industry growing up. | 应该停止, 但这只是一个行业成长。 |
| 416 | You probably remember the good old days of the internet, and the | 你可能还记得 互联网的好时光, |
| 417 | industry grew up. | 工业成长。 |
| 418 | I think that our industry is, many of the young companies are going through | 我认为,我们的工业, 许多年轻的公司正在经历 |
| 419 | their growing pains and they're growing, of course, incredibly fast. | 他们的成长痛苦 他们正在增长,当然, 惊人的快速。 |
| 420 | There's an accountability, responsibility that goes along with building | 有问责制,责任 与建设 |
| 421 | companies that I think they're just growing into. | 公司,我认为他们只是成长。 |
| 422 | NVIDIA keeps adding new products, buying up companies. | NVIDIA不断添加新产品,收购公司. |
| 423 | In some cases, you're competing with your own customers. | 在某些情况下,你和自己的顾客竞争. |
| 424 | Why do you need to have such a broad range of offerings? | 为什么你需要如此广泛的祭品? |
| 425 | We do as little as we can. | 我们尽量少做点 |
| 426 | It seems like you're doing a lot. | 看来你做了很多 |
| 427 | Surprisingly little. | 令人惊讶的是小。 |
| 428 | I think we've done a good job generalizing the work that we do so that | 我认为我们已经做了一个很好的工作 概括我们做的工作 |
| 429 | it could be applied to many different things. | 它可以适用于许多不同的事物。 |
| 430 | Not because we're doing many different things, it's because we | 不是因为我们在做很多不同的事情 而是因为我们 |
| 431 | generalized one thing. | 泛泛地说一件事 |
| 432 | And so if you look at our architecture, we're the only architecture that's | 所以,如果你看看我们的建筑, 我们是唯一的建筑是 |
| 433 | general purpose, meaning we run every closed model, we run every open model. | 一般目的,意味着我们运行每一个封闭模型, 我们运行每一个开放模型。 |
| 434 | We run biology models, the language models, autoregressive models, the | 我们运行生物学模型,语言模型, 自动递归模型, |
| 435 | transformer models, the diffusion models, and all the different various technologies. | 变压器模型,扩散模型,以及所有不同的技术. |
| 436 | The challenge of generalization is, of course, performance. | 当然,一般化的挑战在于业绩。 |
| 437 | How can you be the world's fastest F1 race car and also be a really great | 你怎么会是世界上最快的F1赛车 同时也是很棒的 |
| 438 | sedan? | 轿车? |
| 439 | And so that's the miracle of the technology that we've created. | 这就是我们创造的技术的奇迹。 |
| 440 | On the one hand, we're the world's fastest F1 racing cars, and we're the | 一方面,我们是世界上最快的F1赛车,我们是 |
| 441 | fastest computers, and we're always at the top of the leaderboard. | 最快的电脑, 我们总是在领导板的顶端。 |
| 442 | You benchmark us, we're always the best. | 你以我们为基准 我们永远是最好的 |
| 443 | On the other hand, because we're fungible, we're also the lowest cost. | 另一方面,因为我们是可互换的,我们也是成本最低的. |
| 444 | NVIDIA recently committed to spending about half a trillion dollars across | NVIDIA最近承诺将花费大约5万亿美元 |
| 445 | the global AI industry, and that number is quickly climbing. | 全球AI产业, 这个数字正在迅速攀升。 |
| 446 | There's concern that in doing so, NVIDIA is creating false demand, giving | 人们担心的是,通过这样做,NVIDIA正在制造虚假的需求,给予 |
| 447 | companies the money they need to turn around and buy products from NVIDIA. | 公司需要资金 转身 从NVIDIA购买产品。 |
| 448 | It's no secret that critics have accused NVIDIA of circular financing. | 批评者指责NVIDIA为循环融资,这不是秘密. |
| 449 | How do you toe the line between boosting promising startups and blowing up a | 如何在有前途的初创企业 和爆炸之间 |
| 450 | bubble? | 泡泡吗? |
| 451 | If AI is such a goldmine, why do you have to underwrite so much of it? | 如果AI是这样一个金矿,你为什么要承载这么多呢? |
| 452 | Yeah, this technology requires high capex. | 是啊,这个技术需要高额的盖帽。 |
| 453 | If you look at this generation of AI startups, it's very, very rare that you | 如果你看看这代人工智能的启动, 这是非常,非常罕见的你 |
| 454 | find a company that can be funded on 10 million dollars, and it's not unusual to | 找到一家可以以1000万美元出资的公司 这并非不寻常 |
| 455 | find a company that requires 10 billion dollars. | 找到一个需要100亿美元的公司 |
| 456 | And the reason for that is because it needs this computation resource in order | 原因是它需要这个计算资源 |
| 457 | to deliver the intelligence. | 来传递情报 |
| 458 | Some of the highest quality companies in the AI ecosystem across each one of the | 人工智能生态系统中一些质量最高的公司 |
| 459 | layers of the five-layer cake, it connects us to the best and the most talented | 层层的五层蛋糕, 它连接我们与最好的和最有才华的 |
| 460 | companies. | 企业。 |
| 461 | It helps us continue to advance NVIDIA's platform, if you will, | 它帮助我们继续推进NVIDIA的平台,如果你愿意的话, |
| 462 | because these companies are going to be built on NVIDIA's technology. | 因为这些公司将建立在NVIDIA的技术之上。 |
| 463 | And it helps us capitalize and be one of the early investors in many of them. | 它帮助我们利用 并成为其中众多的早期投资者之一。 |
| 464 | We have a highly disciplined approach of making sure that we are, although one of | 我们有一个非常有纪律的办法,确保我们是,尽管我们是其中之一。 |
| 465 | the earliest investors and many times even the anchor investor, we're not the | 最早的投资者 以及很多次 甚至锚地投资者,我们不是 |
| 466 | lead investor. | 首席投资者。 |
| 467 | We're not the largest investor. | 我们不是最大的投资者。 |
| 468 | In all cases, we own less than 20 percent of the company. | 总之,我们拥有的公司不到20%。 |
| 469 | We've invested some, I guess, in the last five years, maybe close to 100 billion | 我们投资了一些,我想, 在过去五年里, 也许接近1000亿 |
| 470 | dollars, something like that. | 美元,类似的东西。 |
| 471 | That 100 billion dollars resulted in, well, at | 1000亿美元的结果是 |
| 472 | least five times as much other institutional investors into this ecosystem. | 进入这一生态系统的其他机构投资者至少5倍。 |
| 473 | Most of the world's leading AI companies depend on NVIDIA, not just your hardware, | 世界上大部分领先的AI公司都依赖于NVIDIA, 不仅仅是你的硬件, |
| 474 | but your software as well. | 但你的软件也是。 |
| 475 | We're probably the only AI company in the world that works with every AI company | 我们可能是世界上唯一一家 与每个AI公司合作的AI公司 |
| 476 | in the world. | 在这个世界。 |
| 477 | So at what point does that dependence on one single company become a systemic | 所以,在什么时候,对单一公司的依赖 成为一个系统 |
| 478 | risk? | 冒险? |
| 479 | Well, they don't depend on us. | 他们不依赖我们 |
| 480 | They choose us. | 他们选择我们 |
| 481 | I don't worry about that. | 我不担心这一点。 |
| 482 | And the reason for that is because we don't take our position for granted. | 之所以如此,是因为我们不认为我们的立场是理所当然的。 |
| 483 | We came from nothing and we have to earn our keep every day. | 我们一无所有,我们每天都要挣钱 |
| 484 | We have competitors in every company, every industry, every country. | 我们在每一个公司,每一个行业,每一个国家都有竞争对手. |
| 485 | With respect to the size, the size of the market, just remember, it's just a | 关于市场的规模,只要记住,这只是一个 |
| 486 | number. | 数字。 |
| 487 | NVIDIA can only be as large as the, if you will, the pond that we're in, | NVIDIA只能和我们所在的池塘一样大 |
| 488 | the market that we're in. | 我们的市场。 |
| 489 | When somebody asks me, America needs to win the AI race. | 当有人问我, 美国需要赢得AI比赛。 |
| 490 | What I hear is not that one or two or three companies wins the AI race. | 我听到的不是一两三家公司赢得AI比赛. |
| 491 | What I mean is America's companies win the AI race. | 我的意思是 美国的公司赢得了AI比赛。 |
| 492 | If we do the opposite, we scare people and we cause people anxiety, | 如果我们这样做相反,我们吓唬人 并引起人们的焦虑, |
| 493 | then America will lose. | 这样美国就会输了 |
| 494 | In a new administration, will NVIDIA's dominance be at risk? | 在新政府中,NVIDIA的统治地位是否会面临风险? |
| 495 | Are you prepared? | 你准备好了吗? |
| 496 | Are we prepared? | 准备好了吗? |
| 497 | Yeah, sure. | 当然 |
| 498 | I'm in service of our policymakers and our leaders and whatever they want to talk to | 我为我们的决策者和领袖们服务 不管他们想说什么 |
| 499 | me about. | 约我. |
| 500 | I'm more than happy to share with them what I know. | 我很乐意和他们分享我所知道的 |
| 501 | I know a lot about this technology. | 我对这个技术很了解 |
| 502 | I know a lot about engineering and I know what it takes to build this and I know | 我对工程学很了解 我知道建造这个需要什么 |
| 503 | what we're doing to build it. | 我们为建造它所做的一切。 |
| 504 | And if you ask me my thoughts on whether we can advance AI with prosperity for the | 如果你问我关于我们能否为 |
| 505 | future while doing so, every single step the way, safely and responsibly, | 未来,在这样做的时候,每一个步骤,安全和负责任地, |
| 506 | I have absolutely zero concern that we have the ability to do so. | 我毫不担心我们有能力这样做。 |
| 507 | Now, whether we would do so is a completely different thing, | 现在,如果我们这样做是完全不同的事情, |
| 508 | but we have the technology to do so. | 但我们有这样做的技术。 |
| 509 | You said AI supply and demand will balance out in maybe five to ten years. | 你说人工智能的供求 在5到10年内会平衡 |
| 510 | What happens when the world doesn't need as many new chips? | 当世界不需要那么多新筹码时会怎样? |
| 511 | New chips will replace the old chips, just kind of like the personal computer industry. | 新的芯片将取代旧的芯片,就像个人电脑工业一样. |
| 512 | We've got, what, a billion users in the world today. | 我们今天有十亿用户 |
| 513 | In 1993, when NVIDIA was first founded, the personal computer industry was doubling | 1993年,NVIDIA成立之初,个人计算机产业翻了一番. |
| 514 | every single year and, of course, it doesn't double anymore. | 每一年,当然,它不再加倍。 |
| 515 | At the moment, we're kind of doubling every year and my sense is that we're going to | 目前,我们每年翻一番,我的感觉是,我们将 |
| 516 | keep doing that as we build out the infrastructure of the world. | 在我们建设世界基础设施时,继续这样做。 |
| 517 | And how far out can you see that doubling? | 你还能看多远? |
| 518 | You know, we're going to do that for several years and then we'll re-evaluate. | 你知道,我们要做到这一点 几年,然后我们 重新评价。 |
| 519 | And so we're kind of two years into an infrastructure build-out and if you look at | 所以,我们有 两年的基础设施建设 如果你看看 |
| 520 | previous technology cycles, it takes about ten years or so to reach some kind of maturity | 以前的技术周期,大约需要十年左右才能达到某种成熟 |
| 521 | and then after that you're in replacement cycles. | 之后你就进入了替换周期 |
| 522 | And so did that with phones, did that with the internet, did that with PCs. | 还有电话,互联网,个人电脑 |
| 523 | I think it'll happen here too. | 我想这也会发生在这里。 |
| 524 | Could there be a breakthrough that means AI is powered by a fraction of the energy in chips? | 是否有突破意味着AI由芯片中一小部分能量提供动力?. |
| 525 | Oh yeah, yeah, yeah. | 哦,对,对,对。 |
| 526 | And then what does that mean for NVIDIA? | 那对NVIDIA意味着什么? |
| 527 | No, we're just going to keep making things more efficient. | 不,我们只是要 不断提高效率。 |
| 528 | There's a weird dynamic. | 有一个奇怪的动态。 |
| 529 | People used to tell me, hey, Jensen, if you make your chips that much better every generation, | 人们常告诉我,嘿,詹森, 如果你做你的筹码这么好的每一代, |
| 530 | doesn't it say that customers just buy less? | 上面不是说顾客少买一点吗? |
| 531 | And that same exact logic goes, if we make companies more productive, | 同样的逻辑是 如果我们让公司更有生产力 |
| 532 | don't we hire less people? | 我们不雇更少的人吗? |
| 533 | And there's this weird thing called ambition. | 有个怪怪的叫野心 |
| 534 | If you gave me 36 hours a day, I'm not going to sleep more, | 如果你每天给我36小时 我不会再睡觉了 |
| 535 | because I've got ideas, I've got things to do. | 因为我有主意,我有事情要做 |
| 536 | Well, I end up doing more. | 好吧,我最后会做更多。 |
| 537 | And so if AI makes us all more intelligent, I become more ambitious. | 所以如果AI让我们更加聪明, 我就会更加雄心勃勃。 |
| 538 | If the cost of doing something becomes marginally close to zero, | 如果做一些事情的成本 微乎其微接近零, |
| 539 | it turns out that resource becomes widely abundant. | 结果资源变得非常丰富。 |
| 540 | I think the economy is going to get larger. | 我觉得经济会变大 |
| 541 | I think the industries will get larger. | 我认为这些行业会变得更大。 |
| 542 | I think the semiconductor industry will be, at this point, | 我认为半导体工业 此时此刻 |
| 543 | very clearly one of the largest industries in the world. | 很明显是世界上最大的工业之一 |
| 544 | Chipmakers set for their longest losing streak since 2008, | 芯片制作商为2008年以来最长的亏损做准备。 |
| 545 | leaving investors questioning the prospects for the AI trade. | 让投资者质疑AI贸易的前景. |
| 546 | Finally, we're in Berlin. | 我们终于在柏林了 |
| 547 | This is where NVIDIA is holding its third big conference of the year. | 这是NVIDIA举行今年第三次大型会议的地方. |
| 548 | Germany built its economy on the last industrial revolution, | 德国在上一次工业革命的基础上建立了经济 |
| 549 | but the economy's been having kind of a rough time. | 但经济一直很艰难 |
| 550 | Jensen likes to say AI is the new industrial revolution, | 詹森喜欢说AI是新的工业革命, |
| 551 | and if anyone could use it, it's here. | 如果有人可以使用它, 它在这里。 |
| 552 | Germany is rich with the most important part of AI, | 德国拥有AI最重要的部分 |
| 553 | which is the science, the math, the computer scientists, | 这是科学,数学,计算机科学家, |
| 554 | the physical AI, the deep science capabilities. | 物理AI, 深科学能力。 |
| 555 | You have the spirit of innovation, you have the spirit of entrepreneurship. | 有创新精神,有创业精神. |
| 556 | We now have to find a way to get this flywheel going, | 我们现在得想办法把飞轮开走 |
| 557 | and we have to start by building the computers, the infrastructure. | 我们首先要建立电脑和基础设施。 |
| 558 | In Berlin, the line to see Huang was out the door and around the block. | 在柏林,看到黄帝的线是在门外和街区周围. |
| 559 | The event was held at Saturn, which is like a German Best Buy on steroids. | 活动在土星举行,土星就像德国的类固醇最佳买家. |
| 560 | Jensen is in the building, and he's here. | 詹森在大楼里 他来了 |
| 561 | He's here to see everybody tonight. | 他今晚来见大家 |
| 562 | Hey, everybody! | 嘿,大家来! |
| 563 | How do you say Saturn? | 土星怎么样? |
| 564 | Saturn. | 土卫六. |
| 565 | Everything in German, you've got to say really serious. | 用德语说,你得说真的 |
| 566 | Saturn. | 土卫六. |
| 567 | Obviously, you've had this big week. | 显然,你有一个大星期。 |
| 568 | You've gone to so many countries, so much exciting stuff, | 你去过这么多国家, 这么多令人兴奋的东西, |
| 569 | but at the same time, there's like constant headlines. | 但与此同时,有 像是不断的头条新闻。 |
| 570 | This massive market sell-off, chip stocks under pressure. | 这种巨大的市场销售,芯片股票在压力下。 |
| 571 | As this trip has been going on, what have you been watching in the markets? | 这趟旅程一直以来 你在市场上看什么? |
| 572 | What's your take on what's happening with 4Weave? | 你对4Weave的事有什么看法? |
| 573 | All of this is going to pass, and the reason for that is very simple. | 所有这一切将过去,原因很简单。 |
| 574 | We're going through a new platform transition. | 我们正在经历一个新的平台过渡. |
| 575 | News, as you know, happens all the time, but it won't change the fundamentals. | 新闻,如你所知,经常发生,但不会改变基本情况. |
| 576 | 你说A. I. | |
| 577 | will create more jobs than it does. | 创造更多的工作机会。 |
| 578 | A.I. (英语). | |
| 579 | does create more jobs. | 创造更多的工作机会。 |
| 580 | You've said people worried about losing their jobs should learn A.I. | 你说过那些担心失业的人 应该学会人工智能 |
| 581 | What do you say to the person who has learned A.I., | 你对学会人工智能的人怎么说? |
| 582 | who is doing exactly that but still watching their job slip away? | 到底是谁在做那些事... ...但还是看着他们的工作消失? |
| 583 | Well, first of all, I think you're asking a theoretical question, | 首先,我想你问了一个理论问题 |
| 584 | but let's apply the practical answer first. | 但让我们先应用实际答案。 |
| 585 | 记住,A. I. | |
| 586 | is unlike previous technologies in the sense that | 与以前的技术不同,因为 |
| 587 | this requires manufacturing. | 这需要制造。 |
| 588 | It's creating construction jobs, electrician jobs, plumbing jobs. | 它创造了建筑工作,电工工作,管道工作. |
| 589 | In the last six months, something like $400 to $600 billion were invested into A.I. | 在过去的六个月里,大约4000到6000亿美元被投资到A. I. |
| 590 | startups, | 启动, |
| 591 | and the reason for that is people see A.I. | 原因是人们看到A.I. |
| 592 | is productive, it's effective, | 很有生产力,很有效 |
| 593 | and people want to invest in it. | 人们想投资。 |
| 594 | Those $400 to $600 billion has just created how many jobs? | 这四千到六千亿美元刚刚创造了多少工作? |
| 595 | Of course, this technology is a new tool, and this new tool is therefore going to cause | 当然,这种技术是一种新工具,因此这种新工具将导致 |
| 596 | everybody's jobs to change. | 每个人都需要改变工作 |
| 597 | Resisting the change won't make it go away. | 抗拒改变不会让它消失. |
| 598 | The superpowers that the tech companies have are now made available to literally everybody. | 科技公司拥有的超能力 现在已经提供给所有人了 |
| 599 | It democratized the technology. | 它使技术民主化。 |
| 600 | It's elevated everybody's capabilities, and so I see the benefits of A.I. | 这提升了每个人的能力 所以我看到了人工智能的好处 |
| 601 | I'm obviously very aware of the risks and the technology risks, | 我很清楚风险和科技风险 |
| 602 | but in the final analysis, if I don't see the optimistic, | 但归根结底, 如果我不看到乐观, |
| 603 | if we don't see the optimistic, incredibly wonderful future that we're going to create, | 如果我们不看到乐观 令人难以置信的美好未来 我们将创造, |
| 604 | for what reason are we building it? | 为什么我们要建它? |
| 605 | The final act in this dash across Europe was NVIDIA's own flagship conference, GTC. | 整个欧洲的最后行动是NVIDIA自己的旗舰会议GTC. |
| 606 | The CEO stuck to his bullish pitch, recapping his new European investments and partnerships, | 首席执行官坚持他的粗鲁投注, 收回他新的欧洲投资和伙伴关系, |
| 607 | announcing new A.I. | 宣布新的人工智能。 |
| 608 | security and data center products, | 安全和数据中心产品, |
| 609 | and playing around with some A.I.-trained robots. | 和一些训练有素的机器人玩耍 |
| 610 | At one point, he was interrupted by a protest over the reported use of NVIDIA chips in Russian drones. | 有一次,他因据报道在俄罗斯无人机中使用NVIDIA芯片而被抗议打断. |
| 611 | Quite confident in the agency of people. | 颇自信众生之机. |
| 612 | I'm quite confident that today is a safer world than it was 100 years ago, | 我相信今天的世界比一百年前更安全 |
| 613 | 50 years ago, 100 years ago, and 250 years ago. | 50年前,100年前,250年前。 |
| 614 | I'm quite confident in that because the facts would show it. | 我很有信心,因为事实会证明这一点。 |
| 615 | Although the news headlines are more pessimistic and more bad news than ever, | 虽然新闻头条比以往更悲观 更坏消息 |
| 616 | but that's just because bad news is easier to sell. | 但那只是因为坏消息更容易卖掉 |
| 617 | Many people appreciate your optimism and get excited by it, | 很多人都欣赏你的乐观 并对此感到兴奋, |
| 618 | but what if your optimism proves wrong in the end? | 但是如果你的乐观态度最终被证明是错误的呢? |
| 619 | Then I can say that I did everything I could. | 然后我可以说,我尽我所能。 |
| 620 | And the reason for that is because not only am I optimistic, | 原因 是因为我不仅乐观 |
| 621 | I'm doing everything I can to create the technology necessary to keep the industry safe, secure. | 我竭尽所能创造必要的技术,使产业安全可靠. |
| 622 | And so let's not talk about the future in a pessimistic way while we do nothing about it | 所以,让我们不要以悲观的方式谈论未来 而我们却什么都不做 |
| 623 | to prevent that from happening. | 防止发生这种情况。 |
| 624 | And so, you know, maybe I'm older, I'm more parental. | 所以,你知道,也许我年纪大了, 我的父母更多。 |
| 625 | Don't just deliver the bad news. | 别只是传达坏消息 |
| 626 | Tell me what you're going to do about it. | 告阿难言. |
| 627 | And so, so long as we can prove to ourselves that we're doing both sides of that coin | 所以,只要我们能向自己证明 我们在做硬币的两边 |
| 628 | the optimistic future while safeguarding our way in doing so. | 乐观的未来,同时保障我们这样做的方式。 |
| 629 | So long as we do both sides of that, I'm perfectly fine. | 只要我们做两边,我完全没事。 |
| 630 | After a whirlwind week of AI pump, one thing was clear. | 经过一周的人工智能泵, 有一件事是清楚的。 |
| 631 | Jensen Huang has been the leading evangelist for an extraordinary new age of artificial intelligence. | Jensen Huang是人工智能新时代的主要传教士。 |
| 632 | NVIDIA, the U.S. economy, maybe even the entire global financial system | NVIDIA 美国经济 甚至整个全球金融系统 |
| 633 | might just rest on his optimism and accomplishments. | 他的乐观和成就或许只能靠他。 |
| 634 | But now Huang himself is under scrutiny. | 但现在黄帝本人正在接受审查. |
| 635 | He has led the business since its very beginning 33 years ago. | 他从33年前开始就领导了这个生意 |
| 636 | And even as NVIDIA grew into one of the world's most powerful companies, | 即使NVIDIA成长为世界上最强大的公司之一 |
| 637 | there's still no one designated to potentially replace him. | 现在还没有人可以替代他 |
| 638 | And the recent allegations show just how tenuous it is | 最近的指控表明,这是多么脆弱 |
| 639 | to pin the hope and future of an industry on one person. | 将一个行业的希望和未来寄托在一个人身上。 |
| 640 | You famously run NVIDIA in a very unconventional way. | 你以非常不寻常的方式管理NVIDIA |
| 641 | Lots of direct reports, no meetings. | 很多直接报告,没有会议。 |
| 642 | As the company has exploded in size and the problems have become more complex, | 随着公司规模的扩大 问题变得更加复杂 |
| 643 | how has your management style and your mentality changed? | 你的管理风格和心态如何改变? |
| 644 | Well, my management style is what made it possible for the company to grow. | 我的管理风格使公司得以发展 |
| 645 | Our organization is what made the company to thrive and innovate. | 我们的组织使公司得以繁荣和创新。 |
| 646 | I wanted to create an architecture of a company with as few silos as possible. | 我想建立一个公司建筑 尽可能少的筒仓。 |
| 647 | You know, the problem with silos, if the organizations are too rigid, | 仓子的问题 如果组织太僵硬 |
| 648 | then the company has a very difficult time adapting to change. | 公司很难适应变化 |
| 649 | So I think a modern company really wants to be more and more agile and less and less structured. | 所以,我认为一个现代公司 真的希望 越来越敏捷 和更少的结构。 |
| 650 | What about the CEO who runs that company? | 那经营那家公司的首席执行官呢? |
| 651 | You should be as mentally flexible as possible. | 你应该尽量灵活 |
| 652 | I think you have to hold yourself responsible. | 我想你得为自己负责 |
| 653 | You have agency. | 你有机构。 |
| 654 | You're the CEO of the company, for God's sakes. | 看在上帝的份上 你是公司的首席执行官 |
| 655 | You have to have high standards and expectations. | 你必须有很高的标准和期望。 |
| 656 | Your son and your daughter are working at NVIDIA. | 你儿子和你女儿在NVIDIA工作. |
| 657 | I think your son-in-law also works at NVIDIA. | 我想你的女婿也在NVIDIA工作 |
| 658 | Robotics, yeah. | 机器人,对。 |
| 659 | There is talk that your daughter is in line to be your successor. | 有人说你女儿会成为你的继任者 |
| 660 | Is that true? | 是吗? |
| 661 | Well, first of all, I don't hope to job on her. | 好吧,首先,我不希望 工作给她。 |
| 662 | And the reason for that is because that's their aspiration. | 其原因是,这是他们的愿望。 |
| 663 | Spencer and Madison's, they have to go earn it. | 斯宾塞和麦迪逊的, 他们必须去挣钱。 |
| 664 | And it's really, really hard work. | 这真的很辛苦 |
| 665 | It takes a lot of sacrifice. | 这需要大量牺牲。 |
| 666 | For many people, they can focus on their job and their families. | 对许多人来说,他们可以专注于工作和家庭. |
| 667 | When you're the CEO of the company, you have to focus on, of course, your family. | 当你是公司的首席执行官时,你必须专注于你的家人。 |
| 668 | But I've got a whole bunch of other families I'm caring for. | 但我有一大堆其他家庭 我关心。 |
| 669 | I'm responsible for an entire industry. | 我负责整个行业 |
| 670 | And so I think the pressure of being the CEO is high, and it's supposed to be high. | 所以我认为当首席执行官的压力很大,应该很高. |
| 671 | It shouldn't be an easy job. | 这不应该是一个容易的工作。 |
| 672 | It shouldn't be a job that you take for granted. | 它不应该是一个你认为理所当然的工作。 |
| 673 | When I decided to be CEO, I didn't realize the magnitude of the pressure and responsibilities. | 当我决定成为CEO时,我并没有意识到压力和职责的大小. |
| 674 | But look, that's what I signed up for. | 但是,你看,这是我报名的。 |
| 675 | And so now that my family and all of the people that are near me | 现在我的家人和所有接近我的人 |
| 676 | see the responsibilities of the job and what it takes, | 看工作的责任和花费, |
| 677 | then they just got to ask themselves whether this is the level of sacrifice | 然后他们只好问自己 这是不是牺牲的程度 |
| 678 | and responsibility you want. | 还有你想要的责任 |
| 679 | And I'm really hoping that the next CEO of the company is already at the company. | 我真的希望公司下一任首席执行官 已经到公司了。 |
| 680 | But 20 years from now, I hope I stay healthy, and I hope I continue to earn my keep. | 但20年后,我希望我保持健康,希望我继续维持生计。 |
| 681 | But so long as I can do that 20 years from now, | 但只要我20年后能做到这一点 |
| 682 | I hope they're already here, but they have to earn it. | 我希望他们已经在这里,但他们必须挣钱。 |
| 683 | For the next generation of leaders at NVIDIA, what will their challenge look like? | 对NVIDIA的下一代领导人来说,他们的挑战将是什么样子? |
| 684 | I don't know. | 师曰. |
| 685 | And that's the beauty. | 这就是美。 |
| 686 | In fact, that's the exactly right answer. | 事实上,这是完全正确的答案。 |
| 687 | We should create the conditions by which the company continues to thrive. | 我们应该为公司继续繁荣创造条件。 |
| 688 | And we should create the conditions by which the leaders of our company continues to grow. | 我们应该为公司领导人继续成长创造条件。 |
| 689 | And that the next leader of the company is already at NVIDIA. | 公司下一任领导已经到了NVIDIA. |
| 690 | That's my job, is to create that condition. | 这是我的工作,就是创造这个条件。 |
| 691 | And just as I can be fired, the next CEO has to be hired. | 就像我被炒鱿鱼一样 下一任首席执行官必须被雇用 |
| 692 | And so I like it that way. | 所以我喜欢这样 |
| 693 | I like that we can't be overly complacent about our position inside the company, | 我喜欢,我们不能太自满 我们的地位在公司, |
| 694 | and that we can't take it for granted. | 而我们不能把它视为理所当然。 |
| 695 | And you're always vulnerable. | 你总是很脆弱 |
| 696 | You're always earning it. | 你总是在赚钱 |
| 697 | You're always striving to be better. | 你总是努力变得更好 |
| 698 | That's a condition that's good. | 这条件很好 |
| 699 | And we can't take it for granted. | 我们不能把它视为理所当然。 |
该视频共有字幕 701 条。解锁更多字幕为会员功能，请移动到 价格