“芯片女王”买下“AI教母”！AMD豪掷550亿，苏姿丰把李
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:00 | AMD is acquiring WorldLabs. | AMD收购世界实验室. |
| 0:02 | Congratulations to you both. | 恭喜你们 |
| 0:05 | Thank you. | 谢谢 |
| 0:05 | Lisa, I think a pretty obvious place to start is why. | 丽莎,我想一个非常明显的开始是原因。 |
| 0:08 | Absolutely. | 当然 |
| 0:09 | AMD为什么要收购WorldLabs? | |
| 0:11 | Well, first, Ed, thanks for having us. | 首先 艾德 谢谢你邀请我们 |
| 0:12 | And today is a super exciting day. | 今天是令人兴奋的一天。 |
| 0:15 | I can tell you that, you know, we are super excited to be acquiring WorldLabs. | 我可以告诉你,你知道, 我们超级兴奋 获得世界实验室。 |
| 0:19 | I have known Fei-Fei for a long time. | 我认识飞飞很久了 |
展开字幕全文（330 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | AMD is acquiring WorldLabs. | AMD收购世界实验室. |
| 2 | Congratulations to you both. | 恭喜你们 |
| 3 | Thank you. | 谢谢 |
| 4 | Lisa, I think a pretty obvious place to start is why. | 丽莎,我想一个非常明显的开始是原因。 |
| 5 | Absolutely. | 当然 |
| 6 | AMD为什么要收购WorldLabs? | |
| 7 | Well, first, Ed, thanks for having us. | 首先 艾德 谢谢你邀请我们 |
| 8 | And today is a super exciting day. | 今天是令人兴奋的一天。 |
| 9 | I can tell you that, you know, we are super excited to be acquiring WorldLabs. | 我可以告诉你,你知道, 我们超级兴奋 获得世界实验室。 |
| 10 | I have known Fei-Fei for a long time. | 我认识飞飞很久了 |
| 11 | I've had tremendous respect, you know, | 我非常尊敬你,你知道的, |
| 12 | for her as a researcher and as a visionary. | 作为她的研究员和远见 |
| 13 | And if you think about where AI is today, I mean, we are, you know, | 如果你想到AI今天在哪里, 我的意思是,我们,你知道, |
| 14 | I've said often, Ed, that we're still in the very early innings of AI, | 我经常说,Ed 我们还在AI的早期, |
| 15 | and I truly believe that. | 我真心相信这一点。 |
| 16 | And what's becoming more and more clear is that the interaction between, | 越来越清楚的是, 相互作用, |
| 17 | you know, the hardware and the software and the systems and the model layer | 硬件 软件 系统 模型层 |
| 18 | are such that the more you understand end-to-end, | 你越了解末端, |
| 19 | the better system you're going to build. | 你会建立更好的系统 |
| 20 | And so that's why we're acquiring WorldLabs, really, | 所以我们才要收购世界实验室 |
| 21 | is to have access to the world-class talent that Fei-Fei has brought together | 即获得飞飞带来的世界一流人才 |
| 22 | and really bring it together with AMD's complete capabilities around, | 并把它和AMD的完全能力结合在一起, |
| 23 | you know, hardware, software and systems and, you know, build the future of AI. | 你知道,硬件,软件和系统 以及,你知道,构建AI的未来。 |
| 24 | I think we should get to the back story with both of you, | 我觉得我们该和你们俩谈谈后面的故事 |
| 25 | but would you talk a little bit, Lisa, how it came about, you know, | 但是,你会说话一点点,丽莎, 它是如何来的,你知道, |
| 26 | the catalyst for doing the actual transaction? | 实际交易的催化剂? |
| 27 | Well, as I said, I've had tremendous respect for Fei-Fei. | 正如我所说,我非常尊重飞飞 |
| 28 | I mean, you know, everyone knows just how much she has contributed to AI | 我的意思是,你知道,每个人都知道 她贡献了多少AI |
| 29 | over the last, actually, couple of decades. | 过去,实际上, 几十年。 |
| 30 | And from, you know, our personal relationship, | 我们的私人关系 |
| 31 | we've always had that, you know, both mutual respect and trust. | 我们一直有,你知道, 相互尊重和信任。 |
| 32 | But what we saw is, you know, AMD was an early investor in WorldLabs. | 但我们看到的是 AMD是WorldLabs的早期投资者 |
| 33 | So, you know, we saw Fei-Fei build when there were only, you know, | 所以,你知道,我们看到飞飞建造 当时只有,你知道, |
| 34 | just a very few researchers and really building to the team with Marble | 只有极少数的研究人员 真正建设团队与大理石 |
| 35 | and then with now the release of Atlas. | 然后现在发布阿特拉斯。 |
| 36 | And what we see is just a tremendous nucleus of capability | 而我们看到的只是 巨大的核心能力 |
| 37 | that would be a huge addition to, you know, already our very talented, | 这将是一个巨大的额外, 你知道,我们已经非常有才华, |
| 38 | you know, AMD workforce. | 你知道,AMD劳动力。 |
| 39 | So I was thrilled, thrilled that, you know, Fei-Fei and her founders | 所以我很高兴,很高兴,你知道,飞飞和她的创始人 |
| 40 | were excited about joining AMD. | 对加入AMD很兴奋 |
| 41 | And Fei-Fei, reflect on that. | 飞飞,好好想想 |
| 42 | I mean, you and I have spoken very recently. | 我是说,我们最近谈过了 |
| 43 | Yes. | 对 |
| 44 | This, of course, did not come up. | 这当然没有出现。 |
| 45 | But I suppose, you know, a good starting point | 但我想,你知道,一个好的起点 |
| 46 | other than how this came together is what opportunity did you see | 你所看到的机会 |
| 47 | for your work and the team at WorldLabs work to be a part of AMD? | 你的工作和WorldLabs的团队 要成为AMD的一部分? |
| 48 | Yes. | 对 |
| 49 | So just to echo Lisa, this relationship has been going for many, many years. | 所以为了回响Lisa, 这种关系已经持续了很多年了。 |
| 50 | It started with mutual respect and really a genuine friendship | 开始于相互尊重和真正的友谊 |
| 51 | at the personal and professional level. | 在个人和专业一级。 |
| 52 | We're very grateful when AMD was an early believer and supporter of WorldLabs, | 我们非常感谢AMD 早期的信徒和支持者 WorldLabs, |
| 53 | but really what's most fundamental and important and exciting about this | 但是,真的什么是 最重要的和令人兴奋的这一点 |
| 54 | merger is the shared vision for the future of AI. | 合并是大赦国际未来的共同愿景。 |
| 55 | WorldLabs has been a frontier model building company. | WorldLabs一直是一家前沿模型建筑公司. |
| 56 | We have been running as fast as we can to build the next generation | 我们一直在运行 尽可能快,建设下一代 |
| 57 | multimodal, omni-modal for spatial and physical intelligence. | 空间和物理智能的多式联运、全方位模式。 |
| 58 | And as our company grows, so is our ambition, so is our need for compute. | 随着我们公司的发展,我们的野心也随之增长,我们计算的需要也随之增长. |
| 59 | And at this point, if we can join force with AMD, we really are giving ourselves | 此时此刻,如果我们能与AMD联手, 我们真的给了自己 |
| 60 | and AMD and the ecosystem an opportunity to accelerate this flywheel | 和AMD 和生态系统的机会 加速这个飞轮 |
| 61 | between software and hardware development, which is critical, | 在软件和硬件开发之间,这一点至关重要, |
| 62 | critical for today's AI development. | 对今天的人工智能发展至关重要。 |
| 63 | Lisa, I think let's get into how that would work in practice. | Lisa,我想我们来谈谈这在实际中会怎样 |
| 64 | This is an eight point two billion dollar all stock transaction. | 这是820亿美元的所有股票交易。 |
| 65 | I kind of see two distinct things that AMD is acquiring. | 我看到了AMD获得的两个不同的东西。 |
| 66 | There is the insight into the development of models and what the future of AI looks like. | 对模型的开发以及AI的未来外观有深刻的见解. |
| 67 | That helps you inform future generations of hardware, | 这能帮助你了解未来世代的硬件 |
| 68 | separate from just getting a great team of researchers who can also then offer models. | 而不是仅仅得到一个伟大的研究团队 他们也可以提供模型。 |
| 69 | Let's start with the hardware piece. | 先从硬件片开始. |
| 70 | What was your thinking there? | 你当时在想什么? |
| 71 | Yeah, well, I think if you take a step back and just look at what we're trying to achieve | 嗯,我想如果你退一步 看看我们想要达到的目标 |
| 72 | across this AI ecosystem, | 穿过人工智能生态系统, |
| 73 | I mean, it's a tremendous opportunity. | 我的意思是,这是一个巨大的机会。 |
| 74 | A lot of the focus today has been on large language models, | 今天的焦点很多是大语言模型, |
| 75 | really the large foundational model labs. | 真正的基础模型实验室。 |
| 76 | 我们和OpenAI、Anthropic和Meta等人有着非常深厚的特殊关系。 | |
| 77 | And then you have labs like what Feifei has built with World Labs, | 然后你有实验室 就像飞飞公司和世界实验室一样 |
| 78 | which is actually pushing the envelope on sort of the next big thing | 这其实是把信封推到下一件大事上 |
| 79 | around world models and physical intelligence. | 环绕世界模型和物理智能 |
| 80 | We think that's extremely important as well. | 我们也认为这很重要 |
| 81 | Also, the foundational work that Feifei and her team have done in this area | 还有,飞飞和她的团队在这方面所做的基础工作 |
| 82 | are things that we will continue. | 这些都是我们将继续下去的事情。 |
| 83 | But I also see it as a nucleus for the size and scope of our ambition at AMD. | 但我也把它看作是我们AMD野心的规模和范围的核心. |
| 84 | Having a world-class research team, | 有一个世界级的研究团队, |
| 85 | and Feifei's team will be the nucleus of our AI models team, | 飞飞的团队将是我们AI模型团队的核心 |
| 86 | I think is absolutely critical when we think about all of the innovation | 我认为当我们考虑所有的创新时 绝对是关键 |
| 87 | that is to be done over the next decade. | 今后十年将开展这项工作。 |
| 88 | I think we want to do it really with innovation across the entire stack. | 我想我们真的想用创新来完成它。 |
| 89 | And so I view this as both a big vote of confidence for just the work that's already been done, | 所以我认为这既是对刚刚完成的工作 的一次大信任投票, |
| 90 | but even more importantly, what we will do together across the entire AI ecosystem. | 更重要的是,我们将一起 在整个人工智能生态系统。 |
| 91 | A lot of people will not have seen this coming, Feifei. | 很多人不会料到的,费非 |
| 92 | I actually think there's a lot of value in saying where did World Labs get to to date, | 其实我觉得说世界实验室到现在为止 |
| 93 | from founding to present day. | 从创立至今。 |
| 94 | You've been through some of your own M&A activity to grow the team. | 你经历了一些自己的并购活动来培养队伍. |
| 95 | So where do we sit today with World Labs? | 那么我们今天和世界实验室坐在哪里? |
| 96 | What is the specialism and the work you're focused on? | 你所关注的专业和工作是什么? |
| 97 | World Labs is a frontier model company specializing in spatial and physical intelligence. | 世界实验室(World Labs)是一家前沿模型公司,专门研究空间和物理情报. |
| 98 | I was just here talking about our latest model called Atlas, | 我只是在这里谈论 我们最新的模型叫阿特拉斯, |
| 99 | which is a first-of-its-kind multi-model, omni-model that can generate beautiful pixels | 这是第一种能产生美丽像素的多型、全能模型 |
| 100 | but also have the incredible 3D, 4D consistency | 也有不可思议的3D,4D一致性 |
| 101 | and allows cameras and agents to interact within these worlds. | 并且允许摄像头和代理商在这些世界中互动. |
| 102 | So we were founded in 2024, so we're a little more than two years old. | 因此,我们成立于2024年,所以我们的年龄略高于两年. |
| 103 | With really the founding principle that AI is a civilizational technology, | 以AI是一种文明技术的创始原则 |
| 104 | we really believe in the positive impact of AI | 我们确实相信大赦国际的积极影响 |
| 105 | and want to push the envelope beyond just the language modeling | 并想把信封推到 语言模型之外 |
| 106 | and going into spatial and physical because we're seeing the needs for human creators, | 进入空间和物理 因为我们看到了人类创造者的需求, |
| 107 | designers for education, health care, but also the imminent revolution of robotics. | 教育,医疗的设计者, 同时也是机器人革命即将到来的。 |
| 108 | So all this is the foundation of world models. | 所以这一切都是世界模型的基础。 |
| 109 | Over the past two years, we have built models and we have built teams. | 在过去两年里,我们建立了模型,建立了团队。 |
| 110 | We started with the kernel of computer vision, deep learning, large models, as you said. | 我们从电脑视觉的内核开始, 深刻的学习,大的模型,就像你说的。 |
| 111 | 我们还推出了几代模型,即Marvel First和Atlas。 | |
| 112 | And we also expanded our team with robotics talents. | 我们还以机器人的才能扩大了我们的团队。 |
| 113 | We recently acquired Scenics where the technologists there have built | 我们最近得到了Scenics 在那里的技术人员已经建立 |
| 114 | a first of its kind robotic simulation workflow and cutting edge robotic policy training models. | 这是第一种机器人模拟工作流程和前沿机器人政策培训模式。 |
| 115 | So WorldLab is still a startup small team, very ambitious. | 因此WorldLab仍然是一支初创小队,非常雄心勃勃. |
| 116 | Like Lisa said, we are focusing on building the next generation technology. | 正如丽莎所说,我们正在集中建设下一代技术。 |
| 117 | And we cannot be more thrilled to join force with AMD because, as I said, | 和AMD联手,我们再高兴不过了 因为,正如我所说, |
| 118 | I really believe that the future of AI is this co-evolution of AI software and AI hardware. | 我真的相信AI的未来是AI软件和AI硬件的这种共同演变. |
| 119 | Models will gain so much by using the most cutting edge hardware to accelerate the training, the inference, the servicing. | 通过使用最尖端的硬件来加速训练,推论,服务,模型将获得如此巨大的收益. |
| 120 | But in the meantime, hardware would benefit so much from the insight | 但与此同时,硬件会从洞察力中受益匪浅 |
| 121 | and the peek into the future of the workflows and needs of hardware. | 并窥视硬件工作流程的未来和需求。 |
| 122 | Lisa, is it as simple as AMD taking that work and selling models, offering software as a product? | Lisa 难道它和AMD一样简单吗? 接受这个工作并销售模型,提供软件作为产品? |
| 123 | Can I first say like ditto to everything that she just said, | 我可以先说得像她刚才说的一样吗? |
| 124 | because I think she described perfectly both the tremendous work and the opportunity. | 因为我认为她完美地描述了 巨大的工作和机会。 |
| 125 | But look, from our standpoint, AMD continues to be a platform and solutions company. | 但从我们的观点来看,AMD仍然是个平台和解决方案公司. |
| 126 | It's hardware. | 这是硬件。 |
| 127 | We have software as an enabling capability. | 我们有软件作为辅助能力。 |
| 128 | We have systems work. | 我们有系统工作。 |
| 129 | We're bringing that all together. | 我们把这些都集中起来。 |
| 130 | We work very, very closely with our customers who are some of the leading model labs out there. | 我们非常,非常密切地与我们的客户合作, 他们就是那里一些主要的模型实验室。 |
| 131 | So I wouldn't necessarily say that we're changing our model to be selling software. | 所以我不一定会说,我们正在改变我们的模型 以出售软件。 |
| 132 | I really view this as, first of all, a tremendous foundation that WorldLabs has built. | 我认为这是,首先, 一个巨大的基础,世界实验室。 |
| 133 | I think we've always said we believe in the open ecosystem. | 我认为我们总是说 我们相信开放的生态系统。 |
| 134 | And I continue to believe that that's one of the foundations of the AI innovation race going forward, | 我仍然相信这是AI创新竞赛的基础之一 |
| 135 | is you're going to have both open as well as proprietary models. | 你将会有开放和专有的模型。 |
| 136 | AMD has been a contributor there. | AMD是那里的一个贡献者。 |
| 137 | I think we are going to expand that contribution with WorldLabs as part of the equation. | 我认为,作为方程式的一部分,我们将与世界实验室一起扩大这一贡献。 |
| 138 | And the ultimate goal is, with WorldLabs, we build better AI. | 最终目标是,随着WorldLabs,我们建立更好的AI. |
| 139 | I mean, that is AMD's goal, right? | 我的意思是,这是AMD的目标,对不对? |
| 140 | Our goal is, our ambition is really to define the future of AI compute | 我们的目标是 我们的雄心壮志是确定AI计算的未来 |
| 141 | and be the best partner to the largest consumers out there. | 成为最大的消费者的最佳伙伴 |
| 142 | And I think AI WorldLabs will make us much, much better at that because of the insights | 我认为AI WorldLabs 将使我们在这点上更好得多 因为有洞察力 |
| 143 | and the opportunity to offer tremendous innovation across the stack. | 以及提供巨大创新的机会 |
| 144 | This deal is expected to close by the end of the year, which isn't too far away, subject to regulatory approvals. | 这笔交易预计年底结束,不远处,须经监管批准. |
| 145 | How is Dr. Lee going to be integrated into the AMD team? | 李博士将如何融入AMD团队?. |
| 146 | What will her role be? | 她的角色会是什么? |
| 147 | What do you picture on? | 你画什么? |
| 148 | Well, I picture on, you know, join the most exciting AI company in the world. | 好吧,我想象,你知道, 加入最刺激的AI公司 在世界上。 |
| 149 | That's what I picture on. | 这就是我想象的 |
| 150 | But look, Fei-Fei, in addition to the WorldLabs team, | 但看,飞飞 除了世界洛杉矶队之外, |
| 151 | I think Fei-Fei has a tremendous portfolio and ability to add across AMD. | 我认为飞飞拥有巨大的投资组合 和跨AMD增加的能力。 |
| 152 | So I'm very happy to say that she will be AMD's chief scientist, | 我很高兴地说 她将成为AMD的首席科学家 |
| 153 | so executive vice president and chief scientist reporting to me. | 所以,执行副总裁和首席科学家 向我汇报。 |
| 154 | And we will really want and welcome, you know, | 我们会真正想要和欢迎,你知道, |
| 155 | Fei-Fei和WorldLabs团队在AMD路线图上的贡献. | |
| 156 | And if you think about everything that we do, | 如果你想到我们所做的一切 |
| 157 | Ed, one of the things I've always said is there's no one size fits all when it comes to AI, right? | 艾德,我一直说过的一件事是 当涉及到AI的时候,没有一个尺寸的合适,对吧? |
| 158 | I mean, we talk a lot about the big cloud compute and what we're doing in terms of CPUs and GPUs, | 我的意思是,我们谈论了很多 关于大云计算 和我们在做什么 在CPU和GPU, |
| 159 | but the physical AI world, the opportunity in robotics, | 但物理AI世界,机器人的机会, |
| 160 | the end-to-end AI capability across cloud, edge and client, | 跨越云端、边缘和客户端的端对端AI能力, |
| 161 | you know, those continue to be places where I believe we're going to grow a lot over the next, | 你知道,那些继续是 地方,我相信我们将 成长在下一个, |
| 162 | you know, three, five, 10 years. | 你知道,三,五,十年。 |
| 163 | And those are all places that the WorldLabs expertise is really in our sweet spot. | 这些都是世界实验室的专长 真正存在于我们的甜点。 |
| 164 | Fei-Fei, I do need to ask, you know, the WorldLabs had raised significant money earlier in the year. | Fei -Fei,我确实需要问,你知道, 世界实验室已经筹集了大量资金 在年初。 |
| 165 | You have had a lot of momentum. | 你已经有了很多动力。 |
| 166 | What options did you have? | 你有什么选择? |
| 167 | You know, was it always a path towards joining AMD? | 你知道,它总是一条道路 加入AMD? |
| 168 | What did you consider? | 你觉得呢? |
| 169 | I think from the get-go, we build a company to really stick to our mission. | 我认为从开始,我们建立一个公司 真正坚持我们的任务。 |
| 170 | We're very mission-driven and our mission is really to make a positive impact with the technology that we believe in. | 我们的任务非常受任务驱动,我们的任务真正是为了用我们所相信的技术产生积极的影响. |
| 171 | So it's not a matter of option, it's a matter of belief. | 故曰非选非信非信. |
| 172 | And having taken this journey with AMD since the beginning, | 从一开始就跟AMD一起旅行了 |
| 173 | them being our investor as well as chip provider, | 他们既是我们的投资者也是芯片供应商, |
| 174 | and then gradually talking to Lisa about our shared vision, | 然后逐渐和丽莎谈我们的共同愿景, |
| 175 | it truly is the merging of the shared vision and mission. | 这确实是共同愿景和使命的结合。 |
| 176 | So it's not about option, it's about what we can do in the future. | 因此,这不是关于选择,而是关于我们未来所能做的事情. |
| 177 | And I genuinely believe this is going to be one plus one bigger than two. | 我真心相信这将会是1+1大于2。 |
| 178 | 公平地说,你比WorldLabs还大 | |
| 179 | But it's really going to be a very, very exciting journey together. | 但它真的会是一个非常, 非常令人兴奋的旅程在一起。 |
| 180 | In that shared vision, clearly emphasis on open source | 在这个共同愿景中,明确强调开放源码 |
| 181 | and is a non-negotiable in the direction of travel for AI. | 并且是AI在旅行方向上不可谈判的. |
| 182 | Let's speak a little bit more about that, Lisa. | 我们再多谈谈,丽莎 |
| 183 | I mean, why is it important for AMD from the company's perspective | 我的意思是,为什么它是重要的AMD 从公司的角度 |
| 184 | to advocate for an open source market and environment, | 倡导开放源码市场和环境, |
| 185 | but also some of the models that the WorldLabs have worked on have been open, | 但也有一些世界实验室的模型是开放的, |
| 186 | some have been closed or proprietary, and how you balance that going forward? | 有的已经关闭或专有, 你怎么平衡 前进? |
| 187 | Yeah, I think the key is what Fei-Fei said. | 我想关键是飞飞说的 |
| 188 | It is about shared vision for what we're trying to accomplish. | 这是关于共同的愿景 我们正在努力实现。 |
| 189 | And we are trying to be a very big contributor to the expansion, | 我们正在尝试成为 一个非常大的贡献 扩张, |
| 190 | the acceleration, the innovation in the AI ecosystem. | 加速 AI生态系统的创新 |
| 191 | Yes. | 对 |
| 192 | And to do that, I think you're going to have open models, | 为了做到这一点,我认为 你会有公开的模型, |
| 193 | you're going to have proprietary models, you're going to have large models, | 你会有专有的模型, 你会有大的模型, |
| 194 | you're going to have small models, they're going to run on different hardware. | 你会有小模型 它们会用不同的硬件运行 |
| 195 | And that's been a premise for AMD in our entire high-performance computing | 这一直是AMD的前提 在我们的整个高性能计算 |
| 196 | and AI roadmap going forward. | 和AI的路线图前进。 |
| 197 | So I think what you're going to see is more emphasis on having optionality. | 所以,我认为你将要看到的是 更强调有选择性。 |
| 198 | Enterprises want optionality. | 企业希望有选择性。 |
| 199 | You know, when I talk to enterprise CEOs and CIOs, | 我和企业首席执行官和首席执行官谈话时 |
| 200 | nobody wants to be locked into either one hardware ecosystem or one software ecosystem. | 没有人想被锁在一个硬件生态系统或一个软件生态系统。 |
| 201 | People want to be able to choose what's best at a given time. | 众生欲能于一时选择最上品. |
| 202 | And we also want to enable all of these entrepreneurs | 我们还想让所有这些企业家 |
| 203 | and new companies out there that are coming up with great ideas. | 和新的公司 在那里,正在产生 伟大的想法。 |
| 204 | So you're going to see us continue to lean into open ecosystems | 所以你会看到我们继续 向开放的生态系统倾斜 |
| 205 | as a key way of enabling the entire AI ecosystem. | 作为整个人工智能生态系统的关键手段。 |
| 206 | And with that, I do think part of what Fei-Fei and I have talked about | 我和飞飞谈过的 也是我的想法 |
| 207 | is how to leverage some of the great work that WorldLabs has already done | 如何利用世界实验室已经做的一些伟大工作 |
| 208 | to actually accelerate some of the open ecosystem going forward. | 以加速一些开放的生态系统前进。 |
| 209 | Please. | 请 |
| 210 | Can I add? | 我能补充一下吗? |
| 211 | So the history of modern AI cannot be possible without open ecosystem | 所以没有开放的生态系统 现代AI的历史是不可能的 |
| 212 | from the open sharing of neural network algorithms through scientific communities | 通过科学界公开分享神经网络算法 |
| 213 | to my own work, ImageNet, right? | 对自己的作品,ImageNet,对不对? |
| 214 | Yes. | 对 |
| 215 | ImageNet was the first open source data set that was part of the driving force | ImageNet是第一个作为驱动力的一部分的开源数据集 |
| 216 | for modern AI revolution. | 为现代AI革命。 |
| 217 | So I think just to add on to what Lisa said, we need optionality. | 所以我想补充一下丽莎的话 我们需要可选性 |
| 218 | And it's so important that an open ecosystem helped to cross pollinate discovery, | 这很重要,一个开放的生态系统 有助于跨越授粉发现, |
| 219 | innovation, entrepreneurship, business opportunities. | 创新、创业、商业机会。 |
| 220 | This is the meaning and impact of an open ecosystem. | 这是开放生态系统的意义和影响。 |
| 221 | I think a lot of people will say, Fei-Fei, WorldLabs has focused on what we've called | 我想很多人会说,飞飞,世界实验室 专注于我们所说的 |
| 222 | spatial intelligence, a focus on physical AI, but other areas of the market | 空间智能,注重物理AI,但市场其他领域 |
| 223 | where a world model is more appropriate than text and large language model. | 世界模式比文本和大语言模式更合适。 |
| 224 | Do you now go broader in your work within AMD and have not such a, | 你现在在AMD的作品中走得更宽,没有这样的, |
| 225 | narrow is a strange way to put it, but yes, a broader field of work across model types? | 狭义是一个奇怪的说法, 但是是的,一个更广泛的工作领域 跨模式类型? |
| 226 | Yeah, look, we have a lot to look forward to in the future. | 是啊,你看,我们有很多 期待在未来。 |
| 227 | Lisa and I barely got time to talk about all the plans. | Lisa和我几乎没时间谈所有计划 |
| 228 | Obviously, this is going to accelerate and expand our roadmap, right? | 显然,这将加快和扩大我们的路线图,对不对? |
| 229 | We're going to have a lot of conversations. | 我们将有很多对话。 |
| 230 | We're definitely going to keep doing some of the most exciting work that WorldLab | 我们一定会继续做一些 最令人兴奋的工作,世界拉布 |
| 231 | is already doing, but I do envision that we will be expanding and accelerating. | 已经在做,但我确实设想我们将扩大和加速。 |
| 232 | Lisa, this is the first opportunity I've had to speak to you in the era of debate | 丽莎 这是我在辩论时代第一次和你说话 |
| 233 | around AI safety. | 环绕AI安全。 |
| 234 | It's hard to make it succinct, but the argument that the capabilities | 很难简洁,但关于能力的论点 |
| 235 | of models are advancing at a pace greater than safety and alignment are advancing | 以比安全更快的速度推进模式,协调性也正在推进 |
| 236 | to keep up with that development. | 以跟上这一发展。 |
| 237 | What is AMD in your position on how best to address | AMD在如何最好地解决 |
| 238 | all of the concerns that are out in the world at the moment? | 目前世界上所有的问题? |
| 239 | Yes, Ed, this is an extremely important point, so I'm actually glad you brought it up | 是的,埃德,这是一个极为重要的问题, 所以我很高兴你提出来。 |
| 240 | because it is one that we have to be super clear about as an industry. | 因为它是一个我们必须非常清楚 作为一个行业。 |
| 241 | There is no optionality in this. | 这一点是没有选择性的。 |
| 242 | We must have great technology and it must be safe. | 我们必须有伟大的技术,它必须是安全的。 |
| 243 | That's just unequivocal. | 这很明确 |
| 244 | When I look at just history of how we develop technologies, | 当我看看我们开发技术的历史 |
| 245 | it is one of those things where safety has to become a first-class citizen | 这是其中之一 安全必须成为一流公民 |
| 246 | when you're doing development. | 当你在做开发。 |
| 247 | We've seen that in our own work when we think about | 我们从自己的作品中看到过 当我们想到 |
| 248 | all of the security and the confidentiality things that you have to do around | 所有你必须做的安全和保密的事 |
| 249 | the highest performing processors. | 最高的处理器。 |
| 250 | I think we're going to have to bring that | и稱и璶盿ㄓ |
| 251 | into the AI ecosystem and I know that that's front and center in the conversation. | 进入人工智能生态系统 我知道这是对话的前沿和中心 |
| 252 | I am actually an AI optimist. | 我其实是一个AI乐观主义者。 |
| 253 | I believe that we will put the right procedures, processes, | 我相信我们会把正确的程序,过程, |
| 254 | guardrails, I think an industry-wide effort where we actually share best practices | 护栏,我想是整个行业的努力 我们实际上分享最佳做法 |
| 255 | and that's one of the things that we do on the hardware side today, for example, | 这就是我们今天在硬件方面做的之一 比如说, |
| 256 | around security. | 周围的安全。 |
| 257 | Those are all important aspects of it but I think we need to be | 这些都是重要的方面,但我认为我们需要 |
| 258 | unequivocal to those who are the users of AI. | 对大赦国际使用者毫不含糊。 |
| 259 | We absolutely are going to deliver | 我们绝对会实现 |
| 260 | great technology that is safe, that is aligned with its purposes and it requires | 安全、符合其宗旨和要求的伟大技术 |
| 261 | a tremendous amount of work so it's not going to be easy but it is definitely | 大量的工作,所以它不会是容易的,但它肯定 |
| 262 | something that we are committed to as an industry and we must work together | 作为产业,我们致力于的东西 我们必须共同努力 |
| 263 | {\fn黑体\fs22\bord1\shad0\3aHBE\4aH00\fscx67\fscy66\2cHFFFFFF\3cH808080}为了实现这个目标 | |
| 264 | When we spoke last week, Fei-Fei, I listed to you regulation, third-party evaluation, | 上星期我们谈话时,飞飞 我向你们列举了规章,第三方评估 |
| 265 | the Frontier Labs taking responsibility and your response was all of the above | 边疆实验室负责,你们的答复就是以上所述。 |
| 266 | but how do you see your work within AMD on this issue of AI safety, also bearing in mind | 但你如何看待你在AMD内的工作 关于AI的安全问题, 也铭记着 |
| 267 | your long history with this industry and this field of research? | 你对这个行业和这个研究领域有悠久的历史吗? |
| 268 | Yeah, first of all, I totally agree with Lisa and like I said, this entire happening | 是的,首先,我完全同意 丽莎,就像我说的,这一切发生 |
| 269 | is because of the shared vision and shared mission so safety is shared responsibility. | 安全是共同的责任。 |
| 270 | AI is a civilizational technology and like all civilizational technologies, | AI是一种文明技术 就像所有文明技术一样 |
| 271 | we start with the intention of making human lives and work better but along the way | 我们一开始的意图是让人的生命和工作更好 但一路 |
| 272 | we learn that we have to use it, develop it and deploy it responsibly because every | 我们了解到我们必须使用、发展和负责任地部署它,因为 |
| 273 | technology is a double-edged sword so I do think whether it was at World Labs | 科技是一把双刃剑 所以我想它是否在世界实验室 |
| 274 | or Stanford or now joining AMD, I really want to work with Lisa, her team and the rest | 或斯坦福或现在加入AMD, 我真的想和丽莎,她的团队和其余的一起工作 |
| 275 | of the ecosystem to see how we can bring out this shared responsibility at individual level, | 如何在个人层面提出这一共同责任, |
| 276 | individual company level, industry level as well as societal level. | 公司一级、行业一级和社会一级。 |
| 277 | I find this fascinating because I want to go back to AMD's acquiring World Labs | 我觉得这很有趣 因为我想回到AMD收购世界实验室 |
| 278 | that gives deep insight into the development of industry and the models. | 从而深入了解工业和模型的发展。 |
| 279 | The capabilities of the models and training runs improve also because the compute, | 模型和培训的能力提高也因为计算 |
| 280 | its capabilities is improving generation to generation. | 它的能力正在不断提高。 |
| 281 | I guess the root of the question is what responsibility then the compute platform | 我想问题的根源是 计算平台的责任是什么 |
| 282 | providers take for how that training is conducted safely and then on the inference side, | 提供商如何安全地进行培训,然后在推论方面, |
| 283 | the running of those models that they are aligned, they behave as we intended them to do. | 这些模式的运行, 它们是对齐的, 他们的行为就像我们想要他们做的。 |
| 284 | Well, it's a great point and I think every one of us in the ecosystem has a responsibility | 嗯,这是一个伟大的观点 我认为我们每个人在生态系统 有责任 |
| 285 | and we are accountable for our actions. | 我们对自己的行动负责。 |
| 286 | So from us as a hardware platform, we need to make sure that our hardware is secure, | 所以作为一个硬件平台 我们需要确保我们的硬件安全 |
| 287 | that it does what it's meant to do, that when models run on it that the data is secure, | 当模型运行时 数据是安全的 |
| 288 | that is something that is very much front and center in what we do. | 这正是我们所做的事情的前沿和中心所在。 |
| 289 | We need to work with the model companies in terms of how the hardware and software interact. | 我们需要在硬件和软件的相互作用方面与模型公司合作。 |
| 290 | I think Fei-Fei's point here is the thought leadership aspect of it is we must step up as an industry | 我认为飞飞的要点是 领导思想的方面是 我们必须作为一个行业加强 |
| 291 | and I do think this is an industry led thing in terms of we have the accountability to put out safe technology, | 我认为这是一个行业领导的事情, 因为我们有责任 推出安全技术, |
| 292 | but there are ways for us to work more cohesively across the industry because this is kind of new | 但是,我们有办法 更一致地在全行业工作 因为这是一种新的 |
| 293 | and whenever you have something that's kind of new, you're actually learning. | 每当你有了新的东西, 你实际上正在学习。 |
| 294 | Each party might be learning at a different pace and so what I really welcome is the opportunity | 每个派对可能学习的速度不同 所以我真正欢迎的是机会 |
| 295 | for more industry collaboration as well as some third party evaluation to ensure that every part of the stack | 进行更多的行业合作,以及一些第三方评价,以确保堆栈的每一部分 |
| 296 | is working together towards this objective of phenomenal technology that is safe and aligned to its objectives. | 正在共同努力实现这一安全并与其目标相一致的惊人技术目标。 |
| 297 | Lisa, this is a massive moment for AMD. | 丽莎,这是AMD的巨大时刻。 |
| 298 | Actually, it's a moment after some sustained momentum. | 实际上,这是在保持势头之后的一刻。 |
| 299 | One trillion dollars of market cap, a big moment for the company. | 1兆元的市场封顶,对公司来说是一个重大的时刻. |
| 300 | You have been very visible alongside the President of the United States, a leader in industry. | 你和美国总统 工业界的领袖 一起非常引人注目 |
| 301 | What happens next for AMD? | AMD接下来会怎么样? |
| 302 | I appreciate that World Labs is still to close, but what you see is AMD in the future. | 我很感激世界实验室仍然关闭 但你们看到的是未来的AMD |
| 303 | Well, first I'd like to say, Ed, I love this industry. | 首先我想说 Ed 我喜欢这个行业 |
| 304 | I love what we're able to do. | 我喜欢我们能做的 |
| 305 | I love the comments that Fei-Fei made about first and foremost, | 我喜欢飞飞首先说的话 |
| 306 | we're here to make technology better for society, better for people, and that's what I get excited about every day. | 我们在这里是为了让技术 对社会更好,对人更好, 这就是我每天兴奋。 |
| 307 | The fact that our execution has been strong, that we've continued to really work with the leaders of the industry, | 我们的行刑力度很强 我们继续和业界领袖合作 |
| 308 | both large companies as well as we've done a number of acquisitions that have made a large difference to AMD, | 无论是大公司还是我们 已经做了很多收购 给AMD带来了很大的影响, |
| 309 | and World Labs will be the next and an extremely strategic and important one. | 和世界实验室将是下一个,并且是一个极具战略性和重要性的实验室。 |
| 310 | I think this is all part of recognizing that technology is such an important part of society today, | 我认为这一切都是承认 科技是当今社会的重要组成部分的一部分 |
| 311 | whether you're talking about the economy or you're talking about national security, | 不管是经济还是国家安全 |
| 312 | and AMD has a chance to truly influence and direct where the future of computing is going. | AMD有机会真正影响和引导 计算机的未来走向。 |
| 313 | So no better place to be. | 所以没有更好的地方了 |
| 314 | And Fei-Fei, to finish our conversation with you, what happens next? | 飞飞 为了结束我们和你的对话 接下来会怎样? |
| 315 | You know, you and I have discussed this in various mediums, the idea that you're the godmother of AI. | 你知道,你和我在各种媒介上讨论了这个问题, 认为你是AI的教母。 |
| 316 | I'm going to keep her very busy. | 我会让她很忙的 |
| 317 | You know, you have been researching in the space for a long time. | 你知道,你一直在空间研究 很长时间。 |
| 318 | Now, this is the next space for World Labs as part of AMD. | 这是世界实验室作为AMD的一部分的下一个空间。 |
| 319 | But I want that sort of crystal ball of what you think will the next major development in AI will be, | 但我希望你认为AI的下一个重大发展是 |
| 320 | something that you stay up late thinking about, that you're writing about, that you pitched Lisa back on, | 有些事你晚点才想到, 你正在写, 你让丽莎回到, |
| 321 | saying, I think this is what's going to happen. | 说,我想这是会发生什么。 |
| 322 | What happened next is Lisa and I, AMD and World Labs, will build the future together. | 接下来发生的是Lisa和I,AMD和World Labs,它们将共同构建未来. |
| 323 | And I believe so strongly this future has to be benevolent to the human society and human lives. | 我坚信,这个未来必须有利于人类社会和人的生命。 |
| 324 | And this is our shared responsibility. | 这是我们的共同责任。 |
| 325 | And it's what makes us so excited. | 这也是我们兴奋的原因 |
| 326 | AMD首席执行官 Lisa Su,世界实验室创始人,CEO Fei-Fei Li,不久成为AMD首席科学家. | |
| 327 | and a part of the models teams at AMD, among other things. | 以及AMD的模型团队的一部分,等等。 |
| 328 | Thank you both very much. | 非常感谢你们 |
| 329 | Congratulations. | 祝贺你 |
| 330 | Thank you. | 谢谢 |
该视频共有字幕 331 条。解锁更多字幕为会员功能，请移动到 价格