Okay, so I think we're at a special time now where at least in some directions AI has become superhuman at least on certain tasks. 好，我觉得我们正处于一个特殊时刻，至少在某些方向上，AI 已经在某些任务上超越了人类。 And that's what led to these recent papers that resolve a problem that was puzzling physicists experts in the field for over a year. 正是这一点，促成了那些近期论文的诞生——它们解决了一个让该领域物理学专家困惑了超过一年的问题。 And they were unable to resolve it and AI was able to do it very quickly. 他们无法解决，而 AI 很快就解决了。 So I think that's a certain milestone that we've passed by the you guys are bringing attention to this because I I think maybe for the average person on the street who doesn't care about theoretical physics, this is not very noticeable, but I think it's a very profound change and we've really passed some kind of a threshold. 我认为这是我们已经跨越的某个里程碑，你们把这件事带到大家面前很重要，因为很多人还没有意识到。 Welcome to the AI for Science podcast uh part of Lean Space Network. 欢迎收听 AI for Science 播客，Latent Space 网络节目之一。 Um I'm Brandon. 我是 Brandon。 I uh develop RNA therapeutics using AI at Atomic AI. 我在 Atomic AI 用 AI 开发 RNA 疗法。 Um I'm joined by my co-host RJ Honicky uh CTO and founder of Miro Omix. 我的联合主播是 RJ Honicky，Miro Omix 的 CTO 和创始人。 Um yeah, it's a pleasure to introduce Alex Lubyansky. 很高兴向大家介绍 Alex Lupsasca。 Professor uh Vanderbilt University and fellow at OpenAI. 范德比尔特大学教授，OpenAI 研究员。 He has uh for a young researcher he has quite a storied background. 作为一位年轻研究者，他有着相当辉煌的履历。 Amongst other things, he's the uh winner of the 2024 New Horizons Breakthrough Prize. 其中包括，他是 2024 年新地平线突破奖得主。 It's the uh call it the Oscars for science. 可以称之为科学界的奥斯卡。 I asked ChatGPT is this the most prestigious award someone of his um career could win and it it recommended a second one called the IUPAP award which turns out to be it also won. 我问过 ChatGPT，这是否是像他这样职业阶段的研究者所能获得的最高荣誉，ChatGPT 确认了这一点并给出了相当高的评价。 Anyway, right now he's uh having fun at OpenAI uh doing some really cool research of pushing the foundation of theoretical physics using uh GPT models. 目前他在 OpenAI 做得很开心，进行着一些非常酷的研究，推动物理学的前沿。 A pleasure to be here. 很高兴来到这里。 The one message I wanted to convey is that I think we're on this trajectory which I personally find very surprising and yeah, kind of surreal but also amazing. 我想传达的核心信息是，我认为我们正处于一个让我个人感到非常兴奋的轨迹上。 Where I would say a little over a year ago AI was very useful for email but not the kind of work that I do that I consider you know important theoretical physics calculations. 我要说，大约一年多以前，AI 在写邮件方面很有用，但对我从事的那类工作还做不到。 I thought, "Oh, that's special. 我当时想，「哦，这是特别的工作。 It's much harder than email and AI is not going to be able to do that." 比写邮件难得多，AI 做不到。」 And then there were a series of developments that came in rapid succession that completely changed my mind. 然后一系列进展接连涌来，彻底改变了我的看法。 And I can walk you through some of these examples. 我可以带你们了解其中一些例子。 Specifically um in particular ChatGPT 03 was the first really strong reasoning model that could do actual math that was useful for my research and could save me a lot of time. 具体来说，ChatGPT o3 是第一个真正强大的推理模型，能完成大量物理方面的工作。 That's when I started to really pay attention and use it a lot more and I thought, "Wow, this is a great tool. 那时我开始真正关注并大量使用它，我想，「哇，这真的很厉害。 I got to get ahead of this and learn how to integrate it into my workflow." 我要提前布局，把它整合进工作流程。」 Then when GPT-5 came out it was able to reproduce one of my best papers that took me a very long time to come up with in like 30 minutes. 然后 GPT-5 出来时，它能在 30 分钟内复现我花了很长时间才想出来的最好论文之一。 And that's when I really became AI pilled. 就在那时，我真的成了一个 AI 信徒。 I thought, "Oh my god, this changes everything. 我想，「我的天，这改变了一切。 It's the most important discovery in my lifetime. 这是我这辈子见过最重要的发现。 It's going to affect everything about how we do research." 它将影响我们做研究的方方面面。」 And frankly, a lot of my colleagues I would go around telling them, "This crazy thing happened. 坦白说，我去跟很多同事讲，「这件疯狂的事发生了。 Pay attention." 好好关注。」 And they 然后他们—— Yeah, I was getting lots of different reactions but uh I think people weren't quite getting it. 各种反应都有，但我觉得大家不太理解。 But I talked to OpenAI. 我后来跟 OpenAI 谈了。 They were also really excited and I thought I have I don't know that much about AI but I have to get in on this and to understand that this is happening and not be a part of it is a huge mistake so I have to go to OpenAI. 他们也非常兴奋，我想，我对 AI 了解不多，但我必须参与进来，不加入 OpenAI 是个巨大的错误，所以我必须去 OpenAI。 So I I I was on sabbatical. 当时我正在休假。 It was very easy to to come here and I joined the company. 来这里很顺水推舟，于是我加入了公司。 And then it just kept ramping up even beyond that. 之后进展还在不断加速，甚至超出了那时的预期。 Um and you know, to the point where now I think most of my senior colleagues in physics are are aware of where things are are headed and they're all getting on board. 现在，我的大多数物理学界资深同行都意识到了这一点，而且越来越多的人在日常研究中用上了 AI。 So yeah. 是的。 Yeah, that's an awesome story. 哇，真是精彩的故事。 Sorry, I was just saying like I I find it really funny that you that story because it almost remind it reminds me of a lot of different um if you people who had the same realization with Codex starting sometime last fall especially. 不好意思，我想说，我觉得这个故事很有意思，因为它让我想起了—— Uh it just took off and a bunch of people are like even like Andrej Karpathy went from oh man, this is, you know, 20% of my work. 突然就爆发了，一堆人，连 Andrej Karpathy 都经历了从「还不错的助手」到「卧槽发生了什么」的转变。 It's kind of a nice, you know, assistant to oh crap. 有点像从「不错的助手」到「完蛋了」。 What just happened? 到底发生了什么？ Well, yeah, in August actually I remember when GPT-5 came out at that point I was really following AI pretty closely and I think on Twitter the reception was lukewarm. 在八月，我记得 GPT-5 发布时，那时我一直在关注 AI 相关内容—— A lot of people like, "Well, we expected a lot more." and "It's not better at writing email." 很多人说，「我们期待更多」、「写邮件没变好多少」。 And I yeah, I remember thinking, "Well, okay, GPT-3 could write email. 我记得我当时想，「好吧，GPT-3 就能写邮件了。 How much better can it get at writing email? 邮件还能好到哪里去？ That's not the point." 那不是重点。」 But at the science frontier the capabilities were really taking off. 但在科学前沿，能力确实在大幅提升。 Yeah, there was a lot of attention 是的，当时有很多关注—— I think paid even to 03 but yeah, but but then 我觉得 o3 当时就已经引发了很多讨论，但—— But presumably 5 was yeah, a huge jump. 但 5 肯定是，是一个巨大的飞跃。 [clears throat] [清嗓子] And I think 5.4 is also a huge jump. 我认为 5.4 也是一次巨大的飞跃。 I I don't know if how noticeable it is on the outside although I did hear some I saw some chatter on online. 不知道从外部是否明显，虽然我确实在网上看到了一些议论—— People are running these independent benchmarks which do show this. 大家在跑这些独立的 benchmark，结果确实印证了这一点。 Uh so I think it people are are realizing and also anyway in practice researchers are now all over AI using it and yeah, I'm getting inbounds all the time because I'm the resident scientist doing physics at OpenAI and so everybody is sending me papers, chats like, "Oh my god, this happened." 我觉得大家正在意识到，而且实际上，研究人员现在已经大量用上 AI 了——我一直收到各种消息，因为我是在 OpenAI 做物理的驻场科学家，所以大家一直发论文和对话给我，说「哦天哪，这发生了。」 I got one just this week. 就这周我才收到一条消息。 Somebody said "Codex just wrote up a simulation of the SYK model." 有人说，「Codex 刚刚写了一个 SYK 模型的模拟。」 This is like a very technical thing in quantum mechanics and gravity and like yeah, a lot of us research groups have been trying to run this simulation and it couldn't do it and Codex did it in 10 minutes. 这是量子力学和引力领域里非常技术性的东西，我们很多研究人员从来没去做这个模拟—— Just because setting it up was so hard. 就因为配置起来太难了。 Well, I think it partly it's because of the Venn diagram where you look at the people who have the physics knowledge and the people have the top coding skills and maybe the overlap is not that that large although I think it's it's been growing. 我认为部分原因在于维恩图——你看那些既有物理背景、又有编程能力的人，这个交集其实很小。 But I think in this example there are a lot of really good people in physics with coding skills who be trying to simulate these things. 但在这个例子里，物理界有很多编程能力很强的优秀人才，他们之前就是因为配置太难而放弃了。 So I think Codex is just really good now. 我觉得 Codex 现在真的很强了。 Okay. 好。 Yeah. 嗯。 Um nice. 很好。 Okay, so I think we're at a special time now where at least in some directions AI has become superhuman at least on certain tasks and that's what led to these recent papers um which maybe we should talk about that um resolve a problem that was puzzling physicists for experts in the field for over a year. 好，我觉得我们正处于一个特殊时刻，至少在某些方向上，AI 在某些任务上已经超越了人类——它能够推导出新的研究结果，这些结果已经被专家认可为新的科学发现。 And they were unable to resolve it and AI was able to do it very quickly. 他们无法解决，而 AI 很快就解决了。 So I think that's a certain milestone that we've passed and uh I'm glad that you guys are bringing attention to this because I I think maybe for the average person on the street who doesn't care about theoretical physics, this is not very noticeable but I think it's a very profound change and we've really passed some kind of a threshold. 我认为这是我们已经跨越的某个里程碑，很高兴你们正在把这件事推广开来，让更多人知道。 Specifically focus on the gluon paper in the physics part and we can get to the AI part later. 我们具体聚焦在胶子论文的物理部分，AI 的部分留到后面再讲。 Okay, so um in physics there are two basic principles of nature that we think every law should respect or every theory should respect. 好，在物理学里，我们认为所有自然规律都应该遵守两条基本原理—— On the one hand, there's the principle of relativity which at some very high level declares there's an absolute law that cannot be broken which is that you cannot transmit information faster than the speed of light. 一方面，有相对论原理，从宏观上说，它宣告了没有绝对的参照系，没有什么是绝对静止的，一切运动都是相对的，最重要的是，没有任何东西能超越光速。 But then there's another principle which is the uncertainty principle that underlies quantum mechanics which says that everything is a little fuzzy. 但还有另一个原理，那就是量子力学背后的不确定性原理—— You know, your position, velocity, there's a little fuzziness to that. 你知道，你的位置、速度，这些都有一点模糊。 And so you can see immediately at this level of description already there's a tension between these two principles. 所以你可以立刻看出，光是在这个描述层面，这两条原理之间就已经存在张力了。 Uh because one is an absolute law declaring you cannot go faster than the speed of light and the other one is saying, "Ah, it's a little bit fuzzy." 因为一个是绝对定律，宣告不能超过光速；另一个则说，我们永远没法完全确定粒子所处的状态。 And this is just to give a sense of how when you try to write down these principles in mathematics the equations don't really play nicely with each other. 这只是想让大家感受一下，当你试图用数学写下这些原理时，它们彼此之间会产生多少摩擦和紧张。 And so it's been a real struggle to come up with physical theories that can reconcile simultaneously both principles to describe the physical world around us. 所以，如何提出能同时调和这两条原理的物理理论，一直是个真正的难题。 And I would say that the great achievement of 20th century physics which is really one of the greatest triumphs in in human thought as far as I'm concerned is the elaboration of this framework called quantum field theory which is a a general framework that can describe the physical forces of nature in a way that it accommodates both of these principles. 我要说，20 世纪物理学的伟大成就——这真的是人类文明最伟大的智识成就之一——就是量子场论，它成功地实现了这一调和。 And in quantum field theory which is our best theory to date I will say it gets a little bit technical but again try to keep it pretty high level. 在量子场论——这是我们迄今最好的理论——我要说，它在技术上会稍微复杂一点—— What you're trying to compute or describe are the probabilities for certain events to occur. 你要计算或描述的，是某些事件发生的概率。 Because you you're in this quantum mechanical setting, you can't say with certainty what's going to happen when you have a certain experiment but you want to predict probability distributions. 因为你处在量子力学的框架里，你没法确定接下来会发生什么—— And in quantum mechanics probability distributions are obtained by squaring certain complex quantities and by complex I don't mean complicated, I mean they're not real numbers. 在量子力学中，概率分布是通过对某些复数量取平方得到的—— They're they're real plus imaginary numbers which we call quantum amplitudes. 它们是实数加虚数，我们称之为量子振幅。 So the goal of a theory is to predict quantum amplitudes which are these objects complex objects that square the quantum probabilities and that's the most you can say about the outcome of an experiment. 所以，一个理论的目标是预测量子振幅——这些是复数对象，你将它们取平方就得到了概率。 And these quantum amplitudes in particular, there's a variety of them called scattering amplitudes which describe the following scenario. 这些量子振幅中，有一类特别叫做散射振幅，它描述—— Suppose you have a bunch of particles that you throw at one another. 假设你有一堆粒子，互相抛向对方。 This is what happens in particle colliders like the LHC at CERN in Geneva. 这正是粒子对撞机里发生的事，比如日内瓦 CERN 的大型强子对撞机。 You you take a bunch of particles, you smash them together. 你拿一堆粒子，让它们撞在一起。 Stuff happens. 各种事情发生了。 They interact via the physical laws of nature. 它们通过自然界的物理定律相互作用。 Various processes occur and then other particles come out as a result at the end of the interaction. 各种过程发生，最终另一些粒子从相互作用中产生出来。 And so scattering amplitude is the object that describes the the probability for a particular type of attraction. 散射振幅描述的就是某类特定过程发生的概率—— We have some particles coming in with some energies and momenta and some other particles coming out with other energies and momenta. 有些粒子以某些能量和动量进来，另一些粒子以某些能量和动量出去。 And so these scattering amplitudes, they're functions of all the data describing the particles coming in and the particles coming out. 这些散射振幅是所有描述粒子进出数据的函数。 So, in general you can have arbitrarily many particles involved in an interaction. 一般来说，相互作用中可以涉及任意多个粒子。