Claude在湿实验室取得生物发现，OpenAI解决100道数学题

欢迎收听本期播客。

Welcome to the podcast.

今天我们来聊聊Anthropic在生物学领域的最新突破之一。

Today we wanna talk about the one of the newest breakthroughs that anthropic has had in biology.

当然，Anthropic和Dario之间有不少戏剧性的事情，他们说需要控制前沿技术的发展速度，但同时他们又宣布在旧金山开设了一个湿实验室。

There's a bunch of drama, of course, with anthropic and Dario saying they need to pace the frontier, and at the same time they were saying they were opening a wet lab in San Francisco.

湿实验室开放21小时后，就已经有了第一个发现，Dario也一直在谈论这件事，这引起了很多兴奋，但也招来了批评，因为有人担心Anthropic会拖慢许多伟大科学进展的步伐。

After twenty one hours of the wet lab being opened, it has already made its first discovery and Dario has been talking about it, which has led to a lot of excitement, but also backlash for people concerned that anthropic is gonna slow down the development of a lot of great scientific things.

有人担心他们会，你知道的，制造出下一个新冠。

There's people worried that they're gonna, you know, create the next COVID.

所以这个话题有很多不同的看法。

So there's all sorts of sides of the conversation.

我们将深入分析正在发生的事情以及他们在生物学上的发现，因为我认为这是最前沿的进展之一，当像Anthropic这样的公司把全部研究力量投入到AI和科学上时。

We're gonna be breaking down what's going on and what they're discovering in biology, as I think this is one of our most advanced movements when you have a company like Anthropic putting the full force of their research power behind AI and science.

今天节目会很有意思。

This is gonna get interesting today on the show.

在聊这个之前，我想提一下，如果你在用Claude，并且用像Claude CoWork这样的工具，它真的很棒，能帮你在电脑上搞定很多事情，但你可能觉得Claude CoWork不能生成图片、音频或视频，挺遗憾的。那我特别希望你能看看我自己的创业项目，叫AI Box。

Before we talk about that, I wanted to mention, if you use Claude, and you use something like Claude CoWork, which is an amazing tool for getting stuff done on your computer, and you're frustrated that Claude Cowork doesn't have the ability to generate images, audio, or video, I would love for you to check out my own startup, which is called AI Box.

我们做了一个连接器，让所有其他AI工具都能在Claude里面生成内容。

We created a connector that lets all of the other AI tools be generated inside of Claude.

所以，我最近有个经历，我想为一个项目生成大量图片，我在Claude里用了这个连接器，告诉它我想要的图片类型，但那些图片都是要OpenAI生成的。

So, I recently had an experience where I wanted to get a ton of images generated for a project I was working on, and I had this connector inside of Claude and I told it, you know, the types of images I was wanting, but they were all OpenAI images to be generated.

Claude居然能生成超过600张图片，并且全部放进项目里，结果那个网站的流量现在暴涨，过去三个月涨了大概10倍。

Claude was able to go generate over 600 images and put them all into the project, and the traffic on that site consequently is now skyrocketing, and it's up like 10 x in the last three months.

所以总的来说，Claude有很多厉害的功能，但它也不是万能的。

So, overall, there's so many incredible capabilities, Claude doesn't have them all.

如果你想在你用Claude或其他AI工具（比如Muse或ChatGPT work）做的项目里加上音频、图片和视频生成功能，你可以获取AI Box MCP连接器。

If you wanna add all of the audio image and video generation capabilities into the projects that you're working on with Claude or with any other AI tool, like Muse or ChatGPT work, you can get the AI Box MCP connector.

链接在描述里。

There's a link in the description.

起步价只要8.99美元，我们尽量让所有人都能负担得起。

It's only $8.99 to get started, try to keep this affordable for everyone.

链接在描述里，让所有AI模型协同工作。

Link in the description, get all the AI models to work together.

好，我们来谈谈达里奥发布的重磅消息，这是对Anthropic两小时前发的一条推文的回应，推文说Claude发现了一个以前未知的酶系统，隐藏在噬菌体的DNA中，基于酶基因，或者酶基因旁边有一长串重复DNA，结构有点像CRISPR。

Okay, let's talk about the the big announcement that Dario made, it was in response to Anthropic coming out with a tweet, this is just two hours ago, where they said Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriographs based on the enzymes gene, or beside the enzymes gene sits a long array of repeated DNA, a structure that looks somewhat similar to CRISPR.

好，达里奥随后在X上发布了一段解释，说明了实际发生了什么。

Okay, so Dario came on and gave kind of a breakdown of what's actually happening over on X.

他说，他们宣布的是Claude主导的发现，所以显然不是Claude完全独立完成的。

He said that, you know, they were announcing the Claude lead discovery, so obviously it's not completely just Claude running all by itself.

Claude主导的意思是Anthropic资助了这个项目，有真正的科学家参与，但他们所做的是让一群Claude代理运行大量实验来尝试弄清楚问题。

Claude lead meaning like Anthropic is funding this, there are real scientists involved, but what they're doing is having a bunch of Claude agents run a bunch of experiments to try to figure stuff out.

他说，我们怀疑这可能代表一种新的基因编辑机制，其精确功能、生物技术用途（如果有的话）或重要性水平尚不清楚。

He said that, he said, we suspect this could represent a new gene editing mechanism, its precise function, biotechnological utility if any, or level of significance is not yet clear.

所以，你知道，他们发现了新东西，他们说，看，我们发现了没人真正知道的新东西，我们不知道这是否重要，是否意味着什么。

So, you know, they discovered something new, they're like, look, we found something new that no one really knew about, we don't know if this is important, if it means anything.

这是在湿实验室上线大约21小时后发生的，所以并不是说，你知道，还有很多研究需要做，但至少，我的意思是，能加速这个过程真的很令人兴奋。

This came after I think twenty one hours of wet lab being online, so it's not like, you know, there's a lot of research that still needs to happen, but at a minimum, I mean, is really exciting to have sped up.

达里奥说，这项工作大部分是由克劳德完成的，虽然不是全部。

Dario said, The work was done mostly, though not entirely, by Claude.

我们的生命科学团队提出了一个大致的研究方向。

Our life sciences team suggested a broad area of research.

克劳德阅读了文献和大量基因组数据，发现了一些有趣的东西。

Claude read through the literature and a bunch of genome data and discovered something interesting.

然后克劳德提出了验证这一发现的实验方案，我们的团队执行了这些实验。

Then Claude proposed experiments to verify the discovery and our team carried them out.

好，所以在这里，我想很多人都在担心，顺便说一句，确实有一些公司在做类似把机器人和AI模型结合起来的事情，那些机器人正在混合试管里的化学物质，自己完成所有实验，有些人对此感到害怕。

Okay, so right there, I think a lot of people are concerned, and by the way, there are actual companies working on like basically getting robots attached to AI models, and the robots are, you know, mixing the chemicals in the vials and doing all the experiments themselves, and that some people are freaked out about that.

在这种情况下，有一个研究团队，他们说，嘿，这个领域可能有一些东西值得探索，也许那些研究人员已经在这个领域做过研究，知道那里有很多值得挖掘的东西。

In this case, you know, there's a research team and they're like, hey, there's probably some things to discover in this area, perhaps those researchers have already done research in those areas and know that there's a lot going on there.

所以他们给克劳德指了个方向，让它去研究这些内容。

So they point Claude at the right direction, go look at these things.

我的意思是，这就像你或我，深入一个我们非常熟悉的领域，然后让克劳德或ChatGPT帮我们做一些研究，以便我们弄清楚一些事情。

I mean, it's just like you or I, you know, going into a domain that we understand deeply and we're asking Claude or ChatGPT to help us with and do some research so we can figure stuff out.

其实概念是一样的，就是说，所谓“克劳德主导”，克劳德是大脑，但现在很多模型都在帮人做研究，反正就是这样。

It's the same concept, so like it's, it's quote unquote Claude led, Claude's the brain, but like, a lot of these models are doing the research for everything nowadays, but whatever.

我跑题了。

I digress.

不过我觉得这确实挺酷的。

I do think it's cool.

总之，克劳德去做研究，然后说，嘿，我觉得我可能发现了一些东西，你们去做这些实验，然后人类就去做了。

Anyways, Claude goes, does some research, says, hey, look, I think I might've found some things, go go do these experiments, and they have humans going and doing the experiments.

好的。

Okay.

达里奥说，很容易把这当成一次性的或好奇心驱使的成果，但我们反复看到一种模式，就是AI在新的知识领域里的表现，从很弱到超越人类，只需要几年的时间。

Dario said, it's easy to dismiss this as a one off or curiosity, but we've repeatedly seen a pattern where AI performance in our in new intellectual domains goes from weak to superhuman in a matter of a few years.

2023年的时候，模型在数学上还达不到普通高中生的水平。

In 2023, models struggled to do math at the level of an average high school student.

到了2024年，它们开始在全国顶尖高中生的数学竞赛中表现出色。

In 2024, they started to do well at math competitions for the best high schoolers in the country.

到了2025年，他们开始解决一些小的开放问题。

In 2025, they started to solve minor open problems.

2026年初，解决了更重要的开放问题，而到了2026年底，他们开始攻克数学领域最顶尖的几个开放问题。

In early twenty twenty six, more significant open problems, and in late twenty twenty six, they're beginning to solve the top few open problems in all of mathematics.

我们认为AI在生物学领域也正沿着类似的指数增长趋势发展。

We believe AI for biology is on a similar exponential trend.

他们这里具体指的是，OpenAI那边确实搞了个大动作，他们成立了一个数学顾问委员会，设在高等研究院，因为据说OpenAI已经解决了大约100个开放问题。

What they're specifically referring to here is that we actually have a, you know, a big kind of a big kind of thing that came out from OpenAI where they've actually put together like a math advisory board, they're calling it at IAS, after they had OpenAI apparently resolved about a 100 open problems.

所以，你知道，OpenAI解决了一个其他数学家一直在研究的大名鼎鼎的数学难题时，引起了很多争议，有人猜测他们获取了人们的聊天记录，但OpenAI否认了这一点。

So, you know, OpenAI, there's a lot of drama when they solve a big famous math problem that other mathematicians had been studying or working on, and people speculated that they were getting people's chat conversations and OpenAI denies it.

不管怎样，现在很多数学家中，有些人说，嘿，AI正在毁掉数学这个领域。

Regardless, a lot of mathematicians right now are, some of them are saying, Hey, you know, AI is ruining the field of math.

你知道，我们没法再靠脑力去研究东西了。

You know, we can't intellectually work on stuff.

我记得有九位数学家，或者更准确地说，这个月早些时候有25位菲尔兹奖得主签署了一封公开信，他们警告说，AI实验室正在威胁数学家的智力工作，因为AI可以做很多这样的工作。

I think there was nine mathematicians that were, you know, recently there's a whole, or I think 25 field medalist winners signed an open letter earlier this month, they were warning that AI labs are quote, threatening mathematicians' intellectual work, because AI could do a lot of the work.

所以，这里显然有些争议。

So, there's obviously drama here.

但为了回应这些，OpenAI 成立了一个数学顾问小组，他们说，嘿，各位，我们解决了很多数学问题，我们很乐意把这些结果交给真正的数学家去验证和检查。

But in response to all of this, OpenAI has formed kind of a math advisor group where they're like, hey look guys, we are getting a lot of math done, we're solving a lot of problems, we would love to put them back to actual mathematicians to go verify and check what's going on.

他们特别组建的这个小组，他们说这个小组不能要求OpenAI放慢速度，也不能限制他们的研究，它只是一个顾问小组，用来帮助监督这些问题的验证工作。

This group that they have in particular, they put together, they said like the group isn't allowed to tell OpenAI to slow down, or they're not allowed to do as much research, it's just an advisory group that's gonna help oversee the checking of a lot of these things.

所以显然OpenAI说，看，有一个著名的大问题，我们还有100个其他著名的数学难题已经解决了。

So apparently OpenAI says, look, you know, there's one big famous one, we have a 100 other math problems that are big famous math problems that we've solved.

而且其中很多问题都有百万美元的奖金，所以很多数学家都很兴奋，正在努力研究。

And a lot of these, like there's, you know, million dollar prizes, so a lot of mathematicians are excited and are working on this.

有些人也在抱怨，就像我提到的，有25个人签署了那封公开信，他们说，你们这是在破坏数学家做复杂工作的能力。

And some people are, you know, complaining, like I mentioned, 25 people signed that open letter, they're look, you're you're ruining mathematicians' ability to do like complex stuff.

但归根结底，这些AI工具只是计算器。

But at the end of the day, these AI tools are just calculators.

所以，在我看来，即使它们能解决一些问题，它们也只是非常复杂的计算器，这应该意味着数学家能够解决更多的问题。

And so, just because they figured out how to do something, they're a really complex calculator, in my opinion, that just means mathematicians should be able to solve more problems.

如果这些问题已经存在了一百年都没解决，天哪，直接解决掉然后继续生活不是很好吗？

And if these are problems that have been around for a hundred years unsolved, gee, wouldn't it be awesome to just solve them and move on with their life?

说实话，我不太在乎有没有一个所谓的数学领域，人们的工作就是解决这些问题，花一辈子去研究，试图赢得一百万美元的奖金。

Like, I don't really care if there's a quote unquote field of mathematics where it's people's job to solve these problems and figure and spend a lifetime figuring out, trying to win a million dollar prize.

如果我们找到了一个之前不知道答案的问题的答案，直接发表出来，搞定它，然后把它应用到有用的地方，让社会继续前进。

If we found the answer to a problem we didn't know the answer to, just publish it and get it over with, now implement it into whatever it's useful for, and let's move on with society.

这就是我的看法，我知道这可能会惹恼一些人什么的。

So, that's my opinion and I'm sure triggers people or whatever.

我觉得有趣的一点是，Dario显然非常清楚数学领域正在发生的这一切，他拿这个作为参考点说，看，生物学也会发生同样的事情。

One thing that I do think is interesting is Dario's obviously very aware of this whole, you know, situation that's happening in math, and he's using this as, you know, kind of a reference point saying, look, the same thing is gonna happen with biology.

Dario在他的推文中继续说，当然，生物学和数学的主要区别在于，数学可以纯粹在理论上进行，而生物学需要实验。

Dario goes on in his tweet to say, the main difference between biology and mathematics of course is that math can be done purely theoretically, while biology requires experimentation.

有些人因此得出结论，认为AI在生物学中的用处会有限。

Some have used this to draw the conclusion that AI's utility in biology will be limited.

我们认为这是错误的，正如我们今天所展示的，人类可以与AI合作进行实验，在几周内验证关键结果，如果有必要，还可以与AI一起迭代他们的发现。

We think this is wrong, as we've demonstrated today humans can collaborate with AI to perform the experiments, validate key results in a few weeks, and if necessary work with AI to iterate on what they find.

最终，甚至可能让Claude自己安全地进行实验，通过自主控制实验室设备，并采取适当的安全措施，但我们今天还不会这么做。

Eventually, it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment with appropriate safeguards in place, but we aren't doing that today.

我们的实验室也是BSL一级、BSL二级设施，不处理对人类有害的材料。

Our lab is also a BSL one, BSL two facility that doesn't handle materials dangerous to humans.

好的。

Okay.

我觉得他其实是想展示，一方面，这些可能性和机会正在涌现，但另一方面，也在试图克制。

I think he's really trying to show like A, the possibility and the opportunities that are coming out of this, but also, you know, try to hold back.

所以他说，看，这进展非常快，会很了不起，有很多酷炫的事情可以做。

So he's like, look, like this is advancing really quickly, it's gonna be amazing, there's lots of cool things we can do.

Claude最终可以完成所有这些，就像我之前说的，用自主控制的实验室设备。

Claude could eventually perform all of these, like I was talking with autonomously controlled lab equipment.

他又说，但别担心，我们今天不会做这些。

He's like, but don't worry, we're not doing any of that today.

因为显然，几周前他刚吓坏了所有人，说AI可能用生物武器毁灭世界，然后现在他又搞起了湿实验室。

Because obviously he just freaked everyone out a couple of weeks ago saying that AI, you know, could end the world with bioweapons, and then he's like, you know, got this wet lab going.

达里奥接着说，更广泛地看，生物医学的进步有很多阶段，从基础生物学发现，到转化研究，再到药物发现、临床试验，最后才是真正把药物和医疗服务送到病人手里。

Dario then goes on to say, more broadly, biomedical advancement has many stages, from fundamental biology discoveries to translation research, to drug discovery, clinical trials, and finally, the actual delivery of medicines and healthcare to patients.

我们对这些后期阶段也很感兴趣，但哪怕只是加速第一阶段的基础生物学发现，也有可能加快并拓宽整个流程。

We are also interested in these later stages, but even simply accelerating the first stage of fundamental biological discoveries has the potential to speed up and broaden the entire pipeline.

加深我们对生物学的理解，提升生物学家的工具，就能推动所有后续阶段的发展。

Improving our understanding of biology and sharpening biologists tools can drive forward all of the later stages.

比如，通过识别新的药物靶点，找到新的治疗方式，让测量更精确，加快实验或实验循环的速度，这本身又会进一步加速我们对生物学的理解。

For example, by identifying new drug targets, finding new therapeutic modalities, allowing for more precise measurements and speeding up the experimentation or experimental loop, which itself further accelerates our understanding of biology.

这本身不会缩短临床试验的时间，但如果成功了，可以大大增加进入流程的有希望的候选药物数量，即使延迟还在，吞吐量也能提高。

This will not in itself speed up clinical trial times, but if it succeeds, it could greatly increase the number of promising candidates that go into the pipeline and increase in throughput even though latency remains.

基本上他在说，你看，我们很多工作还处在发现阶段，人们会说，那得经过临床试验，那可能要花好几年。

Basically he's saying, look, like just because we're in the discovery phase of a lot of this stuff and people are like, well, you should've got go through clinical trials and that could take years.

他说，对，但如果我们今天就能有90%的把握知道一种药会有效，而不是要试90种不同的药，做所有那些实验，然后找出哪个真的有效。

He's like, yeah, but if we can get, you know, 90% certain that a drug is gonna work today instead of having to pitch 90 different drugs and do all the experiments experiments and figure out which one actually works.

我们就是在加速，因为我们在更快地测试更多高质量的药物。

We are speeding it up because we're getting more high quality drugs tested faster.

我的意思是，这挺好的，他基本上是在说，把药物送到人们手中，比如救命药，有一个巨大的流程，但即使我们只是加速了流程的开头，整个流程也会跟着加速。

I mean, it's a good, like he's basically saying there's a huge pipeline to getting drugs into people's hands, like, know, life saving drugs, but even if we're just speeding up the beginning of the pipeline, the whole pipeline does speed up.

他还说，最终他们会往下游走，尝试处理流程的其他环节，加速整个事情，对此大家有各种看法，但说到底，如果我们能真正发现新药、有真正的医学发现、有真正对人类有帮助的生物学突破，我们为什么不尽可能地去加速它呢？

And he says eventually they will go downstream and try to work on other elements of the pipeline and speed up the whole thing, which, yeah, there's all sorts of opinions on, but, know, at the end of the day, if we can get actual drug discoveries, actual medical discoveries, actual breakthroughs in biology that are helpful for humanity, why wouldn't we want to accelerate that to the absolute utmost?

我相信每个听众都有家人或认识的人，你知道，患有疾病、癌症，各种可怕的事情。

I'm sure everybody listening has a family member or knows someone that has, you know, suffered with a disease, a sickness, cancers, all sorts of horrible things.

如果我们能治好其中一些，我个人希望尽快做到，为了我的家人。

If we can fix and cure some of those, I would want that done as soon as possible, personally for my family.

好的。

Okay.

他说，在《爱的机器》中，这是Dario在他的推文中，你知道，很长的推文，但他说，在《爱的机器》中，那是我写的一篇博客文章，我写道，AI有潜力在五到十年内治愈大多数疾病。

He says, in Machines of, this is Dario again in his tweets, you know, very long tweet, but he said, in Machines of Loving Grace, which is a blog post he wrote, said, I wrote about AI AI's potential to cure most diseases in five to ten years.

顺便说一句，我要在这里停一下。

By the way, I'm gonna stop for a second here.

我就是喜欢Dario引用他自己的博客文章，就像他在引用一本大家都该知道的名著，比如《基督山伯爵》，就像在《爱的机器》里，总之，你得知道那是他的博客文章，他说，一个听起来不可能的目标，对吧，在五到十年内治愈疾病，但我相信，如果AI被应用到流程的每一个阶段，这个目标勉强有可能实现。

I just love how like Dario quoting his blog post, it's like, he's quoting like, you know, like a famous book everyone should know about, like in the Count of Monte Cristo, it's like, in Machines of Loving Grace, anyways, you gotta know that it's his blog post, he said, a goal that sounded impossible, right, curing disease in five to ten years, but one I believe is just barely possible if AI is applied to every stage of the pipeline.

第一步是证明AI首先能帮助推动生物学发现，然后能主导这些发现。

The first step is showing that AI can first help with and then drive biological discoveries.

所以，这就是他们正在做的事情，也就是这个流程的第一步。

So, this is what they're working on, this kind of first step in the pipeline.

他说，Claude的发现是一系列相关前期工作的最新成果，这些工作可以追溯到几十年前，从CRISPR这样的系统开始，再到Bridge、Recombinase和VIPR等发现。

He said, Claude's discovery is the latest in a line of related prior work that goes back decades, beginning with systems like CRISPR and continuing with discoveries like the Bridge, Recombinase, and VIPR.

然后他说，在过去几年里，人们对基于逆转录酶（RT酶）的系统兴趣大增，而Claude发现的系统所依赖的酶就是RT酶。

In the past few years, he then says, recently there has been heightened interest in systems based on reverse transcriptase, RT enzymes, and the enzyme underlying the system clot identified.

最近，斯坦福的一个团队独立描述了一种新型RT系统，带有一个相关的非编码阵列，在某些方面与Claude发现的系统相似，但它们是独立进化出来的不同系统。

And most recently, a Stanford team working independently described a novel RT system with an associated non coding array that is in some ways similar to the one Claude found, though they are distinct systems that evolved independently from each other.

我相信我们正处于发现这类系统并将其开发成强大生物技术工具的起步阶段。

I believe that we are at the very beginning of finding such systems and developing them into powerful tools for biotechnology.

这里我想说的是，斯坦福那边有一个团队发表了一篇论文，讲的是他们发现了一种不同的酶，一种类型的酶，不是同一种，你知道，它们是独立进化的，但你可以看到，基本上可以回溯Claude是怎么做到的，对吧？

So the one thing that I will say here, there was a group over at Stanford that essentially published a paper on how they found like a different enzyme, a type of enzyme, not the same one, know, they evolved differently, but you can see like, you can basically backtrack how Claude did this, right?

他们看了斯坦福研究团队的做法，在很窄的领域里做了研究，然后说，如果我们把那个发现应用到所有领域，还有什么没被发现呢？

They looked at the Stanford researchers team, looked at what they did in a very narrow area and said, okay, if we applied that discovery to everything, what's left to discover?

我不想说这是轻而易举的事，你知道，这毕竟不是已经运行了很久的东西，但大概就是这个概念，而且我不觉得这有什么不好。

And I don't wanna say low hanging fruit, I mean, you know, this isn't something that's been running for a long time, but that's kind of the concept here, and I don't think that's a bad thing.

这绝不是贬低Claude。

That's not derogatory to Claude at all.

我敢肯定有些人会说，你看，Claude就是在抄袭别人的成果。

I'm sure some people I bet you will say like, look, you know, Claude's just ripping off of other people's work.

我的意思是，当你发现能救人性命的药物时，我才不在乎谁抄袭谁呢。

I mean, when you're discovering drugs to save people's lives, I don't care who's ripping off of whose work.

咱们先把药发现出来，说实话，我也不在乎谁得到荣誉，但这是我个人的看法。

Let's just get the drugs discovered and I don't really care who gets the credit either, if I'm being honest, but that's just me.

我只是希望，你知道，人们的生命能得到拯救和改善。

Like, just want, you know, people's lives to be saved and improved.

但这确实很有趣，对吧？

But it is interesting, right?

就像人们在狭窄的领域做出了发现，有时是因为资金原因，或者是因为，你知道，他们就是人手不够。

Like people have made discoveries in narrow areas, and sometimes for funding reasons, or for, you know, just the, they just don't have enough manpower.

有很多不同的原因让他们在这些发现上保持狭窄。

There's a lot of different reasons why they stay narrow on those discoveries.

他们没有把发现应用到整个领域。

They don't apply it to the entire field.

所以如果在一个非常小的狭窄领域有发现，AI可以抓住它，并在很多不同的领域进行实验，你知道，拓宽发现的视野，让它更有价值。

And so if there's a discovery in a very small narrow area, AI can grab that and run experiments across a ton of different fields and you know, kind of brighten the horizons of what was found and you know, it's more worthwhile.

所以，你知道，还有更多的发现等着去发掘。

So, you know, there's more discoveries to be found.

所以我觉得我们可能会看到很多这样的情况，人们会批评说，嘿，看，AI实际上不能自己发现任何东西，等等等等。

So I think that we'll probably see a lot of this and people will criticize it and say like, hey, look, AI is not able to actually discover anything on its own, yada yada.

也许这其中有真实的成分，也许它会像Gary O说的数学那样，随着时间改进。

And maybe there's elements of truth to that, maybe it will improve over time like Gary O says math has done.

但归根结底，我真的不担心AI是否能真正拿别人的发现，应用到整个领域，找到我们没见过的新的领域，我不知道为什么我们不想让它尽可能多地发现。

At the end of the day though, I'm really not concerned if AI truly can go take other people's discoveries, apply them to the entire field and find new areas that we haven't seen, I don't know why we don't want it to discover as much as humanly possible.

嗯，不是人力所能及的，而是机器所能及的。

Well, not humanly possible, as much as machinely possible.

好，我们来说说达里奥的那条推文。他说，我为Anthropic在生命科学领域加速AI的公共效益所投入的资源感到自豪，我们的目标不仅是扩大生命科学团队，还要与其他科学家合作，把这种方法应用到更广泛的问题上。

Okay, finishing up the tweet from Dario, he said, I'm proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life science, and we're aiming both to grow our life science team and to work with other scientists to extend this approach to a broad range of problems.

如果你有研究合作的提案，或者有兴趣加入我们的生命科学团队，请联系我们。

If you have a proposal for a research collaboration or are interested in joining our life sciences team, please reach out.

我觉得这可能是他打广告的最佳时机，就像在说，嘿，各位，如果你想加入团队，或者你有一些即将取得的好发现，我们很乐意帮你。

I mean, this is probably the best place for him to put a plug like, hey guys, look, like if you wanna join the team, if you have any other good discoveries that you're on the verge of making, we would love to help you with it.

我觉得这就像一把双刃剑，不过说实话，我一点都不在乎。

I think it's two sides of the sword, which I again, do not give a crap about to smallest degree.

一方面，如果你像斯坦福团队那样在做一些很酷的事情，而且你快要取得发现了，或者你觉得自己发现了什么，然后你去找达里奥说，嘿，你能资助这个吗？

One is, if you're working on something cool like the Stanford team was and you're close to a discovery, you have a discovery, you think you figured something out and you go to Dario like, Hey, can you fund this?

因为他们不会随便给任何人提供资金。

Because they're not just gonna, you know, fund budget to anyone.

他们不仅提供资金，还给你实验室，帮你推进项目，让你加入一个很棒的团队，给你一些好工具，还给你很多Anthropic的资源。

Like they're funding the budget, they're giving you the lab, they're working on it, they're putting you in around an amazing team, they're giving you some great tools, they're giving you resources to a lot of anthropics kind of tools.

从某种程度上说，这就像你带着一个发现去找一家大型制药公司，然后他们资助你，你就是在找一个资助者。

In a way, I would just say like, it's the same thing as if you went to a big pharmaceutical company with a, you know, perhaps a discovery and they were funding it, like you're finding a funder.

现在我觉得有些人说的缺点是，比如，你辛辛苦苦把东西研究出来了，然后Anthropic肯定会像这样发个大新闻，把功劳都揽到自己身上。

Now the downside that I think some people say is like, well you did all the work to figure that out and now Anthropics most definitely gonna do a big press release like this and take their credit.

对此我想说，谁在乎呢？

To which I say, who cares?

如果这是能帮助人类的事情，如果Dario想揽这个功劳，我是说，如果能把你的名字写上去，说这是你发起的，那当然好，但咱们还是先把这些发现公之于众，造福人类吧。而且如果他出钱，你出力一起做，你应该感到兴奋才对，不过也无所谓了。

If it's something that's gonna help humanity, if Dario wants to take credit for it, I mean, it'd be great if you could get your name on there for getting it started, but like, let's just get the discoveries out there for humanity, and if he's funding it and you're helping and working on it, like you should be excited about it, but whatever.

说到底，各位，这真的很令人兴奋，有很多不可思议的东西正在被发现，很多药物。

At the end of the day guys, this is really exciting, there's lot of incredible things being discovered, a lot of drugs.

我知道我一直挺悲观的，而且我觉得网上很多人，如果你看那些评论，都是说，哦不，OpenAI、Anthropic要放出下一个新冠了，我们现在可以用Anthropic以十倍速度制造新冠之类的。

I know I've been pessimistic, and I think a lot of the internet, if you read the comments on there, it was like, oh no, you know, OpenAI's, Anthropic is gonna release the next COVID, we can now create COVID 10 times faster with Anthropic and stuff.

我觉得说实话，如果我们认真看待的话，Anthropic团队很优秀，他们的生命科学团队都是好人，他们想做出好的发现。

I think honestly, like if we're being serious, Anthropic is full of a good team, their life sciences team are good people, they're trying to make good discoveries.

我不觉得有什么见不得人的勾当。

I don't think there's anything nefarious going on.

而且说到底，他们也是活生生的人。

And at the end of the day, really these are real humans.

我的意思是，目前来看，也许未来会有所改进和进步，但现在这仍然只是真实的人类团队在引导AI模型和Anthropic的资源，朝着他们认为能找到东西的方向去探索。

I mean, at this point, maybe it'll improve and progress in the future, but right now this is still just real human teams pointing the AI model and Anthropics resource in a direction they believe that they can find something.

一旦它发现了什么，真正的人类团队就会去做实验。

Once it finds something, it's a real human team going and doing the experiment.

所以总的来说，我觉得有很多人在参与其中，有很多监督，确保这一切都做得正确且安全。

So overall, I mean, I think there's a ton of human in the loop, a ton of oversight, a ton of making sure that this is done correctly and safe.

Anthropic知道，如果出了任何差错，后果会有多严重。

Anthropic knows how bad, you know, it would be if anything were to go wrong.

所以我觉得我们会看到最高级别的安全保障，也会看到一些非常令人兴奋的发现从这一切中涌现出来。

So I think we're gonna see some, you're gonna see top level safety, and I think we're gonna see some really exciting discoveries coming out of all of this.

各位，非常感谢你们收听本期播客。

Guys, thanks so much for tuning into the podcast.

如果你还没有给这个节目留过评论，那真的能帮到我很多，希望你能给我一个评论。

If you haven't ever left a review on this show, it would help me out so much to, get a review from you.

所以如果你喜欢这个节目，或者学到了什么新东西，我非常希望能收到你的评论。

So if you appreciate it and if you learned anything new, I would love to get a review on the show from you.

留个五星好评，再留个评论吧。

Drop a five star review and drop a comment.

每条我都会看，真的很感谢大家。

I read all of them, so I really do appreciate them.

如果有什么其他话题、公司或者内容想让我在节目里聊的，尽管告诉我，我们下期再见。

Let me know if, there's any other topics or companies or things that I should be covering on the show, and I'll catch you all in the next episode.