OpenAI Astra 的特点

欢迎收听今天的播客。

Welcome to the podcast today.

今天几乎所有主要的人工智能实验室都有大新闻，还有一些关于人工智能在科学领域的有趣故事。

We have some big news from almost all of the major AI labs today, as well as some interesting stories when it comes to AI in science.

首先我想讲的是，Anthropic 发布了 Fable 和 Mythos 5.1。

The first thing I wanna cover is that Anthropic is shipping Fable and Mythos 5.1.

它们的 token 更便宜，而且护栏也更宽松一些。

They have cheaper tokens and they have some looser guardrails.

OpenAI 推出了 Astra 模型，它在漏洞利用基准测试中几乎拿到了满分，所以它发现了一些零日漏洞或攻击，说到他们最新的模型和黑客攻击，这真是让人大开眼界。

OpenAI is putting out the Astra model and it hit a near perfect exploit bench score, so it's finding a bunch of two, it found a couple zero day hacks or exploits, so when it comes to their latest model and hacking, this is kind of a this is a wild ride.

另外，好莱坞现在正在推销，或者我应该说，谷歌正在以每个角色 4000 万美元的价格向好莱坞制片厂推销 AI 授权协议。

In addition, Hollywood right now is pitching or Google, I should say, is pitching Hollywood studios at $40,000,000 per character for AI licensing deals.

你可以想象一下像米老鼠，或者你知道的，也许其他一些大牌，比如海绵宝宝。

So you can imagine something like Mickey Mouse or, you know, maybe some of the other big names, SpongeBob SquarePants.

对吧？

Right?

他们愿意花四千万美元来获得其中一个角色的授权。

They're putting $40,000,000 to get one of those characters licensed.

有意思的是，OpenAI 之前也做过类似的授权方案。

It's interesting because OpenAI was doing a similar licensing scheme a while back.

后来他们准备推出 Sora 的时候，就把这个方案搁置了，把 AI 模型的重点更多放在逻辑上。

They actually dropped it when they were gonna drop Sora and kinda focus their AI model more on the logic.

五角大楼推出了 ChatGPT Mill 和 Grok 的政府版，放在他们的 genai.mil 网站上。

The Pentagon has launched ChatGPT Mill and Grok for government and is on their genai.mil site.

我们稍后会聊一下这个。

We'll talk a little bit about that.

美国农业部打算用卫星和 AI 来估算农作物产量。

And the USDA is gonna test satellites and AI for crop estimates.

他们希望这将来能帮到农民。

This is something that they're hoping will help farmers in the future.

在进入正题之前，我想说，如果你想在一个地方访问全球 80 或 90 个顶尖 AI 模型，我建议你去看看 aibox.ai，那是我自己的创业项目。

Before we get into all of that, I wanted to say if you wanna get access to the 80 or 90 top AI models in the world in one place, I'd love for you to go check out aibox.ai, which is my own startup.

我们打造了一个在过去一个月里非常受欢迎的东西，就是 AI Box MCP，它基本上可以让你把世界上所有的 AI 模型，包括音频、视频、图像，都集成到你常用的任何 AI 工具里。

We have created something that's been incredibly popular over the last month, which is the AI Box MCP, and it basically lets you put every AI model in the world, audio video image, into any of the popular AI tools you use.

所以如果你在用 Cloud Code，或者 ChatGPT Codecs，甚至 ChatGPT Work、Claude Cowork，你都可以让 Claude 在它内部生成图像、音频、视频。

So if you're using Cloud Code, if you're using ChatGPT Codecs or really ChatGPT Work, Chat Claude Cowork, you can get Claude to generate images, audio, video, all inside of Claude.

不管是你最常用的工作区还是 Cursor，你都可以让所有不同的 AI 模型在那个工作区里生成内容。

Whatever workspace or cursor, whatever workspace you you most frequently use, you can get all of the different AI models generating inside of that workspace.

所以如果你想看看，就去 aibox.ai/mcp 吧，那里可以获取那个连接器，让你在里面生成所有这些不同的素材。

So if you wanna go check it out, it is aibox.ai/mcp if you wanna get access to that connector that lets you generate all of those different assets inside.

今天我想先聊聊新的 Fable 和 Mythos 五部曲，他们在这方面发布了什么。

The first thing I wanna talk about today is the new Fable and Mythos five story, what they've kinda put out in regards to this.

他们发了一篇博客文章，我挺喜欢他们的博客文章的。

There's a blog post they dropped that I I'm liking their blog posts.

他们现在做得很有美感。

They're very aesthetic these days.

你可以进去把博客横幅图片的配色方案换掉，这完全没必要，但真的很酷。

You can, like, go in and switch the color palette on the the banner image on their blog, which is completely unnecessary, but it's very cool.

他们的博客标题是《Claude Fable 5.1 和 Mythos 5.1》。

Their blog is titled, Claude Fable 5.1 and Mythos 5.1.

这挺有意思的。

It's kind of interesting.

他们在这里把命名方式统一了，都用 Fable 5.1 和 Mythos 5.1。

They're syncing up the naming conventions here with Fable 5.1 and Mythos 5.1.

我觉得过去几年 OpenAI 在命名上真是放飞自我，各种编号体系乱七八糟，现在感觉稍微稳定了一点，但命名方式一直挺混乱的。

I feel like that's something that OpenAI really just let just went wild with in the last couple years, all of their different numbering schemes, and it feels like they're getting a little bit more stable now, but the naming conventions have been something that's been pretty wild.

这次让我特别兴奋的一点是，如果你去看他们的一些科学研究，里面有一大堆很酷的演示，展示了他们如何帮助科学研究。

One thing in particular that I'm excited about with this, if you go into some of the scientific research they have, a whole bunch of very cool demos of of ways that they're helping with scientific research.

他们在展示他们做的各种事情，比如他们一直在帮研究人员观察金星上的一个小盾状火山。

They're showing off different things that they're doing, you know, they have like a small shield volcano in on Venus that they've been helping researchers look at.

他们还有很多关于心脏和动脉的研究，显然这些能力在变得越来越强。

They have a lot of work around hearts and arteries, and it's just like the capabilities of this obviously are getting a lot better.

他们会重点展示它如何帮助科学，而不是如何帮助黑客。

They're gonna go and focus on how it can help science instead of how it can help hacking.

Fable 5.1 今天已经在所有云平台和 Anthropic API 上发布了。

It is Fable 5.1 is available today on the any cloud platforms and on the Anthropic API.

我很欣赏它同时发布在 API 上，因为这样你就能在所有不同的平台上使用它了。

I really appreciate it when it comes out on the API at the same time, because then, you know, you're gonna get it on all of the different platforms.

Mythos 5.1 仍然仅限于网络安全和生命科学合作伙伴使用。

Mythos 5.1 is still restricted to cybersecurity and life sciences partners.

所以某些公司会获得访问权限，但因为它太强大了，能用于黑客攻击等，所以还没有向公众开放。

So certain companies are gonna get access to that, but because it's, so capable for hacking and all that other stuff, it's still not available to the general public.

企业级前沿防护，也就是 Anthropic 为 Fable 提供的零日保留服务，将向用户推出，而 Mythos 5.1 在 Terminal Bench 4.0 和 Humanities Last Exam 上创下了多项纪录，它还生成了我之前提到的定制 GPU 优化和高分辨率金星地图预发布版。

Enterprise Frontier Safeguards, which is Anthropic's zero day retention tier for Fable, is gonna roll out to users in this rolling out earlier or rolled out, and Mythos 5.1 is setting a bunch of different records on Terminal Bench four point o in Humanities Last Exam, and it also produced a custom GPU optimization and high resolution Venus map prerelease that I was kind of talking about earlier.

目前，系统卡将 Mythos 5.1 评为自动化 AI 研发加速风险极低。

Right now, the system card rates, Mythos 5.1, it's very low risk for automated AI R and D acceleration.

这符合当前的一些趋势。

This is kind of in line with some current trends.

总的来说，我非常兴奋。

Overall, I'm really excited.

在他们的公告中，他们表示Anthropic从未在未经明确许可的情况下使用企业数据进行训练，而且永远不会这样做。

On their announcement, they said Anthropic has never trained on enterprise data without explicit permission, and we never will.

这是他们一直在强调的一点。

This is something that they've been kind of really putting out there.

你知道，很多人担心这会获取他们公司的数据之类的。

They're you know, a lot of people are concerned that this is gonna be getting their their company's data and whatnot.

但总的来说，真的很兴奋。

But overall, really excited.

这是一个大模型，而且有很多重大更新。

This is a big model and there's a lot of big updates.

你可以去看看他们的发布和博客文章，了解所有不同的内容。

You can go check out their release and their blog post to see kind of all of the different things.

他们有一个关于安全、安保和一致性的部分，而且他们也谈了很多新的性能基准之类的，但相比基准，我对他们想出的一些酷炫用例以及人们用它做的事情更感兴趣。

They have a section on safety, security, and alignment, and they're also talking a lot about some of the new performance, you know, benchmarks and stuff, which I'm I'm less excited about the benchmarks compared to, just some of the some of the cool use cases they've come up with and some of the things that people are doing with it.

OpenAI正准备发布他们的Astra模型，为了应对这个，基本上这就是他们对Anthropic推出新Mythos 5.1或5的回应，那可是上了新闻的大事。

OpenAI is getting ready to release their Astra model, and in in anticipation of this, I mean, is basically their response to what, Anthropic has come out with their new with their Mythos 5.1 or, you know, five, which was kind of the big thing that got in the news.

但这确实是个大事，他们就是说，看，这个模型在安全研究方面同样厉害，我是说，技术上就是黑客攻击，但他们不喜欢这么说。

But this is really something that that, you know, they're saying, look, this is just as capable at security research and, I mean, technically hacking, but they don't like to put it that way.

它在漏洞利用基准测试上基本拿了满分，这个基准是衡量这些AI模型如何利用已知系统漏洞进行攻击的。

It had basically a perfect score on exploit bench, which is benchmark measuring how well these AI models are hacking known system vulnerabilities.

我想说的是，比这更重要的是，你可以说，好，我们知道有这个漏洞。

And what I'll say is more important than this is like, you could say, okay, we know there's this vulnerability.

我们把它给这个模型。

We're gonna give it.

我们看看它能不能黑进去。

We're gonna see if it can hack it.

然后，好，它拿了满分。

And, okay, it got a perfect score.

但有很多安全漏洞是零日漏洞，也就是说，没人知道它们的存在。

But there's a lot of security vulnerabilities that are day zero, meaning, you know, no one knows about them.

而这个模型实际上也涉及到了这些漏洞。

And this model has actually gone into that as well.

在这个测试的修改版本中，由OpenAI工程师构建，Astra实际上在没有人类指导的情况下发现并利用了零日漏洞，也就是说，他们给了它代码库或工具，让它去攻击，结果它找到了以前没有人能找到的东西。

In a modified version of this test, was built by OpenAI engineers, Astra actually discovered and exploited two zero day vulnerabilities without any human guidance, meaning they said, you know, here's the code base or here's the tool, go and hack it, and it found stuff that no human had been able to find before.

所以这不像，哦，我们知道这些漏洞，然后它就能搞定。

So it's not like, oh, we we know these vulnerabilities and it gets it.

它是去发现全新的东西。

It's going and finding brand new stuff.

OpenAI表示，Astra是第一个跨越他们所谓的“关键网络安全门槛”的LLM。

OpenAI says that Astra is the first LLM to cross their quote critical cybersecurity threshold.

所以这基本上是他们最高的内部风险等级，他们说，看，这个模型足够好，已经排在最前面了，这是非常高风险的事情。

So this is basically their highest internal risk tier, and they're like, look, this model is good enough that it is at the absolute front, front of the line for, like, this is very high risk.

因此，他们给它加了很多防护措施。

And so because of that, they're putting a bunch of guardrails on it.

他们在训练的时候，如果它能成功攻击某个系统，他们就会回去修复并加上防护。

They're, you know, when they're trained, if it's able to hack something, they go back and and, like, try to go fix it and add guardrails.

所以Astra最先进的网络安全能力在发布时将会受到限制。

So access to Astra's most advanced cybersecurity capabilities are gonna be restricted at launch.

这基本上就是Anthropic用他们的Mythos在做的事情，也是他们今年早些时候做的类似工作。

This is basically what Anthropic is doing with their Mythos and kinda what they did earlier this year.

OpenAI正在部署Astra，并增加了思维链监控和账户级风险评分，以便捕捉滥用行为。

OpenAI is deploying Astra with additional chain of thought monitoring and account level risk scoring to catch misuse.

这让我很感兴趣。

That's interesting to me.

对吧？

Right?

账户级风险评分。

Account level risk scoring.

就是说，你的公司或你的账户风险有多高？

It's like how high risk is your company or is your account?

然后根据这个，就决定了我们会怎么处理这个模型，或者给你什么样的访问权限。

And depending on that is like, that that kind of, like, determines what we're gonna do with the model or what we'll give you access to.

所以，是的，我的意思是，这确实挺疯狂的。

And so, yeah, I mean, this is, this is pretty wild.

如果你去看他们关于这一切的博客文章，你会看到他们在漏洞利用测试上的表现好了多少。顺便说一句，如果你用苹果设备，现在也可以在苹果播客和Spotify上看视频了。

If you go over and look at their blog post on all of this, you can see just how much better they did on the exploit bench, which by the way, if you're on Apple, you can now watch this is also videos now published on Apple and Spotify.

所以如果你对我说的那些屏幕共享内容感兴趣，你可以看到相关视频。

So you can see the video of this if you're interested in the screen shares of all the stuff I'm talking about.

但是，如果你去看的话，如果他们给模型8万个输出令牌，Astra模型在漏洞利用基准测试上的完成率大约是39%或40%，而之前最好的模型，比如GPT 5.6 Sol，表现就没那么好，完成率接近11.5%。

But yeah, if you go and look at it, if you're if they give it, you know, 80,000 output tokens, the Astra model got like a 39 or 40% completion rate on exploit benchmark, whereas when you're using something like the GPT 5.6 Sol, which was their previous best model, it was doing something not quite as not quite as good, which was closer to they had it closer to 11.5% of completion.

所以我们真的在提升。

So we're really bumping up.

我的意思是，这基本上是在漏洞利用基准测试上的内部端口上提升了四倍。

I mean, that's basically four x on on this internal port on exploit benchmark.

所以图表在向右上方上升，这些模型变得好多了。

So the graph is going up and to the right and these models are getting a lot better.

谷歌正在向好莱坞制片厂推销，在某些AI授权交易中，每个角色要价4000万美元，但这不会适用于每一个角色。

Google is pitching Hollywood Studios $40,000,000 per character when it comes to certain AI licensing deals, and this isn't gonna be like every single one.

不是每个角色都会得到4000万美元的报酬，但我觉得对于最受欢迎的那些角色，他们会这么做。

It's not like every character is gonna is be paid $40,000,000, but I think for the most popular ones, this is something that they're gonna do.

迪士尼、华纳兄弟、探索频道和环球影业，他们都被提供了数亿美元，用来在自家的片库上训练Gemini，而且还有按角色定价的标准。

Disney, Warner Brothers, Discovery, and Universal all have hundreds of millions of dollars that they're being offered to train Gemini on their libraries, and they have the per character price tag.

我觉得这特别有意思，因为这感觉很像你平时去买数据或者任何其他数据集一样，对吧？

It's so interesting to me because it feels a lot like how you would, I mean, go and buy data any other dataset, right?

就像你会根据数据的价值、稀有程度，还有数据集里的数量来决定买哪些数据。

Like you you would go buy datasets based off of the value, based off of how rare it is, based off of the quantity inside of the data set.

所以现在，我们好像是在给迪士尼之类的角色赋予内在价值。

And so now that's like we're we're putting these intrinsic values on the characters of of, you know, Disney or something like that.

当然，这肯定是针对受版权保护的角色。

And this is gonna be for, of course, copyrighted characters.

多个角色的总交易价值可能达到数十亿美元，而且制片公司还有可能从YouTube的AI广告收入中分一杯羹。

The total deal value across multiple characters could run into the billions with studios potentially getting a cut of YouTube AI ad revenue.

对吧？

Right?

所以如果他们就像，嘿。

So if they are like, hey.

你看。

Look.

我猜他们会把它作为YouTube AI推出，让YouTube能够生成AI视频。

We're gonna I'm assuming they're gonna roll this out as YouTube AI where YouTube's able to generate AI videos.

如果很多人用AI生成视频，里面出现了米老鼠、唐老鸭、彼得潘或者其他受商标保护的角色，而且这些视频带来了广告收入，那么制片公司就能分到一部分收入。

And if it's making ad revenue from a lot of people generating these AI, you know, generated videos and it's got Mickey Mouse on there or Donald Duck or it's, you know, they've got Peter Pan or whatever their trademarked characters are, the studios are gonna get a percentage of the revenue.

谷歌的DeepMind今年夏天早些时候已经和A24达成了7500万美元的投资协议，而迪士尼、华纳兄弟、探索频道和环球，顺便说一句，他们都没有表态。

Google's DeepMind already made a $75,000,000 investment deal with A24, which was earlier this summer, and Disney, Warner Brothers, Discovery, and Universal, none of them are, by the way, saying anything.

截至本月初，还没有人签署任何协议，所以这些都还在谈判中。

Nobody has signed anything as of, the beginning of this month, so this is all in talks.

这些报价都还在谈判中。

These offers are in talks.

我很好奇他们会不会回来要求更多，或者接下来会发生什么。

I'd be curious to see if they come back wanting more or if, you know, what kind of happens there.

狮门影业2024年与Runway的授权协议至今还没有推出任何产品。

Lions Gate's 2024 licensing deal with Runway has yet to produce any sort of released product.

这是几年前的事了，他们当时签了这个授权协议，但到现在也没看到什么成果。不过我们知道这里面有很多钱。

This was, you know, a couple years ago, they had this licensing deal and we haven't seen much come out of it, but we know that, you know, there's there's a lot of money in this.

在《洛杉矶时报》上有一篇文章讲了这个故事，汤姆·努南，他以前是制片厂和电视网的高管，他在谈论这件事时说，谷歌正在召集好莱坞最聪明的一些人来试图搞清楚一件不可避免的事情，那就是好莱坞和AI将会相互依赖，以及我们如何能最好地理解这一点，而不只是让人们感到恐慌。

Over on the LA Times, there was an article kinda talking about the story, Tom Noonan, who's a former studio and network executive, talking about this story said, Google's enlisting some of the smartest people in Hollywood to try to unpack what's inevitable, which is Hollywood and AI are gonna become reliant on each other and how best we can understand that without just terrifying people.

所以他认为这件事是必然会发生的，我们只能等着看它会怎么发展，但基本上不可避免，这就是他对这件事的看法。

So he views this as something that is, you know, it's it's gonna happen and we're gonna just have to wait and see, you know, how this rolls out, but it's kind of inevitable is what his take is on all of this.

谷歌正在推出ChatGPT Mill，也就是Military的缩写，还有Grok，是给政府用的，在genai.mil上。如果你不知道的话，这个网站是专门给军方用的。

Google is launching ChatGPT Mill or short for military and Grok for government on genai.mil, which if you don't know, this is kind of a special site only for the military.

如果你好奇，去访问genai.mil，它会提示你，你访问这个页面是因为你没有被授权从DOW网络之外访问genai.mil。

If you go there, you know, if you're curious and you go to genai.mil, it's gonna hit you with a you have reached this page because you're not authorized to visit genai.mil from outside of the DOW networks.

如果你认为这是个错误，请联系你们公司的IT，也就是你们指挥部的IT服务台。

If you believe this is an error, please contact your company's IT you know, your command's IT service desk.

基本上，是的，如果你是平民，你是进不去的。

And basically, yeah, you can you're not gonna be able to get into this if you're a civilian.

这个只对军方开放。

This is only for the military.

这是国防部正在推行的事情。

This And is something that the Department of Defense is rolling out.

他们推出了IGFD Mill和Grok for government，大约有300万军人和文职人员可以使用。

They launched IGFD Mill in Grok for government, which is accessible to about 3,000,000 military and civilian personnel.

当然，也包括文职人员，对吧？

Of course, even civilians, right?

这些人必须在部门内部工作。

These are people that have to be inside of the department.

我们的genai.mil已经注册了170万用户，这超过了国防部300万员工总数的一半。

Our genai.mil has already onboarded 1,700,000 users, which is more than half of the Department of Defense's 3,000,000 person workforce.

Anthropic的Claude不在那个门户里。

Anthropic's Claude is not in that portal.

我们知道特朗普和整个政府把他们标记为供应链风险，并且指定了Anthropic，然后Anthropic就打了场官司。

We know that Trump and the whole administration kind of labeled them as supply chain risk, and they designated Anthropic as such, and then Anthropic kind of had the whole court battle.

我觉得后来被推翻了，但我不确定这个案子现在进展如何，不过目前它确实不在那个名单上。

I think it got overturned at some point, and I'm not sure kind of where we are back and forth on this case, but it's, you know, missing right now from that list.

政府版Grok是通过SpaceX的Star Shield AI来交付的，这算是一个安全的卫星网络，建立在Starlink的基础设施之上。

Grok for government is shipping under SpaceX's Star Shield AI, which is kind of a secure satellite network, which is built on Starlink infrastructure.

这挺有意思的。

This is kind of interesting.

这是埃隆进入政府供应领域的方式。

This is Elon's way of getting into supplying the government.

说实话，我觉得这是个非常聪明的举措，不仅把Grok的AI模型，还把Starlink的基础设施都结合起来了。

And honestly, mean, I think it's a it's a very smart move kind of pairing not just the AI model of Grok, but also the the Starlink infrastructure.

对吧？

Right?

因为如果你在某个偏远地区，没有手机信号，或者很难接入互联网，用ChatGPT可能会有点困难，但理论上你可以通过Starlink用上Grok。

Because if you're in some remote area that doesn't have cell reception or that's really hard to get or it doesn't have internet service, accessing ChatGPT might be a bit difficult, but you could get Grok, you know, theoretically through Starlink.

而且我相信军方肯定有自己的卫星连接设备，所以对他们来说这可能不像普通老百姓那么重要，但我觉得他们既有很好的基础设施层，又有AI模型，这让他们成为一个有趣的合作伙伴。

And I'm sure the military's got all of this their own satellite, you know, connectivity stuff, so it's probably not as big of a deal as it would be for your average civilian, but I think that they have a really good infrastructure layer as well as the AI model that makes them an interesting partner.

五角大楼还和亚马逊网络服务、微软、英伟达以及Reflection AI签了协议，所以有很多不同的AI交易在进行，还有很多其他AI公司也在参与。

The Pentagon has also signed deals with Amazon Web Services, Microsoft Nvidia, and Reflection AI, So there's a bunch of different AI deals going down, many other AI companies as well.

美国农业部打算用卫星和人工智能来帮助估算农作物产量。

The USDA is gonna test satellites and AI to help with crop estimates.

之前农民们有不少抱怨，所以现在这个机构开始采用新的遥感技术和机器学习，目的是提高产量预测的准确性。

There was a bunch of backlash from farmers, so now the agency is piling this kind of new remote sensing machine learning and their idea is to get better at their yield forecasting.

农民们说，市场对他们很不利，因为他们觉得产量预测不准确，导致商品价格或食品价格波动，这对农民很不利。

Farmers say that the markets have moved against them a lot because they say that the the yield forecasting is not accurate and so it makes it, you know, the price of commodities or the price of food changes and it's bad for the farmers, they're saying.

所以政府就说，好吧，也许我们可以介入，让预测更准确一些。

So the government's like, okay, well maybe we can come in and make this more accurate.

这个试点项目是在农民们批评美国农业部的产量预测导致谷物市场波动、损害生产者利益之后推出的。

The pilot follows a bunch of criticism from farmers who say that the USDA yield forecasts have moved grain markets in ways that hurt producers.

这个机构每月发布一份报告，叫做WASD，基本上大部分影响市场的数据都通过这个报告发布，他们为全球农业发布这些数据，但主要影响玉米、大豆和小麦的价格。

The agency's monthly, it's called the WASD and it's a report, and it basically shows that most market moving data is kind of going through this, and they release this for global agriculture, but it basically drives corn, soybean, and wheat pricing.

所以，当他们说产量会是多少，比如会有大量过剩，价格可能突然下跌，或者他们说，嘿，有严重短缺，那么玉米价格可能上涨。

So, you know, when they say this is what the yield is gonna be, there's gonna be this huge oversupply, all of a sudden the price might drop, or if they're like, hey, there's a huge shortage, then the price for corn might go up.

所以无论如何，如果有任何政府机构影响玉米或大豆等商品的价格，显然，种植这些作物的农民如果觉得这对他们不利，就会抱怨。

So anyways, if there's any sort of government agency that's affecting the price of something like corn or soybeans, obviously, the farmers that are growing that, if they feel like it's negatively impacting them, are gonna complain about it.

所以归根结底，我能做的最好的事情就是让它尽可能准确。

So at the end of the day, the best thing I can do is make it as accurate as possible.

用卫星和人工智能可能是做到这一点的最好方式。

Using satellites and using AI is probably the best way to do this.

这也在模仿私营部门的做法。

This is also mirroring what the private sector does.

他们使用卫星加人工智能的产量建模。

They use satellite plus AI yield modeling.

所以有点好笑的是，美国农业部有自己的预测。

So it's kind of funny that USDA has their own forecast.

很多对冲基金有自己的内部卫星或无人机机队，去测量所有这些数据。

A lot of hedge funds have their own, you know, in house satellite or drone fleets that are going and measuring all this stuff.

他们把数据交给人工智能来正确建模，寻找套利机会，比如政府高估或低估了供应量，他们就会在这些不同的赌注上做多或做空。

They're bringing it back to AI to model it correctly, and they're looking for arbitrage opportunities where maybe the government overestimated the supplier, underestimated the supply, and they'll take the over under on those different bets.

他们会利用这些套利机会赚钱，基本上就是利用政府预测错误的地方，以及市场提前预测的情况。

They'll make money off of the arbitrage opportunities of, you know, basically what the government got wrong on that and what the what the markets were forecasting ahead of time.

总之，对冲基金已经这么做了很长时间了，我很高兴政府和美国农业部也要开始采用这种最佳实践，用人工智能让数据尽可能准确，我希望这对农民来说是个巨大的胜利。

So, anyways, hedge funds have been doing this for a long time, and I'm excited that the government and the USDA is gonna be getting in on basically the best practice using AI to make this as accurate as possible, and I hope that this is a great win for farmers.

如果你对这个故事感兴趣，47 ABC WMDT 做了一个很棒的报道，一个很棒的采访，谈到了这对农民的影响以及正在发生的事情，还有一个很棒的名言，来自博士。

If you're interested in this story, forty seven ABC WMDT did a great story on this, a great interview talking about how this is impacting farmers and what's going on, and a great quote from that, Doctor.

阿尔法埃尔·阿尔卡拉德，他是 UME 精准农业的副教授，他说：“最终，这些进步有可能支持开发更有韧性和生产力的作物，并为农民提供更好的决策支持工具。”

Alfael Alcalad, who is UME's associate professor in precision agriculture said, Ultimately, these advancements have the potential to support the development of more resilient and productive crops and provide better decision support tools for farmers.

所以总的来说，这对农民和所有涉足这个行业的人来说都是好事。

So overall, this should be great for farmers and everyone involved in the industry.

非常感谢你收听这个播客。

Thank you so much for tuning to the podcast.

如果你喜欢这一集，朋友们，如果你能在你收听播客的地方留下评分和评论，那对节目帮助太大了。

If you enjoyed this episode, guys, it would help the show out so much if you could leave a rating and review wherever you get your podcast.

我会阅读所有的评论，它们能帮助节目在算法中被发现。

I read all of the reviews, and they help the show out to be found in the algorithm.

我正在努力达到200条评论。

I'm pushing to get to 200 reviews.

我在苹果上已经有165条评论了，所以如果你还没留过评论，那真的会帮大忙。

I have a 165 on Apple, so if you haven't dropped a review already, it would help to show a ton.

我会超级超级感激的。

I would be super, super grateful.

我知道只需要一秒钟，但对节目帮助很大。

I know it only takes a second, but it helps to show out a ton.

我最近也开始给播客加视频了，这期就是视频。

I've recently also been adding video to the podcast, which this is a video.

所以如果你有兴趣看视频，留个评论告诉我你感兴趣。

And so if you're interested in seeing that, leave a review and let me know that you're interested in the video.

我们会继续为你带来这些内容。

We'll get these we'll keep these coming for you.

好的。

Alright.

非常感谢你的收听，我们下期再见。

Thanks so much for tuning in, and we'll see you in the next episode.