← 返回任务列表

The Most Important New AI Tools from OpenAI DevDay

240 段 · 1 位说话人 · 原片 23:14
M1
M10:00

OpenAI 团队表示,内部发生的重大变化之一是,像 Astra 这样的最新一代模型的能力极大地提升了他们的开发速度,以至于很多他们原本以为要到 2027 年才能实现的事情,实际上在这次 Dev Day 发布中就推出了。

The team at OpenAI said that one of the big changes that had happened internally was that the power of the latest generation of models like Astra had increased their speed of development so significantly that many things that they thought were only going to come in 2027 were actually coming as part of this Dev Day announcement.

结果就是超过 20 项不同的发布和公告,其中包括一些重磅产品,比如 OpenAI 对 Muse 和 GrokBot 的回应,还有一些低调的亮点,比如企业账户现在可以在一个开放市场上使用 OpenAI 的积分来购买对开源模型的访问权。

The result of that was more than 20 different launches and announcements, including some big headliners like OpenAI's answer to Muse and GrokBot, and some sleeper hits like the fact that enterprise accounts can now use OpenAI credits on an open marketplace to buy access to open-source models as well.

总的来说,我们在 OpenAI Dev Day 上看到的东西,并没有改变我们一直在观察的行业大趋势。

Overall, what we got at OpenAI Dev Day does not change the big patterns and trends that we've been seeing in the industry.

向更具成本效益的模型迈进,这些模型走向更持久的工作方式,以及其中一部分持久工作从单人体验转向多人体验。

A move to more cost-efficient models, those models moving to more persistent work, and some amount of that persistent work moving from a solo to a multiplayer experience.

相反,Dev Day 所强调的恰恰是,这些才是值得关注的趋势,也是将在未来几个月重塑我们所有人使用 AI 方式的趋势。

Instead, what Dev Day reinforced is that these are the trends to pay attention to and the ones that will reshape how we all use AI in the months to come.

AI Daily Brief 是一档每日播客和视频节目,涵盖 AI 领域最重要的新闻和讨论。

The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.

好了朋友们,在我们开始之前,先快速说几个通知。

Alright friends, quick announcements before we dive in.
M1
M11:51

那么,今天节目的最后要说一点。

Now, last note on today's episode.

本来呢,我打算在 OpenAI Dev Day 之外,把昨天白宫发生的所有事情也完整地塞进来,但内容实在太多了,装不下。

Originally, I was going to try to jam in a full accounting of everything that happened at the White House yesterday on top of OpenAI Dev Day, but there was just too much to fit.

所以我决定今天做一期完整的 Dev Day 回顾,然后我们会深入探讨特朗普与各大前沿实验室负责人会面的成果,以及这些是否会改变 AI 未来发展的方向。

So I've decided to go all Dev Day recap today, and then we will dig much deeper into what came out of Trump's meeting with the leaders of all the major frontier labs, and if and how it changes anything about the future development of AI.

不过现在,让我们进入 Dev Day,看看它说明了什么——不仅仅是关于 OpenAI,也是关于我们整体在 AI 领域所处的位置。

For now though, let's get into Dev Day and find out what it says, not just about OpenAI, but about where we are with AI in general.

OpenAI 的公告日真是发布量巨大的一天。

It was an absolutely monster day of announcements from OpenAI.

事实上,公告太多了,所以在今天的节目中,我们只来得及把那些重要的都过一遍,讨论一下公告的内容是什么、是否在预期之中、大家的第一印象,以及它是否会改变 AI 竞赛的格局。

In fact, there were so many that on today's episode, we are just going to get through all of the big ones, discussing what the announcement was, whether it was expected or not, people's first impressions, and how it shifts the AI race, if at all.

这周晚些时候,我们会更深入地探讨其中一些最重要的内容,但现在,光是过完这些就已经要花我们所有精力了。

Later on in the week, we'll go deeper on some of the most important ones, but for now, it is gonna take all that we have just to get through it.

首先登场的是 Dots,这是 OpenAI 对 Muse、GrokBot 以及正在席卷 AI 领域的个人智能体浪潮的回应。

First up is Dots, OpenAI's answer to Muse, GrokBot, and the wave of personal agents that are taking over the AI space.

Sam Altman 将 Dots 描述为能力非常强、始终在线的智能体,旨在处理你能想到的任何事情。

Sam Altman presented Dots as remarkably capable, always-on agents built to handle really anything you can think of.

它们把 AI 带到了一个全新的形态中。

They bring AI to a whole new form factor.

通过 Dots,用户可以创建一个持久的智能体,通过类似短信的界面与之交流,然后让它们使用自己的云电脑来执行任务,支持超过四万个不同的应用。

With Dots, users can create a persistent agent, communicate with it through a text message style interface, and then set them to work on tasks using their own cloud computer with support for 40,000 different apps.

Dots 还可以通过语音通话或者 Microsoft Teams 和 Slack 进行通信,并且在不同服务和会话之间保持持久的上下文。

Dots can also communicate via voice call or across Microsoft Teams and Slack, with persistent context carried over between services and sessions.

在功能上,它和 Muse 有很多相似之处,甚至包括那个可爱的吉祥物。

Functionally, it shares a lot with Muse, right down to the cutesy mascot.

不过,至少在发布的时候,还是有几个需要注意的小问题。

However, at least at the time of release, there are a few caveats worth noting.

目前,用户一次只能创建一个 Dot,但 OpenAI 计划在未来扩展到智能体团队。

For now, users can only create one Dot at a time, but OpenAI plans to expand to teams of agents in the future.

与 Muse 的另一大区别是,OpenAI 只向专业版、商业版和企业版客户提供 Dots。

The other big difference from Muse is that OpenAI is only offering Dots to pro, business, and enterprise customers.

Muse 之所以如此成功的一个很大原因是它完全免费,这意味着 OpenAI 自然限制了他们在这一方面的竞争能力。

One of the big reasons Muse has been so successful is making it completely free, meaning OpenAI is naturally limiting their ability to compete on that front.

首席财务官 Sarah Friar 表示,愿景是最终把 Dots 带给整个用户群体,但就目前而言,这是一个面向专业用户的產品。

CFO Sarah Friar said that the vision is to bring Dots to the whole consumer base, but for now, this is a prosumer product.

OpenAI 最大的卖点是他们使用 GPT-4o 来驱动这些智能体,所以这是我们第一次看到真正的前沿模型来驾驭第一方的个人智能体。

OpenAI's big selling point is that they're using GPT-4o to power the agents, so this is the first time we're seeing a truly frontier model pilot a first-party personal agent.

那么,从很多方面来说,这是最不令人意外的公告之一。

Now, this was in many ways the least surprising of the announcements.

自从 OpenAI 推出以来,这种个人智能体形态基本上是必然会出现的,即使是对 GrokBot 这类面向商业的版本的关注也是如此,OpenAI 进入这个特定领域基本上是板上钉钉的事。

This personal agent form factor has been basically inevitable since the launch of OpenAI, and even the interest in business-focused versions like GrokBot, OpenAI getting into this particular area was pretty much inevitable.

那么,Dots 到底准备好进入黄金时段了没有,还有待观察。

Now, how ready for primetime Dots is remains to be seen.

现场演示出了一些问题,虽然我对这些问题的在意程度远低于 X 上的那些评论家们。演示出问题,基本上只是让你知道这确实是现场直播的一种方式。

The live demo had some issues, although I would put way less in that than the chattering classes on X. Demos going wrong is basically just a way for you to know that they are actually live.

不幸的是,人们一开始用上GPTs,就遇到了其他一些初期问题。

Unfortunately, once people got their hands on GPTs, they ran into a number of other teething problems as well.

有人发帖说,GPTs丢失了所有的工作内容。

Posted a session where GPTs lost all of its work.

Tenebra说,他们本来很兴奋想试试,但发现没办法同时连接多台电脑来共享完整的上下文。

And Tenebra said that they were excited to try it out, but found that they couldn't connect multiple computers at once to share their full context.

其他人有更好的体验。

Others had a better experience.

分析师Max Weinbach写道,我测试这些东西好不好用的基本标准是,它能不能在访问我的邮箱和Google Drive的情况下,自主帮我处理报销。

Analyst Max Weinbach wrote, my basic test for how well one of these works is can it autonomously do my expenses with access to my email and Google Drive?

Grokbot搞得一团糟,没有按我说的做。

Grokbot made a mess and didn't do what I told it.

前三天还行,后来就开始出错了。

It did the first like 3 days and then started to mess up.

Dot第一次就搞定了,而且一直表现得很好。

Dot did it after the first time and has been good.

Justin Schroeder说,他觉得Dot比Grokbot或Muse更专注于工作。

Justin Schroeder said that he thought that Dot was even more work-focused than Grokbot or Muse.

他写道,感觉更不像个人助理,而更像Codex的编排者,以及Slack上的协作者。

Think a bit less personal assistant, he writes, and a bit more Codex orchestrator and Slack collaborator.

事实上,Justin写道,Slack才是它的真正亮点。

In fact, Justin writes, Slack is where it really shines.

你的团队可以直接给你的机器人发消息,机器人可以执行操作,比如打开新的面板。

Your team can message your bot directly and your bot can take actions, like spin up new panes.

另外,这也涉及到一个我觉得会到处都有的争论,Justin认为,这本来应该做成一个单独的App。

Also getting at a debate which I think will be everywhere, Justin suggests that he thinks that this should have been a separate app.

他写道,ChatGPT很好,Codex也很好,在我看来它们是完全不同的产品,而Dot更像Codex,而不是ChatGPT。

ChatGPT is great, he writes, Codex is great, they are very different products in my opinion, and Dot is way more Codex than ChatGPT.

Every的Vibe Check几乎把它留作待定,他们的用户反馈说,有时它确实非常出色,但有时又让人极其沮丧。

The Every Vibe Check almost leaves it as a TBD, with their users reporting flashes where it was truly excellent, but then other times where it was just extremely frustrating.

当Brandon Chu提出这是大家都在趋同的形态时,Nate B. Jones写道,我觉得我们需要区分这里的形态和实用性。

When Brandon Chu suggested that this was the form factor that everyone was converging on, Nate B. Jones wrote, I think we need to distinguish between form factor and utility here.

没错,眼下形态确实在趋同,那可能会变,但实用性并没有。

Yes, form factor is converging at the moment, that may change, but utility is not.

Muse擅长特定的事情,比如打电话和实际的邮件处理。

Muse is good at specific things, like phone calls and practical email work.

Instinct擅长其他特定的事情,比如旅行。

Instinct is good at specific different things, like travel.

Dot擅长把AI上下文作为工作台。

Dot is good at AI context as a work surface.

不管怎么定位,这都是不同的实用性。

Regardless of positioning, that's distinct utility.

这个价值万亿美元的问题是,你选对了要搞定的实用性吗?

The question with a T for trillions is, did you pick the right utility to get right?

我们这周晚些时候会再聊DOTS,但我觉得最重要的启示是,继续关注这个领域。

We'll come back to DOTS later in the week, but I think the big takeaway is continue to watch this space.

虽然承认现在还有点小毛病,但Allie K. Miller认为,这里有一些相当重大的区别。

While acknowledging that it's a bit buggy right now, Allie K. Miller argues that there are some big differences here that are fairly significant.

比如每个DOT都有自己专用的虚拟机,而且是常驻开启的,这让它能更主动,还有其他一些值得关注的变化。

Things like each DOT getting its own dedicated virtual machine, it being always-on, which allows it to be more proactive, and some other changes that are worth watching.

无论如何,榜单上的第二名是Decisions API,这是OpenAI对Jev的回应。

In any case, number 2 on the list is the Decisions API, which is OpenAI's answer to Jev.

记住,Jev是一个判断模型。

Now remember, Jev is a judgment model.

它不擅长输出文本。

It's not good at outputting text.

它擅长对事物进行分类,给出0到1之间的置信度分数,换句话说,就是做判断,而判断是决策的前奏,这大概就是这个功能名字的由来。

It's good at classifying things, giving confidence scores between 0 and 1, making judgments in other words, which are the precursors to decisions, which presumably is where this feature got its name.

Decisions API允许用户调用一个模仿Jev快速分类和决策能力的Luna版本。

The Decisions API allows users to call a version of Luna that mimics Jev's quick classification and decision-making abilities.

用户定义一系列问题和可能的答案,模型就会快速输出响应。

Users define a list of questions and possible answers, and the model outputs fast responses.

OpenAI 表示,这个 API 可以用于分类、路由请求,或者为智能体选择下一步操作。

OpenAI said that the API could be used for things like classification, routing requests, or choosing an agent's next action.

他们声称,决策速度比 Responses API 快十倍。

They claimed 10 times faster decision-making compared to the Responses API.

不过,一个大问题是,这比直接用 Jev 好在哪里?

Still, one of the big questions was how is this better than just using Jev?

实际上,用例的一个区别是 Luna 支持视觉输入,而 Jev 做不到。

One difference in actual use cases comes from Luna's support for visual inputs which aren't possible with Jev.

在过去的几周“Jev 热”中,测试用户展示了 Jev 快速分类视觉营销等场景,但这需要底层进行图像到文本的转换。

Over the past couple weeks of Jevmania, beta users have shown off Jev making rapid classification for things like visual marketing, but that requires an image-to-text transformation under the hood.

Decisions API 去掉了这一步,从而扩展了可能的用例。

The Decisions API removes that step, and in that way expands its set of possible use cases.

目前,Decisions API 还只是面向有限群体的预览测试版。

Now, right now, the Decisions API is just in preview for testing with a limited group.

从重要性来说,我认为这更多是承认 Jev 所代表的不仅仅是新模型。

And in terms of significance, I think more than anything, this is a recognition that what Jev represents is more than just a new model.

它是一个极其有用的、而且我敢说是即将成为生态系统中基础的构件。

It is an extremely useful, and dare I might say soon-to-be fundamental primitive to have in the ecosystem.

Jev 之所以成功,不是因为它花哨或性感。

Part of why Jev has hit wasn't that it was flashy or sexy.

而是因为,当你看到它比传统 LLM 做得好的那些事,实际上,当你看到 LLM 被硬塞进那些它们并不擅长的场景,就像方榫插圆孔,那么有个判断模型和生成模型并排,显然是显而易见的。

It's just that as soon as you see all the things that it does better than a traditional LLM, in fact, when you see that LLMs were being rammed like square pegs into round holes, into use cases that they weren't great at, it just seems obvious that having that sort of judgment model to sit alongside your generative model is pretty obvious in retrospect.

再说一遍,现在比较还为时过早,但我倾向于认为,这个领域有空间容纳不止一个模型。

Once again, it's too early to really do comparisons, but I tend to think that this is a space where there is room for more than one model available.

实际上,我认为几乎每个前沿实验室都会很快推出某种版本。

In fact, I think that pretty much every frontier lab will have some version of this available very, very soon.

接下来是 OpenAI Space,他们的新共享文档工作区,让人和智能体可以协作。

Next up on our list is OpenAI Space, their new shared document workspace that gives human teams and agents a place to collaborate.

你可以把这个功能想象成增强版 AI 的 Google Drive。

You can think of the feature sort of like an AI-enhanced Google Drive.

团队可以在共享电子表格、演示文稿和其他文档上工作,但 Space 也为团队提供了存放工作场所自动化的地方。

Teams can work on shared spreadsheets, slide decks, and other documents, but Space also gives teams a place to house workplace automations.

Dots 可以在 Space 中原生处理文档,但团队也可以设置定时任务,在 Space 中生成交付物。

Dots can natively work on documents in Space, but teams can also set up scheduled tasks to produce a deliverable in their Space.

很多人关注的是 OpenAI 用 Space 对标 Microsoft 365、Google Drive 或 Notion,虽然这确实属于这些领域,但那些服务越来越不适合智能体工作,OpenAI 推出真正 AI 原生的生产力套件只是时间问题。

Many focused on OpenAI going after Microsoft 365, Google Drive, or Notion with Space, and while this is certainly generally in those spaces, each of those services have felt increasingly like a bad fit for agentic work, and it was only a matter of time before OpenAI launched a truly AI-native productivity suite.

然而,几乎其他所有公告都有很大分歧,但 Space 特别受重度用户喜爱,他们立刻爱上了。

Now, whereas almost every other announcement had a pretty wide diversity of opinions, Space was one where especially the power users were in love right away.

Ray Fernando 写道,我提前体验了 OpenAI Dots,我可不是来跟风吹捧的。

Ray Fernando wrote, I've had early access to OpenAI Dots and I'm not here to join in on the glaze fest.

这能杀死 Grokbot 或 Hermes 吗?

Is this a Grokbot or Hermes killer right now?

不。

No.

但是 Spaces 以及让智能体在应用里蓬勃发展,感觉是这些智能体的正确方向。

But Spaces and allowing agents to thrive in apps feels like the right directions for these agents.

在我看来,Bots 有点被高估,而 Spaces 被低估了。

In my opinion, Bots slightly overhyped and Spaces underhyped.

我认识的每家公司都想要一个 AI 原生的协作工作区。

Every company I know wants an AI-native collaboration workspace.

我们将拭目以待,看这是否会吸引更多企业加入 OpenAI 生态系统。

We'll be watching to see if this pulls more enterprises to the OpenAI ecosystem.

明显的好处是,你写作时不用再在 ChatGPT 和 Notion 这样的应用之间切换窗口。

The obvious win is that you stop switching windows between ChatGPT and another app like Notion while you write.

不那么明显的是速度。

The less obvious one is speed.

当机器人在 Google Docs 中通过浏览器构建文档时,编辑速度几乎慢如蜗牛,因为 Docs 不是为智能体设计的。

When a bot builds a document through its browser in Google Docs, the edits practically crawl in, because Docs wasn't built for agents.

原生文档没有那种延迟。

Native documents don't have that lag.

你可以在文档里直接@你的bot,而不是回到聊天窗口告诉它去检查文件。

You can tag your bot inside the document itself instead of going back to the chat and tell it to check the file.

我早就预料到公司会这么做。

I've long expected the company to do this.

2024年的时候,我就写过ChatGPT里文档、幻灯片和表格的潜力,现在这个产品终于上线了,我很高兴。

In 2024, I wrote about the potential for documents, slides, and sheets inside ChatGPT, and I'm glad the product is finally here.

我现在大部分工作已经在ChatGPT的应用内浏览器里完成了,而这项功能让整个流程更顺畅。

I'm already doing most of my work in ChatGPT's in-app browser, and this makes that process smoother.

我可以在文档的评论里@我的bot,然后边工作边直接得到内联的修改。

I can tag my bot in a comment on a document and get a revision inline as I'm working.

这种来回互动感觉更像是真正的协作。

The back and forth feels more like collaboration.

当天的大模型发布是GPT-6.1 Sol,OpenAI把它宣传为接近Astra的智能,但价格只有五分之一。

The big model launch for the day was GPT-6.1 Sol, which OpenAI pitched as near Astra intelligence for a fifth of the price.

这次发布距离GPT-6 Sol只有一个星期,而且带来了一些相当显著的升级。

The release comes just a week after GPT-6 Sol and provides some fairly notable upgrades.

在所有的effort levels上,编码基准测试都有了显著提升,也就是说,6.1 Sol在中档设置下的表现超过了6 Sol在最高设置下的表现。

Coding benchmarks are up significantly across all effort levels, meaning that 6.1 Sol on medium settings outperforms 6 Sol on max settings.

这个模型在DeepSWE上的最高分是高设置下的75.2%,而超高设置和最高设置反而出现了性能下降。

The model's top score on DeepSWE was 75.2% on high settings, with extra high and max actually seeing a degradation in performance.

我们第一次看到这种现象是在Opus-5上,看起来最高设置开始迫使模型过度思考,并对自己原本正确的回答产生怀疑,反而弄巧成拙。

We first saw this phenomenon with Opus-5, and it looks like top-effort settings are starting to force models to overthink and second-guess correct responses, to their detriment.

而且,高设置下的得分实际上比Astra的最佳表现还要高一点,这支持了OpenAI所称的“接近Astra的智能”。

Now, that high setting's score on DeepSWE was actually a touch higher than Astra's best performance, supporting OpenAI's claim of near-Astra intelligence.

这个模式在大多数主要基准测试中都很相似——相比6 Sol是很大的提升,让6.1 Sol以便宜得多的价格达到了和Astra同一水平。

The pattern was similar across most major benchmarks— a big improvement over 6 Sol that landed 6.1 Sol in the same ballpark as Astra at a much cheaper price.

其中一个值得注意的基准测试是OSWorld,它测试的是长时程的计算机使用能力。

One of the notable benchmarks was OSWorld, which tests long-horizon computer use.

6.1 Sol在最高设置下的最佳成绩是71.4%,超过了6 Sol的64.4%,并接近Astra最佳的73.5%。

6.1 Sol's best performance scored 71.4% on max settings, beating 6 Sol at 64.4% and coming close to Astra's best at 73.5%.

6.1 Sol的价格也明显更低,只有6 Sol的三分之一,以及Astra的13%。

6.1 Sol was also significantly cheaper than the others, at a third the cost of 6 Sol and 13% the cost of Astra.

更重要的是,提高effort level几乎不会改变成本,这表明OpenAI在这个模型的计算机使用效率上取得了重大突破。

What's more, increases in the effort level barely changed the cost, suggesting that OpenAI has made some big breakthroughs in the efficiency of computer use with this model.

Artificial Analysis给这个模型打了52分,比Astra低一分,也落后于Opus、Sonnet和Fable,排名第五。

Artificial Analysis scored the model at 52, one point shy of Astra and also behind Opus, Sonnet, and Fable, coming in 5th place.

他们还发现了令人难以置信的成本效益,把Pareto frontier又往前推了。6.1 Sol完成基准测试的成本只有Astra的四分之一,比6 Sol便宜31%。

They also found incredible cost efficiency that pushes the Pareto frontier, with 6.1 Sol's benchmark run completed at a quarter the cost of Astra and 31% cheaper than 6 Sol.

如果你不需要用最高设置运行它,6.1 Sol似乎可以达到更低的成本,比如GLM-53 Flash——那是ZAI开源模型更便宜的版本。

If you don't need to run it at max settings, 6.1 Sol seems capable of hitting smaller costs like GLM-53 Flash, which is the cheaper version of ZAI's open-source model.

OpenAI称它是目前性价比最高的模型。

OpenAI called it the most cost-efficient model for its performance available today.

那么,这个发布是意料之中还是意料之外呢?一方面,从这些实验室拿到一个新模型从来都不算什么太意外的事。

Now, as to whether this one was expected or unexpected, on the one hand, it's never all that surprising to get a new model from one of these labs.

尤其是在像Dev Day这样的大日子。

Especially on a big day like Dev Day.

意料之外的是我们上周才刚拿到o1。

Surprising in that we only got o1 last week.

所以说到第一印象,我们只能说,还得等我们自己上手试过才知道。

So when it comes to first impressions, let's just say we're gonna have to wait to get our hands on it ourselves.

因为,你每看到一条像Dropout Layer在X上发的帖子——说“GPT-6 o1让Opus 5.5看起来都变贵了”——就会看到一条像BridgeBench这样的帖子。

Because for every post you can find like this one from Dropout Layer on X, GPT-6 o1 just made Opus 5.5 look expensive, you also get one like this one from BridgeBench.

我们用Sunset Ocean测试跑了一下,和GPT-6 o1一样的成本,但速度慢一倍,而且它几乎没能渲染出一片海洋。

We ran it through our Sunset Ocean test, same cost as GPT-6 o1, twice as slow, and it barely rendered an ocean.

大家好。

Hello everyone.

关于AI的一个重大变化是,我们的思路已经从“如何给我们的页面排名”转变为“如何成为AI足够信任、愿意引用来回答问题的来源”。

One big change around AI is we've shifted our thinking from how we rank our pages to how do we become the source that AI trusts enough to answer with.

在 KPMG,他们亲眼看到了这一点。

At KPMG, they're seeing this firsthand.

AI 生成的结果现在直接呈现答案,通常一次点击都不需要。

AI-generated results now surface answers directly, often without a single click.

这就是为什么他们越来越关注生成式引擎优化,也就是 GEO,通过结构化内容,让 AI 系统能够检索、理解,并将其作为可信权威来引用。

That's why they are increasingly focused on generative engine optimization, or GEO, structuring content so AI systems can retrieve it, understand it, and cite it as trusted authority.

这不仅是 SEO 的演进,更是一种可见性的要求。

This is not just an SEO evolution, but a visibility mandate.

确实,KPMG 对 GEO 的要求很简单。

And indeed, the GEO mandate from KPMG is simple.

如果 AI 在塑造决策,你的专业知识就需要出现在答案里面。

If AI is shaping decisions, your expertise needs to show up inside the answer.

请访问 kpmg.com/us/geo 了解详情。

Read all about it at kpmg.com/us/geo.

再说一遍,网址是 kpmg.com/us/geo。

Again, that is kpmg.com/us/geo.

这就是为什么大多数遗留系统现代化项目会失败。

Here's why most legacy modernization projects fail.

做这项工作的 AI 无法大规模理解代码库。

The AI doing the work can't understand codebases at scale.

它只看到一小部分上下文,检查语法,却错过了分布在全球应用生态系统中多年的决策。

It sees a small slice of context, examines syntax, and misses years of decisions distributed across the global application ecosystem.

Blitzy 像解决所有问题一样解决这个问题。

Blitzy solves this the way it solves everything.

在任何迁移开始之前,Blitzy 的代理基于你的代码,将整个遗留系统逆向工程为一个持久的知识图谱。

Grounded in your code before any migration begins, Blitzy's agents reverse engineer the entire legacy system into a persistent knowledge graph.

每一个依赖、每一个约束、每一条曾经存在于某位工程师脑中的部落知识。

Every dependency, every constraint, every piece of tribal knowledge that used to live in one engineer's head.

基于这种理解,Blitzy 自主执行语言迁移、框架升级以及单体到微服务的转型,并且全部经过端到端验证。

From that understanding, Blitzy autonomously executes language migrations, framework upgrades, and monolith-to-microservices transformations, all validated end-to-end.

Blitzy 的一个客户使用编码代理,以 137 周为基线,用 16 周时间将一套价值 1000 万美元的单体保险技术栈进行了现代化改造。

One Blitzy customer modernized a $10 million monolithic insurance stack in 16 weeks against a 137-week baseline with coding agents.

那是 9 倍的压缩。

That's 9x compression.

在加速路线图的同时,清除技术债务。

Retire technical debt while accelerating your roadmap.

请访问 blitzy.com 了解具体做法。

See how at blitzy.com.

就是 B-L-I-T-Z-Y 点 com。

That's B-L-I-T-Z-Y dot com.

最好的团队不是靠一个明星扛起所有人。

The best teams don't have a single star carrying everyone else.

他们了解自己的长处和对方的短处,并加以利用。

They know their own strengths and each other's weaknesses and play to both.
M1
M115:12

不过,希望这些 o1 级别的模型足够好,因为我们可能永远看不到 GPT-6.1 Astra 了。

Now, hopefully these o1-class models are good enough though, because we may never see GPT-6.1 Astra.

The Wall Street Journal 报道说,OpenAI 出于安全考虑,已经取消了发布下一代旗舰模型的计划。

The Wall Street Journal reports that OpenAI have scrapped plans to release the next version of their flagship model over safety concerns.

OpenAI 的安全系统负责人 Sachi Jain 在一份声明中表示,凡是涉及安全和一致性的事情,都存在权衡取舍。

Sachi Jain, OpenAI's head of safety systems, said in a statement, for anything regarding safety and alignment, there's a trade-off.

你确实需要找到一条正确的界线,既要保持在范围内,又要避免模型在执行任务时偷懒,哪怕遇到阻碍。

You really do need to find what's the right line between staying within scope, but also avoiding laziness in terms of how the model actually pursues tasks, even when it hits friction.

听起来,OpenAI 试图降低那种在最近几个月里导致多起事件的执着程度,但没能找到一个理想的中间地带。

It sounds like OpenAI tried to turn down the tenaciousness that had led to multiple incidents over recent months, but couldn't find a happy medium.

Jain 补充说,虽然这个模型比之前的版本少了些懒惰,但在保持在范围内、遵守授权,以及它如何向用户反馈已完成工作类型这些方面,还不太达标。

Jain added that although the model was less lazy than its predecessors, it didn't quite meet the bar in terms of staying within scope and authorization, and how it communicates back to the user about the type of work it's done.

显然,当我说我们永远不会看到 6.1 Astra 时,我是有点夸张了,但实际上,看起来他们确实需要取得一些技术突破,才能达到下一个前沿水平,并以他们认为足够安全的方式发布。

Now obviously I'm being a bit hyperbolic when I say that we'll never see 6.1 Astra, but it actually does seem like they're going to need to hit some technical breakthroughs before we get to some of the next levels of the frontier in a way that they deem safe enough to release.

接下来,我们要讲的是活动中那些被认为比较小的公告,其中许多仍然被大家认为意义深远。

Now from here we get into what are considered the smaller announcements from the event, many of which people are still noting as quietly significant.

OpenAI向我们展示了其平台计划的最新版本,向外部开发者开放了ChatGPT。

OpenAI showed us the latest version of its platform plans, opening ChatGPT to outside developers.

OpenAI团队的Thibault写道,你现在可以构建完整的原生应用,通过插件扩展直接发布到ChatGPT中。

Thibault from the OpenAI team wrote, you can now build full native apps with plugin extensions and ship them right in ChatGPT.

我们每周有超过12亿用户,会在对话中直接呈现相关的插件。

We have over 1.2 billion weekly users and will surface relevant plugins right in the conversations.

这个功能在技术上叫做插件扩展。

This feature is technically called plugin extensions.

其实OpenAI在这一年多来一直试图在这个功能上找到平衡,测试了原生集成、插件和MCP来扩展生态系统。

Now, OpenAI has been trying to get the balance right on this feature for more than a year, testing native integrations, plugins, and MCP as a way to expand the ecosystem.

那么,这最终会不会成为让ChatGPT像应用商店一样运作的功能呢?这似乎一直是OpenAI的重大目标之一。

So could this finally be the feature that lets ChatGPT function as an app store, which always seemed like one of the big goals for OpenAI?

当然,这里肯定会出现一股争相尝试的热潮。

Certainly there is going to be a rush to experiment here.

我已经看到像Jana Wellender做的个人造型师应用这样的东西了。

I'm already seeing things like this personal stylist app from Jana Wellender.

但也有人在想,对应用开发者来说,这些取舍真的值得吗?

But others are wondering if the trade-offs for the app developer are really worth it.

麻省理工学院的Christian Catalini写道,你带来应用,OpenAI带来智能。

MIT's Christian Catalini writes, You bring the app, OpenAI brings the intelligence.

那OpenAI是不是也拿到了用户的使用痕迹?

Does OpenAI also get the user traces?

如果你的贡献是教你的合作伙伴怎么做你的工作,那这个合作还真有意思。

An interesting partnership if your contribution is teaching your partner how to do your job.

对此,Signal回应说,长尾应用可能会选择加入,但除此之外,这对其他任何人来说都是一个糟糕的主意。

To which Signal responded, long tail of apps might opt in, but otherwise this is a terrible idea for anyone else.

这个的另一面是用ChatGPT登录,功能正如其名。

A flip side of this is Sign In with ChatGPT, which is exactly what it sounds like.

它让你可以用ChatGPT登录其他账户。

It allows you to use ChatGPT to sign into other accounts.

OpenAI应用研究主管Boris Power写道,用ChatGPT登录可以让你使用任何基于我们API构建的应用。

OpenAI's head of applied research Boris Power wrote, Sign in with ChatGPT lets you use any app built with our API.

这对应用开发者来说好多了,不用费那么大劲去说服用户为单独的订阅付费。

Makes it much better for app developers not needing to jump a huge hill justifying paying a separate subscription.

ChatGPT就是一站式超级智能果汁。

ChatGPT is your one-stop superintelligence juice.

这个想法基本上是,开发者允许用户随身携带他们的智能订阅,降低了应用的门槛,也让他们可以利用这样一个事实:很多人已经选择使用ChatGPT作为他们的超级智能解决方案了。

The idea is basically that developers allow people to carry their intelligence subscription with them, lowering the barrier to entry for their apps and allowing them to take advantage of the fact that so many people have already opted into using ChatGPT for their superintelligence solution.

用ChatGPT登录是一件大事,我不明白他们为什么把它框定为身份验证,实际上这是OpenAI在把优惠定价扩展到第三方,并把服务捆绑进订阅里。

Sign-in with ChatGPT is a huge deal, and I don't understand why they framed it as auth when it's really OpenAI expanding favored pricing to third parties and bundling services into the subscription.

这终于开始让他们的激励与客户保持一致,个人不用再为token重复付费,应用也无需在API成本之上再为一切定价。

It finally starts to align their incentives with customers so that individuals stop double-paying for tokens, and apps stop having to price everything on top of API costs.

我想这会扩展到更多应用,甚至所有应用,Anthropic也将不得不跟进,为他们的订阅增加价值。

I'd imagine that expands to many more if not all apps, and Anthropic will need to follow to add value to their subscription.

在那种情况下,应用可以绕过token成本,重新只为应用层收费,这对所有人来说可能都是好事。

In that world, apps could bypass token costs and charge for the app layer alone again, which is probably good for everyone.

换句话说,这个思路是,应用不再需要为了让人使用智能而收取看起来很贵的价格,他们只需收那20美元,或者任何他们想给应用体验定的价格,然后让API成本直接流向底层。

In other words, the idea here is that instead of apps having to charge what looks like huge prices to get access to intelligence, they can just charge their $20 or whatever they wanted to price their app experience at and let the API costs flow directly to the underlying.

是的,那意味着他们可能没法加价从中赚一笔了,但好处是,当用户已经在付ChatGPT订阅费时,应用不用再费劲说服用户去付那些额外费用,这很可能超过他们本可以赚到的钱。

While yes, that means they might not get to scalp those costs and add a little premium to them, the benefits of not having to convince people to pay those additional fees when they're already paying for a ChatGPT subscription likely outweighs the money that they could make otherwise.

对于我的重度用户们,OpenAI为他们的订阅推出了一些新选项。

For my power users out there, OpenAI has introduced a few new options for their subscriptions.

第一个是UltraFast模式,在Codex中token生成速度快8倍,在应用里快6倍。

The first is UltraFast mode, which offers 8x faster token production in Codex and 6x speed in the app.

目前,UltraFast模式只适用于ChatGPT Work中的Astra和Codex,但OpenAI表示很快会加入对o1和Sol的支持。

Right now, UltraFast mode is only available for Astra and Codex in ChatGPT Work, but OpenAI say they will add support for o1 and Sol soon.

此外,OpenAI还推出了一个500美元的新订阅等级,使用量是Plus等级的25倍。

In addition, OpenAI is introducing a new $500 subscription tier that offers 25 times the usage of the Plus tier.

这也是目前唯一一个能使用UltraFast模式的等级。

It's also the only tier that gets access to UltraFast mode for the time being.

活动开始前,Thibault宣布OpenAI会重新开放他们那个每月200美元的Pro tier,几周前他们关掉了这个选项,不过这次做了一些调整。

Heading into the event, Thibault announced that OpenAI would reopen access to their $200 a month Pro tier, which they had turned off a few weeks ago, but with a few tweaks.

使用量的计算方式改了,算下来API成本降低了50%。

Usage calculations have been changed, netting out to a 50% reduction in terms of API cost.

Thibault解释说,这是在一堆糟糕的选择里最好的那个。

Thibault explained that this was the best of a bad set of choices.

OpenAI优先考虑不引入5小时使用限制,并且认为API成本降低加上更高效的模型,意味着用户还是能完成大致相同的工作量。

OpenAI prioritized not having to introduce the 5-hour usage limit and argued that API cost reductions and more efficient models would mean users can still get roughly the same amount of work done.

一方面,大家多少都预料到了这种事,但另一方面,你可以想象,同一个模型回归但价值降低了,反响会怎么样。

Now, on the one hand, people were sort of expecting something like this, but at the same time, you can imagine how well it went over to have the same model coming back with reduced value.

对我们想了解AI现状来说,你看,算力限制是真实存在的,当下的,而且会一直持续下去。

For our purposes here of trying to understand what it says about the state of AI, look, man, compute constraints are real, present, and permanent.

前沿模型在现有算力下已经快跑到极限了,而新算力上线的速度远远跟不上人们对这种数字智能使用量的增长。

Frontier models are running up against the walls of what they can do with the compute that we have, and new compute isn't coming online at anywhere near the speed that people are increasing their use of this digital intelligence.

不过好的一面是,我们会看到效率方面的大推动。

The upside of that, though, is that we're going to see a big push around efficiency.

长远来看,这最终会让我们所有人享受到更具性价比、更好的体验,但过程中肯定会有一些波折。

Which in the long run should net out to more cost-effective, better experiences for all of us, but it's gonna have some bumps along the way.

在开发者和vibecoder那边,Codex现在有了专门的云环境,让用户合上笔记本电脑后还能继续工作。

Over in developer and vibecoder land, Codex now has a dedicated cloud environment, allowing users to keep working after they close their laptop.

这个功能迟早会来,因为之前旧金山到处都是人把大拇指塞在笔记本上走来走去,都成了一个梗了。

This one was always coming after people walking around with their thumbs jammed into laptops became so common across San Francisco that it became a meme.

但看更广泛的趋势,从GrokBot开始,整个行业都在往云端实例移动,它已成为所有代理产品的必要组成部分。

But looking for broader patterns, starting with GrokBot, we've seen the entire industry move towards a cloud instance being a necessary part of all agentic products.

这延续了一个重大转变:AI不再是你像使用软件那样交互的东西,而是变成一种更持久、始终在线的应用,无论你通过什么方式访问它。

This is a continuation in the big shift from AI being something you engage with like software, to becoming a more persistent always-on application regardless of how you access it.

Codex CLI也迎来重大更新,有了新外观和新功能。

The Codex CLI is also getting a major refresh with a new look and new capabilities.

滚动时历史记录可以追溯到更早,支持单线程的同志们——我绝对是其中之一——而输入框始终固定,随时随地可以输入下一行。

History goes back further as you scroll, supporting the mono-thread masses, of which I am absolutely one, while the composer stays pinned so you can always type your next line.

Codex CLI现在能让你更好地管理并行工作。

The Codex CLI now gives you a better way to manage parallel work.

使用/agents查看正在运行的任务、检查进度并在不同任务之间切换。

Use /agents to see what's running, check progress, and jump between tasks.

在企业方面,OpenAI推出了Private Intelligence,保证零数据留存,甚至在推理阶段也是如此。

On the enterprise side, OpenAI has launched Private Intelligence, which guarantees zero data retention even at inference time.

过去几个月,隐私一直是对企业至关重要的问题,现在也成了为企业提供服务的公司最关心的问题——大家越来越担心OpenAI和Anthropic从他们的客户那里窃取数据。

Over the past few months, privacy, which was always a huge issue for enterprises, has become a huge issue for the companies supplying the enterprises, with growing concerns that OpenAI and Anthropic are skimming data from their customers.

这基本就是微软最近所有沟通的潜台词或者明说:你为什么不该信任他们的竞争对手。

This is basically the subtext or the main text of every communication from Microsoft these days about why you should be not trusting their competitors.

所有这些意味着,对于有需要的人来说,更强有力、更清晰的端到端数据隐私保证是一件大事。

All that means that a stronger, clearer end-to-end data privacy guarantee is a big deal for those who need it.

OpenAI还推出了一个模型市场。

OpenAI has also launched a model marketplace.

它允许用户通过OpenAI的Responses API和Codex从Base10购买开源权重模型的推理服务。

It allows users to buy open-weight model inference from Base10 through OpenAI's Responses API and Codex.

这个功能在另一天我们可以深入聊,因为它对OpenAI的护城河有一些重大影响。

Now this is one that on another day we could go way deep into because it has some big implications for OpenAI's moat.

它保护他们免受开源颠覆,也给顾客更多选择。

It protects them from open-source disruption and gives customers a lot more choice.

这也意味着企业可以放心地对OpenAI做出大笔开支承诺,因为他们知道这笔钱可以轻松用于多模型策略,包括开源权重模型。

It also means enterprises can feel comfortable making large spending commitments with OpenAI, knowing that they can easily use that spend on a multi-model strategy that includes open-weight models.

请相信我,我们还会再聊到这个,因为我觉得它比人们第一印象中以为的要重要得多。

Trust me when I say that we are going to come back to that one because I think it is bigger than people are giving it credit for at first blush.

我们只是刚刚触及表面,现在快速总结一下所有这一切意味着什么,以及OpenAI DevDay揭示了AI下一阶段的什么。

So we have barely scratched the surface here, but let's try to quickly sum up what all of this amounts to and what OpenAI Dev Day revealed about the next phase of AI.

我们没有看到什么全新的爆炸性东西。

We didn't get something blisteringly new.

我们看到的是对过去几个月我们在这个节目中一直在记录的许多趋势的确认。

What we got was confirmation of a lot of the trends that we've been cataloging on this show over the past several months.

首先,通过 SOL 6.1 和 Decisions API,我们看到这个趋势在延续和加深,就是模型既够好又够便宜,可以用它们来做所有事情。

First, with SOL 6.1 and the Decisions API, we got a continuation and a deepening of the trend of models that are good enough and cheap enough that they can be used to do everything.

也就是说,你不用关掉它们,也不用为它们该用来做什么而做决定,它们就能把一切都做了。

I.e., you don't have to turn them off, you don't have to make decisions about what they're used for, they can just do it all.

有了 Dots,我们有了持久且主动的 agent,它们利用这一点去真正做所有事情。

With Dots, we have persistent and proactive agents that take advantage of that to actually do everything.

说到做所有事情,插件扩展表明,这意味着把一切都接入进来,同时也去往任何地方,也就是用 ChatGPT 登录其他应用。

And when it comes to doing everything, plugin extensions shows that that means bringing everything in, as well as going everywhere, which is ChatGPT sign-on for other apps.

最后,通过 Space,我们得到更多确认,下一代 AI 将不仅仅是单机模式,还会是原生的团队和多人模式,即使目前仍处于早期阶段。

Finally, with Space, we have more confirmation that the next generation of AI will not just be single-player mode, but will also be team and multiplayer mode in a native way, even if that remains nascent so far.

这里还有很多很多可以探索的,但今天的 AI Daily Brief 就到这里。

Lots and lots more to explore here, but that is going to do it for today's AI Daily Brief.

一如既往感谢你的收听或收看,下次再见,peace。

Appreciate you listening or watching as always, and until next time, peace.
已剔除 2 处广告(点击展开查看)
M1
M11:05广告 · 已剔除

First of all, thank you to today's sponsors: KPMG, Blitzy, Robots and Pencils, and HyperAgent.

To get an ad-free version of the show, go to patreon.com/AIDailyBrief, or you can subscribe on Apple Podcasts.

To learn more about sponsoring the show, send us a note at [email protected].

There are also all sorts of other goodies at aidailybrief.ai.

On our website, you can get access to all of our free AI training programs.

Those include free multi-week self-directed programs like the Multiplayer AI Sprint, which is live right now, but also free live webinars like the one that is coming up on Thursday, October 1st at noon that is all about building your personal AI benchmark.

You can also find links to our paid trainings, such as the Superintelligent Executive Catch-Up and the Superintelligent Executive Agent Leadership Program.

Both of which are registering new cohorts that start over the next couple of weeks.

M1
M114:15广告 · 已剔除

That's the team Robots and Pencils has built on purpose.

Nobody there is grinding through busywork to pad a headcount number.

People come for the hard problems, and they stay because everyone around them is leveling up at the same time.

In a market full of companies that are just trying to hire fast, that's worth a look.

Check out robotsandpencils.com/careers.

This episode of the AI Daily Brief is brought to you by HyperAgent, where you run fleets of agents your team can manage together.

Forget local agents and chat workflows waiting on your laptop to be prompted.

HyperAgent deploys always-on agents in the cloud, doing real work across the tools your team already uses.

Marketing agents turn competitor moves into landing pages.

Sales agents enrich leads, draft emails, and update the CRM.

Ops agent chases the paperwork and tracks the budget.

Every agent has access to shared context and follows your rules about scope and approvals.

It's time you had agents that feel like teammates.

Hire yours at HyperAgent.

Get $100 in credits at hyperagent.com/AIDailyBrief.