OpenAI 为何竞相让电脑操作超越人类

欢迎来到 A16Z 播客。我是 Sema Amble，和我的同事 Fabrizio Saraphiniti 一起。八月份我们写过关于计算机使用智能体的文章，以及它们如何开始在公司内部做实际工作。今天请到了 OpenAI 的 Ari Weinstein 和 James Sun，他们将帮助解析计算机使用智能体的现状，并分享对未来的看法。Ari、James，感谢你们的到来。

Welcome to the A16Z podcast. I'm Sema Amble. I'm joined by my colleague Fabrizio Saraphiniti. In August, we wrote about computer using agents and how they're starting to do the real work inside companies. Today I'm joined by Ari Weinstein and James Sun from OpenAI who are going to help unpack what is going on with computer using agents from their perspective and share some thoughts on the future. Ari, James, thanks for joining us.

当然，很高兴来到这里。

Of course, happy to be here.

是的，感谢邀请我们。

Yeah, thanks for having us.

Ari，先请你讲讲 Astra 中的哪些特性推动了计算机使用智能体的进步，以及真正解锁了什么。

Ari, kick it off and tell us a little bit about what in Astra has made computer using agents move forward and what's really unlocked.

是的。Astra 是一个很棒的模型，擅长很多事情。我们的计算机使用模型在 Astra 之前就已经很好了，比如 Soul 本身在计算机使用方面就很出色。所以根据我的经验，Astra 主要在几个边缘方面有帮助。第一，它平均稍微快一点，因为完成同样的任务使用的思考 token 更少。它在某些创造性工作上更好，比如我们在发布视频中有个演示，你和它对话，它在 Microsoft Paint 上画画。它在这方面好多了，我认为这可能也会迁移到其他类型的创造性任务上。最后我发现它在边缘情况下更可靠。我感觉我更信任它一点。就像那种难以察觉但非常重要的差异，比如 99% 可靠和 97% 可靠之间的区别。我觉得它稍微高一点，感觉非常具体。

Yeah. So Astra's an amazing model. It's good at so many things. Our computer use models were already so good before Astra, you know, like Soul itself is amazing at computer use. So Astra really helps with a few things around the edges in my sort of anecdotal experience. One is it is a little bit faster on average because it uses fewer thinking tokens to do the same task. It's better at certain kinds of creative work like for example we had this demo where in our launch video where you were talking to it and it's drawing things on Microsoft Paint. It's way better at that sort of thing which I think probably transfers to other kinds of creative tasks also. And then the last thing I find is it's just more reliable around the edges. Like you know it's kind of I feel like I trust it a little more. like it's sort of the kind of thing where there's a little difference that is hard to perceive but really important between something being 99% reliable and 97% reliable. And I feel like it's sort of just a little a little higher in a way that feels pretty pretty concrete.

是的。我想补充一些，Astra 也是我们最对齐的模型。对。所以谈到信任，如果你看它的系统卡，它在抵抗提示注入方面好多了。它在不采取任何意外行动方面也好多了。所以当考虑到企业工作，特别是公司用它处理敏感事务，比如采购或与供应商、客户沟通，你会感觉使用它安全得多，这非常令人兴奋。然后在创造性方面也有一些有趣的事情。我想内部我们让 Astra 和 5.6 玩电子游戏，它每次都赢。所以有一些很酷的推文，比如 Astra 能在 30 小时内击败《辐射》之类的。所以它在玩电子游戏方面也很棒。我不知道这可能会转化为什么，但非常令人兴奋。

Yeah. And I think adding on to some of those things, Astra is also our most aligned model. Right. And so, you know, talking about trust a little bit. Um you know, if you look at the system card for it, it is much better at being resistant to prompt injection. um it's much better at not taking any kind of um unintended actions. And so when you think about it for like enterprise work, especially for companies using it for sensitive things, whether it's like procurement or you know talking to vendors or talking to customers, it's going to be just a you're going to feel a lot safer kind of using it and that's really really exciting. And then so some some fun things on the creative side. I think internally we had um you know Astra play video games against uh you know 5.6 and it wins every time. And so there are some cool tweets that people have made where, you know, Astro is able to beat Fallout in like 30 hours or something like that. And so it's also just really amazing at playing video games. So I don't know that probably translates to something, but yeah, it's very exciting.

也许就这一点深入一下，消费者与企业使用计量智能体的用例，围绕计算机使用智能体的所有脚手架，影响非常不同。如果让你们选，你们认为消费者和企业之间哪个更令人兴奋？

Maybe just to double click on that point, consumer versus enterprise use cases of metering agents, very different implications around kind of like all the scaffolding around I would say the computer using agents. How do you guys think about where things are more exciting between consumer and enterprise if you were to pick?

我认为我们在 OpenAI 的目标是构建对每个人都有用的产品。有趣的是，Ari 在这里更偏向消费者，而我在企业方面花的时间多一点。我认为消费者始终是我们尝试推动可能性边界的地方。就像 Ari 提到的，Astra 比之前的模型快得多，所以我们经常思考用更快更好的模型能实现哪些新的消费者体验。然后在企业方面，安全自然变得更有趣，因为每当我们与大公司交谈时，计算机使用的用例实际上相当明显。企业内部有很多遗留软件没有 MCP 工具，这些软件已经使用了 15 到 20 年，而且很多时候对普通人来说很难使用。很多 UI 来自过去十年。你甚至不知道如何操作一些关键功能。所以我们看到公司构建技能来教模型更好地使用它们。所以也许思考方式是：企业是能力可能已经存在更久的地方，公司正在追赶并理解如何处理它；而消费者是我们试图找出今天软件可能无法实现的下一类用例。也许你可以分享更多。

I think you know our goal at OpenAI is really about building products that's useful for everyone. And so um you know actually it's interesting because Ari is definitely more of a consumer guy here and I spend a little bit more time on the enterprise side. Um I think the consumer is always the place where we try to push the frontier on what's possible. uh you know uh with Astra like Ari mentioned there is it's a much faster compared to the previous models and so we think a lot about like what new consumer experiences can we enable with a faster better model and then on the enterprise side that's kind of naturally why the safety becomes more interesting because I think the enterprise whenever we talk to large companies I think the use cases for comput is actually fairly obvious right inside enterprises there's lots of legacy software that doesn't have any MCP tools they are software that's been used for like 15 20 years and so there's and and and a lot of times it's actually kind of hard to use for a normal person. right? Lots of like UI that is like from the last, you know, decade. And so you don't even know how to operate some of the key key key things. And so we've seen companies build skills that kind of teach the model to be able to use them better. And so I think maybe the way the way to think about it is enterprise is where, you know, the capabilities have been there maybe for a little bit longer and the companies are catching up trying to understand how to deal with it and consumers where we're trying to figure out the next types of use cases that's not maybe possible today with software. Maybe you can share more.

是的，消费者用例对我来说很有趣，因为就像如何让你的智能体去做日常生活中烦人的事情，节省大量时间，让你有更多时间专注于你兴奋的事情。我觉得看到 dot 真的很有趣，每个 dot 都有自己的计算机，你可以让它在云端为你做事。它有完整的网络浏览器，可以为你买东西，可以预约理发等等任何你想做的事情。所以这真的很有趣，我认为未来还有更多有趣的事情。

Yeah, I mean the consumer use cases are so fun for me because it's kind of like how can you put your agent to work doing the things in your daily life that are so annoying and save you tons of time and just have more time to focus on the things that um that you're excited about. Um I think it's been really fun to see dot uh where you know each dot gets its own computer and you can ask it to do things for you in the cloud. Um it gets a full web browser. It can buy things for you. can um you know book haircut appointments and uh you know anything else you wanted to do. So um so that that's been really fun and I think there's a lot more on the horizon. That's going to be really fun.

是的。我知道那会是我关于 dot 的问题。我想 James 你可能用 dot 来走查产品反馈。

Yeah. Know that was going to be my question on on on dots. I think James you probably use a dot to walk through product feedback.

如果你能多分享一些就太好了。我很好奇你们那边是怎么使用 Dots 的，以及计算机使用功能如何融入你们用它们做的事情？

It would be cool if you could share more about it. I'm just so curious to know on your end guys, like how are you using Dots and how does computer use play into what you can do with them?

是的，我觉得 Dots 真的很酷，原因有几个。一是它非常主动。所以很多次我都很惊讶，无论是在个人生活还是工作中，Dots 会标记出一些我完全遗漏的事情。就在前几天，我们内部的邮件系统里，每当我们收到外部邀请邮件时，它实际上不会显示在我们的日历上。所以我最终会错过很多外部会议。现在 Dots 会帮我捕捉这些。它会说，哦，你的收件箱里有个邀请，但不在你的日历上。你要我在你的 Gmail 里接受它，让它显示出来吗？这真是救星。这是小事，但很重要。但我也让它主动寻找——我觉得它非常擅长主动寻找那些形状不是每次都很具体的任务。所以我还设置了一个例子：我们内部使用 Slack，随时都有大量消息。我以前有一个 Slack 任务，使用自动化和 codecs，每小时或每 30 分钟去拉取一些东西。结果我最终忽略了很多通知，因为有时有用有时没用。有了 Dots，我可以给出一个高层次的目标。比如，嘿，如果我汇报链中的任何人说了什么，就标记给我。如果有人给出一些关键反馈，对某些产品用你的判断，去吧。然后因为它有云计算机，我有一个很酷的工作流程：当有人报告时，他们通常有 Sentry 日志。我可以让它去使用云计算机，实际上驱动我的本地浏览器——而 Dots 会同时控制你的本地计算机和云计算机——登录 Sentry，拉取日志，分析它，给我反馈。所以它节省了整个循环：弄清楚哪些反馈重要、查看它、然后去 Sentry、下载日志、喂给 codecs、询问想法，所有这些现在都可以自动化。所以，是的，Dots 至少对于这些主动和长期运行的任务非常棒，这些任务不是每次都有相同的形状。

Yeah, I mean I think Dots is really cool for a few reasons, right? One is it's really proactive. So I've been very surprised a lot of times when in both my personal life and work, Dots will just flag something that I just completely slipped through the cracks, right? Just the other day—so internally on our email system, whenever we get an external email for an invite, it doesn't actually show up on our calendars. And so I end up missing a bunch of external meetings. And so Dots now is catching these things for me. It's like, oh, there's this invite that's in your inbox. It's not on your calendar. Do you want me to accept it in your Gmail so it shows up? So that's been a lifesaver. It's a small thing but it matters a lot. But I also have it proactively looking at—I think it's really good at proactively looking for tasks where the shape of it is not super concrete every single time. So an example that I also have set up to do is—we use Slack internally and there's a ton of messages going on at all times. And I used to have this Slack job using automations and codecs that's like every hour or every 30 minutes go and try to pull things. And I've just ended up tuning out a lot of the notifications because it's just like sometimes useful, sometimes not. With Dots, I can give a high-level goal. It's like, hey, if anybody in my reporting chain says anything, flag it to me, right? If anybody is giving some critical feedback, use your judgment about certain products, go for it. And then because it has a cloud computer, I've had this pretty cool workflow where when someone reports, they usually have a Sentry log. I could have it go use the cloud computer, actually drive my local browser—while Dots is going to control both your local computer as well as the cloud computer—log into Sentry, pull the logs, analyze it, give me feedback. And so it's saving that entire loop of figuring out which feedback is important, looking at it, then going to Sentry, downloading the logs, feeding into codecs, asking for thoughts, and all that can now just kind of be automated. And so, yeah, Dots is great for at least these proactive and long-running tasks where they're not the same shape every single time.

我们稍微谈到了可靠性、速度、token 使用和效率。你们在持续改进发布的产品时，如何平衡所有这些方面？

We talked a little bit about reliability, speed, token usage, and efficiency. How do you guys think about balancing all these things as you keep making improvements on what you're releasing?

我认为我们在 OpenAI 对计算机使用功能非常兴奋，我们投入了很多。我们已经达到了一个我认为我们拥有最快、最便宜、最可靠的计算机使用产品的地步。而且我认为我们实际上还可以走得更远。我们花了很多时间研究速度。速度真的很有趣，因为一方面，计算机使用的力量在于它可以在后台执行任务。所以在某些情况下，花费多长时间并不重要，因为你可以继续做其他事情，而它为你完成任务。但有时你也想看着它以确保它正确执行。你需要等待它，因为它需要在其他事情发生之前完成。所以速度非常重要，因为你不希望有这种感觉：哦，我本可以自己做的，那样会容易得多。所以我们现在已经把计算机使用功能提升到了比普通人使用计算机更快的水平。我认为我们正接近超人的水平，即比专家计算机用户（比如我们桌上的某个人）更快。我认为当那发生时，将是一个巨大的变革。我认为这将改变我们所有人与计算机交互的方式，因为突然之间，任何你可以手动做的事情，它都可以更快地完成。所以除非你出于某种原因真的想自己做——我认为人们可能有很多理由想自己做——但在默认情况下，你可能只是想让智能体为你做。我认为这将让我们构建各种酷炫的用户体验。所以速度是我们经常考虑的事情。然后显然，我们的策略是提供多个不同质量、不同价位的模型。所以对于不同的用例，你可能想使用不同的模型。同时，我们较便宜的模型在计算机使用方面随着时间的推移变得越来越好。所以我认为随着时间的推移，人们想要用于哪些任务的模型实际上会有更多变化。

I think we're really excited about computer use at OpenAI and we've invested in it a lot. And we've gotten to a place where I think we have the fastest, cheapest, and most reliable computer use product. And I think there's actually so much further we can take that. We spent a lot of time looking at speed. You know, speed is really interesting because on one hand, the power of computer use is it can do tasks in the background. So in some cases it doesn't matter how long it takes because you can be going and doing other things while it's accomplishing the task for you. But also sometimes you want to watch it to make sure it does it correctly. You need to wait for it because it needs to complete this before something else can happen. And so speed is so important because you don't want to get this feeling like, oh, I could have just done that myself and it would have been so much easier. So we've gotten computer use now to a place where it's faster than the average human at using a computer. And I think we're getting really close to a place where it is superhuman, meaning it is faster than an expert computer user like probably one of us at this table. And I think that's going to actually be a sea change when that happens. I think it's going to change the way we all interact with computers because all of a sudden anything that you could do manually, it could do faster. So unless you really want to do it yourself for some reason, which I think there may be many reasons people would want to do that, but in a default case, you probably just want the agent to do it for you. And I think that's going to let us build all kinds of cool user experiences. So speed is something we think about a lot. And then obviously we have a strategy of offering multiple different models at multiple different qualities and price points. And so for different use cases you may want to use a different model. At the same time, our less expensive models are getting much better at computer use over time. And so I think there will actually be more variation over time in which models people want to use for which tasks.

补充一下，我认为另一个想法是：我们如何让客户控制安全性、可靠性、自主性与速度之间的平衡？对吧？比如，当模型很慢时，你可以看着它做事，但然后你受限于你能做的任务，因为你不能只是坐在这里看着它做——实际上效率更低，因为它比你慢。但一旦你达到一个无法再观察和审计它所做的每一件事的地步，我们如何让特别是企业和商业用户能够控制其中有多少人工参与，以及可审计性？对吧？就像几个月前——这是 Astra 的一部分——几个月前，这是 Astra 的一部分，你知道，我们推出了一套企业控制功能，让企业选择：第一，有多少人工参与？比如他们是否希望智能体自动批准一切？我们有一个非常棒的系统叫做 Guardian，对吧？或者自动——我认为应用中的自动批准——那是一个非常聪明的模型，它会查看计算机使用智能体采取的每一个动作，以决定它是否安全、是否对齐，所以没有意图。消费者可能会说，我只想让它做一切，我不想被要求批准。而企业则能够决定哪些类型的动作实际通过。所以我认为，随着我们的计算智能体在做事方面也变得超人，控制部分变得越来越重要。

Adding on, I think another thought is: how do we give customers control over safety, reliability, autonomy versus speed? Right? Like if you have—when the model is slow, you can kind of watch it do stuff, but then you're limited to that task that you can do because you can't just sit here watching the thing doing it—like it's actually less effective because it's slower than you. But once you can get to a point where you can no longer observe and audit everything it's doing, how do we give especially enterprises and businesses the ability to control the amount of human in the loop there is and then the auditability? Right? Like I think a few months ago—it's part of Astra—a few months ago it's part of Astra, you know, we kind of launched a set of enterprise controls that let businesses choose: one, how much human in the loop is there? Like do they want the agent to automatically approve everything? We have this really awesome system called Guardian, right? Or auto—I think auto approve in the app—that is a really smart model looking at every single action that the computer use agent is taking to decide if it's safe, if it's aligned, so there's no intent. A consumer might be like, I just want it to do everything and I don't want to be asked for approval. Whereas an enterprise is able to decide which type of actions actually go through. And so I think the control part becomes more and more important as our computing agents become also superhuman at doing things.

完全同意。我认为当你尝试推出更多企业用例时，雇主们——他们对效率感到兴奋，但也警惕着什么东西被释放出来了。

Totally. I think as you try to roll out more enterprise use cases, I think employers—they're excited about the efficiency, but they're also wary of what is sort of being unleashed.

也许甚至从另一面看，你们有没有看到应用软件那边的反应——好吧，现在有使用计算机的智能体在访问我的遗留工作流软件了？你们怎么看待这个问题？你们是更专注于被消费者或企业用户使用的那个智能体吗？

Maybe even on the other side, have you guys seen the response from application software around, okay, now there's computer-using agents accessing my legacy workflow software? And how do you guys think at all about that? Are you more focused on just the agent that's being used by the consumer or enterprise user?

是的，我觉得这个问题在开放网络上其实变得非常有意思。当它是在公司批准并使用的软件内部时，其实更直接，因为就像，哦对，100% 我们希望让人们在这里自动化完成工作，而不必自己枯燥地做。但说到互联网，我觉得大概有几类不同的群体，对吧？一类是，我们的商业模式与让智能体使用我们的产品并不一致，对吧？你想想广告商、出版商，他们的商业模式完全依赖广告。如果智能体替他们使用网站，那广告收入从哪来？而且我认为在大多数地方，智能体的广告展示实际上并不计入真实收入。所以这是一个我们一直在处理、并与合作伙伴团队一起解决的有趣问题。还有另一个问题，对于那些真的想拥抱计算机使用智能体访问其网站的公司，很多公司正在采用让人类和智能体能够一起使用的标准。比如 WebMCP 就是一个我们非常兴奋的标准。我们在应用浏览器中添加了对它的支持。所以你可以想象这些场景：智能体提升了使用网站的体验，但并没有完全取代它。所以我觉得，在我看来浏览互联网仍然很有趣，对吧？我仍然喜欢购物。我仍然喜欢看所有的酒店、航班和地点。所以你可以想象这样的场景：智能体做所有枯燥的工作，但它开辟出一种用户体验，让你来浏览，比如“好，现在我想完成交易，然后输入我的信用卡，做所有这些事，然后你可以把它交还给智能体。”所以 WebMCP 是一种有趣的方式，你与网站的界面仍然是网站本身，而不是 API，但然后人们是一起使用它。所以这些是一些想法。

Yeah, I think this problem becomes really interesting on the open web actually. I think when it's within software that is approved and used by the company, it's actually more straightforward because it's like, oh yeah, 100% we want to have people automate and do work in here that they don't have to do tediously themselves. But when it comes to the internet, I think you kind of have a few different camps of folks, right? One is, our business model is not in line with having agents use our product, right? You think about advertisers, publishers whose business model is entirely on advertisement. If the agents are using the website for them, then where's the ad revenue coming from? And I think in most places an agent's ad impression does not actually count towards real revenue. And so that's an interesting question that we're always working through and working with our partnerships team on. There's another question around, for the companies who do really want to embrace having computer-use agents using their sites, a lot of them are adopting standards that let humans and agents be able to use it together. So like WebMCP is a standard that we're really excited about. We added support for it in the app browser. And so you can imagine these scenarios where the agent is elevating the experience of using the site, but it's not entirely replacing it. So like, I think browsing the internet is still very fun in my opinion, right? I still like doing shopping. I still like looking at all the hotels and flights and places. And so you can imagine places where the agent does all the tedious work, but it kind of carves out a user experience where you're doing the browsing, like, "Okay, now I want to finish the transaction and then put in my credit card and do all this stuff, then you can hand it back to the agent." And so WebMCP is kind of an interesting way where your interface with the website is still the site itself versus an API, but then the person is kind of using it together. And so those are some thoughts.

是的。换个话题，我们想稍微深入探讨一下完成工作或端到端完成工作的含义。我们很想听听你们的一些看法，关于，你知道，什么效果好。你们如何定义合适的狭窄可重复任务，以及你们认为哪些类型的任务已经可以交给智能体，而哪些类型则像，好吧，它仍然需要用户这边做一些检查和纠正？

Yeah. Shifting gears, we want to double click a little bit on what it means to do the job or get the end-to-end job done. We'd love to hear a little bit from you guys around, you know, what works well. How do you guys define the right narrow repeatable task, and what sorts of tasks do you think are ready to hand over to the agent versus the ones like, okay, it still requires some checking and correcting on the user side?

我的意思是，总的来说，我发现这个模型现在相当有能力做你在计算机上能做的任何事情。你知道，我试图强迫自己做的一个练习是，每当我有一个任务，我首先想的是我能不能直接让智能体来做。现在显然有一些敏感的地方，比如如果我要给别人发消息，我希望感觉它来自我。但总的来说，我的思考方式是，现在使用计算机的智能体，自从 Astra 以来，已经能够做人能做的任何事情。以前有一些问题，比如在具体性和需求方面有些模糊。我通常发现，模型在当你要求它做某事并且有些模糊时，它实际上会，你知道，有了 Astra，我想我们推出了这种灵活输入，对吧，所以它实际上会问你，嘿，我想要一些细节，如果你不及时回答，它就会继续，然后拉取必要的上下文。我认为有些事情，比如如果你有可重复的工作流，最好能够尽早定义需求，类似于写产品规格。但总的来说，我通常觉得除非有非常敏感的事情你想参与其中，否则我现在做任何事情都直接找计算机模型。我很好奇你是否也一样。

I mean generally I've found the model to be pretty capable to do anything that you can do on a computer now. And you know one exercise that I try to force myself to do is that whenever I have a task the first thing I think about is just can I just have the agent do it. Now there are obviously some sensitivities around, you know, if I'm sending a message to someone I want to feel like it's coming from me. But in general I think the way I think about it is the computer-using agents now since Astra is now capable of doing anything that a person can do. And there are questions around, I think before of like oh there are some like squishiness around like specificity and requirements. I've generally found that the model is getting pretty good at whenever you're asking it to do something and there's some vagueness it will actually, you know, with Astra I think we launched this like flexible input right and so it will actually ask you about hey like I want some details and if you don't answer in time it'll just keep going and then pull the context that's necessary. I think there are some things around, you know, if you have repeatable workflows it's nice to be able to kind of define, you know, the requirements early, similar to like writing a product spec. But in general, I generally feel like unless there is something incredibly sensitive that you want to be in the loop for, I just reach for the computer model for everything I do now. I'm curious if the same for you.

是的，确实如此。我正在努力培养那种直接去用它的冲动，因为，你知道，它太好了，有时候很容易陷入旧的做法。我认为计算机使用作为一种技术，真正酷的地方在于它的普遍性。你知道，它适用于每一个应用程序。它不像我们构建的任何其他东西，开发者必须做一些工作来采用它。有人必须构建一个连接器 MCP 服务器。你知道，计算机使用适用于一切。它通过像人类一样使用应用程序来工作。我认为如果你没有花太多时间在这项技术上，这是值得理解的。它就像计算机使用像人类一样使用计算机。它让智能体拥有你所有的能力。我认为这非常强大。

Yeah, it is. I'm trying to build that impulse to reach for it because, you know, it's so good and sometimes it's so easy to get locked into the old way of doing things. I think the really cool thing about computer use as a technology is the universality of it. You know it works with every application. It's unlike any other thing that we've built where you know the developer has to do some work to adopt it. Someone has to build a connector MCP server. You know computer use works with everything. And it works by using the application the same way a human uses it. Which I think is something worth understanding if you haven't spent much time with this technology. It like computer use uses computers the same way a human does. It makes the agent have all the capabilities you do. Which I think is so powerful.

是的。我想补充 Ari 说的，部分原因是我觉得随着时间的推移，我已经不再把它想成我是在用它们做计算机使用任务还是非计算机使用任务。它们经常是相连的，对吧？就像我们对计算机使用的思考方式是，计算机使用不是你拿起就说我想用它做点什么的东西。它实际上是模型必须不惜一切代价去做的通用 API。所以如果你有一个雄心勃勃的任务，你想让智能体去做，很可能它越复杂，就越可能在那段旅程的某个时刻触及计算机使用。所以也许唯一我不使用计算机使用的情况是，如果任务太简单，我直接去点两个按钮做起来更容易。但你知道，随着计算机使用变得更快，甚至更快，那么问题就变成也许我不需要那样做，我只需用语音驱动，它就和我一起使用计算机。所以这些也是一些有趣的想法。

Yeah. And I guess adding on to what Ari said, part of it is I think over time I've stopped thinking about it as am I using them on a computer-use task or a non-computer-use task. Frequently they're connected, right? Like the way we think about computer use is that computer use is not a thing you pick up and just say I want to do something with it. It's really the universal API that the model has to do whatever it takes. And so if you have an ambitious task and you want the agent to do it, likely the more complex it becomes, the more likely it will probably touch computer use at some point in that journey. And so maybe the only time that I don't reach for computer use for some cases is if the task is so simple that it's easier for me to just go and click two buttons and do it. But you know as computer use gets faster, even faster, then the question becomes maybe I don't need to do that, I just drive voice and it just uses the computer with me. So those are some interesting thoughts there too.

是的。你们觉得有什么需要与企业用户分享，让他们对此感到放心吗？显然，你们对计算机使用完全着迷，但你知道，我认为企业用户有时显然会犹豫，比如，好吧，是的，人类会点击完成它，可能做得更差，但那种感觉是，有问责制，而且，你知道，那个人受过训练，以及所有经典的反对意见。

Yeah. Do you think there's anything you have to share with an enterprise user to get them comfortable with that? Obviously, you guys are totally pilled on computer use, but you know, I think there's obviously hesitance sometimes on enterprise user of like, okay, well, yes, the human would click through it and probably do it worse, but there is that feeling of like there's accountability and, you know, that person's trained and all the classic rebuttals.

我很好奇——得说服周围的人。

I'm curious—have to convince people around.

当然。最近和很多企业有过交流，我认为主要的一点是我们把它看作两层。第一层是我们如何创建确定性控制。我认为这真正给企业带来很多安全感——即使我们的模型是有史以来最对齐的，在防止提示注入等方面比大多数人做得更好，因为智能如此聪明，但当企业管理员能够基本上说某个任务或某些数据源太敏感而不能使用时，就会有一种安全感。所以我们看到企业——我们推出了两件事。一是能够确定性地为你的智能体创建一个沙盒。所以能够说它可以使用某些应用，不能使用其他应用；它可以使用某些网站，不能使用其他网站。然后二是这种人在回路的部分——比如哪些类型的操作需要人工干预？所以我们可能会说，任何时候有下载和上传，必须有人批准。任何时候它访问新网站，必须有人批准。再说一次，这些都是企业可以调整的旋钮。所以这是一。然后二是向企业管理员展示我们在内部运行的所有评估，关于这些模型有多对齐，甚至与内部和第三方评估相比。我认为还有一些我们全面做的基础工作，只是为了帮助人们建立对这些类型产品的信任。我们知道它太新了。对人们来说，这是一种神奇有时是陌生的体验。所以我们投入了很多工作在产品设计上。所以如果你愿意，你总是可以看到光标执行任务，它以有趣的方式动画化，所以看起来不无聊。然后就像 Jame 说的，这对企业非常可定制，但对普通用户也是如此。在使用新应用或网站之前，我们会征求你的许可，取决于你处于哪种模式。我认为这有助于人们不觉得它会在他们的电脑上做一些失控的事情。你有点想要有这种建立信任的时刻，你弄清楚它感觉如何，你想让它为你做什么，然后随着时间的推移你会感觉更舒服。

For sure. Lots of conversation with enterprises over the last little bit, and I think the main thing is we think about it as two layers, right? The first layer is like how do we create deterministic controls. I think that's really what gives enterprises a lot of feeling of safety—when they can, even though our model is the most aligned ever and it's better than most people at protecting against prompt injections and things because the intelligence is so smart, I think there is just a feeling of safety when the enterprise admins can basically say for a certain task or for certain data sources, it's just too sensitive to be used. So we see kind of enterprise—so we launched two things, right? One is the ability to deterministically create a sandbox around your agent. So being able to say it can use certain apps, it can't use other apps; it can use certain websites and not other websites. And then two is this kind of part of a human in the loop—like what type of actions does it require a human to be intervened? And so we might say, okay, for anytime there's a download and upload, a person has to approve it. Anytime it goes to a new website, a person has to approve it. Now, again, these are all like dials an enterprise can go through. And so that's one. And then two is really just showing the enterprise admins all the evals that we've run internally around how aligned these models are, even compared to both internal and third-party evals. I think there's also some baseline stuff that we do across the board to just try to help people build trust with these kinds of products. We know it's so new. It's such a magical and sometimes alien experience for people. And so we put a lot of work into the design of the product. So you can always see the cursor if you want to watch it perform a task, and it animates in a fun way so it's not boring to watch. And then like Jame was saying, this is very customizable for enterprises, but also for a regular user. We ask you for permission before using a new application or website, depending on which modes you're in. And I think that helps people not feel like it's going to go do some runaway stuff on their computer. You kind of want to have this trust-building moment where you figure out what it feels like and what you want it to do for you, and then you sort of feel more comfortable over time.

Codex 的光标太可爱了。我喜欢它——这是一个小细节，但看着它使用电脑给你信心，让体验变得美好。所以我认为有一个用户体验点绝对值得投资。我想更广泛地好奇一下：好吧，现在我们有能力的模型，它们运行成本越来越低，可以大规模运行。在你看来，公司应该考虑自动化什么？你认为企业内最有前景的用例在哪里？有一些工作流程吗？也许围绕 ERP，也许围绕人类和智能体如何合作完成任务。但如果你必须选择你最兴奋看到你的模型在世界上表现的地方，比如——

The Codex cursor is so cute. I love it—it's such a small detail, but watching it use the computer gives you confidence and it makes the experience nice. So I think there's a user experience point that's definitely worth investing in. And I guess I'm curious in a broader sense: okay, now we have capable models, they're getting cheap enough to run at scale. In your view, what should companies think of automating? Where do you see the most promising use cases within enterprises? Are there some workflows? Maybe something around the ERP, maybe something around how humans and agents can work together to get tasks done. But if you had to pick where you are most excited to see your models perform in the world, like—

是的，我想从我的角度，我很想听听我们的。就像两大类别。第一类是自动化繁琐的工作流程。那是比较无聊的一个，我会给一些无聊的例子，比如——

Yeah, I guess from my perspective, I'd love to hear ours. It's like two big categories. The first category is automating tedious workflows. That's the more boring one and I'll give some boring examples like—

无聊即机会。

Boring as opportunity.

机会。是的，没错。比如 ERP 就是一个很大的。所有你能想到的缩写——ERP、CRM，

Opportunity. Yeah, exactly. Like ERPs is a really big one. And so all the acronyms you can think of—ERPs, CRM,

所有记录系统。

All the systems of record.

是的。每一个记录系统，对吧？我们见过人们——你知道采购软件——比如你去供应商那里，每年你订购一堆实验室设备，而不是有一个巨大的电子表格，比如一千行，然后运营团队——所以公司通常的做法是，你有一个巨大的电子表格，由团队自己填写。然后运营团队拿那个巨大的电子表格，放入内部采购软件，然后它被最终确定为一堆行，然后销售团队——然后采购团队——拿那个巨大的软件输出，然后出去联系供应商。计算机可以自动化所有这些，对吧？所以采购是我们看到的一个大领域。财务是我们看到的另一个非常有趣的类别。公司内有很多不同的账本，对吧？所以你有费用账本，所有这些东西的账本，所以月底有很多对账，很多这些对账财务软件没有 API 能够快速拉取东西。所以财务团队能够使用计算机使用查看所有这些账本，对账，提取任何有趣的东西，要么标记给人，要么做一些调查。也是一个非常无聊但有用的工作流程。对于我们的销售团队，很多销售团队使用很多不同的软件来处理转录、获取数据，所有这些。所以让计算机使用去做 prospecting，然后放入 CRM 或某种外联软件等等。所以所有记录系统是无聊的部分。然后市场研究是我们经常看到的另一个。在我的空闲时间，我一直在为 Bmon 构建这个视频注释软件。我从 Bmon 中提取，我喜欢拍摄自己玩游戏，然后让智能体去看东西，我想好吧，我如何让这个东西真正有用——有一些软件用于匹克球、壁球、网球——所以我只是让计算机使用去登录，在这些软件上创建账户，尝试它，截图，给我反馈，然后尝试将其构建回应用。所以这些都是可以完成的繁琐工作流程，然后我认为一些更有趣的创造性工作，即实现更雄心勃勃的工作，就像，你知道，我们谈过模型现在在视觉设计方面非常擅长。

Yeah. Every single system of record, right? And so we've seen folks—you know procurement software—like you're going to vendors and every year you're ordering a bunch of lab equipment instead of having this giant spreadsheet with like a thousand rows that then the ops team—so it's like work companies normally what they have done is you have this giant spreadsheet that gets filled out by the teams themselves. Then an ops team takes that giant spreadsheet, puts it into an internal procurement software, and then that gets finalized as a bunch of lines, and then a sales team—then procurement team—takes that giant software output and then goes out and reaches out to vendors. Computers can automate all of that, right? And so procurement is a big one we see. Finance is another really interesting category we've seen. And so there's a lot of different ledgers within a company, right? And so you have ledgers for expenses, ledgers for all these things, and so at the end of the month there's a lot of reconciliation and a lot of these recon financial software don't have APIs to be able to quickly pull things. And so the finance team being able to have computer use look through all these ledgers, reconcile, pull out anything that's interesting to flag either to a person or then do some investigation. Also a very boring but useful workflow. For our sales team, a lot of sales teams use a lot of different software for transcripts, for getting data, all these kind of things. And so having computer use go and do prospecting and then putting that into a CRM or some kind of outreach software and then all that. So all the systems of record is the boring part. And then market research is another one that we see a lot. In my free time I've been building this video annotation software for Bmon. I pull out of Bmon and I like to film myself playing the game and then having the agent go and look at stuff and I was like okay how do I make this thing really useful—there's some software out there for pickleball, for squash, for tennis—and so I just had computer use go log in, create an account on these softwares, try it, take screenshots, give me feedback, and then try to build it back into the app. And so these are all kind of tedious workflows that could be done, and then I think some of the more interesting creative work of enabling more ambitious work is like, you know, we talked about the model now being really good with visual design.

所以电路设计、架构工作、硬件工程、游戏设计——这些也是我认为我们非常兴奋的领域，Astra 可以涉足。企业现在不太常考虑这些，因为他们在想那些无聊的机会。但我认为随着人们更了解模型的能力，接下来我们会看到更多使用。但好奇你们有什么想法。

And so circuit design, architecture work, hardware engineering, game design—these are also areas that I think we're really excited about, where Astra can take. Less common for enterprises to think about those right now because they're thinking about the boring opportunities. But I think as folks understand the capabilities of the model more, that's where we'll see more usage next. But curious if any thoughts.

是的，我觉得在消费者端也类似，有很多事情是自动化苦差事，或者只是做你不想做的事情。它帮你填写 DMV 表格。它帮你从收件箱里的一堆邮件中退订。

Yeah, I think kind of similarly on the consumer side, there's a lot of things where it's sort of automating drudgery or just doing things you don't want to do. It's filling out DMV forms for you. It's helping you unsubscribe from a bunch of emails in your inbox.

然后我觉得还有更多类似的事情——我听到很多人的故事，人们做他们本来没有时间或不会想到去做的事情。比如，我不知道，就我个人而言，我在维护关于财务的电子表格和关于膳食计划的电子表格。我有一个朋友是私人教练，他为客户制作定制的东西。其中一些事情在个人生活中会非常耗时。但这让事情变得如此容易。我用 Apple Numbers，这是一个比较冷门的电子表格程序，但模型仍然能搞定。

And then I think there's also things that are more like—I'm hearing from a lot of people stories of people doing things that they just wouldn't have had time or wouldn't have thought to do otherwise. Like, I don't know, for me personally, I'm maintaining spreadsheets about my finance and spreadsheets about meal planning. And I have a friend who's a personal trainer who's making customized things for his clients. And some of these things would just be so time prohibitive to do in your personal life. But this makes it so easy. I use Apple Numbers, which is like an obscure spreadsheet program, but the model still knocks it out.

我想我从未听说过有人用 Apple Numbers。

I don't think I've ever heard someone use Apple Numbers.

这是一个出奇地好的电子表格软件。然后我觉得还有第三类事情，我最为兴奋，才刚刚开始，我认为即将到来，那就是真正拥有一种不同的方式来使用整个电脑或与软件交互。我想你在我们的 Astra 广告中看到了非常酷的例子，人们只是用语音与电脑交谈。这是使用电脑如此自然的方式。我认为随着时间的推移，我们会看到软件本身的形式因素因这项技术而改变，我对此非常兴奋。

It's a surprisingly great spreadsheet software. And then I think there's a third category of thing which I'm kind of the most excited about, which is just starting and I think on the horizon, which is really just having a different way to use your entire computer or interface with software. I think you saw in our Astra ad really cool examples of people just talking to their computer with their voice. It's such a natural way to use your computer. And I think over time we're going to see the form factor of software itself change due to this technology, and I'm really excited about that.

关于这一点——好吧，软件的形式因素在改变。想象一下，如果你是企业用户，或者你习惯使用 30 年前那种笨重的 ERP。有一种情况是，你现在释放一组智能体来操作那个软件，你并不真的在意。另一种情况是，新的 AI 原生玩家在使用计算机使用智能体。我很好奇，你认为世界会更倾向于保留传统玩家并在上面加一层，还是你认为计算机使用智能体绝对可以成为新软件玩家的一部分？这是我经常思考的事情，所以好奇你们的想法。

On that point—okay, form factor of software changing. If imagine if you're an enterprise user or you're used to using whatever clunky ERP from 30 years ago. There's a world in which you now unleash a set of agents that just operate that software and you don't really care. There's another world in which a new AI-native player is using computer-using agents. And I'm curious, do you think the world is going to go more in the direction of keep the legacy folks and just have this layer on top, or do you think computer-using agents can absolutely be part of new software players? And this is something I think a lot about, so I'm curious to get your guys' thoughts.

这是一个如此有趣的时刻，同时创建新软件的成本和复杂性大大降低，所以你无法想象软件变化如此之大，但世界上也存在如此多的传统软件。计算机使用非常适合与之交互。所以我预计这两件事都会持续很长时间。如果你看看今天在生产中的软件，很多是 10、20、30、40 年前写的，我不认为那些东西会一夜之间消失。我确实认为，在某种程度上，计算机使用的存在可能会给公司带来更多压力，使软件更好地与智能体协作，因为所有软件默认都会与智能体协作，我认为 webc 是一个很好的例子，我认为我们会看到更多类似的东西。所以是的，我有点预计两者都非常重要。但我确实认为它会随着时间的推移而转变。你怎么看，James？

It is such an interesting moment where simultaneously the cost and complexity of creating new software is so much lower, so you can't imagine software changing so much, but also there exists so much legacy software in the world. The computer use is so well suited to interact with. So I expect both of those things to continue for a long time. If you look at the software that's in production today, a lot of it was written 10, 20, 30, 40 years ago, and I don't think that stuff's going to go away overnight. I do think that in some way the existence of computer use may put more pressure on companies to make software work better with agents, because all the software is going to work with agents by default, and I think webc is a great example of this, and I think we'll see more stuff like that. So yeah, I sort of expect both to be super important. But I do think it'll kind of shift over time. What do you think, James?

是的，我同意你。我觉得就像——有很多艺术和工艺，你理解这一点，因为 Ari 和他的团队打造了美丽的产品。Cursor 全力以赴。所以我认为构建人们可以拿在手里的伟大产品需要很多艺术和工艺。对我来说，如果互联网变成这个没有面孔的东西，只是智能体互相交谈，人们不再使用任何东西，我会非常难过。我不认为会那样，因为手工艺和手工软件有一种让人使用起来感觉很好的东西，因为设计它的人真正理解客户想要体验什么。比如我妻子最近开始玩这个棋盘游戏。我忘了名字。就是连接五个石子来赢。我想名字大概是 Omach。总之，她试图玩得好，因为她在网上和朋友玩，她被虐得很惨。所以我说，‘好吧，我要给你建一个战术训练器，让你练习不同的开局策略，这样你可以变得更好。’例如，我创建了第一个版本。我说，‘哦，我知道她喜欢以某种方式做。她喜欢这些类型的学习方式胜过其他。’所以我能够把软件打造得对她来说感觉非常好。这种个性化软件，我知道今天可能仍然感觉只有真正的技术人会做。但我认为如果你考虑几年后，如果任何人都能构建任何东西，他们可以在很短的时间内为小众用例构建这些真正有工艺的软件，这些软件如此适合人们，我认为那是一个美丽的世界。再说，在她的训练结束时，她学到的教训和犯的错误，可以是一个智能体总结并以文本格式给她，但当她真正与 AI 一起玩并试图变得更好时，那是人必须做的事情。所以我确实认为——我乐观地认为 UI 不会消失。我觉得这有点偏离你原来的问题，但我确实认为廉价软件的压力会使软件必须变得更好。但我不认为会只是这样：是的，一切都会有 MCP 和 API，你知道计算永远不会有用。

Yeah, I agree with you. I think it's like—there's a lot of arts and craft, and you understand this because Ari built beautiful product with his team. The cursor was all in. And so I think there's a lot of art and craft into building great products that people can hold in their hand. And for me, if the internet becomes just this faceless thing that is just agents talking to each other and people are not using anything anymore, I would be very sad. And I don't think it will be like that because there's something about craftsmanship and artisanal software that just feels nice to use because the person who designed it really understands the experience that the customer wants to have. So like my wife recently got into playing this board game. I forgot the name of it. It's where you connect five pebbles to win. I think it's like Omach is the name of it. And so anyway, she's trying to get good at it because she's playing online with friends and it's getting her ass kicked. And so I was like, 'All right, I'm going to build you a tactical trainer for going through different gambits so you can get better at it.' That's for example something that I created a first version. I was like, 'Oh, I know she likes to do it in a certain way. She likes these type of ways to learn more than others.' And so I was able to craft the software to be something that feels really good for her. And these kind of personalized software, I know it's maybe today still feels like only really tech people will do. But I think if you think about this a few years out, it's like if anybody can build anything and they can build these really craft software in a very short amount of time for niche use cases that are so well tailored for people, I think that's a beautiful world. And again, at the end of her training, the lessons and mistakes she made, that could be an agent that summarizes it and gives her in text format, but when she's actually playing with the AI and trying to get better, that's something that the person has to do. So I do think it's—I'm bullish that I don't think UI is going away. And I think kind of a little tangent to your original question, but I do think that the pressure of cheap software will make it so that software has to be better. But I don't think it will just be like yeah everything will have an MCP and an API and you know compute will never be useful.

我非常喜欢这个观点。我认为人类天生就习惯于直接操作。你真的希望能够用自己的手移动和做事。

I love this point so much. I think that humans are so hardwired for direct manipulation. You really want to be able to move and do things with your own hands.

我觉得思考过去软件如何一刀切，以及不得不如此，真的很令人兴奋。我们每天交互的很多软件都是为数百万人、数千万人、数亿人、数十亿人使用而设计的。我认为我们将进入一个时代，任何软件都可以为一百个人或一个人设计，那种软件将更加个性化和独特。

I think it's really exciting to think about how software in the past was so one-size-fits-all and had to be. A lot of the software we interact with every day is designed for millions, tens of millions, hundreds of millions, billions of people to use. And I think we will enter an era where any piece of software could be designed for a hundred people or one person, and just how much more personal and unique that kind of software will be.

是的。我的心理模型是，大概有两个类别。一个是消费端，你要么从中获得乐趣，喜欢使用或学习。不是它的实用性，而是互动或学习的乐趣，你愿意花时间在上面。在 B2B 端，你可能希望你的智能体完成一堆繁琐的工作，但也有审计日志的概念，你可以个性化你想要的交互方式，但产品实际上是我作为人类用户设定目标，然后能够理解逻辑，但不一定需要跑完整个命令链来了解它是如何发生的。所以任何 B2B 用户仍然需要弄清楚，好吧，为什么那笔付款被支付了，然后回去并获取发生情况的摘要，并能够说，好吧，在这个时间支付给这个人，是的，30 天后，应该应用这么多折扣等等。他们希望智能体负责执行，但他们希望在某种程度上，仍然需要放入一些控制，我认为这确实需要某种 UI。而且它必须是，我认为如果我们越来越相信 AI 会做更多这样的事情，那部分也必须使用起来令人愉快，对吧？为什么不呢？我们为什么要满足于糟糕的软件？

Yeah. My mental model is I think there's probably two buckets. One is on the consumer side you've got either something you just get joy out of, you enjoy using or you learn from. It's not the utility of it. It's the joy of interacting or learning or whatever it is and you want to spend time on it. On the B2B side, you may want your agent to complete a bunch of the tedious work, but there's also this concept of like the audit log and you could be personalized in how you want to interact, but the product really is me as the human user setting the goal and then being able to understand the logic but not necessarily need to run through the entire chain of command of how it happened. And so any B2B user still needs to figure out, okay, why did that payment get made and go back in and get a summary of what happened and be able to say, okay, pay this person at this time, yeah, at 30 days, this much discount should be applied, etc. They want the agent to take care of the execution of it but they want to be able to at some level, some level of control still needs to be put in there and I think that's something that does require a UI of some sort. And it has to be, I think if we're moving more towards believing that AI will do more of this, that part has to be enjoyable to use too, right? Like why not? Like why should we settle for shitty software?

我确实认为，如果我们能看到新型软件，我会很乐观，这些软件有传统的用户界面，但底层也使用智能体。然后也许有时这些智能体本身使用计算机使用来自动化遗留软件，通过给你一些更好、更专用的体验。我对这一类非常兴奋。

I do think it'll be really interesting to see if we, and I'm optimistic that we will, see new kinds of software where it has traditional user interfaces but is also using agents under the hood. And then maybe sometimes those agents are themselves using computer use to automate legacy software by giving you some much nicer more purpose-built experience. I'm really excited for that class.

是的，这有点像我们在内部喜欢使用的共享智能层，对吧？就像你的智能体在其他地方使用遗留软件并在其他地方做事。如果你在使用一个应用程序，该应用程序想要访问那个上下文，以便能够提供更好的体验。我认为如果每个网站都有自己的智能体，而不与你的智能体对话，那是一个非常碎片化的世界。因此，弄清楚如何协同工作，让智能体可以插入并使用你的其他软件，我认为这将在未来几个月内非常有趣地展开。

Yeah, it's kind of like the one where we like to use internally like the shared intelligence layer, right? Like you have your agent using legacy software somewhere else and doing things other places. You like if you're using an application, the application wants to get access to that context so it's able to deliver a better experience. It's like I think if we had every website has its own agent that doesn't talk to your agent, it's a very fragmented world. And so figuring out like how that plays together where the agent can plug in and use your other software, I think it's going to be a really interesting thing to play out over the next few months.

我喜欢你们谈论这个，因为我认为人们大大低估了人类与 AI 交互的重要性，尤其是如果你想让产品成功。所以要考虑人类习惯如何协作。你希望任何智能体都能融入团队。我只是好奇，如果你把这个应用到一家新初创公司，假设你明天要创办一家公司，你想使用计算机使用来解决企业内的一个工作流程。你会怎么考虑？你会，哪些事情会浮现在脑海中，比如哦，我想知道它什么时候失败。我希望它能够与人沟通，或者我希望它像一个，我不知道，一个黑盒，就像如果你现在必须与创始人交谈，当他们尝试使用你的模型和能力来解决企业问题时，你会给他们什么建议？

I love that you guys talk about this because I think it's so underappreciated, the importance of human-AI interaction, especially if you want a product to be successful. So being thoughtful of how humans are used to collaborate. So you want any agent to be able to fit into the team. I'm just curious, if you apply that to a new startup, let's say you're starting a company tomorrow and you want to use computer use to solve a workflow within the enterprise. How would you think about it? And would you be, what would be the things that jump to mind in terms of, oh I want to know when it fails well. I want it to be able to communicate with people or I want it to be like a, I don't know, a black box just like maybe if you had to talk to founders right now, what advice would you give them as they try to use your models and capabilities to just solve problems for enterprises?

在我看来，这很大程度上取决于模态。你是一个小团队，希望团队中的个人能够以更高效的方式使用团队正在使用的某个软件？或者你正在构建的东西本身想要与现有软件交互，在这种情况下，也许你在云端的某个机器在后台进行计算机使用。我认为这会真正影响你希望人类与智能体协作的模态，对吧？在前一种情况下，我认为我们的很多产品都在服务这个用例，你坐在 ChatGPT 前，可以进行很多很棒的协作。在后一种情况下，也许你正在用 API 构建一些东西。你在做自己的计算机使用实现。然后真正有趣的是弄清楚你希望人类在什么时候参与进来？智能体是在出错时告诉你吗？是在成功时告诉你吗？是否有东西在总结，这样你就不会收到大量通知说‘嘿，发生了这个。发生了这个。发生了这个。’而你实际上会忽略它们。我不知道。你怎么看，James？

In my mind, it depends so much on the modality. Are you a small team and you want an individual on the team to be able to use some piece of software that your team is using but in a more efficient way? Or are you maybe building something that itself wants to interact with existing software, in which case maybe you have some machine in the cloud somewhere that's doing computer use in the background. I think that will really affect the modality of how you want the human-agent collaboration to be, right? And in the former case, I think a lot of our products are serving this use case where you sit down at ChatGPT and there's a lot of great collaboration you can have there. And in the other case, maybe you're building something with the API. You're doing your own computer use implementation. And then it really is interesting to figure out when do you want the human to be in the loop? Is the agent telling you when something's wrong? Is it telling you when something went right? Is something summarizing that so that you don't get bombarded with notifications saying, 'Hey, this happened. This happened. This happened.' Which actually you're going to ignore. I don't know. What do you think, James?

是的，我的意思是，你知道，Ari 和我在加入 OpenAI 之前都是初创公司创始人，所以我仍然非常看好初创公司。我认为问题归结为是否还有护城河，对吧？在实验室做事情的同时，初创公司还能解决什么问题？总的来说，我认为有两件事。一是可构建的东西太多了，对吧？即使我们现在规模很大，有数千人，我们内部仍然感觉非常捉襟见肘，团队对于所有问题来说并不大。所以如果你有一个你真正深入理解的问题，你有行业专业知识，你了解监管如何运作、市场推广动作、如何服务那些客户，我认为那仍然是一个非常绿的领域。而且通过我们的 API 获得的工具，就像 Ari 提到的计算机使用，我认为仍然会让你加速。我认为归根结底，速度和支付是当今存在的模式。我经常思考的另一件事是，在我看来，仍然有很多垂直行业的 AI 扩散速度要慢得多。

Yeah, I mean, you know, both Ari and I were startup founders before we joined OpenAI and so I'm still very bullish on startups. I think the question kind of comes down to what is there still a moat, right? Is there something that startups can still tackle while labs are doing things? I mean in general I just think there's like two things. One is I think there's so many things to build, right? And even though we're quite big now, thousands of people, we are still feeling very stretched thin internally and the teams are not very big for all the problems. So if you have a problem that you really deeply understand, you have industry expertise, you understand how regulatory works, the go-to-market motion, how to service those customers, it is still I think a very green field out there. And the tools that you have through our APIs like Ari mentioned with computer use, I think we'll still let you accelerate and speed. I think at the end of the day speed and pay is kind of the mode that exists today. I think the other thing I think about a lot is like there are in my opinion still a lot of vertical industries where the diffusion of AI is much slower.

所以，你知道，我前几天和一家公司聊过，他们用 computer use 帮助餐饮集团拍摄后厨仓库的照片，然后利用视觉技术，根据对明天客流量的预测，看看会缺什么，接着用 computer use 去像 shopppee 这样的采购软件，为你订购第二天的食材。基于这些，它还会进入你的 ERP 系统，调整员工的排班表，因为去年这个时间点，由于假期等原因，客流量要大得多。所以那些餐饮集团可能不是科技公司，对现有工具并不那么熟悉。因此，我认为在旧金山，我们经常看到很多人想‘哇，实验室可以搞定一切，那些代码工具，为什么不是每个人都安装代码工具然后让它干活呢？’但对于普通美国人、中小型企业主、普通企业来说，甚至理解这些工具能做什么是一回事，以合理的方式连接工具是第二回事，部署它是第三回事，维护服务是第四回事。所以在我看来，还有很多事情要做，如果有复杂的工作流程需要大量不同的软件，computer use 是连接它们、完成整个预期流程的好方法，然后以让那些客户觉得手感好的格式和 UI 来构建，这仍然是一个非常大、非常大的机会，我认为这个机会仍然存在。

And so, you know, I was talking to a company the other day who uses computer use to help restaurant groups take pictures of their back storage, and then using that, the vision looks at what is going to be missing based on a projection of tomorrow's traffic, and then it takes that to then use computer use to go to a shopppee or some kind of procurement software to then order your groceries for the next day. And then based on that, it will go into your ERP to adjust the timetable for your employees who might need more people to help because in the last year at this time of the day there was a lot more traffic coming in from a holiday or something like that. And so those restaurant groups are probably not tech companies that are super familiar with all the tools available. And so I think living in San Francisco we see a lot of thinking 'wow, labs can do everything, the codecs, and why doesn't everybody just install codecs and just have it do stuff?' But for the average American and the average small to medium business owner, the average enterprise, even understanding what the tools are capable of is one, connecting tools in a way that makes sense is two, deploying it is three, servicing is four. And so there's still a lot, in my opinion, where if there are complex workflows that require a lot of disparate software, computer use is a great way to join them, complete the entire intended flow, and then building it in a format and UI that feels good to hold for those customers feels still like a really, really big opportunity that I think is still open.

请讲讲接下来有什么，你对前沿的哪些方面感到兴奋，可以分享关于 computer use 智能体的？

Tell us about what's next, what are you excited about that's on the frontier that you can share around computer using agents?

嗯，我已经大致提到过了，但我们非常兴奋于速度，我们非常兴奋于让 computer use 超越人类。我们还非常兴奋于融合不同种类接口的东西，将多模态输入和输出结合在一起。这样你可以有更流畅的计算机体验。我认为现在我们的 computer use 产品很多都非常注重后台。就像，让我给智能体一个任务，这样它可以在我在做其他事情时完成，这实际上是产品的一个巨大优势。但我认为未来我们还会看到更多这种你和智能体之间的实时协作。所以我对这个领域非常兴奋。除此之外，就是更好、更便宜、更可用、更易用。

Well, I've sort of gestured towards it already, but we're really excited about speed, we're really excited about making computer use superhuman. And we're really excited about things that blend together different kinds of interfaces that bring together multimodal inputs and outputs. So you can have more fluid experiences with your computer. I think right now a lot of our computer use product is very background focused. It's very much like, let me give the agent a task so it can do that while I do something else, which is actually a tremendous strength of the product. But I think that in the future we'll also see more of this real-time collaboration between you and the agent. So I'm really excited about this area. And otherwise just better, cheaper, more available, easier to use.

更多。是的，没错。能够真正访问一切。是的。

More. Yeah, exactly. Being able to actually access everything. Yeah.

我认为有，你知道，我们和 DOTS 一起，对吧，我们给人们这些云中的计算机。Tragic work 也有一个云中的浏览器。老实说，对我来说，我兴奋的一件事就是让体验感觉更连贯一些。就像今天你可以安装 Chrome 扩展，它使用你的密码管理器和 Chrome 中的所有东西，然后我们在桌面应用中有一个内置浏览器，它有一套不同的 cookie 和不同的密码。你的云中计算机也有不同的东西，对普通消费者来说这很令人困惑。就像为什么我有这四个不同的浏览器，它们彼此不通信？如果我在一个中打开一个标签页，它会在另一个中出现吗？所以我认为我兴奋的很多方面是，我们如何让 computer use 从这个被企业和真正硬核尖端用户使用的产品，变成人们真正能理解的东西。这有点矛盾，因为我认为 computer use 实际上是最容易理解的 AI 产品，因为它是 AI 在使用你的计算机，但由于你必须管理的不同状态和在不同地方使用它的复杂性，实现细节有时会渗透出来，然后它就变得有点难以理解。所以能够让它达到消费级、更容易理解、更连贯，是我下一季度非常兴奋的事情。

I think there is, you know, we with DOTS, right, we're giving folks these computer in the cloud. Tragic work also has a browser in the cloud. I think honestly, for me, one thing I'm excited about is just making the experience feel a little bit more coherent. Like today you can install the Chrome extension and it uses your password manager and everything in Chrome, and then we have an in-app browser in the desktop app which has a different set of cookies and different passwords. Your computer in the cloud also has different things, and for the average consumer it's confusing. It's like why do I have these four different browsers that all don't talk to each other? If I open up a tab in one, does it show up in the other? And so I think a lot of what I'm excited about is how do we make computer use go from this product that is used by enterprises and really hardcore tip-of-the-spear kind of users to something that's actually understandable for people. It's kind of like an oxymoron because I think computer use is actually the most easy to understand AI product because it's an AI using your computer, but then because of the complexities of the different states you have to manage and the different places you're using it, the implementation details sometimes seep through and then it becomes kind of hard to grok. And so being able to make that consumer grade and easier to understand, more coherent, is something I'm really excited about next quarter.

是的，我认为这回到了一个点：仍然是由人类来启动智能体，对吧？而且他们必须——UI 和控制越好，使用起来越简单、越愉快，他们从中得到的就越多。这绝对是原则。

Yeah, and I think it comes back to the point that there is still a human that's setting off the agent, right? And they have to—the better the UI and controls and the easier and more enjoyable it is to use, the more they're going to get out of it. And that's absolutely principle.

关于这一点，我还想提另一件事：我认为今天产品的另一个挑战是，你必须花点心思考虑如何使用它，你必须记得使用它，你必须说服自己去做与以往习惯不同的事情。我确实认为，嗯，DOTS，就像 James 之前谈到的，有这种非常强大的主动性，它们可以主动来找你，识别情况，帮助你解决问题。我认为我们还会看到更多的主动性。我非常兴奋于一个智能体能适应你而不是你适应它的世界。我认为我们会看到很多这样的情况，我认为这将使人们更容易真正与这项技术产生联系、理解和融入，同时仍然是由他们来设定和瞄准目标，如果这说得通的话。

On that point, one other thing I want to gesture towards is I think another challenge of the product today is you have to put some thought into how you're going to use it and you have to remember to use it and you have to convince yourself to do things that are different from how you're used to doing them. And I do think, um, dots, like James was talking about earlier, have this really powerful proactivity where they can sort of come to you and identify situations and help you solve problems. I think we'll also see a lot more proactivity. I'm really excited about a world where the agent can sort of adapt to you instead of you adapting to it. And I think that we're going to see a lot of that and I think it will make it a lot easier for people to actually relate to and understand and incorporate this technology while also still being the ones sort of setting and searing the goals, if that makes sense.

我们还没谈到的一件事是 computer use 背后的势头以及发生了多大变化。Fitzi 和我在我们的文章中聊过一点。很想请你谈谈它现在的情况，你知道，到目前为止的轨迹以及未来的方向。

One of the things we haven't talked about yet is the momentum behind computer use and how much has changed. Fitzi and I chatted a little bit about this in our piece. Would love for you to just kind of talk about where it's, you know, how the trajectories so far and where things are headed.

是的，我的意思是 James 和我在 OpenAI 研究 computer use 已经有一段时间了，过去六到九个月非常令人兴奋，对吧？因为我们开始研究这些 computer use 产品时，很少有人在使用它们。

Yeah, I mean James and I have been working on computer use for some time now at OpenAI and it's been an incredibly exciting past six to nine months, right? Because we started working on these computer use products and very few people were using them.

今年早些时候，由于研究团队的出色工作，模型突然变得非常擅长这类任务。然后我们一直与内部多个团队紧密合作，围绕这一点打造真正优秀的产品体验。就在过去几个月里，产品突然疯狂增长，人们真正看到了它的价值。但我认为这仍然非常早期。目前仍然只有一部分用户在使用计算机使用功能，我认为我们会看到它大幅增长。

The models all of a sudden became great at these kinds of tasks earlier this year because of amazing work the research team did. Then we've been working super hard with a bunch of teams internally on how to create actually good product experiences around this. Just in the past couple of months, all of a sudden the product's growing like crazy and people are really seeing the value in it. But I think it's still so early. It's still the case that only some portion of our user base is using computer use, and I think we're going to see it grow quite a bit.

是的，完全同意。我是今年一月加入 OpenAI 的，参与 Atlas 产品，我很喜欢 Atlas，愿它安息。但我们意识到的一件事是，在 Atlas 产品中，智能体体验实际上并不是一个大卖点，原因有几个。用的人不多，部分原因是大多数人把它当作普通浏览器。过去几个月，由于模型能力变得更强，我们现在已经把浏览器和浏览器使用功能整合到所有已有智能体使用场景的现有产品中。我们看到了每周 YC 级别的增长，并且总是对它的速度感到惊讶。我认为下一个大的拐点就像云一样，对吧？桌面端仍然很大程度上是高级用户的东西。我知道我们可能一直在用桌面端，但大多数消费者都在手机上，对吧？那么，你如何给每一个聊天用户，如何给每一个只是用手机处理日常生活的人一台强大的计算机，他们甚至不需要理解如何使用，它就能为他们把事情搞定。所以非常兴奋这如何将其带给十亿用户。

Yeah, totally. I mean, I joined OpenAI in January and I joined working on the Atlas product and loved Atlas, RIP Atlas. But one of the things that we realized was in the Atlas product, the agentic experience wasn't actually a big selling point for a few reasons. Not many people used it, partly because most people treated it like a normal browser. Over the last few months, because the model capabilities have gotten better, we've now incorporated browser and browser use into all of our existing products where there was already agentic usage. We've just seen YC-level growth week over week and are always surprised by how fast it can go. And I think this next big inflection point is like cloud, right? Desktop is still very much a power user thing. I know we probably use our desktop all the time, but most consumers are on their phones, right? So how do you give every single chat user, how do you give everybody who's just using their phone for their day-to-day life a powerful computer that they don't even need to understand how to use, that it just kind of makes things work for them. And so really excited about how that kind of brings it to billion users.

感觉计算机使用智能体背后有巨大的势头，现在正是时候：好的，我们如何让它进入更多人的手中？他们信任它，他们觉得他们真的可以使用它。你之前关于连贯体验的观点——那是真实存在的。我认为一旦那变得更好，消费者会感到更舒适地使用它，并且感觉更自然。

It feels like computer using agents have had a ton of momentum behind them and now it's the moment of like, okay, how do we get it into more people's hands? They trust it, they feel like they can really use it. And your point earlier about coherent experiences—that's a real thing. I think once that gets a lot better, consumers will feel more comfortable using it and it'll feel more natural.

我认为这是对的。我认为我们正处于计算机使用的 Scaling（规模扩张）时刻，同时面向消费者和企业。我非常兴奋地看到它会走向何方，以及人们会用它们构建什么。

I think that's right. I think we're at the scaling moment of computer use both for consumer and enterprise. At the same time. And I'm really excited to see where it goes and what people build with it.

很好。我们也是。

Nice. Us too.

好的。非常感谢你们加入我们。这太棒了。我们喜欢谈论这个话题，直接从你们那里听到真的很有趣。

All right. Well, thank you so much for joining us. This was awesome. We love talking about this topic and it was really fun to hear it from you guys directly.

谢谢。也非常有趣。

Thank you. A lot of fun as well.

有趣。是的。谢谢。

Fun. Yeah. Thanks.

谢谢。

Thank you.

谢谢你们。

Thank you guys.