当每个人口袋里都有一位健康专家时会发生什么？| OpenAI的Karan Singhal

我们在医疗保健上花了5.3万亿多美元。

5.3-plus trillion we spend on healthcare.

如果这算一个国家，它的GDP能排到世界第三，可我们还是面临大量资源短缺。

If it were its own country, it'd be the third-largest GDP, and yet we have a ton of scarcity.

OpenAI的使命是确保通用人工智能惠及全人类，让人们获得可信、充足、可靠的健康智能。

OpenAI's mission is to ensure AGI benefits all of humanity and give people access to trustworthy, abundant, reliable health intelligence.

最打动我的故事之一是Nathan Labenz的故事。

One of the stories that I was most touched by was the story of Nathan Labenz.

他儿子经历了一段漫长的波折，有一阵子他儿子非常疲惫、易怒，AI帮他们理清了整个过程，最后发现他儿子得了一种侵袭性白血病。

He basically had this long saga with his son where at some point, he was suffering a lot of fatigue and irritableness, and AI helped them navigate the process, and it turns out he was diagnosed with an aggressive form of leukemia.

你在心理健康方面做的有些工作，其实非常复杂。

Some of the work that you've done around mental health, there's a lot of complexity.

面对这么多可能性，你怎么引导用户呢？

How do you help guide a user on that, given that could mean so many things?

我们一直在和全球多位持证心理健康专家合作，思考如何在各种棘手场景中正确定义模型行为，并确保我们能达成共识，从而真正衡量表现。

We've just been working with a bunch of licensed mental health experts across the world to think about the right ways to define model behavior in a bunch of tricky scenarios and ensure that we can get to a level of consensus where we can actually measure performance.

我们来谈谈信任。

Let's talk about trust.

一个人可能正在跟大语言模型对话，真心希望听到一切都没问题，但从医学角度看，他们可能需要听到完全不同的答案。

A person can be speaking to an LLM, really hoping to hear that everything is okay, but from a medical perspective, they need to hear perhaps something very different.

你怎么教人工智能分辨这两者的区别？

How do you teach an AI to recognize the difference?

每周已经有超过3亿人在用ChatGPT咨询健康问题。

More than 300 million people are already using ChatGPT for health and wellness questions every week.

当这种AI变得主动、个性化，并且越来越融入你的整个医疗体系时，会发生什么？

What happens when that AI becomes proactive, personalized, and increasingly connected to the rest of your healthcare?

在本期节目中，a16z合伙人Jay Rughani与OpenAI健康AI研究负责人Karan Singhal对话，探讨公司让可信赖的健康智能广泛可及的愿景。

In this episode, a16z partner Jay Rughani sits down with Karan Singhal, who leads health AI research at OpenAI, to talk about the company's vision for making trustworthy health intelligence widely accessible.

Karan解释了为什么OpenAI将医学专业知识融入其通用模型，而不是单独构建一个医疗模型，以及如何与数百名医生合作评估这些系统，还有为什么目标不是取代医生。

Karan explains why OpenAI is training medical expertise into its general purpose models rather than building a separate healthcare model, how it works with hundreds of physicians to evaluate those systems, and why the goal isn't to replace doctors.

相反，他认为AI将成为安全网和四分卫，帮助患者了解自己的健康状况，帮助临床医生发现可能遗漏的问题，并最终在就诊之间处理更多碎片化的工作。

Instead, he sees AI becoming a safety net and a quarterback, helping patients understand their health, helping clinicians catch things they might otherwise miss, and eventually navigating more of the fragmented work between visits.

他们还会讨论当医疗智能变得极其便宜时会发生什么，AI如何缩短新医学知识进入临床实践所需的数年时间，以及医疗领域出现“第37步”的可能性——一个最初让人类专家感到惊讶，但最终改变我们对疾病理解和治疗方式的洞见。

They also get into what happens as medical intelligence becomes dramatically cheaper, how AI could shorten the years it can take for new medical knowledge to reach clinical practice, and the possibility of a healthcare Move 37, an insight that initially surprises human experts but ultimately changes how we understand or treat disease.

大家好。

Hey, everyone.

欢迎回到播客。

Welcome back to the podcast.

我是Jay Rughani，a16z的合伙人，负责投资医疗和生物技术公司。

I'm Jay Rughani, a partner at a16z, leading investments in healthcare and biotechnology companies.

今天和我一起的是来自OpenAI的Karan Singhal。

I'm joined by Karan Singhal of OpenAI.

他负责OpenAI的健康AI研究，我们有很多话题要聊。

He leads health AI research at OpenAI, and we've got a lot to talk about.

感谢你的到来。

Thanks for being here.

谢谢你的邀请。

Thanks for having me.

我想从一个大的问题开始。

I want to start with the big question.

今年早些时候有一份报告，你们也公开讨论过。

There was a report earlier this year, you all have talked about it publicly.

每周有数亿人使用ChatGPT询问健康相关的问题。

Hundreds of millions of people each week use ChatGPT for health-related questions.

他们都在问些什么呢？

What are they asking?

他们问的那些问题，是不是从我们主要的医疗体系里得不到答案的？

What are the kinds of questions they're asking that they're not getting from our main healthcare system?

对，让我先退一步，谈谈我们为什么一开始要投资健康领域。

Yeah, let me zoom out and talk a little bit about why we're investing in health in the first place.

OpenAI的使命是确保通用人工智能惠及全人类，所以从宏观上看，我们主要通过三种方式来实现。

So OpenAI's mission is to ensure AGI benefits all of humanity, and so we see that at a high level happening in three ways.

第一种是构建和部署通用人工智能，也就是开发强大的模型，围绕它们打造产品，进行部署，并建立合作伙伴关系。

The first is building and deploying AGI, so building capable models, building products around them, deploying them, partnerships.

第二点是防止负面影响，也就是各种安全问题。

The second is preventing downside, safety of various kinds.

第三点是推动积极成果的实现。

The third is making upside happen.

所以我们把健康看作是一个非常非常具体的途径，既能推动积极成果，也能在这个过程中解决医疗保健中出现的一些安全问题。

And so we view health as a really, really concrete way to make upside happen and also work on some of the safety problems that come up in healthcare along the way.

因此，我们真正致力于确保两件事。

And so really, we're really invested in making sure, basically two things.

第一，我们能够提升人类健康的下限，让人们免费或几乎免费地获得可信、丰富、可靠的健康信息，同时也要提升上限，让以前不可能的事情成为可能。

One is that we can raise the floor of human health and give people access to trustworthy, abundant, reliable health intelligence for free or nearly free, and really raise the ceiling as well and make things possible that weren't possible before.

所以，要看到这种提升下限的最好方式，就是想想我们在消费领域的做法。

And so the best way to see the kind of raising the floor is really to think about how we've been working in the consumer space.

正如我们之前报告的，每周有超过三亿人使用ChatGPT来咨询健康和保健类问题，我们把这看作是一个极其重要的责任，必须做好，而且这个数字还在快速增长。

And so as we've reported, there's more than 300 million people weekly using ChatGPT for health and wellness types of questions, and we take that as an incredibly important responsibility to get that right, and that's been rapidly growing over time.

我觉得你看到的一个现象是，如果你想想获得医疗服务的体验，即使在美国，即使在湾区，想预约一个专科医生，有时候也要等上几周甚至几个月。

I think one of the things that you see is just if you think about the experience of getting healthcare, even in the US, even in the Bay Area, if you're looking to get a specialist appointment, even here in the Bay Area, sometimes it can take weeks or months.

你可能会花很多时间琢磨怎么预约、安排时间，忙活半天，最后可能只见到医生15分钟，还没准备好，出来的时候还在想：“我还有六个问题想问呢。”

You might spend a lot of time figuring out how to get an appointment and scheduling and all these kinds of things, and after all that time, maybe get 15 minutes with a doctor where you don't feel adequately prepared, and you maybe walk out of it thinking, "Oh, I had six more questions that I wanted to ask."

我们希望AI能成为你整个就医过程中的伙伴，健康不应该只局限于那15分钟的看诊时间，AI能真正帮你走好整个就医之路。

And our hope is that AI can be part of this care journey with you, and health is not something that is just kind of restricted to these kind of 15-minute intervals, and AI can actually help you navigate your care journey.

这其实也反映了现在消费者是怎么用AI的。

And so that actually reflects how people are using AI on the consumer side today.

大家带着问题和疑虑来，想获取一些信息。

People are coming in with questions and concerns, information that they'd like to have.

他们想知道怎么安排就诊，或者问问题，或者搞清楚医生交代的事情怎么跟进，诸如此类。

They want to figure out how to navigate their care visits or ask questions or figure out how to follow up things that the doctor told them or things like this.

我们发现，在两次就诊之间，AI对消费者的帮助最大，同时我们也在医生端和企业端做类似的投入。

And we're finding that in between these visits is when AI can help people most on the consumer side, and we're making similar investments on the clinician side and on the enterprise side as well.

对，这太棒了。

Yeah, it's amazing.

我们有一个非常出色的医疗体系。

We have an extraordinary healthcare system.

如果我得了重病，没有哪个国家比美国更让我想待着，但我们却有很多资源短缺的问题。

If I got really sick, there's no country I'd rather be in than the US, and yet we have a ton of scarcity.

你也看到过一些这样的数据。

You've seen some of these statistics.

超过一亿人没有固定的初级保健医生。

Over 100 million people don't have access to a regular primary care doc.

即使有固定医生的人，也常常要等四十多天才能见到医生。

Even the ones who do often have to wait 40-plus days to see that doctor.

有时候你见到医生，就像你说的，只有短短十五分钟，而你还有更多问题。

Sometimes you meet the doctor, exactly as you said, and it's a quick 15-minute visit, and you have more questions.

上周我因为皮肤问题去看了一位皮肤科医生。

I saw a dermatologist last week about something.

我提前在苹果备忘录里写下了所有问题，但还是忘了问一个，现在我在预约复诊，又要等四个月。

I wrote down all my questions beforehand in an Apple Notes, and I still forgot to ask something, and now I'm scheduling a follow-up, and it's another four months out.

我们都有过这样的经历。

It happens to all of us.

我们都有过这种感受。

We've all felt that.

有没有什么具体的病人故事或者最近的案例，能让我们想象一下未来的方向，以及你们的工作会带我们去哪里？

Is there a specific patient story or recent case that you all have seen that kind of help us imagine where the world is going and where your work is taking us?

我们团队的工作让人感到欣慰的一个方面，就是看到那些数字，看到有多少人在使用这项技术，以及它对他们来说似乎有多有用。

One way in which the work of our team is gratifying is just seeing the numbers and seeing how many people are using the tech and how useful it seems to be for them.

另一个方面就是听到人们的反馈，对吧？

One other way is just hearing from people, right?

所以我们每周都会收到很多人发来的邮件，或者通过其他各种方式联系我们——

So we get people coming in, sending us emails or whatever it is, reaching out in various ways multiple times a week-

他们说：“这就是它如何帮助了我。”

... saying, "Here's how this has helped me."

我觉得最让我感动的故事之一，是Nathan Labenz和他儿子的故事。

I think one of the stories that I was most touched by was the story of Nathan Labenz and his son.

他主持《认知革命》播客，他也在那个播客里讲过自己的故事。

So he hosts the "Cognitive Revolution" podcast, and he's kind of told his story in that podcast.

但他基本上和他儿子经历了一段很长的波折，当时他儿子很小，却一直很疲劳、易怒，还有类似的情况。

But He basically had this long saga with his son where at some point he was suffering a lot of fatigue and irritableness and things like this, his very young son.

AI帮他们理清了思路，意识到这其实是个挺严重的问题，当时他们正在度假，于是决定带孩子去做检查，AI还反复提醒他们：“嘿，要认真对待这件事。”

And AI helped them to navigate the process of realizing that this was a pretty serious thing while they were on vacation and deciding to take them for testing and repeatedly telling them, "Hey, take this seriously."

结果孩子被诊断出患有一种侵袭性的白血病。

And it turns out he was diagnosed with an aggressive form of leukemia.

在整个治疗过程中，Nathan说他们当时经常用GPT-5 Pro来帮助理清流程、理解化验报告之类的，甚至在专家有空解读之前就用了，我觉得这给他们带来了很多安心。

And throughout his care journey, Nathan reports using GPT-5 Pro at that time quite a bit for helping navigate the process, helping understand lab reports and things like this, even before a specialist had time to help interpret, and I think it brought them a lot of peace of mind.

所以OpenAI正在逐步把健康智能融入到模型的每一个环节，但同一个模型也被用来写代码或规划假期。

So OpenAI is building health intelligence and health intelligence into your model at every step, but it's the same model that folks use to write code or plan a vacation.

你能帮大家理解一下，从架构的角度来看，为什么你们要这样构建吗？

Can you help folks understand why you're building it that way from an architectural standpoint?

为什么一个通用型AI适合作为像医学这么复杂的领域的基础呢？

Why is a general-purpose AI the right foundation for something as complicated as medicine?

人们经常问我们通用模型和专用模型哪个更好的问题。

People often ask us this general versus specialized model debate.

其实我以前在上一份工作里就是做专用模型的，所以这个问题我两边都有经验。

And I used to work on specialized models actually in a previous role, and so I definitely have experience on both sides of this question.

我觉得最关键的是，我们说的其实是在通用模型里做专门的训练，这可以说是两全其美。

I think the most important thing here is what we're talking about here is actually specialized training in a general model, which is, I think, actually the best of both worlds.

所以我们在ChatGPT里做的，以及所有我们发布的模型，在训练的每个阶段，我们都加入了专门的医疗数据，过去一年里所有模型都是这样。

So what we do for ChatGPT and all the models that we ship in ChatGPT, across every stage of model training for ChatGPT, we actually have specialized healthcare data added to each phase of training since all of our models in the last year.

这样一来，你得到的模型既有前沿大模型在智能和推理上的规模和通用提升，又能从我们和全球很多医生合作得来的专门训练中受益。

As a result is now you have a model that benefits from all the scale and general improvements in intelligence and reasoning of the big frontier models, and also the benefits from the specialized training that we get from, for example, integrating data that we get from working with a bunch of physicians across the globe.

所以我觉得这种方法真的是两全其美。

And so I think you get the best of both worlds with this approach.

如果你分出来，说“嘿，我们要训练一个专门用于医疗的小型专用模型”，

If you forked and said, "Hey, we're going to train a smaller, specialized model that is just for healthcare,"

像我以前那样，或者像很多人那样，你会发现那个模型很快就会落后于前沿。

like I used to, and like many people have, what you'd find is that model would just fall behind the frontier relatively quickly.

因为前沿和通用模型的进步速度太快了。

Because the pace of the frontier and the general model improvements is so fast.

而且，如果你押注在规模上，我们作为OpenAI，作为整个行业，发现事情会变得更好。

And if you bet on scale, what we found as OpenAI, as an industry, is that things will get better.

所以我们做的就是搭上这个规模前沿的顺风车，利用我们能把评估和训练数据整合进模型的能力，让模型既专业又能受益于通用能力。

And so what we're doing is hitching a ride on that scale frontier and leveraging the fact that we have the ability to integrate evals and training data into the models to make the models both specialized and also benefit from the general capability.

我想开始想象一下你们正在构建的技术能带来什么可能性。

I want to almost start imagining what's possible with the technology that you're building.

显然，今天的技术已经很棒了，但你们不会止步，还有很多工作要做。

Obviously, where it is today is amazing, but you're not stopping, and there's a lot of work to go.

我觉得很多人都被口袋里有个私人医生的想法吸引，这个医生知识渊博，涵盖所有医学专科，这样你就不必为了不同的问题去看心脏病专家和皮肤科医生。

I think many people are allured by the idea of a personal doctor in your pocket, something that is perfectly knowledgeable, has knowledge across all medical specialties, so you don't necessarily need to go see a cardiologist and a dermatologist for different questions.

它完全个性化，拥有你所有的病史和背景，记得你说过的每句话，记得你的病史，还有你的偏好，而且还能激励你、推动你，让你在合适的时间得到合适的信息，从而引导你养成更好的健康习惯，或者遵守医生给你的治疗方案。

Something that is perfectly personalized to you, has all your medical history and context, remembers everything that you've said, remembers your medical history, but also your preferences, and is also able to motivate you and nudge you so that the information that it gives you is given to you at the right time in a way that can nudge you towards better health behaviors or adhering to the protocol that your doctor gave you.

这是你们OpenAI最终想要打造的那种体验吗？

Is that sort of the experience you're building at OpenAI in the fullness of time?

我们离那个目标有多近？

And how close are we to that?

是的，这是我们非常期待的事情。

Yeah, that's something that we're really excited about.

我们对这个主动式个人健康助手的愿景感到非常兴奋。

We are very excited about this vision of a proactive personal health assistant.

我认为这并不会取代真正的医生或看医生的机会，但真正重要的是，我们如何将获取健康信息的体验变得更加个性化，更贴合你的具体情况，这是我们投资ChatGPT健康领域的重要部分。ChatGPT是我们面向消费者的产品，让用户能够连接他们的健康信息，这些信息会被安全使用，不会用于训练数据或广告定向投放等。

I don't think that's any kind of replacement for actually having a doctor or having access to a doctor, but really thinking about the ways in which we can take that experience of getting health information and making it more personal to your context is a really important part of our investment in health in ChatGPT, which is this consumer offering that we have that enables users to connect their health information for that information to be used securely, not to be used for training data or for targeting ads or anything like this.

这是我们思考如何迈向这个越来越个性化、主动化访问世界的重要方式。

And that's a big part of how we think about how we're going to move towards this world of really increasingly personal access and proactive access.

这也是我们一直在思考如何部署的方式。

It's also been how we've been thinking about deployment.

实际上，我们从一开始就尽可能广泛地部署，不仅面向付费用户，也面向免费用户。

So we've actually from the beginning, been deploying this as widely as possible to our free users as well as our pro users.

在3亿每周活跃的健康用户中，绝大多数是免费用户，而且大部分来自美国以外。

And so we have been of the 300 million weekly active health users, the vast majority of them are free users, and we take that, and most of them are outside the US as well.

因此，我们非常重视这种分发问题，把它放在核心位置。

And so we take that kind of question of distribution here really, really centrally.

所以当我们展望这个主动式个人健康助手的下一前沿时，我们也会把这一点付诸实践。

And so as we're thinking about this next frontier of the proactive personal health assistant, we're bringing that to bear as well.

我觉得用例能帮助人们想象真正可能实现的事情，以及所有这些工作所贡献的成果。

I think use cases help us imagine, help people imagine what's actually possible, what all this work is contributing to.

口袋里装个私人医生，这个想法我觉得挺让人兴奋的。

The idea of a personal doctor in your pocket, I think is exciting.

当格雷格·布罗克曼来播客做客时，他提到过他朋友的一个情况，她当时在医院，她的医生正准备给她用某种药。

When Greg Brockman came on the podcast, he spoke a little bit about this situation he had with a friend of his was in the hospital, and her physician was about to administer a certain medication.

她在ChatGPT里核对了那种药，根据她的病史发现可能会有药物相互作用或不良反应。

She double-checked that medication in ChatGPT, realized based on her medical history that she would have some drug interaction or some adverse event that may come from that.

她把结果给医生看了。

She showed that to her doctor.

医生也同意，说：“我当时没时间仔细翻你的病史，所以没注意到这一点。

Her doctor agreed and said, "I didn't have enough time to look through your medical history to notice that.

非常感谢你提出来。”

So thank you so much for raising that."

我们难道不应该一直为每个人这么做吗？

Shouldn't we do that for everyone all the time?

如果成本够低、够方便，我们能不能对每个临床场景都做第二、第三、第四次意见？

If it's cheap enough and convenient enough, can't we do a second, third, fourth opinion for every clinical scenario?

这就是我们未来的方向吗？

Is that where we're going?

我非常喜欢这个想法，就是为患者和临床医生提供一个安全网，它就在诊疗过程中，不总是干扰或充当中间人，而是真正思考如何增强现有体验并在此基础上发展。

I love this idea of a safety net for both patients and clinicians that is just there in the care journey and not always interfering or being the middleman, but really thinking about the ways in which it can augment the existing experience and build on it.

我们实际上在临床医生方面更直接地研究了这项工作，在与Panda Health合作的一系列实验中，我们研究了部署的临床医生副驾驶，它正是以这种方式部署的，作为临床医生工作时的安全网。

We did this work actually more directly studying this on the clinician side, in a series of experiments that we ran with Panda Health, where we studied this idea of a clinician copilot that was deployed, and it was exactly deployed in this way where it was a safety net for the clinician as they were doing their work.

它会主动提出建议，比如：“嘿，这可能有问题。”

It came in and proactively suggested, like, "Hey, this might be off."

可能有你未注意到的药物相互作用。

There might be this medication interaction that you might not be tracking.

我们能否调整一下，诸如此类。

Can we potentially adjust that, and things like this.

我觉得在患者这边也会看到同样的情况。我对这个主动式代理的想法非常兴奋，它能在你的医疗旅程中真正帮助你，确保如果你的糖化血红蛋白升高而且你是糖尿病患者，你的医生会重新审查你的药物，诸如此类，确保这些情况被标记出来进行适当的审查。

I think we're going to see the same thing happen on the patient side, and I'm very excited about this idea of a proactive agent in your care journey really helping you and making sure that if your A1C rises and you're diabetic, that you get a review of your medications by your doctor and things like this, and making sure that that's flagged for appropriate review.

你提到了个性化的主动式代理。

You mentioned personalized proactive agents.

个人代理在过去几个月里从受欢迎程度的角度来看已经崭露头角。

Personal agents in general have kind of emerged on the scene from a popularity perspective over the last few months.

比如Instinct、Muse、GrokBot、OpenAI的Dots。

Instinct, Muse, GrokBot, OpenAI's Dots.

你可以给这些AI应用一个目标，对于不熟悉的人来说，它可以执行一系列任务。

You can give these AI apps an objective, for folks who are not familiar, and it can execute a series of tasks.

所以它不仅仅是“嘿，做这个任务”，然后它用一个循环就执行完。

So it's not just, "Hey, do this task," and then it executes the task with a single loop.

它可以使用你的偏好。

It can use your preferences.

它可以访问你授权给它的所有应用程序。

It can access all of your applications that you give it permission to.

如果你把信用卡信息给它，它就能访问，然后替你完成一个目标。

It can access your credit card information if you give that to them, and it can execute an objective on your behalf.

想到我们那个虽然很棒但常常很混乱的医疗系统，真的很难不让人想象各种可能性。

One can't help but imagine the possibilities there in navigating our wonderful but often disjointed healthcare system.

比如预约挂号，确保保险能覆盖这次就诊，帮你按照医生出院后给的护理方案去执行和指导你。

Scheduling an appointment and making sure your insurance is covering that appointment, helping you and guiding you to follow a care protocol that your doctor gave you post-discharge.

甚至还能跟保险公司协商处理医疗账单错误，各种在医疗系统里造成麻烦的事情都能搞定。

Even negotiating with your insurance company on a medical billing error, all kinds of tasks that just create friction in our healthcare system.

这就是你工作的方向吗？

Is that where you're going with your work?

我们是不是可以期待OpenAI在这方面有所作为？

Should we look forward to that from OpenAI?

我对这个愿景非常兴奋。

I'm very excited about that vision.

我觉得，从患者和医生的角度来看，医疗系统最让人头疼的问题之一就是碎片化，整个生态系统缺乏连接。

I think if you look at, from both a patient and clinician perspective, one of the most difficult things about the healthcare system and how it serves people is this question of fragmentation and lack of connection of the ecosystem.

而且医疗行业也是那种你不能指望明天所有东西都有MCP连接器的行业，对吧？

And healthcare is also one of those industries where you shouldn't expect everything to have an MCP connector tomorrow, right?

这里有很多遗留软件。

It is a place where there's legacy software.

一开始就不一定有最好的AI或软件连接性，而且很多人是在虚拟机上运行他们的主要EHR软件的，比如在医疗生态系统中。

There's not necessarily the best connectivity with AI or software in the first place, and many people are running their primary EHR software in the healthcare ecosystem on a VM, for example.

所以就像——

So like-

虚拟机，他们在云端运行的，对。

Virtual machine that they're running in the cloud, yeah.

没错。

Exactly.

所以实际上有一个有趣的方面，虚拟机在某些方面让AI代理更容易操作，但这对于代理和计算机使用来说绝对是对的。

And so there's actually an interesting component where a virtual machine actually makes things easier in some ways for AI agents, but this is absolutely right for agents and computer use.

所以我确实认为，这里有很多传统上存在于患者和获得所需护理之间、临床医生和提供所需护理之间，以及医疗数据本身之间的障碍，这些数据往往是整个系统为患者良好运作的瓶颈。

So I definitely think that this is a place where a lot of the barriers that have traditionally existed between patients and getting the care they need, and clinicians and delivering the care that they need, and healthcare data in general, which is often kind of a bottleneck to the whole system working well for patients.

我觉得很多这些障碍都会消失。

I think a lot of those will dissolve.

好的，那我们是在聊应用场景。

Okay, so we're talking about the use cases.

你在帮我们描绘愿景。

You're helping us dream the dream.

在继续深入之前，我们先聊聊信任。

Before we keep going down that road, let's talk about trust.

显然，信任在这里是最重要的。

Obviously, trust is the most important here.

同一个底层模型，既能帮我理解化验结果或者医生说的话，也能帮一位训练有素的专业医生分析非常复杂的病例。

The same underlying model can help me understand my lab results or something that I heard from a doctor, but it can also help a highly trained specialized physician make sense of a very complicated case.

那在回答时，你会做哪些调整，让这两种不同的人，或者任何其他可能使用同一个底层模型的人，都能看懂呢？

What do you change in the answer so that it's legible to each of those two personas or any other persona that might be engaging with the same underlying model?

就像你说的，医疗领域是信任驱动的，所以不仅要让模型更好、能评估它们，还要做好沟通，让整个行业跟上步伐。

As you said, health is a place where it moves at the speed of trust, and so you need to be able to not only make the models better and measure them, but also communicate and bring the world along.

所以这一直是我们思考和策略中非常重要的一部分，无论是我们如何与患者合作、如何与最终用户合作，还是如何与临床医生合作。

And so that's been a very important part of our thinking and our strategy for both how we work with patients and how we work with end users, and how we work with clinicians.

所以对你问题的简短回答是，我们通过与世界各地的专家合作，以各种方式获取反馈，来评估和训练模型，以调整它们的行为，使其适应不同的人群——无论是普通用户，还是已经处理过复杂护理流程的有经验的普通用户，或者是临床医生，以及其他类型的用户角色。

And so, the brief answer to your question is basically, we evaluate and train the models by working with experts across the world and getting feedback in various ways to tailor their behavior, to whether a person is a lay user or an expert lay user that has been navigating their care for a while, a complex care journey, or a clinician, or other kinds of user personas.

甚至在我们2025年5月发布的原始Health Bench评估中，就有一个主题或部分专门关注模型如何根据用户是普通用户还是临床医生来调整其行为，以及使用技术术语的程度等。

And so back even for the original Health Bench evaluation that we released back in May 2025, there was actually a theme or part of the evaluation that was really focused on this distinction of how the model behaved and tailored its level of use of technical jargon and things like this to whether a person was a lay user or clinician.

从那以后，我们不仅在衡量模型能否满足不同用户需求方面做得更好，也在朝着这个方向训练模型。

And since then, we've both gotten better at measuring whether the models are able to cater to different kinds of users, and also training the models towards that.

所以我们的很多方法不仅仅是让模型更好地服务这些不同的用户，还包括如何让他们跟上步伐。

And so a lot of our approach has really been, not only how do we make the models better for these different users, but how do we bring them along?

其中一部分也涉及我们如何进行评估以及如何传达这些评估结果。

And part of that is how we do evaluations and communicate about them as well.

所以信任有几个组成部分。

So there's a few components to trust.

我们已经讨论了根据用户来调整答案。

So we've talked about tailoring the answer to the user.

另一个重要的方面是意见分歧。

Another big component is disagreement.

一个人可能正在和LLM或聊天机器人对话，分享他们所有的症状，但内心其实希望听到一切都没问题。

A person can be speaking to an LLM or speaking to chat and be sharing all of their symptoms, but really hoping to hear that everything is okay.

所以从消费者或用户的净推荐值角度来看，他们想要听到安慰的话，但从医学角度来看，他们可能需要听到完全不同的内容。

So from a consumer/user NPS perspective, net promoter score perspective, they want to hear reassurance, but from a medical perspective, they need to hear perhaps something very different.

你怎么教AI识别这种差异，并管理不同的产品目标，比如消费者参与度和留存率，与正确性和临床质量之间的平衡？

How do you teach an AI to recognize the difference and manage different product objectives there, either from consumer engagement and retention versus correctness and clinical quality?

这在我们如何进行评估和训练中非常核心。稍微放大一点来看，关于我们如何回应和评估中的同意与分歧问题。

This is super central to how we do both evaluation and training, and so zooming out a little bit and the question of agreement and disagreement around how we respond and how we evaluate.

一种做法是，你可以找几位临床医生，比如加入OpenAI健康团队，然后对他们说：‘嘿，你们来搞清楚如何处理这些不确定的病例，我会向你们咨询这些不确定的病例，你们帮我们构建评估和训练数据等等。’

One thing you could do and one thing that you could base your approach on here is we could bring in a few clinicians into, let's say, the OpenAI health team and say, "Hey, you guys figure out how to navigate these kind of uncertain cases, and I'll ask you about these uncertain cases, and you'll help us build evaluations and training data and stuff like this."

然后让那几位医生去解决这个问题。

And leave it to those few physicians to figure it out.

但实际上，这并不是最好的方法。

This is actually not the best approach.

结果发现，你要是问五个医生同一个问题，能得到八个不同的答案。

So it turns out if you ask five physicians or something this any question, you'll get eight different answers.

所以我们在OpenAI刚开始关注健康领域时，做的第一件事就是建立一个全球医生网络——

And so one of the first things that we did when we first started focusing on health at OpenAI was really build this global physician network-

——我们和全球数百位专家合作，涵盖几十个专科、语言和执业国家，来思考如何正确汇总专家的智慧，这样我们就不依赖一两个或三个医生来裁决，尤其是在可能存在分歧的领域。

... of hundreds of experts that we work with across the globe, across dozens of specialties and languages and countries of practice, to think about the right ways to aggregate the wisdom of experts so that we weren't relying, especially in areas where there could be disagreement, we weren't relying on one or two or three physicians to adjudicate.

这是其中一部分，就是考虑评估和训练数据来自不同专家，而不是依赖一两个人。

So that's one part of it, is thinking about evaluation and training data coming in from different experts rather than relying on one or two people.

但你问题的第二部分是，如何平衡相互冲突的目标，对吧？

But then the second part of your question is, how do you balance competing objectives, right?

比如说，医生们和他们的集体智慧指向一个方向，而用户可能希望你即使情况不乐观也说些安慰的话。

Let's say the physicians, and the aggregated wisdom of the physicians is pointing in one direction, and then maybe the user actually wants you to say something comforting, even when the situation's not comforting.

这基本上就是安全优先的问题。

It's basically a prioritization of safety.

所以对我们来说，当有数亿次健康对话时，认真对待我们的责任非常重要。

And so it's very important to us that we take our obligation very seriously when there's hundreds of millions of health conversations coming in.

所以在整个评估和训练体系中，我们设置了一些评估来追踪这类棘手的情况，即模型应该主动说：‘嘿，这不太安全，你应该去看医生。’

And so across the evaluation stack and training stack, we have evaluations that track these kind of difficult situations where the model should actually go and say, "Hey, this is unsafe, you should seek care."

即使用户不想听，或者想要某种替代疗法之类的。

Even if the user doesn't want to hear it, or wants to get some kind of alternative treatment or something like this.

这是我们评估工作中非常重要的一部分，而且随着时间的推移，我们不仅通过与医生合作改进了评估方法，还看到模型在这些领域的整体表现有了显著提升。

And that's a very important part of how we do evaluation, and over time, we've not only improved how we've done the evaluation with working with physicians, but also seen pretty dramatic model improvements across the board, in those areas.

还有另一个例子，或者说另一种分歧的情况，就是我作为用户，向ChatGPT提出一个医学问题。

There's another example, or another instance of disagreement, which is me as the consumer, I ask a medical question of ChatGPT.

然后我在看医生的时候把那个回答拿给医生看。

I then show that response to my doctor during a visit.

顺便说一句，我的医生很喜欢我。

My doctor loves me, by the way.

然后他们意见不一致。

And then they disagree.

他们不同意那个回答。

They disagree with the response.

那我该怎么处理这种情况呢？

How should I handle that situation?

对，我觉得你应该像处理其他分歧一样来处理它。

Yeah, I think you should handle it similarly, as if disagreement was happening in some other way as well.

人们对不同的事情会有不同的看法，作为病人，你有权要求另一位医生看看，或者去找别的医生寻求第二意见。

People do disagree about different things, and I think you have a right as a patient to ask for another doctor to take a look and get a sense, and go to another doctor to get a second opinion.

是的，这就是我建议的做法，不管你是否同时使用ChatGPT获取健康信息。

And yeah, that's what I would advocate for, whether or not you're also using ChatGPT to get health information as well.

非常有道理。

Makes a ton of sense.

信任还有另一个方面，就是绝对的准确性。你们做的很多工作，至少从我听到的来看，似乎都是把答案建立在世界医学文献和证据的基础上。

There's another element of trust, which is just absolute correctness, and a lot of the work that you all have done, at least from hearing from you, seems to be about grounding answers in the world's medical literature and evidence.

医学的某些部分，大家意见比较一致，但还有很多医学领域仍在探索或研究中，即使在医学界也存在分歧。

Some parts of medicine, there's not a lot of disagreement about, but there is much of medicine that's still being discovered or being studied, and there is disagreement even in the medical field.

对于这类问题，当我们还在探索真相的时候，就很难把答案建立在真相之上。

And for those kinds of questions, it's difficult to ground it in truth when we're still discovering truth.

科学是一个过程，等等。

Science is a process, et cetera.

在这种情况下，你怎么看待正确性？

How do you think about correctness in those kinds of instances?

是的，这是个非常好的问题。

Yeah, it's a really great question.

不管有没有AI，这都很重要。

It's important with or without AI.

所以，如果你放眼全局，看看今天的医学共识是怎么分布的，对吧？

So if you zoom out and look at how is medical consensus distributed today, right?

今天这个过程是怎样的？

What is the process of that today?

基本上，有一大堆人在写各种生物医学论文，提出改进建议、治疗方法或者类似的东西，这些都可能提高护理质量。

You basically have a bunch of people writing biomedical papers of various kinds, suggesting improvements or treatments or things like this that could improve the quality of care.

然后，每年各个专科都会有一个制定临床指南的过程，这些指南会描述，比如：“嘿，在这种情况下，如果你是心脏病专家，看到一个有X、Y、Z症状的病人，你大概应该这样做。”

And then you have this yearly process across different specialties of building clinical guidelines, which describe like, "Hey, in this situation, if you're a cardiologist and you see a patient with X, Y, and Z, here's what you should probably do."

所以这些临床指南，你可以把它们大致看作流程图，每年由世界各地的不同机构发布，根据最新研究，大致描述在不同情况下你应该怎么做。

And so these clinical guidelines are roughly, you can consider them flowcharts that are put out each year by the different kinds of bodies across the world that describe roughly what you should do in different scenarios based on the latest research.

然后最后一步是，你有了生物医学文献，然后有了临床指南，然后就是人们在实践中到底怎么做。

And then the final step is, so you have either the biomedical literature, then you have the clinical guidelines, and then you have what do people actually do in practice.

结果发现，最后这一步是个巨大的瓶颈。

And so the final step, it turns out, is a massive bottleneck.

所以有研究专门探讨，临床指南要花多长时间才能真正渗透到全球各地医生的日常实践中。

So there are these studies on how long does it take for clinical guidelines to penetrate the practice of everyday doctors across the world.

过去人们发现，根据研究不同，这个过程需要五到十七年才能真正实现。

And in the past, people have found that this takes, depending on the study, five to 17 years for that penetration to actually happen.

哇。

Wow.

这时间太长了。

Which is massive.

你肯定不想在实践的时候——

You don't want to be practicing-

太夸张了。

Crazy

……医学落后了五到十七年。

... medicine five to 17 years behind.

所以我们思考的一个问题就是，怎么把这种按过去方式行医的时间，压缩到几周、几天，甚至几个小时，对吧？

And so one of the things that we think about is how do you compress that period of practicing in the past, down to weeks or days, or even hours, right?

这也是我们做医疗版ChatGPT和临床医生版ChatGPT背后的一个重要原则，这些是面向临床医生和医疗专业人士的产品，我们会在诊疗的那一刻引入最新的生物医学文献——

And so this is one very important principle behind our work on ChatGPT for healthcare and ChatGPT for clinicians, which are clinician and healthcare professional facing offerings where we're bringing in the latest biomedical literature in the moment of care-

……让大家在比如十五分钟的门诊过程中，在诊疗现场，很快就能得到问题的答案。

... and allowing people to very quickly, in the course of let's say a 15-minute visit, at the point of care, get answers to their questions.

所以这是回答你问题的一种方式，就是说，在实践中，虽然很多领域缺乏医学共识，但有些情况下是有共识的，只是没有被落实。

So that's one way of answering question, which is like, in practice, while there is a lack of medical consensus in many areas, in some cases there is medical consensus, but it's not being practiced.

所以你能做的一件事，就是大幅减少这种情况的发生，这只是回答你问题的一种方式。

And so one thing you can do is really dramatically reduce the amount that's happening, is just one way of answering your question.

回答你问题的第二种方式，就是如何处理那些仍然存在真正医学不确定性的地方。

A second way of answering your question is how to handle the places where there's still genuine medical uncertainty in there.

我觉得有几件事。

I think there's a couple things.

第一，你要训练模型，让它们能识别自己的不确定性。

So one is you want to train the models to be able to recognize their own uncertainty.

在浏览和获取信息之后，你要让模型知道，嘿，信息不够，要坦诚说出来。

After browsing and getting information and things like this, you want the models to be able to know, hey, there just isn't enough information, and to be upfront about that.

所以，这是我们评估工作中的一个重要部分。

And so, that's been an important part of how we've done evaluation.

这其实是Health Bench公开的一部分，包括对不确定情况下的模糊表述的评估。

That's actually a public aspect of Health Bench, which includes evaluations on hedging, for example, when there is this kind of uncertainty.

所以这是很重要的一点。

So that's one important thing.

第二点，关于如何提高人类健康的上限，就是要真正思考模型如何加速科学研究，推动医学前沿。

The second thing, this goes to this question of how do you raise the ceiling of human health is really thinking about the ways that models can actually accelerate the process of doing science and actually pushing the medical frontier.

这个我们可能待会儿再聊。

And so maybe we'll chat about that later.

听你这么说，感觉你总是在考虑三类人。

Hearing you talk, it feels like you're always thinking about one of three constituents.

要么是患者，要么是注册的临床医生，要么是企业，不管是医疗服务机构还是其他类型的企业。

It's either the patient, it might be the individual clinician that's signed up, or it might be the enterprise, whether it's a healthcare provider enterprise or some other enterprise.

你怎么考虑在架构层面，还有所谓的“框架”层面来设计产品，以满足所有这些不同用户的需求呢？

How do you think about designing the product at the architectural level and at the, let's say, the harness level to meet the needs of all these different users?

我觉得这个问题可能分三部分来回答。

I think maybe there's three parts to that.

第一，让模型对这些用户来说真的非常好用。

One is making the models really, really good for these users.

第二，围绕模型构建合适的框架或产品，第三，就是实际的部署或合作推广，去触达用户。

The second is building the right harness or product around it, and the third is really the deployment or partnership motion to reach out to the users.

而且，说实话，现在很多时候模型本身已经相当不错了。

And by default, I think a lot of the time, the models are often pretty good these days.

所以我们之前聊过我们的评估方法，和临床医生合作，以及类似的事情。

And so we talked about our approach of evaluation, working with clinicians, and things like this.

我们看到性能和可靠性都有了非常显著的提升。

We are seeing pretty dramatic improvements in performance and reliability.

几个月前我们聊过，我们在免费版上做的改进，以及把免费版的性能提升到前沿水平。

We talked a few months ago about the improvements that we've been making on the free tier and bringing free tier performance to the frontier.

所以，每周有12亿人使用ChatGPT——给他们一个在健康领域极其有用的前沿模型，而且是免费的。

So, of the 1.2 billion people who use ChatGPT weekly- giving them a model for health that is just incredibly useful at the frontier for free.

所以这显然是我们让健康智能普及化战略的重要部分。

So that's been a big part of our strategy, obviously, for making health intelligence abundant.

第二点就是我们的产品。

The second thing is our product offerings.

我们的想法是，模型可以很好，但如果使用模型改善健康的门槛太高，人们就不会去用。

And so the way we think about this is you can have a model that's really good, but if the activation energy to use the model to improve health is too high, then people won't do it.

所以我们很多产品工作，都是在考虑你提到的这些用户群体，想办法降低人们使用计算和智能来改善健康的门槛。

So a lot of our product efforts are really thinking about these constituents that you're talking about and thinking about the ways in which we can reduce the activation energy for people to use compute and intelligence to improve human health.

这包括我们在健康领域的ChatGPT工作、面向医疗企业的ChatGPT，以及面向临床医生的ChatGPT。

And so that goes along the lines of our work in health and ChatGPT, ChatGPT for healthcare, on the enterprise side, and ChatGPT for clinicians.

我们设计这些产品时，很大程度上就是在想，人们使用这些模型会遇到哪些障碍、瓶颈和门槛。

A lot of the way we design these offerings is basically to think about what are the barriers and bottlenecks and activation energies for people to be able to use the models.

比如说，向医疗系统提供企业版ChatGPT时，在安全和合规方面有哪些瓶颈？

And so, for example, what are the security and compliance bottlenecks in providing an enterprise offering of ChatGPT to healthcare systems?

这是我们两年前一直在思考的一个问题。

So that was one question that we were thinking about a lot two years ago.

所以我们想，好吧，我们需要在安全和合规方面进行投入。

And so we thought, okay, we need to invest in security and compliance.

我们必须确保能够提供符合HIPAA标准的服务。

We need to make sure we are able to offer HIPAA compliance.

我们必须确保能够提供医疗机构所需的隐私保护，这样我们才能处理他们的医疗数据，并提供他们信任的服务。

We need to make sure that we're able to offer the kinds of privacy protections that healthcare organizations need so that we're able to handle the healthcare data that they need and be able to offer services that they trust.

所以，从大约两年前到大约一年前，这是我们工作中非常重要的一部分。

And so that was a very important part of the work we were doing for roughly two years ago to roughly a year ago.

然后我们就根据看到的采用过程中的瓶颈，一步步往前走。

And we kind of just go from there, based on the bottlenecks that we see to kind of adoption.

第三点是合作伙伴关系和部署。

The third thing is kind of partnerships and deployment.

再说一次，几年前，我觉得OpenAI在医疗领域的话语上起步很早。

And again, it's really like, a few years ago, I think OpenAI was very early to speak in the language of healthcare.

当我们与医疗合作伙伴互动时，我觉得我们很早就开始了这些对话。

And when we interacted with healthcare partners, I think we were early in those conversations.

所以我觉得我们之后带来的一个关键点，就是真正考虑你提到的每一个相关方，真正思考如何用他们的语言说话，如何带领他们一起走这条路。

So I think one of the things that we've brought to bear since then is really thinking about each of those constituents that you're talking about and really thinking about how to speak their language and how to kind of bring them along this journey.

因为即使模型变得非常聪明，如果没有人真正信任或使用它们，那也实现不了我们的使命。

Because even if the models get really smart, if nobody's really trusting them or using them, that's not really achieving our mission.

你提到了一点，就是你们与Epic合作，把产品直接集成到Epic里。

And is that, you're alluding to a little bit some of the work you've done with Epic to kind of integrate the product directly into Epic.

能多分享一些这方面的内容吗？

Maybe share a little bit more about that.

因为很多在医疗行业待了很久的人，对与Epic的这种合作水平非常感兴趣。

Because a lot of the folks who've spent a lot of time in healthcare were very interested to see that level of collaboration with Epic.

这是我们方法中非常重要的一部分。

This is a very important part of our approach.

就像我们之前提到的，医疗行业的碎片化一直是个问题。

Like we've talked about, fragmentation in healthcare has been one of the issues.

数据和流程彼此脱节，我认为这是阻碍AI在医疗领域有效应用的一大障碍。

And the fact that there is kind of data and workflows that are disconnected from each other, I think is a big obstacle to the use of AI productively in healthcare.

所以，我们与Epic以及其他合作伙伴的合作，最近我们宣布了一项新进展，现在医疗提供者或专业人士可以在ChatGPT中使用Epic，同时ChatGPT也能在Epic内部发挥作用。

And so, our work with Epic, our work with other partners, we had a recent announcement where we announced that there's now a way to use Epic in ChatGPT, if you're a provider or a healthcare professional, and also for ChatGPT to be useful in Epic itself.

在Epic的工作流程中，你可以通过侧边栏直接使用ChatGPT。

And so in the Epic workflow itself, you can have a little sidebar where you can actually use ChatGPT as well.

这是其中一方面。

And so that's one thing.

我们还在开发其他类型的连接器，最近也公布了一些，未来还会有更多。

We're also working on different other kinds of connectors that we've also announced recently, and we have more coming.

我们正在做大量工作，努力打破这些壁垒，让医疗生态系统真正连接起来。

And so there is quite a bit of work that we're doing to make sure that we are kind of breaking down these barriers to the kind of the healthcare ecosystem being connected.

而且我觉得Epic只是我们识别瓶颈并努力解决这些问题的众多例子之一。

And I think Epic is just one of many examples where we're just identifying those bottlenecks and trying to fix them.

关于信任，我想最后谈一下你在心理健康方面做的一些工作，这也是一个非常复杂的领域，除了我们之前谈到的临床复杂性之外，可能还有其他目标需要你去解决。

The last thing I want to hit on with respect to trust is some of the work that you've done around mental health, and this is another area where there's a lot of complexity, and there may be other objectives that you need to solve for beyond the clinical complexity that we talked about.

如果我问我的OpenAI Dot，"嘿，我想在接下来的几周里感觉好一些，"你怎么理解这个问题？

If I ask my OpenAI Dot, "Hey, I want to feel better over the next few weeks," how do you even interpret that question?

你怎么帮助我实现这个目标？

How do you help act on that?

考虑到这可能意味着很多事情，你怎么引导用户？

How do you help guide a user on that, given that that could mean so many things?

不管在哪个平台上，我们都非常致力于帮助那些遇到困难或处于危机中的用户。

And regardless of the surface, we are super committed to helping users when they're having a difficult time or when they're in crisis.

比如，对于Dots，我们一直在投入我们在所有产品表面上都有的安全保障措施，这个产品也不例外。

And so, for example, for Dots, we've been investing the same kinds of safeguards that we've been putting in all of our product surfaces in that product surface as well.

所以，我们方法的一个重要部分就是建立一个基本的安全底线。

So that's an important part of our approach is kind of like a baseline safety floor.

然后，在心理健康基准方面，我们确实采取了这种方法，不仅衡量安全底线，还思考模型如何能够主动帮助提高上限——

And then, with Mental Health Bench, we've really taken this approach of not just measuring the safety floor, but thinking about the ways in which the models can actually proactively help kind of raise the ceiling-

……并且真正思考模型如何帮助提供积极的心理健康体验，从仅仅担心安全底线转向衡量那些积极的结果。

... and really think about the ways in which models can help provide positive mental health experiences and move from just worrying about the safety floor to measuring those kinds of upsides.

所以再说一次，这种方法实际上与我们在健康基准上的工作非常相似，我们一直在与世界各地的持牌心理健康专家合作，思考在众多棘手场景中定义模型行为的正确方式，并确保我们能够达成一定程度的共识。

And so again, the approach is actually very much mirroring our work on Health Bench, where we've just been working with a bunch of licensed mental health experts across the world to think about the right ways to define model behavior in a bunch of tricky scenarios and ensure that we can get to a level of consensus.

人们在这里的分歧甚至比身体健康领域更大。

People definitely disagree even more here than physical health.

在这些棘手的情况下达成一定程度的共识，这样我们就能实际衡量性能，并为行业创建这种共享的基准。

Get to a level of consensus in some of these tricky situations where we can actually measure performance and create this kind of shared benchmark from the industry.

所以我们最近发布了那个基准。

And so we released that benchmark very recently.

与健康基准和健康基准专业版类似，目标一直是：我们能否创建这个基准，为所有人提高上限，并在所有模型开发者、医疗行业和心理健康专家之间创建这种真正共享的标尺或衡量标准，用来确保我们朝着正确的方向前进。

And similar to Health Bench and Health Bench Professional, the goal has really been, can we create this benchmark that kind of raises the ceiling for everybody and kind of creates across all model developers and the healthcare industry and mental health experts like this really shared ruler or measuring stick that we can use to kind of make sure that we're moving in the right direction.

好的，我们已经讨论了智能以及围绕智能改进而出现的用例。

Okay, so we've spoken about intelligence and the use cases that are coming around intelligence approving.

我们谈到了信任，这很关键。

We've spoken about trust, which is critical.

接下来聊聊医疗护理，我想区分一下医疗信息和医疗护理。

Let's talk about medical care, and I want to draw the distinction between medical information and medical care.

我觉得对于医疗信息来说，如果你有健康问题，或者我自己遇到什么健康问题，聊天工具和其他大语言模型越来越成为获取答案和解决问题的更好、更快、更便宜的方式。

I think for medical information, I think if you have a health concern, if I have a health concern about myself or something that I'm dealing with, chat and other LLMs are increasingly the better, faster, cheaper way to get an answer and a resolution to that question.

但如果仅仅靠医疗信息解决不了问题，那么至少目前，我就得进入传统的医疗护理系统。

But if in the scenarios where medical information alone doesn't get me to resolution, I then, at least today, am thrusted into the traditional medical care system.

所以如果我需要开处方，需要去看专科医生，需要做后续的血液检查、核磁共振或其他扫描，或者需要医生给我做检查，我还是得在传统的医疗系统里转。

And so if I need to go get a prescription, if I need to go see an in-person specialist, if I need to get some follow-up blood work or some MRI or some other form of scan done, if I need a clinician to examine me in some way, I still need to navigate the traditional healthcare system.

其中很多环节都受到美国医疗执照以及我们医疗系统中其他基础设施的监管和约束。

And a lot of that work- ... is sort of governed and regulated by the US medical license and sort of other infrastructure that we have in our healthcare system.

这是一个多轮对话的过程，不是简单一个问题就能解决的，而是“我得去找几个人聊聊”。

It's the multi-turn conversation, where it's not just a single question resolution, it's, "I got to go talk to a few people.

我得去办几件事。

I need to go get a few things done.

我们需要收集一些信息。

We need to collect some information."

在很多方面，医疗服务是一个打包的产品。

In many ways, medical care is a bundled product.

我们得到诊断、分诊和治疗方案。

We're getting the diagnosis and triage and the care plan.

我们得到医生的同理心、安慰和支持。

We're getting the empathy from the clinician and the assurance and the support.

我们得到相关的法律责任，以及他们承担的责任。

We're getting the legal liability associated with that and what they're taking on.

AI能提供所有这些吗？

Will AI provide all of that?

它会在哪里止步？

Where will it stop short?

你觉得这一切会走向何方？

Where do you think this is all going?

不，我不认为AI能或会提供所有这些。

No, I don't think AI can or will provide all of that.

而且我觉得你说得对，健康信息和医疗护理之间确实有区别。

And I think you're right that there's this distinction between health information and medical care.

所以我认为我们在这里的角色就是赋能整个生态系统，因为未来不可能由一家公司来提供所有的医疗服务，对吧？

And so I think the way we see our role here is really to enable the whole ecosystem, because there's no future where any one company is delivering all the medical care, right?

所以我们的角色就是思考医疗护理的最后一公里对人们来说到底是什么样的。

And so our role here is really to think about what does the last mile of medical care really look like for people.

我们能不能真正理解这一点，并且让每个人都能更容易地获得医疗服务，这样对患者产生实际影响的最后一公里，对我们所有的合作伙伴来说，就像最后100米一样简单。

Can we really understand that, and can we make that easier for everybody, so that the last mile of actually having impact on patients just looks like the last 100 meters for all of our partners.

这就是我们的想法，就是让我们的合作伙伴能够非常容易地使用AI来支持每个人在护理过程中的需求。

And so that's kind of how we think about it, is really making it very easy for our partners to use AI to support everybody in the care journey.

我认为OpenAI在这里有一些相对独特的事情可以做，因为它拥有健康领域的入口。

I think there is stuff that OpenAI can do relatively uniquely here because of the front door to health that it has.

所以，举个例子，它可以帮助你解答很多这类问题，或者帮你导航这个旅程，或者随着我们越来越有用的个人代理的出现，帮你处理一些日程安排和类似的事情，就像你的健康四分卫一样。

And so, for example, it can help you navigate a lot of these kinds of questions or help you navigate this journey, or as we get increasingly helpful personal agents, help you with some of the scheduling logistics and things like this and just act as a quarterback for your health.

话虽如此，我觉得它不会解决我们面临的所有挑战，至少短期内不会。

That said, I don't think it'll resolve all these challenges that we'll have, at least not in the short term.

而且我觉得在很大程度上，我们是在帮助很多提供医疗服务的合作伙伴，他们大多在用我们的模型以各种方式——

And I think to a large extent, we're looking to enable a lot of our partners that are delivering care, that are overwhelmingly using our models in various ways-

——让这些操作更顺畅、更容易，这样患者就不会陷入那种碎片化的体验。

... to make these operations happen smoother and easier so that patients are not left in this kind of fragmented experience.

那么作为消费者，我应该期待从OpenAI那里得到的是你描述的那种“健康总指挥”式的服务吗？

So as a consumer, should the experience that I look forward to from OpenAI be sort of the quarterback care that you described?

我会在界面上和临床医生对话吗，还是你会把我转给医生并帮我找到他们？

Will I be talking to a clinician in the interface at any point, or would you perhaps route me to a doctor and help me find that?

因为光是预约和安排就诊本身就是个大难题。

Because given that's a whole challenge, just scheduling and making those visits happen.

我应该期待什么呢？

What should I look forward to?

我觉得你会看到，我们希望能非常有效地帮你导航寻找医疗服务的过程。

I think you'll see we will be hopefully very helpful in helping you navigate the process of finding care.

我觉得我们不会去做你提到的所有那些事情。

I don't think we'll be in the business of doing all the things that you mentioned.

但这就是我们真正依赖所有初创公司和护理提供机构的地方——

But that's where we're really relying on all of our startup and care provider organizations-

当然。

Sure

还有合作伙伴之类的。

... and partners and things like this.

所以回到我之前提的那个问题，你在构建那个未来，但你是和合作伙伴一起构建的，你是为消费者和患者构建的，你是为临床医生构建的，你是为企业构建的。

So coming back to a question I mentioned earlier, which is you are building that future, but you're building it with partners, you're building it for consumers and patients, you're building it for clinicians, you're building it for enterprises.

对于临床医生和企业，给我们讲讲你正在推出的一些工具，以及一些真正有效的合作吧。

For clinicians and enterprises, give us some texture on some of the tools that you're launching and some of the collaborations that have really worked.

我们有一些非常令人兴奋的合作，而且我们为ChatGPT for Healthcare推出了一些非常棒的初始合作伙伴。

We've had a few really exciting partnerships, and we've had some really amazing launch partners for ChatGPT for Healthcare.

有很多用户在使用ChatGPT for Clinicians，而且非常喜欢它。

We've had lots of users using ChatGPT for clinicians and really enjoying it.

这个产品的用户留存率真的让我们很惊喜。

The retention on the offering has been a very pleasant surprise.

大家看起来都很喜欢它。

People seem to be really enjoying it.

对，我们一直在和很多医疗系统合作。

Yeah, we've been working with a variety of health systems.

我觉得有几个特别突出的，一个是加州大学旧金山分校，我们一直在和他们合作，也聊过这个。他们是Epic、电子病历和ChatGPT之间连接的最早采用者之一，而且他们真的走在前沿，比如考虑让医生用它来做病历回顾之类的事情，并且思考如何让照顾病人的体验不再那么碎片化。

I think a couple that stand out, one is we've been working with UCSF, and we've been chatting about that, and they've been one of our earliest adopters for this connection between Epic and the EHR and ChatGPT, and they've been really on the front lines of thinking about, for example, clinicians using it for chart review or things like this, and thinking about the right ways in which the experience of caring for a patient doesn't need to be this super fragmented.

你不需要切换标签页或者切换不同的系统。

You need to switch tabs or switch experiences and things like this.

所以他们在这方面是我们的重度用户。

And so they've been some of our power users on that front.

另一个是Advent Health，他们实际上也是我们的重度用户，在考虑如何在整个OpenAI的工作区里部署智能体。

I think another one is Advent Health, and they've actually been some of our power users in thinking about deploying agents across the workspace, actually across all of OpenAI.

他们一直是这方面的领导者。

They've been a leader.