德米斯·哈萨比斯谈2030年前实现通用人工智能、治愈所有疾病、AGI之后的生活等

德米斯，非常感谢你再次来到这里。

Demis, thanks so much for being here again.

每次和你聊天都特别愉快。

It's always great talking to you.

能来这儿真的很开心。

It's great to be here.

谷歌I/O大会当然发布了大量新消息。

So Google IO, obviously, tons of announcements.

你最期待的是哪一项？

What are you most excited for?

哇，真的。

Oh, wow.

我们确实公布了很多令人兴奋的新内容。

It's, I mean, I think we announced a lot of really exciting things.

你大概注意到我对Omni特别激动了。

You probably saw me pretty excited about Omni.

我觉得我们在世界模型上的这个方向特别有意思。

I think this is really interesting direction for us on world models.

我一直都很支持这个方向。

I've always been a big proponent of that.

不过显然，Flash 3.5才是重头戏，把它整合进Anti-Gravity 2.0里头。

But, obviously, you know, I think Flash 3.5 is the big story and integrating that into anti gravity two point o.

我觉得大家一定会特别惊喜，这真是个超强的实用型模型。

I think people are gonna be really pleasantly surprised, and I think it's an amazing kind of like workhorse model.

当然，后面还有更多要推出，比如Pro版本。

And then, of course, you got a lot more to come, you know, with Pro.

对。

Mhmm.

是的。

Yeah.

银行类产品正在上升。

Bank products are going up.

现在代理功能已经集成到Spark里了。

Agents are here with Spark.

我觉得在消费端，AI带来的效率提升真的开始显现了。

And I think, like, on the consumer side, we're really starting to see, like, the productivity gains in AI, like, kicking in.

比如用Nano Banana，内容创作者已经在实际使用这些工具，丰富他们的创作工具箱，产出更多内容之类的。

Like, even with Nano Banana, for example, content creators are genuinely using these tools to new neutral in tool kits, make make more content, that sort of stuff.

我想问问，拿现在的Omni和三年前刚发布Gemini的时候相比，你的AGI时间表有变化吗？

I'm curious, like, looking at today with, like, Omni, for example, versus three years ago when when Gemini was first released Yeah.

你的AGI时间表是提前了，还是变慢了？

How have your AGI timelines kinda shifted?

还是说一切都在按原计划进行？

Have have they has has it gone according to plan, or is it gone or slower?

我觉得还是按原计划推进的。

I think it's gone according to plan.

几年前，你可能看到过，甚至在采访里我曾说过，大概需要五到十年，那大概是两三年前的事了。

So a few years ago, you probably saw maybe we even talked about when in interviews, I used to say five to ten years, maybe that's two or three years ago.

我觉得我们现在的进展完全在计划之内，而且我的信心更足了。

I think we're sort of exactly on track, and and I would say my confidence interval is narrowed.

所以现在我觉得，大概在二十年到三十年之间，误差在一两年左右。

So, you know, these days, I'm thinking it's twenty, thirty plus or minus a year.

哇哦。

Wow.

是的。

Yeah.

明白了。

Okay.

二十年到三十年，误差在一两年，这时间已经很近了。

So twenty, thirty plus or minus a year, that's that's soon.

对。

Yeah.

非常近了。

Very soon.

你们内部是怎么准备的，比如说

How are you guys preparing, like, internally

在谷歌？

in Google?

首先，过去一年我们做了大量底层工程工作，等我们发布类似这样的产品时，你们就能看到成果了。

We we you know, first of all, we've got to we've done a lot of engineering work under the hood in the last year, which you'll see you see the surface of when we announce things like this.

但实际上，整个技术栈几乎都得重写，要以AI为先、模型为先，现在还要以智能代理为先。

But, actually, you can almost think that whole stack had to be rewritten to be sort of AI first, model first, and now agent agent first.

没错。

Right.

这跟传统的软件开发相比，变化可太大了。

So that's quite a big change from, like, normal software.

而且，显然还有庞大的基础设施，每天要以闪电速度服务数十亿用户，这些都必须稳定运行。

And, obviously, there's all this huge infrastructure that serves billions of users every day at lightning speed, and that's all got to work.

过去一年，我们在这些方面取得了巨大进步，所以你们才能看到AI模型的成果这么快、这么有力地融入我们的产品和服务中。

And I think we've made massive improvements on that over the past year, and that's why you can see all the goodness of the AI models flowing through into our product services so quickly and in such compelling ways.

我觉得这已经是个很大的突破了。

So I think that's big.

不过说到通用人工智能，我觉得我们目前的进展还算顺利。

But, yeah, in terms of AGI, I think we're we're sort of on track.

实际上，我们现在的进度和2010年刚成立DeepMind时的预期基本一致。

And actually, we're sort of on track from where we thought back in 2010 when we started DeepMind.

那时候，像我的联合创始人谢恩就公开说过，大概2028年能实现，当时我们觉得这是一场二十年的征程，现在看来，我们依然走在正确的轨道上。

You know, people like Shane, my co founder was on record saying around 2028, 2013, you know, we thought of it as a twenty year mission, and I think we're on track for that.

你觉得要达到通用人工智能，目前还缺哪些关键能力？

What do think the capabilities that are missing are still before we hit AGI?

我觉得还缺那么几样东西。

So I think there's still a few things missing.

我觉得现在的智能系统还是有点片面，它们在某些方面表现得极其出色。

I feel like we still have kind of jagged intelligences in the sense that they're amazingly good at certain things.

你看看Omni的表现就明白了。

You're seeing that with things like Omni.

我觉得这简直让人难以置信。

I think it's kind of mind blowing.

我们现在对这些已经有点麻木了，但你可以把第一个版本看作是视频领域的‘纳米香蕉’。

We're a little bit desensitized to all of that now, but it's you can think of it as sort of like nanobanana for video, this first version.

而且

And

像你这样的内容创作者一定会玩得很开心。

I think content creators like yourself should have a load of fun with it.

不过，说到它对物理规律和世界运行方式的理解，虽然还不完美，但只要你静下心来想想，就会觉得太惊人了。

But, yeah, in terms of, like, its understanding of physics and and and and the way the world works, it's not perfect, but it's pretty mind blowing if you stop and think about it.

这些恰恰是我们想象中的通用人工智能应该具备的能力。

And those are hallmarks of what we imagine an AGI to be able to do.

当然，这对机器人和智能眼镜这类应用来说，意义重大。

And, of course, it's gonna be amazingly important for things like robotics, right, and glasses.

但另一方面，这些系统现在还是有很多做不到的地方。

But on the other hand, there's still a lot the systems fail at still.

这一点在智能体的那部分表现得特别明显。

And and you can see that in the agentic side of things.

过去六个月，它真的开始对人们的日常生活产生实际帮助了。

It's it's starting to become, I think, the last six months, actually genuinely really useful for people in their daily lives.

但对普通用户来说，还做不到完全放手不管。

But it's still not quite fire and forget for the average user.

这些系统在可靠性和某些关键环节上还是有明显问题。

I think it's there's still some reliability and some things these systems all get stuck on.

所以还需要提升一致性，以及持续学习的能力。

So those things need to be consistency, continue learning.

这方面还需要一些关键性的突破。

There's still some breakthroughs that are needed in that direction.

明白了。

Gotcha.

所以这本质上还是个研究问题。

So this is like sort of a research problem

或者说是的。

or Yes.

关于这些方面，我们已经有了大量的研究。

So we have a ton of research on things like that.

我觉得，目前任何系统中的记忆功能都还不到位。

Also, memory is still not quite right, I would say, in any of the systems.

现在的方法有点靠蛮力。

It's a little bit brute force.

这里需要更聪明的解决方案。

Something smarter, I think, is needed there.

比如长期推理和规划，还有几个关键领域。

So there's a few, you know, long term reasoning and planning.

我认为，只有少数几个领域取得突破后，我们才能真正称得上是通用人工智能。

There's a few areas, critical areas, only a handful that I still think need to be cracked before, you know, we would have something like AGI Mhmm.

那种能被称为通用人工智能的系统。

That we could call AGI.

是的

Yeah.

说实话，现在看到设备之间的个性化体验，真的挺让人惊喜的。

I mean, we're it's really cool to see the personalization across the devices now.

我记得去年我们还在聊这个，今年你们就在夏天推出了眼镜版。

I think last year, we're talking about this, and this year, you guys are releasing it in the summer with the glasses.

对

Yes.

还有手表也上线了。

And you got the watches.

对

Yes.

而且它还能和

And it connects to

你的手机连接起来。

your phone.

还有Fitbit这些设备。

The the Fitbits and all of these things.

是的。

Yeah.

一切正在慢慢整合，但记忆功能还没完全到位。

It's slowly come together, but the memory isn't, like, fully there.

没错。

Exactly.

所以现在是越来越成型了。

So it's coming together.

我觉得最让人兴奋的是，今年的眼镜相比去年有了巨大进步，很快就会普及到每个人手里。

I mean, I think one of the things that was most exciting to show is the progress we've made with glasses this year compared to last year, and soon it's gonna be in everyone's hands.

我真期待能亲自用上。

I can't wait to, you know, have that myself.

这一直是我心中的梦想——一个能随身陪伴、帮你处理日常琐事的个人助手，比如自动搞定那些烦人的行政事务和邮件，那就太棒了。

It's one of the dreams I've always had, a personal assistant that helps you with you everywhere you go and just helps you in your daily life, you know, get rid of mundane admin tasks and email would be great.

你知道的，处理我的邮件。

You know, dealing with my email.

但这样你就能腾出更多时间去做自己真正想做的事。

But it's so you have more time to do what, you know, what you want in the day.

我觉得我们现在离真正让人眼前一亮的产品已经很近了。

And I think we're really close now to having really compelling things.

在我看来，这个全场景数字助手的概念，才是智能眼镜一直等待的杀手级应用。

And in some ways, I think the this idea of a universal digital assistant, in my view, is is the killer app that the form factor of glasses was always waiting for.

谷歌在眼镜领域当然有很长的历史，但我觉得这才是真正的杀手级应用，我们一直以来都在朝这个方向努力，现在就差临门一脚了。

You know, Google, of course, has a huge history in long history in glasses, but but I think this is the actual killer app, what we've been sort of building towards now, and we're on the cusp of.

这真的让人非常兴奋。

And it's very exciting.

再把它们和你其他设备、健康数据这些整合起来，只要你愿意。

And then also integrating it to all the other devices and your health data and things like this if you want to.

然后由Gemini帮你处理这些事。

And then Gemini being able to help you with those kinds of things.

从这个角度看，Gemini 的订阅确实很有吸引力。

The Gemini subscription is quite compelling in that way.

对吧？

Right?

因为它把所有功能都整合在

Because it really has everything in

一个平台里。

one.

全都打包在一起了。

Bundled into one.

确实如此。

And Yeah.

这正是谷歌全栈能力的精彩之处之一。

That's one of the, I think, beauties and of of the power of the the the full stack we have at Google.

没错。

Right.

从芯片到模型，再到产品，全都涵盖在内。

You know, from the chips to the models to the to the products.

是的。

Yeah.

咱们往前看，一年前我们聊的时候，你说过AI有希望在我们有生之年治愈所有疾病。

Jumping ahead here, I think a year ago, we spoke and you said that AI could help cure all diseases in our lifetimes.

没错。

Yes.

但从那以后，你在AI药物研发领域看到的最大进展有哪些？

But since then, what are some of kinda, like, the biggest developments you've seen in AI, drug discovery?

你当初的时间表有调整吗？

And have your timelines kinda shifted since then?

是的。

Yeah.

我觉得时间表没变，反而更坚定了。

Again, I would say they've not shifted, but they've hardened.

它们变得更加明确了。

They've they've tightened.

所以现在我对这些时间表更有把握了。

So I'm more sure, let's say, about those timelines now.

我原本希望在未来十年到二十年内，医学能取得重大突破。

But I was hoping it would be in the next decade or two, we'd make these big advances in medicine.

现在我非常确信了。

Now I'm very sure.

我不敢说百分之百肯定，但我非常有信心，事情确实会这样发展。

I wouldn't say I'm certain, but I'm very confident that that is the case.

看到过去一年我们在Gemini和DeepMind的科学领域取得的进展，尤其是Isomorphic Labs，我们正在全力推进专门用于理解化学和生物化学的模型——这些正是设计药物所需的关键技术。

Seeing the progress we've made in the last year on our Gemini and for science side of things at DeepMind, but also especially isomorphic labs, where we're really trying to push forward the models, the specialized models that are required to understand chemistry and biochemistry, the kinds of things you'd need to design drugs.

进展非常顺利。

And it's going amazingly well.

我们刚刚筹集了一大笔资金，进一步加速这一进程。

We just raised a bunch of money to accelerate even further.

现在我们已经有了一些处于临床前阶段的试验化合物，这意味着它们离真正成为治愈手段迈出了第一步。

And we have excitingly now some test compounds in preclinical stage, which is the beginning of them becoming real cures.

我认为，当这一天到来时，将是人工智能领域最重要的转折点。

And I think that actually will be the biggest watershed moment in AI when that happens.

我同意。

I agree.

没错。

Yeah.

你说的治愈所有疾病，是一次解决一种病吗？

So when you say like cure all diseases, is it sort of one disease at a time?

有没有哪种疾病你觉得会最先被攻克？

Are there particular diseases you think come first?

是的。

Yeah.

我的意思是，我们在打造一个平台，你可以把它想象成AlphaFold，但AlphaFold只是其中一小部分，或许还需要五六到十来个类似AlphaFold级别的突破。

I'm meaning more like building a platform that and think of it as AlphaFold, but AlphaFold is only one small piece, but, know, maybe like half a dozen to a dozen other AlphaFold level breakthroughs.

这并不容易。

So it's not easy.

我们正在把所有这些环节整合成一个连续的平台，从你关注的靶点、对疾病机制的理解，一直延伸到可以进入临床试验的先导化合物。

You know, we're working on all those Pulled together into one continuous platform that goes from the target of interest, the way you think the disease is working, all the way to a lead compound that you want to test in clinical trials.

我的意思是，把通常需要十年才能完成的流程，压缩到几个月，甚至几周内完成。

So I'm talking about sort of compressing that average of, you know, ten years it can take to do that down to months, maybe even weeks.

当然，之后还有临床试验，以及药物上市前的种种后续步骤。

Of course, there's still the clinical trial stage and there's all the downstream stuff before you get an actual drug someone can take.

这部分又可能需要五到十年的时间。

That can take another ten five to ten years.

我觉得人工智能也能在这些环节上提供帮助。

And I think AI can help with that too.

但我现在主要谈的是药物发现的核心引擎——也就是从疾病特征出发，找到潜在疗法，再进入验证阶段的这一部分。

But I'm really just talking about the drug discovery engine, the bit where you go from the, you know, the the disease profile to a potential cure that you then needs to be tested.

没错。

Right.

而且并没有针对某种特定疾病。

And there's no specific disease.

我希望它能适用于所有疾病，是的。

I hope it could work on everything Yeah.

各种类型的疾病。

Type of disease

我今天在主题演讲中注意到，上面提到癌症是一种免疫相关疾病，没错。

I actually noticed on the keynote today, it said cancer is an immune source Yes.

所以我们先从这些领域入手，免疫学和肿瘤学，部分原因是这些疾病确实非常严重。

So we're starting on those areas first, immunology, oncology, partly because obviously some of those things are so serious.

因为这些疾病本身非常严重，临床推进的速度也更快。

It's more, the clinical side of that is quicker because they're such serious diseases.

在这些领域，任何能提供帮助的手段都至关重要。

Anything you can do to help is essential there.

我深感有必要这么做，我一直觉得，这正是我最初最想用人工智能去实现的目标。

And I feel a real need to I've always felt that that's what I first and most importantly wanted to do with AI.

这真是了不起的成就。

That's really incredible work.

对。

Mhmm.

接下来的问题是，等技术成熟了，你们怎么考虑让它普及开来？

Guess the follow-up question to that is, like, after the technology is ready, how are you thinking about democratization?

我的意思是，难道只有富人才能用吗？

You know, is it available just to the rich?

怎么才能让每个人都用得上？

How can everyone have it?

我们该怎么让

How can we make

这项技术真正造福全人类？

it best for the world?

谷歌在这方面非常重视，这一点毋庸置疑。

We care a lot about that at Google, obviously.

你知道，我们所有的产品都有大量用户，现在一共有13款产品，包括Gemini应用，全球用户达到九亿。

You know, all our products are used by, you know, there's, I don't know, 13 products now, billing user products, Gemini apps 900,000,000.

所以我们非常重视让全世界每个国家的人都能用上这项技术。

So we are really take very seriously this idea of democratizing the technology to everyone around the world in every country.

谷歌一直以来都是这样做的，无论是搜索、YouTube，还是其他核心产品。

And that's what Google's always done with search and all and YouTube and all its main products.

所以我们花大量精力开发那些轻量级、甚至超轻量级的模型，它们效率高、成本低、速度快，能在任何地方运行。

And that's why we spend so much time on these flash level and even flashlight level, very efficient, very performant models, but that that are cheap and fast and can be used everywhere.

当然，我们自己内部也有强烈的需求，所以才会推动所有产品都采用这种模式。

Of course, we have a big need for that ourselves internally, which is why we push that for all of our products.

但我们相信，开发者和创作者们也会特别喜欢这种设计。

But we also think it's the sort of thing that developers and builders are gonna really enjoy too.

明白了。

Okay.

咱们往前看五年吧，假设一下。

Let's let's fast forward five years, maybe let's say.

假设我们实现了通用人工智能，并且攻克了一些重大疾病。

And let's say we hit AGI, and we've cured some of the big diseases.

对。

Yeah.

那你现在最关注的是什么？

Where are you focusing your time?

这就是下一个目标。

It's the next mission.

我觉得我可能会关注几件事。

Well, there's a couple of things I think I would be thinking about.

一是我自己用AI工具，来帮助我更清楚地理解我的整体科学目标——那就是探索现实的本质。

One is using the AI tools myself to help us understand my my kind of overall scientific goal, which is the nature of reality.

对吧？

Right?

从小时候起，这一直是我最着迷的事情。

That's always what I've been sort of obsessed with since a kid.

说白了，就是做些物理研究和实验，但会借助人工智能的帮助。

And really, means doing some physics and physics experiments, but with the help of AI.

所以这是第一点。

So that's one thing.

第二点，我可能会稍微转向一些更哲学的话题，比如人到底意味着什么，人类的独特性在哪里，以及当人类开始迈向星辰大海时，社会该如何调整和适应，这又会对社会和人类的处境带来怎样的影响。

Second thing, I think, is maybe I would pivot a little bit to more philosophical topics around, you know, what it means to be human, what's special about that, how do we want society to change and adapt to this incredible transformation that is gonna happen as we, you know, start to travel to the stars and, you know, what does that mean about society and the human condition.

所以如果可能的话，我想在这方面出一份力。

So I'd like to maybe try and help out there if I can.

说到人类的意义，你有没有一些初步的想法？

Speaking about human meaning, do you have any, like, initial thoughts there?

我知道你在这上面想得特别深。

I know you're very deep in the weeds here.

但你有没有想过，在AGI出现之后，人类的意义还剩下些什么？

But have you thought about, like, what's left for human meaning after AGI at all?

在我看来，这一切都还有待我们去书写。

Well, I think there's it's all still to be written in my view.

你知道吗，我觉得下一代人会从小就自然地使用这些工具，就像我小时候接触电脑和编程那样。

You know, I think the next generation, they're gonna be growing up natively with these tools just like I did with computers and programming in the first place.

我们创造了这样一个世界，我真想看看今天的学生们在这些工具——简直像超能力一样——的陪伴下会长成什么样。

And then we created this world that I can't wait to see what the students of today are gonna do growing up with these sort of tools and almost superpowers in a way.

再把它们和传统的数学、计算机科学技能结合起来，我认为这些能力在未来十年内依然至关重要，这样才能真正理解这些系统的工作原理，从而充分利用它们。

And combining that with traditional skills in math and computer science, I think that's still gonna be relevant, at least as far as I can see the next ten years, so that you can really understand how these systems work so you can take best advantage of them.

然后，我相信会涌现出许多令人惊叹的新事物。

And then I think amazing new things are gonna be created.

我对人类的创造力充满信心，因为我们天生就具有无限的适应能力。

I've I've I've got a lot of faith in human ingenuity where we're sort of infinitely adaptable.

我们自己本身就是通用智能。

We're general intelligences ourselves.

别忘了这一点。

Don't forget.

对吧？

Right?

看看我们周围建起的一切。

Look what we built around us.

用我们原始的狩猎采集者大脑，这简直不可思议。

It's incredible with our hunter gatherer brains.

我们为什么要止步于此？

Why would we stop here?

其实，我们在适应工具方面特别在行。

You know, it's like we're amazing at adapting to tools.

想想我们现在对这些工具的看法——今天这些让我们惊叹的AI工具。

Look at the way we see even the tools we have today, these AI tools we're marveling at today.

就在几年前，这还根本不敢想象。

Just a few years ago, that would have been unthinkable.

可现在，我们早就完全适应了。

But today, we've just totally adapted to it.

就像说，对啊，就是这样。

It's like, oh, yeah.

我们现在跟聊天机器人对话，水平都快赶上图灵测试了，还能用全能模型生成任何视频。

We chat with bots, basically, Turing test level, and we make omni models that can create any video.

然后我们就觉得，嗯，挺正常的。

And we're kinda like, yeah.

这事儿吧，已经没啥稀奇的了。

You know, that's kinda normal.

对啊。

Yeah.

我觉得这正体现了人类适应能力的积极一面，往后肯定也会是这样。

And so I think that's the good side of human adaptability, and I'm sure that will be the case going forwards too.

这就像我常举的Waymo的例子，你第一次坐进去，拍个视频，

It's kind of the Waymo analogy I like where, like, you try Waymo once and, you know, you're taking a video

对。

Yeah.

车里头那场景，简直让人不敢相信。

In the car and it's insane.

是的

Yeah.

无人驾驶汽车，第一次坐的时候觉得挺神奇，但第二次你就觉得，哦，这

Driverless car, car, but then you do it a second time and it's like, oh, this is

对

Right.

挺正常的

Kinda normal.

挺正常的

It's kinda normal.

没错

Exactly.

居然觉得这很正常，反而有点奇怪

It's weird that it's normal.

对吧

Right?

现在这些AI工具也是同样的情况。

And it's the same with these AI tools now.

但这也恰恰说明了一个问题。

But, again, that's the sign.

因为我们的大脑是通用型的，非常通用，尽管它原本是为了狩猎采集这种特定生存方式进化出来的。

That's because our brains are general and so general, even though they were evolved for a particular purpose, you know, hunter gathering.

但我们人类本来就是会制造工具的物种。

But we were tool making species.

这正是我们和其他动物的根本区别。

And that's what separates us from the other animals.

我认为，正是这一点让我们走到了今天这个现代文明的阶段。

And I think that's led us to where we are today with modern civilization.

我觉得，人工智能不过是下一个自然的进化步骤。

I think AI is just the next step.

随着AI越来越强大，你认为哪些人类技能会变得越来越有价值？

What human skills do you think get more valuable as AI just increasingly gets more capable?

我觉得这一点在短期内非常明确，至于十年后就不好说了。

Well, I think it's very clear in this certainly in the short term, it's harder to predict about, you know, ten years from now.

但在我看来，未来五年内，审美、设计感、原创思维，以及能把不同领域知识融会贯通的能力，会变得特别重要。

But I would say in the next five years, it's very obvious to me that taste, design sensibility, original thinking, being able to synthesize different subjects together.

我觉得知识面广、技能全面的人会更有优势。

I think being quite broad minded in your knowledge and skills.

我觉得这会非常棒。

I think it's gonna be amazing.

那些既有创意又懂技术的人，像你这样的，用上这些工具后会占据绝佳的位置——把技术跟创意结合起来，不管是艺术还是科学领域都一样。

People who are kind of creative and technical, people like yourself, I think are gonna be in an amazing position with these technical tools, but then combining it with creativity, whether that's in in in the arts form or the science form.

我一直说，真正优秀的科学家和普通科学家之间的区别，不在于技术能力。

You know, I've always said what separates the great scientists from the good scientists is not their technical capability.

因为作为专业科学家，技术水准肯定都不差。

Because if you're a professional scientist, you're obviously very good technically.

真正关键的是他们的创造力、审美和判断力。

It's their creative capability, their sort of taste and judgment.

我觉得这会在秋季实现。

And I think that's gonna be at the fall.

还有，无论你是做产品、拍电影还是做游戏，都要学会真正与你的受众建立情感连接——你该怎么打动那些你为之创作的人？

Also, connecting with your audience, whether you're building a product or making a film or game, you know, how do you really connect emotionally with the other human beings that you're making this for?

这种能力不会来自工具，也不会来自AI。

That isn't gonna come from the tool, the AI tools.

我认为这必须来自人类创作者和人类专家。

I I think that's gonna come from the from the human creator and the human expert.

没错。

Right.

所以，培养品味和判断力，同时持续学习AI相关的技能，没错。

So taste and judgment and continue to learn the skills, the AI skills needed Yes.

把这两者结合起来。

Combining the two.

对。

Yes.

因为你对这项技术理解得越深，就越能充分利用它，始终站在前沿。

Because the deeper you understand the technology, the more you're gonna be able to take advantage of it and keep at the at the forefront.

没错。

Right.

好的。

Alright.

时间差不多了，但我还想问你一个问题：新一阶段的重点明显放在智能体和AI上，但最近有没有什么你觉得特别重要却被人忽视的方向？

So we're almost out of time, but the one thing I wanna ask you is the new cycle is very much focused around agents and AI, but is there anything specific that you think is just been on your mind recently that is underhyped?

我觉得我们正迎来一个全新的智能体时代，这一点我们已经明显在推进了。

Well, I think we got a I'm very excited about this new AgenTik era, and and you can see us leaning into that.

整个IO战略的核心就是这个，比如Spark、反重力，还有全新的Flash。

The whole of IO is really about that, right, with Spark and antigravity and the the new Flash.

不过，我们当然也得考虑这背后的安全性问题。

But, of course, we've also got to think about the security side of that too.

我觉得我们特别擅长的一点，就是把前沿的强大能力，和稳定可靠的性能结合起来。

You know, that's one of the things I think we can do really well is combine really amazing frontier capabilities, but with reliability and robustness Mhmm.

还要能跟大家已经熟悉并喜爱的那些产品生态无缝连接。

As well as connecting to, you know, the ecosystem of products people already know and love.

所以我觉得，这将是下一阶段的重点。

So I think that's the next phase.

不过我也担心，现在大家对某些模型的安全问题已经开始有顾虑了。

But I do worry about you're seeing it a little bit with the cyber worries about some of the models.

我觉得这仅仅是开始，我们既要为这些令人兴奋的机会欢呼，也得提前防范潜在的风险。

And I think that's just the beginning of some of the issues that we need to make sure we guard against as well as getting excited about all the amazing opportunities.

所以这些事最近一直在我脑子里转悠。

So I think that's playing a little bit on my mind.

也许现在正是推动建立一些行业标准，甚至加强国际合作的好时机。

Maybe this is the time now you know, to try and push some standards and maybe international cooperation.

明白了。

Got it.

非常感谢你抽出时间。

Well, thanks so much for your time.

每次和你聊天都很愉快

It's always a pleasure talking

是啊

to Yeah.

和你聊天总是这么开心

Lovely talking to you as always.

非常感谢你

Thanks so much.

抽出时间来做这次交流

For doing this.

谢谢你

Thank you.