第16集 - 为生命科学构建人工智能

大家好，我是安德鲁·梅恩，欢迎收听OpenAI播客。

Hello, I'm Andrew Main, and this is the OpenAI Podcast.

今天这期节目，我们邀请了研究负责人焦 Joy 和产品负责人王云云，聊聊OpenAI在生命科学领域的应用。

On today's episode, we're talking with research lead Joy Jiao and product lead Yunyun Wang about OpenAI for life sciences.

我们会看看，新一代模型在生物学和医学上能带来哪些以前不敢想的突破，以及如何负责任地部署这些顶尖技术。

We'll explore what new models are making possible in biology and medicine and what it takes to deploy the most advanced capabilities responsibly.

这让它能突破以前根本不敢想的难度和发现边界。

This allows it to kind of reach new levels of difficulty and discovery that we didn't think was even possible before.

把顶尖专家级别的知识，交到更多人手里。

Putting like really capable expert level knowledge in the hands of a greater amount of people.

我们有个口号是：用算力攻克所有疾病。

One of the taglines was to scale test time compute to cure all disease.

这其实就是我们团队的口号。

So that is like our team tagline.

我们一开始只做了个基础API，后来出了ChatGPT，对话式交互特别适合文本处理；接着代码能力上线，再后来是代码模型，再往后是编解码器。

We started off with just a basic API and then we had ChatGPT, which is more conversational, was really good for text as code became a capability, went through basically code models and then codecs.

现在越来越多的科学家和生命科学领域的研究人员在使用这些系统，这意味着我们需要调整工具，更好地适配他们的工作方式吗？

Now that you're getting more scientists and the life sciences working on these systems, does that mean things have to evolve to help with the way researchers might work with these tools?

是的，我们非常期待推出这套专门面向生命科学的模型系列。

Yeah, we're really excited to build and deploy the life sciences model series.

这是一个全新的、聚焦生物化学的模型系列，完全围绕复杂的生命科学研究流程设计。我们目前的重点是提升机制性理解，从基因组学和蛋白质研究入手，优先支持早期发现场景，因为我觉得这正是当前科研的一大瓶颈——只要增加思考时间、算力，再结合更强大的AI模型，就能真正突破这些障碍。另外，如何把这些模型无缝融入工作流程，也是一大关键。我们已经有多种产品形态可以部署，看到ChatGPT上出现了很多优秀的文献综述流程，而这些模型也在推动长周期智能代理工作流的前沿发展。我们在Codex上正全力支持这类应用。

So this is a new biochemistry focused model series that's really anchored on these very complex life science research workflows and we're focused on adding new mechanistic understanding starting with genomics understanding and protein understanding and really focus on early discovery use cases because we feel like that's like one of the core bottlenecks that greater thinking time, greater compute and really like leveraging like more capable AI models can help meaningfully scale some of these like research barriers and I think there's also like a model orchestration piece of actually how to embed this into workflows and it's been really great, first of having all these different product surfaces to deploy to, we're seeing a lot of really great literature synthesis workflows happening on ChatGPT and these models really push the frontier of long trajectory agented workflows and we're really able to empower that on codex.

在模型协同方面，企业级应用特别看重可重复性和稳定性，我们正通过开发一些面向特定转化生物研究用户的插件来解决这个问题。

And more on the model orchestration piece is that I think for enterprise use cases there's like this reproducibility and repeatability element and we are trying to overcome this by working on like some of the life sciences research plugins that we're shipping for very specific translational bio users.

这个生命科学插件已经集成了50多项功能，本质上都是经过模板化的可复用流程——比如你需要跨多篇论文做证据匹配搜索，或者做通路分析这类常做的任务，现在只需一键部署就能完成。这也是我们平衡专业化与规模化的一种方式，我们希望……

So the life sciences research plugin has over 50 skills which are essentially templatized repeatable workflows that if you need to whether do some sort of cross evidence match and search across various different papers or do pathway analysis, something that's repeatable that you often do, we can have almost like a one click deploy option by using our life sciences plugins on top and that's also how we're kind seeing the balance between scaling for very specialized purposes, something we're hoping to

下一步能拓展到临床应用，同时也要让这些工具适用于所有基础生物学研究。

get into is maybe clinical purposes but also make it very general use for all foundational biology.

我觉得，只要善用工具，模型的能力就能走得更远。

I think the models can get quite far by using tools.

比如，我们可以在研究流程里直接集成开源的蛋白质结构预测算法。

So for example, we can use open source protein structure prediction algorithms inside a research stack.

在这种情况下，模型其实就像个普通的计算生物学家在干活。

And in this case, the model was acting kind of like a regular computational biologist.

你平时就是跑这些工具，看看结果，再稍微调一调输入参数。

You will kind of go run these tools on a computer, you will look at the alpha, you will tweak the input a little bit.

我觉得我们的模型现在就已经能做这些了。

So I think that is something our models can already do.

我觉得让模型更强大的关键，是把它培养成真正的生化专家。

I do think what will make the models even more powerful is to start to turn them more into kind of a biochemistry expert.

有了这种直觉和专业经验，你就能更聪明地用这些工具，更快找到正确答案。

And I think with this kind of intuition and expertise, you can use these tools even more intelligently and get at the right answer more quickly.

你当初是怎么对生命科学产生兴趣的？

How did you get your interest in life sciences?

我嘛，说起来最初的专业背景其实就是生命科学。

My, I guess original background was actually in life sciences.

从小我就对生物学特别着迷。

So I've always been interested in biology as a kid.

我大概十年前在哈佛拿到了系统生物学的博士学位。

I got my PhD in systems biology around like a decade ago from Harvard.

我觉得学术界挺有意思的，但节奏有点太慢了，不太合我胃口。

Found academia to be very interesting, but the pace was a little bit more slow moving than I would've liked.

那种天天得亲自动手在实验室里，把一点点液体从这个管子倒到那个管子的经历，真的挺耗人的。

And I think just the experience of kind of like having to physically be in the lab and kind of like transferring small amounts of liquid from the one tube to another.

我更想要个节奏快点的领域，能让我自己掌控进度，不被拖着走。

I think I wanted something a little bit faster paced where I felt like I was more in direct control of like my own velocity.

于是我就转行做了软件，最后来到了OpenAI。

So I went from that to software and I ended up here at OpenAI.

现在重新回看生物学，对我来说简直像回到了原点。

And so this is kind of like a full circle moment for

我正开始重新关注生物学，

me where I'm like starting to look

想想怎么用AI来加速当年那个自己。

at biology again and looking at how to accelerate my previous self with AI.

所以啊，我真的特别期待AI在这个领域能取得什么进展。

So yeah, really excited to see what progress AI can make in this space.

你就觉得：这速度也太慢了吧。

So you're like, yeah, this is too slow.

干脆转投AI，把节奏加快，好让我重新回到这个领域。

Let me go off an AI and speed it up so I can get back into it.

对，不过从我这边看，我真是一点都不想再碰移液枪之类的东西了。

Yeah, except, you know, from this end, I don't really ever want to touch a pipet or anything again.

我更希望让FortiBrowlots替我干这活儿。

So I would prefer FortiBrowlots to do it for me.

是啊。

Yeah.

我们经常拿这个开玩笑。

We joke about that a lot.

我们做这个的一大动力就是，能自动化移液，以后再也不用亲手干这事儿了。

A lot of our motivation for this is we can automate pipetting and never have to do that again.

有意思的是，我一直在看你们和Ginkgo Bioworks的合作，就是把GPT-5和AI系统跟机器人实验室结合起来，居然能把流程提速这么多，能跟我们讲讲吗？

Well, that's what's interesting is I was looking at what you all did with Ginkgo Bioworks and the idea of taking GPD five and taking an AI system and then working with a robotic lab and how it was able to speed things up, could you tell us a little about that?

是的，Ginkgo这个项目挺有意思的，我记得刚启动的时候，大概是2025年7月吧。

Yeah, the Ginkgo work is interesting because I think when it started, I think it was like July, 2025.

那时候GPT-5刚训练完不久。

And at that point, GPT-five had just finished training.

我们其实完全没底，不确定这些模型能不能搞懂生物学。

We were really not sure if the models could do any kind of biology.

我们的训练数据里生物学内容本来就很少。

We didn't really have that much biology on our training data.

主要是数学和计算机科学，这也很合理，毕竟这些问题都有明确答案。

It was mostly math and computer science, which I think makes sense because these things have verifiable solutions.

但生物学不一样，除非你真去实验室做实验，否则根本没法验证对错，对吧？

And this is usually not the case in biology unless you have, you can go and do the experiment in a lab, right?

所以当初跟Ginkgo合作时，我们最根本的问题就是：这模型到底能不能干点生物学的事？

So when we started the collaboration with Ginkgo, it was really, can the model do any biology at all?

它能设计出真正能生成目标产物的实验吗？

Can it design experiments that actually make reactants, like make the product that we want?

结果特别出人意料，当GPT-5和Ginkgo一起设计出第一批实验，数据回来时，我们居然真做出了有量的蛋白质。

So it was actually quite surprising, I think when GPD5 designed the first set of experiments with Ginkgo and the results came back where, oh, we made a non zero amount of protein.

这真的挺让人惊讶的。

That was actually quite surprising.

从那时候到现在，才短短六个月，现在回头看，模型能加速科研这件事，反而觉得再自然不过了——这本身才最让人震惊。

And then I think progressing from that point in time, which is just roughly like six months ago to now, where it actually just feels quite obvious that our models can accelerate science is actually just really surprising.

这恰恰证明了可能性的边界在哪里。

And it really shows that the art of the possible, I think.

在Joy和Ginkgo团队做这个实验之前，我们自己其实也没底。

I think before that experiment led by Joy and the bio Ginkgo team was conducted, I think we really didn't know like for ourselves.

我一直说，只有真正动手做实验，我们才真正明白能做成什么，现在我们还有几个和别的团队合作的实验在进行中，这种加速才是我们真正想要的。

And I always say like we kind of learn that for ourselves when we engage in these experiments and we have a few more in the works with others and I think that is the type of acceleration that we're looking for.

处理高通量实验数据特别难，计算量巨大，而很多科研流程里，真正拖慢进度的，根本不是技术，而是人。

Ingesting high throughput experimental data is really difficult, it's very compute intensive and I think for a lot of these scientific workflows, like the true bottleneck for the speed and progress of scientific acceleration is at almost a human bottleneck.

我觉得我和Joy展望的未来是，瓶颈不再是人，而是算力了。我们可以部署多个小型子代理并行协调，把任务拆解开来，让研究人员能专心分析和解读这些实验产出的最有价值的洞察。

And I think the future that me and Joy see is that it's no longer human bottlenecks but rather maybe compute bottlenecks and we're really able to like deploy mini sub agents doing parallel orchestration to divide and counter all these tasks and the researcher can now spend their time on like really analyzing, interpreting like the most meaningful insights coming out of that.

所以云云，你是怎么进入这个领域的？

So Yunyun, how did you get into this?

说到这个，回头想想，我在OpenAI的大部分时间其实都在做和生物研究相关的工作，形式不同而已。

Yeah, I think reflecting back, I've actually been working on like biology research in some shape or form for a majority of my time here at OpenAI.

我最早是做生物风险防控和生物防御项目的。

So I first started on working on bio risk meditations and a lot of our bio defense initiatives.

现在转到生命科学的研究这边，我才真正体会到这个问题有多难，从两个角度去碰它，让我特别有感触。我最初接触湿实验研究，其实是从传染病和生物学工作开始的，所以我对生物安全的兴趣一直都在。现在模型能力越来越强，又能用在有益的方向上，做生命科学相关的事，真的感觉特别对时机。

And so I feel like coming to now working on the life sciences research side gives me like just appreciation for how difficult this problem is and tackling it from both sides and I've, my initial entries point into wet lab research was actually through doing a lot of infectious disease and biology work, so I think I've always gotten like the interest in biosecurity in that way so this just feels like a really great moment like right now to work on it especially when our models are getting more capable with beneficial use and just general life sciences.

OpenAI专注生命科学有多久了？

How long has OpenAI been focused on life sciences?

我觉得是我们在设计能力评估的时候，才真正意识到这事儿是能做成的。

Yeah, I would say it was really the way we design our capability evals that show us that this is possible.

大概两年来，我们一直在做各种早期实验，现在又加入了像全自动湿实验这样的模型闭环实验。

So it's been, I think for at least like two years now that we have worked on a lot of our early research research experiments and now with the gentle autonomous like wet lab model in the loop experiments.

我们还有几个在这一领域的研究伙伴，真的让人特别期待。

I think we have a few more research partners in the space that we're really excited about.

现在还不能一一列举，但化学设计、蛋白质设计、酶设计这些方向，AI特别有优势，很多人都在关注。

I think I can't actually name everyone right now, but there's a lot of stuff kind of in the chemical design, protein design, enzyme design space that I think is very AI native and a lot of people are interested in.

理解世界怎么运作，理解化学反应、细胞互动、细胞内通路如何协同，最终能不能加速新药研发？

So understanding how the world works, understanding how chemicals react, understanding how cells interact, how pathways inside cells interact all the way to can we accelerate drug discovery?

比如给定一种疾病模型，帮科学家搞清楚它的发病机制。

So given a disease kind of model help scientists understand the mechanism.

一旦确定了靶点，我们能不能快速找到并设计出针对它的药物？

Can we once given a target actually find, design a drug against that target?

甚至能不能加快FDA的审批流程？

Can we even accelerate the FDA approval process?

我觉得AI在这整个链条的每一步都能发挥作用。

So I think there's a role for AI to play kind of at every step of this pipeline.

说真的，AI在方方面面都有巨大潜力。

And yeah, I think just I I think there's a lot of AI possible in in everything.

我去过一些前沿实验室，外面看总觉得那地方特别高大上。

I've been to some of those cutting edge labs and on the outside, have this impression of it.

可一进去，真看到有人摆着一排排培养皿、样本，就一个研究生在那里咔嗒咔嗒地操作。

Then you walk in there and you literally see somebody with, you know, a row of petri dishes, a row of samples, and just some grad student going click, click, click.

我当时就想：原来科学的节奏就是这样的啊。

And I'm like, oh, this is the pace of science.

比我还慢呢。

Is fast as More me.

但没错，就是这样。

But yeah, exactly.

你一看这速度，真受不了，非得想办法提速不可。

You're like, enough of this, I gotta go speed this up.

可我们老忘了，科学的节奏本来就是靠人手能多快做出来决定的。

But we forget that's often the pace of science is, just how fast the human hands can move through that.

而有了这样的工具，就特别让人兴奋。

And with a tool like that, it's kind of exciting.

当你开始用这些工具去探索新的治疗路径，或者只是做评估时，也会带来一种可能性——这些技术可能被用在不太好的地方。

When you start using these tools to maybe think about new pathways for treatments or just evaluate, you also introduce the idea that these could be used for things that maybe are less desirable.

生物武器是个经常被提到的问题，毕竟如果AI能找出代码漏洞，那它也可能找到基因层面的漏洞。

Bioweapons are something that comes up a lot, the fact that if you know an AI can figure out how to do a code exploit, might be able to figure out how to do a gene exploit.

你们是怎么应对这个问题的？

How are you addressing that?

没错，这确实是个好问题，我认为这也是目前我们最关注的、随着AI能力提升而上升的严重风险之一。

Yeah, that's a great question and I think it is just probably one of the most severe risks that we're currently really tracking for rising AI capabilities.

我们最初的应对思路是，怎么去评估信息风险——什么时候AI可能会提供合成危险病原体的最后一步？

Our first approach to that was really thinking about how do we assess for information hazards, at what point does a model now maybe give the final step in a synthesis of a dangerous pathogen?

但我们发现，那些前序步骤看起来都特别普通，根本很难分辨出来，换句话说，一个正当研究者走的步骤，和坏人走的步骤，看起来几乎一模一样。

And what we found is that the precursor steps to that really looks very benign and it's really hard to distinguish between, so another way to put it is like the same steps that a beneficial, like a legitimate actor might take looks very similar to the ones that A.

你从大肠杆菌开始。

You start with E.

你从某种……

Coli, you start with something that's,

对，你说得完全对。

Yes, you exactly.

我依然认为，我们在应对这个问题时采取非常谨慎的风险规避策略是正确的。但现在，我特别看好‘差异化访问’和‘负责任的部署’，这已经成为我们所有安全措施的核心。我们得清楚，用户群体是分化的，我甚至觉得，未来的发展方向就像‘模型即职业’——就像不同模型有不同的性格，你得根据手头的任务，调用最合适的一个。

So I still think that we made the right call for really taking a very risk averse approach to that, but now I'm really excited about differentiated access and responsible deployment as really a core pillar of all of our safeguards work and really understanding that there are different user segments and I almost feel like the future we're going towards is something like models as professions similar to how models have different personalities and sometimes you want to invoke the right one depending on the type of workflow you're looking at.

这其实跟生物医学研究很像：做药物研发的科学家，他们需要访问的数据往往管控得非常严格，而且使用者基本都是博士，具备顶尖的生物学专业知识。

So I think how this translates is similar to how biologists working on therapeutics and their research, they require access to, datasets are often very tightly controlled or they require access to just expert level, like they all have PhDs and have like expert level like biology like knowledge.

那这个逻辑怎么应用到模型上呢？

How does that translate over to models?

我觉得，我们得像对待生物研究那样，既采用同样的培训方式，也要有同等严格的安全措施，把它们部署成企业级的高安全级别系统。

I think that's why we have to kind of similarly take to the same training approach but also the same security approach and deploying that in like a way where we can have those very heightened enterprise grade controls in place.

你刚提到了安全防护。

So you just mentioned safeguards.

能具体说说，这在当前场景下是怎么体现的吗？

Can you explain how that applies here?

在哪些地方我们特别需要这些防护？

Where you would need them?

你为什么需要它们呢？

Why you would need them?

对。

Yeah.

我们为几乎所有模型在不同风险领域都精心设计了新的安全措施，但说到生物领域，这是第一个同时兼具双重用途风险和能力风险的例子——随着我们能力的提升，风险也随之上升。正因为如此，当我们发现模型的推理能力出现显著跃升时，虽然当时没什么先例可循，我们也没第一个启动这么高的安全标准，但还是决定必须谨慎行事。

So we very thoughtfully design and design new safeguards for pretty much all of our models across very different risk areas but I think when it came to bio, this was like the first dual use risk that is both also a capability risk so it very much correlates with how we as capabilities improve, the risk correlates and I think that's why our first approach, when we really, there was no precedent for a lot of this work and we weren't the first to really activate these high safeguards when we saw significant reasoning jump in our model capabilities.

我们特别想确保这事做对，而最好的方式就是逐步推进。

We really wanted to make sure that we did it right and I think the best way to get it right is to incrementally deploy.

是的，我觉得这中间的界限非常微妙：一边是能加速良性科学和有益研究的强模型，另一边是可能被恶意使用者利用的模型。

Yeah, I think it's really a fine line between having a very capable model that's capable of accelerating benign science and beneficial science versus a model that could be used by a bad actor.

而最安全的模型，大概就是完全没能力的那个，对吧？

And I think the safest model here would be a model that just had no capability, right?

你——

You-

那可就太差了。

It's not very good.

是的，这模型不算厉害，但特别安全。

Yeah, it's not very good, but it's very safe.

反过来，如果你有个几乎能洞悉所有实验细节的‘全知型’模型，一旦落到坏人手里，后果可能很严重——比如有人直接让它设计一种新型高传染性病原体。

And then on the other hand, if you had a model that is basically an Oracle of the physical world, that basically knows everything about every experiment, that model could fall into the wrong hands and do potentially very bad things because someone can go and say, hey, design a new novel pandemic potential pathogen.

而这个模型完全能自己动手把这事给办了。

And the model can just go and do that autonomously.

所以我们得搞清楚，这两者之间到底该怎么划线，谁该有权限接触这种超强模型，谁不该有。

So I think we need to kind of figure out where we draw the line in between the two and kind of think about who gets access to a potentially very capable model and who doesn't.

我们在所谓的‘公开访问流量’中发现，光看用户输入的提问，根本没法判断他们真实意图。

And what we found in kind of what we call general access traffic is that it's very difficult to figure out what a user's actual intentions are just from kind of reading a prompt.

举个例子，有人问：‘帮我克隆一个基因。’

And I think as an example of this, let's say someone says, hey, help me clone a gene.

模型甚至可能根本不知道具体是哪个基因，但它照样能给出一套完整的操作流程。

The model might not even be given what the gene is, but it can come up with a protocol for it.

这个基因可能是绿色荧光蛋白，也可能是某种毒素。

And so this gene could just be something like green fluorescent protein, or it could be a toxin.

光从对话上下文里，根本没法判断真实意图。

And there's basically no way to figure that out from the context of the conversation.

所以在实际应用中，这成了一个特别棘手的问题。

And so this becomes a very difficult problem in production.

就像云云说的，我们最终选择偏向安全，一旦觉得可能被滥用，就让模型主动拒绝——比如回复：‘抱歉，这个我帮不了你，但我可以给你一个大致的流程概述。’

And basically, I think, like Yunyun said, we decided to kind of on the side of safety here and basically say that, Okay, if we think that there is a potential for misuse, we either have the model kind of self refuse the user, in which case it tends to say things like, sorry, I can't really help you with that, but I can give you a high level overview of this protocol instead.

这确实让不少专业科学家很恼火，但也能理解。

And this unfortunately very much annoys our kind of professional scientists understandably.

除此之外，我们还叠加了多层防护机制。

And then we also kind of have multiple layers of mitigation on top of that.

但要说真正释放模型的全部能力，关键还是得实现分级访问。

But I think really to unlock full capabilities of our models, what we need is this differentiated access.

也就是说，我们必须清楚用户到底是谁。

And what this means is we know who the user actually is.

他们得是正规研究机构或药企的在职专业人士。

They are a professional working at a legitimate research institution or a pharma company.

由于这些机构有严格的监管规定，我们知道他们用的所有试剂、所有细胞系都全程可追踪。

And because of the regulations around these institutions, we know that for example, all the reagents are being tracked, all the cell lines that they're using are being tracked.

所以我们能确信，这是一位正规用户，而不是某个在地下室胡乱折腾的陌生人。

And so this gives us confidence that this is a legitimate user and not a random person in the basement doing who knows what.

这让我们能给他们提供比普通用户更多的功能权限。

And that allows us to give them basically more capabilities than we are able to provide to the general access traffic.

你现在在实验室里用这些模型时，具体能做些什么？目前的能力到底怎么样？

What can you do right now if you're working with the models, you're working it within a laboratory, what would you say the capability is at this moment?

我觉得大家用模型的方式其实差别挺大的。

So I think people use the models for very different things.

我最近跟一些大实验室的人聊过，他们是怎么用我们的模型做编码相关工作的。

I've talked to people in the bigger lab recently on kind of how they've been using our models on like codecs.

有时候需求简单得就一句话：‘你能帮我做个表格吗？’

And sometimes it's as simple as, Hey, can you write a spreadsheet for me?

我可不是只想减少点移液操作的次数那么简单。

I don't want to just minimize the number of pipetting steps that I have to make.

这让我特别有共鸣，因为我在读研的时候也亲手干过同样的事。

And this hits me very hard because I had done the same thing by hand in grad school.

这其实就是一个很基础的数学软件操作。

So that's like a very simple, just mathematical software operation.

但也有更复杂的任务，比如：你能用这些生物设计工具给我设计一个酶吗？

And then there's much harder tasks like can you design a enzyme for me with all these biological design tools.

我觉得这里面的复杂程度跨度特别大。

So I think there's a huge range of sophistication.

确实，让我特别兴奋的是，我们能用模型当一个更强大的筛选器，真正去检验和评估那些新颖的想法。我注意到，我们的很多研究伙伴和用户在做科研时，往往需要一种不同的对话方式或提问风格——模型得像个挑剔的怀疑者，就像人类科学家评估原创性和可行性那样。

Yeah and something I'm very excited about is how we can use our models to be a more powerful discriminator and like really testing and assessing like new novel ideas and I think something that I've been noticing as a trend with a lot of our research partners and also the users of our models is that models for scientific research and tasks almost require a different persona or a different prompting style so I often feel like a model that is much more scrutinizing or a skeptic at good ideas is it's very similar to how human scientists would go assess like originality and feasibility.

说白了，就是帮我们从海量新论文和前沿假说里，挑出那些真正可行、值得去验证的，才能推动真正的突破。

It's really like I think helping understand like out of all the new papers and new publications out there that push the frontier of all these hypotheses, what are the ones that are really feasible and valid for testing that's going to help lead to new breakthroughs.

再比如把这套思路用在疾病靶点筛选上，潜在的药物靶点多得数不清，关键是怎么把范围收窄。我觉得这时候模型的辅助作用就体现出来了。

And then translating this to something like disease target screening selection, the potentials for these drug targets are endless but it's really like narrowing down the aperture and I feel like that's where the assistance comes.

这种工作要大规模做起来太难了，而有个模型能帮我们加速这个过程，我觉得这正是我们负责任地把模型交给这些用户时，最期待看到的直接价值。

This is extremely difficult work to do at scale and having a model that can empower and accelerate that process I think is kind of like one of the immediate impacts we're hoping to see by responsibly deploying this model to those users.

这轨迹看起来真有意思。

It seems like it's a very interesting trajectory.

你从当初用GPT-3的API、GPT-3.0.5，到后来出了ChatGPT，现在又有了ChatGPT应用，还有Codecs，一路走来变化真快。

You went from there was, you had GPT-three on the API and GPT-three 0.5, then you get ChatGPT, and now we have ChatGPT apps, and now we have Codecs.

感觉能用这些工具干的事儿，是越来越多了。

And it sounds like these things just the number of things you can do with this continues to grow.

你觉得这事儿接下来会怎么发展？

How would you see this building?

你是不是觉得，它最终会变成一个全能的基础设施，啥问题都能靠它来解决？

Do you see this as basically just becoming a complete infrastructure for kind of every kind of inquiry you might wanna pursue?

对，我们的理想就是让科学工作流的那些基础操作，全都在Codex上跑起来。

Yeah, I think the dream is to have a lot of the basic foundations of scientific workflows happen on codex.

目标就是让Codex能搞定电脑上能做的所有事儿。

And I think that the goal is to have codex to pretty much be able to do everything that is possible to do on the computer.

当然，我们还想更进一步，比如把它连到机器人系统上之类的。

Of course, we also want to extend beyond that with kind of hooking it up to robotics and so forth.

但现在我们其实已经能做了，比如我们远程或笔记本上有好几台开发机，可以直接说：‘Codex，去把这些代码在所有远程开发机上跑一遍。’

But I think right now we already do things, for example, if we have a bunch of different dev boxes on our remote, on our laptop, we can actually say, Hey, Codex, go and run this code on all of these different dev boxes that are all remote.

然后Codex就能帮你搞定。

And then Codex can do that.

你可以说：‘帮我监控一下这个。’

We can say, monitor this for me.

它就自己去忙别的事了。

It kind of like go away and do something else.

而Codex会默默帮你盯着所有日志。

And the Codex is like, they're watching all the logs for you.

它还能专门为你写一些小工具，用来分析特定数据、可视化数据。

It can build a lot of just kind of fit for purpose software for analyzing specific pieces data, for visualizing data.

比如我们团队里做实验生物学的，以前都是直接发原始数据，但最近我发现，大家开始发HTML文件了——都是Codex生成的漂亮界面，里面还有旋转的蛋白质模型。

So for example, if we have experimental biology data that we're sending each other on the team, what I've noticed recently is instead of sending the raw data, we've started sending HTML files, are just these kind of like beautiful UI stack Codex has built with kind of like spinning proteins.

这真的彻底改变了我们之间分享和协作的方式。

And it's actually just really it kind of changes the way that we share with each other and collaborate.

是的，我们刚开始规划用户和组织怎么采用这个技术时，原本设想的是每个科学家都能有个专属助手或搭档，帮他们整理和优化团队的整体产出。而下一步的升级，就是把整个科研机构都搬上来——一个项目团队可以部署一整支智能代理队伍，分工协作、并行处理任务，模仿现有的工作模式，慢慢摸索出它们怎么协同解决更大规模的问题。

Yeah, when we first started mapping out how users and organizations might adopt this, I think we envision that each scientist would get their personal assistant or their coworker and this is a way that they can kind of stale their collective output and then the next paradigm of that would be scaling up whole research institutions where a whole program team can actually deploy a workforce of various agents and they can all do parallel task delegation, kind of mimicking a lot of these existing patterns and we can figure out the pieces of like how they can all collectively like work together to solve like larger tasks.

挺有意思的是，OpenAI 一直强调算力的重要性。

It's interesting because OpenAI has talked about the need for compute.

我们有时候就简单觉得，哦，能多聊几句、多问点问题就行。

I think that sometimes we just sort of think like, okay, so I can have more conversations and stuff.

但当你真在说把这些工具打造成完整平台、用于科学探索的时候，就会发现算力优势其实至关重要。

But when you're talking about the idea of building these tools to become entire platforms or scientific exploration, it sounds like the compute advantage is really critical.

对。

Yeah.

我觉得我们可以从两个维度来看算力怎么扩展。

I think there's two different axes we can think about how we are scaling compute.

大家比较熟悉的那个，就是把模型做得更大。

The one that I think everyone's familiar with is just getting bigger models.

从 GPT-2 到 GPT-3，模型规模确实暴涨了一大截。

And I think as we went from GPT-two to three, there was a huge size increase.

而且模型还展现出一些特别惊艳的涌现特性。

And there were just these amazing emergent properties from the model.

想想当年GPT-2刚发布的时候，我们都惊呆了，它居然能写出一篇关于独角兽的通顺文章。

I mean, thinking about when GPT-two was released, we were all kind of collectively amazed that it was able to write a coherent article about unicorns.

但现在，我们已经完全进入了一个新世界，对吧？

And now we're in a completely different world, right?

而这一切，很大程度上得益于模型架构的突破。

And a lot of that is driven by model architecture.

没错。

Yes.

但更重要的是，模型参数量的激增，让它达到了我们以前根本不敢想的智能水平。

But also just the number of parameters in the model just allows it to achieve this incredible intelligence that we never thought was possible before.

另一个维度，就是我们说的推理时计算扩展。

And then on the other axis, we have what we call a test time compute scaling.

也就是模型在生成文本、逐个输出词元的时候。

And this is when you are inferencing a model, when it's kind of spitting out tokens.

最近出现了一种叫‘子推理模型’的东西，它能花上可扩展的时间来思考问题。

And this is a thing that happened fairly recently when we call these sub reasoning models is that you can think for a scalable amount of time.

这个时间长短取决于模型觉得问题有多难。

And this is a variable depending on how difficult it thinks a problem is.

我们甚至可以让模型一直思考，几天几夜都不停，就为了彻底想透一个问题。

But we can have the model think for days where really just kind of ways to just kind of have it think forever about a problem.

这样一来，它就能达到以前根本不敢想的深度和发现层次。

And this allows it to kind of reach new levels of difficulty and discovery that we didn't think was even possible before.

说到数据中心，大家通常只想到生成猫图或者聊天，但其实更准确的理解是，这些系统将来是用来处理那种超长期、超复杂的思考任务的。

When we think about data centers, we often just sort of think about it as generating cat pictures or doing text conversations, but I think that's really the helpful framework to look at is that these are going to be systems for doing extremely long term, big complex processes of thinking about this.

所以像Stargate这样的项目说要建大量算力，我就觉得特别合理。

And it's it's to me, it just makes a lot more sense when you know, projects like Stargate saying we're going to be building a lot of compute.

他们要的不是现在这些用途，而是为这种长远的思考任务做准备。

It's not just for what we're doing now, but it's going to be for things like that.

我们当初在Slack上宣布Teams团队时，有一句标语就是‘扩展推理算力，治愈所有疾病’。

When we had first announced the Teams formation on Slack, I think one of the taglines was to scale test time compute to cure all disease.

这简直就是我们团队的口号。

So that is like our team tagline.

这就是我们团队的模型。

That's our team model.

这目标可真够远大的。

That's ambitious.

没错。

Yeah.

我有个朋友，孩子生下来就得了那种罕见病，她到处筹款，想尽一切办法帮忙。

I had a friend whose child was born with one of those orphan diseases, and she would do fundraisers, do everything she could to try to support.

有些研究人员在努力找治疗方法，但时间不够、人手也不够。我觉得我们现在正处在一个新时代，这类工具或许能让这种情况成为过去。

Some researchers who were trying to find a cure for this but they're just not enough time, not enough people and you know, I'm hopeful that we're kind of in an age now where these kinds of tools are going to make that maybe a thing of the past.

是的，我觉得模型在这些案例里已经帮上大忙了。

Yeah, I think we're already seeing the model help a lot in these cases.

比如药物重新利用这类事情。

I think from things like drug repurposing.

比如，一种已经获得FDA批准用于其他适应症的药物，根据我们对它作用机制的理解，这个模型在很多情况下都建议可以暂时用它来缓解症状。

So for example, a drug that's already been cleared by the FDA for use in one different indication, but for kind of like from mechanistic understandings of how that drug works, the model has suggested in many different cases for maybe you can use this drug to temporarily ameliorate symptoms.

我们还看到个性化医疗方面有很多进展。

We're also seeing a lot of advances in personalized medicine.

比如，设计ASO或其他基于RNA的疗法现在已经很普遍了。

So for example, the design of ASOs or other RNA based treatments is very common.

我觉得，没错，我们其实已经非常非常接近了，能用人工智能大规模推广这种应用。

And I think, yeah, we are actually very, very close to being able to scale this up in a really vast way with AI.

我觉得再过一两年，这里就会发生巨大的变化。

I think just in the next year or two, I think we'll see very big changes here.

我认识的每个研究员，你问他们实验室最缺什么，他们都说：人手不够，人手不够，太需要有人来做这种工作了。

Every researcher I know, when you ask them what they could use in their lab, they always say more hands, more people, more people doing this kind of work.

你也会听到一些人问：AI会不会取代这些人？

And you hear some people talk about, well, is AI going to displace that?

我觉得不会。

And I think, no.

听起来这就像给所有能做的事按了个加速键。

It sounds like it's just a big accelerator for all the things that could be done.

是的，我完全同意。

Yeah, I completely agree.

你想想实验室自动化，很多瓶颈其实出在怎么把实验流程转化成平台能跑的指令上。

I feel like when you think about lab automation, for example, a lot of the bottleneck comes from actually being able to translate a protocol into something that can be run on the platform.

我们有合作伙伴跟我们说，Codex 帮他们解决了这个问题。

And we've had partners tell us about how Codex has been helping them do this.

这本质上就是个编程问题——你得懂 Ylap 是怎么运作的。

And this is kind of fundamentally, I have coding problem, have understanding how Ylap works.

再说到数据分析，我觉得我们的模型能帮那些统计学基础不强的人，也能严谨地分析数据。

And then I think thinking about the data analysis piece, I feel like having our models kind of walks through a user who maybe doesn't have the deepest understanding of statistics, they can still rigorously analyze the data that's coming in.

模型能帮他们测试不同假设，推荐合适的统计方法，还能指出数据里的潜在问题和偏差。

The model can kind of help them probe different hypotheses, or it can suggest different statistical tests, it can point out potential issues and biases in the data.

我觉得这些都能提升科研人员的能力，让他们做出更好的科学。

I think these are all ways of kind of uplifting individual scientists and helping them do better science.

但我认为，科学家始终是不可或缺的一环。

But I don't think we can ever fully replace the scientists in the loop.

所以你现在已经在实验室里了，也找到了帮自动化提效的方法。

So you've been put into the lab, you've figured out how to help with automation.

你觉得六个月后、一年后，我们会走到哪一步？

Where do you think we're gonna be six months from now, twelve months from now?

我特别希望有一天能说，AI设计出了一种新药，或者治愈了某种疾病——六个月能做到吗？我不确定，但我相信未来几年内一定会实现。

Well, I would really love to get to the point where we can say that AI has designed a new drug or cured I the don't know if that can happen in six months, but I will hope in the next few years that's going to happen.

我觉得这种迹象其实已经在整个研发流程的各个阶段出现了。

I think we're seeing signs of this happening kind of all over different stages of the pipeline.

比如在药物发现早期，像文献综述、合成路径规划，或者AI发现新的生物学机制，这些要真正变成上市新药，过程肯定非常漫长，可能得十年左右。

I think obviously earlier in the drug discovery process where you're kind of looking at literature synthesis, or the model is kind of discovering new biology for that to become a new drug on the market is going to be a very long process or possibly like a decade.

但我觉得，我们完全可以在临床试验阶段，甚至更早一点——比如安全性评估或药物设计阶段——就开始加速这个流程。

But I think there's ways that we can really speed up this process by starting at maybe the clinical trial stage for starting a little bit before then and the safety reviews or in the drug design phase.

所以，说到底，未来几年最让我兴奋的，就是这些方向。

So I think, yeah, basically that's what I'm the most excited about coming up in the next few years.

是的，对我来说，最让我兴奋的是我们的用户、我们的科学家能在平台上实现的各种可能性。

Yeah, for me, I think I'm most excited about all the possibilities that our users, our scientists can do on our platforms.

举个例子，如果研究人员在我们的平台上用我们的模型做出了新发现，还能因此申请专利，那将是巨大的突破——所以我们特别重视早期发现，从一开始就教模型理解背后的机制。

So for one, I think a huge win would be if a researcher can patent a new finding or a new discovery on our platform using our models and that's why we really focus on early discovery and starting with building like teaching the models like the mechanistic understanding.

说白了，我们就是想通过生命科学领域的模型，给科学家们提供最强大的工具，帮他们大大加快研究进度。

So this is again like trying to provide the most powerful tools through our life sciences models to these scientists so they can really accelerate the speed of their research.

你觉得未来模型会不会强大到能精准预测细胞反应或者实验结果？

Do you think we'll get to a point where the models are gonna get really good at basically predicting the cell or predicting the outcome?

我觉得肯定能。

I think definitely yes.

这得看系统的复杂程度。

I think it depends on the complexity of a system.

比如，我们现在模型已经很擅长预测化学反应的结果了。

So for example, one thing our models are already very good at is predicting the outcome of a chemical reaction.

但随着生物化学和生物学复杂度上升，最难预测的可能是：某种药物对特定的人或特定系统会不会有毒性？

And I think as you increase in biochemical and biological complexity, some of the hardest things to predict is given a drug, will this be toxic to a specific person or to a specific system?

我们想一步步朝这个方向推进，这确实在我们的规划里，迟早要实现。

And I want to slowly work our way up to that, but that is definitely on a roadmap, that's something we want to do eventually.

当我们看那些处理语言或数学的模型时，评估起来其实挺简单的。

When we're looking at models that do things like language or math, it's pretty easy to put together evals for it.

它答对了还是答错了，一眼就能看出来。

Did they get the problem right or get it wrong?

那对于做生物学的模型，评估又该怎么做呢？

What do evals look like for models that are doing biology?

是的，我们有好几种方式来评估模型的表现。

Yeah, we have various different ways of evaluating model performance.

一个特别好的方法就是用实验数据来检验。

A really nice way to do this is kind of with experimental data.

比如有人已经做了实验，然后你让模型预测一下，它能不能准确猜出实验结果？

So someone has already done the experiment and then you ask the model, can it kind of predict the outcome of these experiments?

所以很多虚拟细胞的工作，其实就是这么个流程，对吧？

So a lot of the kind of virtual cell work, basically it looks like this, right?

有人已经对数百万个单细胞做了RNA测序，然后你把数据喂给模型，让它预测从未见过的扰动结果。

So someone has done single cell RNA seq on millions of different cells, and then you feed this to a model and then you try to get it to predict a unseen perturbation.

我们也可以用大量合成数据来做。

We can also do a lot with synthetic data.

也就是说，你可能生成了一组数据，特意在里面埋了一些容易让模型踩坑的陷阱。

And this means that maybe you have generated a set of data and you put very specific characteristics in this data that could be kind of like foot guns for the model.

这些恰恰是普通计算生物学家每天都会遇到的问题。

And these are things that maybe a typical computational biologist might encounter day to day.

比如数据里的一些奇怪偏差。

So this could be some weird bias in the data.

或者是你必须处理的质量控制问题，或者统计校正。

It could be some QC thing that you have to do or statistical correction.

因为我们自己生成了这些数据，就能真正测试模型能不能像一个经验丰富的计算生物学家那样，把各种常见错误都识别出来。

And because we generated the data ourselves, then we can actually go and test the model's capability as a computational biologist that does a catch all of these different mistakes.

所以评估方法其实有很多可以发挥创意的空间。

So there's a lot of different ways you can be creative with evaluation.

不过话说回来，我觉得Y Lab还是目前最真实的模型评估方式，对吧？

But that being said, I think Y Lab is still kind of the final like real evaluation of the model, right?

就像你常说的，生物学里没什么是真正的‘真实’，除非你能用现实世界的数据证明它。

And as you like to say, nothing in biology is really real until you can prove it in the real world.

所以我们有很多科研合作，目的就是做这种实地验证。

And so we do have a lot of research collaborations where we try to do just that.

是的，评估方法这些年变得越来越复杂和精细，尤其是在设计能同时体现价值创造和解决生命科学复杂问题的评估时更是如此。

Yeah, evals have really become more complex and sophisticated over time and I think that's especially true for designing evals that can really capture both value creation but solving complex problems for life sciences.

所以我们特别关注那些不是简单玩具问题的案例，比如真实世界里预处理数据的混乱情况。设计新评估时，我们通常先试着复现一个已有实验，这样就有基线可循——我们知道当前的前沿水平或真实情况是什么。

So I think we really try to focus on examples that are not like toy problems but really capture that like for example, like the messiness of like pre processed site data and when we design these new evaluations, a starting point is often just trying to recreate an existing experiment, so something that has already a baseline so we already know what the either current state of the art looks like or the current ground truth looks like.

我特别感兴趣的一个评估方向是，看看我们的模型能不能预测抗体结合能力，尤其是针对已知病毒变异株的现有研究。等我们有了这个基线，就可以进一步问：我们能不能把这用到之前没人做过的事情上？这其实是从头设计抗体的关键前奏，比如拓展对新病毒变异株的中和能力，也通往新疗法甚至新疫苗的开发。

So a evaluation I'm really excited about is looking at if our models can assess like the antibody binding predictions and looking at how that's been done for our existing virus variant and then once we have already done that baseline, we can push forward and say can we do this with something that hasn't been done before And I think that is like some of the precursor steps to de novo antibody design, maybe expanding the neutralization for new viral variants and that's also on the path to new treatments and potentially developing new vaccines.

在生命科学领域，尤其是学术会议和同行中，大家对这个的反应怎么样？

What has been the reception in the life sciences, particularly at conferences in the community, people you know?

你有没有发现大家普遍愿意接受，还是持怀疑态度，或者根本觉得这没什么用？

Have you seen a lot of willingness to embrace this or skepticism or people who just don't think it's helpful?

我觉得这可能跟你在哪个地区有关。

I think it probably depends on what part of the country you're in.

我感觉在西海岸，大家对AI都特别热衷。

I feel like kind of being on the West Coast, everyone is pretty AI pulled.

他们特别接受这种AI科学家和智能体工作流，也真真切切看到了AI的未来。

And so they really embrace this AI scientist, agentic workflows, and they really kind of see the future for AI.

但要是去东海岸参加大会，情况就大不一样了。

When I'm at a conference on the East Coast, this changes a lot.

我觉得那边的人普遍更谨慎一些。

I think people are generally a bit more skeptical.

对AI的能力，可能多多少少有点怀疑。

Maybe there's a little bit more doubt around the AI capabilities.

嗯，大概就是文化上的差异吧。

And yeah, think it's just maybe like a cultural difference.

我觉得大部分大型AI实验室都集中在这儿。

I think most of the big AI labs are here.

所以我们亲身体验过这些模型到底能做什么。

So we kind of have a firsthand experience of what the models are capable of.

这在一定程度上改变了我们的看法。

And this kind of changes our perspective a little bit.

那你怎么弥合这个差距呢？

How do you bridge that gap?

你怎么让更多的科学家真正理解呢？

How do you get more scientists to understand?

听起来人越多越好，毕竟这些技术还有不少短板需要完善，要是能让那些持怀疑态度的人也参与进来，找到自己的用武之地，那就更棒了。

Because it sounds like the more people contributing the better because there are weaknesses or areas need to be improved upon and the more you get people who are maybe skeptical about this to sort of figure out how to participate.

是的，我觉得有几种办法。

Yeah, I think there's a few different ways.

最简单的方式就是把我们的模型接入各种平台，比如聊天工具或者Codex。

The easiest way is by launching our models through different platforms like chat or codex.

再就是直接给科研人员展示这些工具有多实用，比如帮一个天天做移液的科学家做个稀释序列的表格。

And I think just by kind of showing individual scientists how useful this could be, maybe just making a serial dilution spreadsheet for someone who's pipetting.

但这确实有实际价值，对吧？

But that has real value, right?

我觉得可以从这儿慢慢推进。

And I think you can kind of build up from there.

从另一个角度说，我们确实和一些实验室有深入的合作，比如抗体设计或者酶的设计。

I think coming from the other end, we do have these more deep research collaborations with labs for, for example, antibody design or enzyme design.

这些合作往往能产出论文，别人看了就会说：哦，这个AI系统做了不少工作。

And these sort of things are they result in publications and that people will read and say, Okay, AI system did a lot of work.

它有真正的生物新颖性。

It has biological novelty.

而且已经在实验里验证过了。

It's been proven out in the wet lab.

所以我觉得这也能增强大家对这个系统的信任。

And so I think that also lends credibility to the system.

是的，我觉得最简单的办法就是用行动说话，发论文，主动和科学界互动。其实怀疑态度是健康的，应该被鼓励。

Yeah, I think the simple answer is you show by doing and you show by publishing and engaging with the scientific community and I think the skepticism is really healthy and should be welcomed.

看到大家这么兴奋，真的特别好，毕竟这技术潜力巨大，要是用对了，真能发挥出全部价值。

Think it's just really great to see people get really excited because and also trying to like disprove maybe because the potential for this technology is so great if we get it right and if we can actually really leverage its full capabilities.

我觉得谨慎一点、认真琢磨怎么把这技术用在实际问题上，完全合理。但每次我们发论文，都会让人意识到：我们需要更严谨的评估方式，得贴合生命科学的实际工作流程和研究难题，这样别人才能看了之后说：‘天啊，我现在有上百个点子，能直接用到我实验室里，解决那些卡脖子的问题。’

I feel like the carefulness about how do we actually make this work for real problems is like very much warranted but yeah, I think when we publish and I think that just also shows a need for more rigorous evaluations that represent like these life science workflows or research problems so people can look at and eval and say yes, like I feel like now I have like a 100 different ideas for how I can implement this into my lab and solve some of the current bottlenecks I'm facing.

我确实遇到过不少人，他们觉得AI太强大了，但又不知道该怎么正确用它，挺有压力的。

I actually think there's a certain amount of stress I've encountered from people who are worried that AI is really powerful, but they don't know how to use it the right way.

所以大家普遍有种感觉：‘我得在工作和生活里多用点AI’，可又说不清AI到底该在哪儿插一脚。

And so there is this general feeling of like, I need more AI in my workflow, in my life, but they don't know where AI should come in.

我觉得产品的一个核心愿景，就是让它简单到不用想，直接就能用。

And I think part of the product vision is to just make it so simple that it just works.

你只要去像Codex这样的工具里说一句：‘我想做我现在正在做的这件事’。

So you can just go to something like Codex and say, hey, I want to do whatever I'm doing today.

然后Codex自己就能搞定所有环节——多智能体协作、工具调用，全都不用你操心。

And then Codex can figure out all the different pieces, the multi agent workflows, the tool calling, all of that.

所以啊，你根本不用纠结怎么从AI里捞好处，它自然就帮你把事儿办了。

And so, yeah, basically you don't have to stress about how to get uplift from AI and it just happens naturally.

每次模型变得更聪明、更懂用户的时候，我们都能看到明显的进步。

We we do see those step changes every time these models become smarter and understand users better.

因为现在不用花太多时间琢磨怎么写提示词，也不用去挖那些小技巧了，实用性自然就上来了。

You get more utility because some people go, I don't have to spend a lot of time trying to prompt it or figure out all the tricks to it.

如果你要给一个正在读高中的学生，打算投身生命科学领域的年轻人提建议，你会说什么？

If you were talking to somebody who was considering getting into the life sciences, maybe a high school student right now, what advice would you give them?

我上高中的时候，参加过美国生物奥林匹克竞赛。

I feel like when I was in high school, so I did the USA Biology Olympiad back when I was a high school student.

在所有那些奥赛里，我觉得生物被看成是最靠死记硬背的，跟数学完全不一样。

And I think out of all the different Olympiads, I think biology was seen as kind of the most like memorization heavy one versus like math, right?

数学更多是考验你解题时的灵活应变和思维扩展，而生物呢，就是记和背。

Where it's kind of more, I don't know, test time confused scaling, whereas biology is more kind of memory and retrieval.

我希望有了AI之后，它已经掌握了所有科研知识点之间的关联，能真正激发人的创造力，让学习从死记硬背变成帮人把不同领域的研究串联起来。

I think my hope is that with AI having kind of like learned all the relationships between all the different research pieces, is that it can really uplift human creativity and just make the process less memorization and more kind of helping people connect different fields of research together.

让生物研究的边界，被推得更远一些。

And just kind of, I guess, furthering the frontiers of what people are able to explore in biology.

所以我觉得，对高中生来说，你其实没必要把生物课本全都背下来。

So, yeah, I feel like my advice to, I guess, a high school student will be that maybe you don't have to kind of go and memorize all the biology books.

你不如多用AI去探索一下。

You should just do more exploration with AI.

你可以直接读论文，然后随时提问。

I think you can definitely read papers and just ask questions.

这样既能深入钻研，也能快速了解全局。

And I think you can do both deeper dives and broader overviews this way.

学习的方式真的不一样了。

And I think just the way of learning really changes.

我在实验室的时候就发现，

I found that when I was in the lab,

做生物研究跟参加CS黑客马拉松完全不一样——那时候我们第一次一起开发应用，那种团队协作的感觉特别有激情。

there wasn't a real solo, like individual aspect of doing biology research compared to like, for example, when I went to my first like CS Hackathon, there was some excitement about just like the collaborative nature when we first built our app together.

所以我希望，未来早期使用者和学生们用我们的模型，甚至用Codex这样的运行环境时，也能感受到这种协作的乐趣。

So I feel like that's really the future I hope to see for early adopters and students using our models and maybe using it in like the Codex runtime because there is like a collaborative nature to it too.

我觉得，比如你可以把你的脚本、对话记录，或者有一天我们每个人都有自己的科研助手，能派出去跟队友协作。

I think like for example, your scripts or sending your conversations or maybe one day we have like, we all have like our own like co scientists or agent and we can like deploy our agent to now work with a teammate in that way.

我觉得这会带来全新的互动方式和体验，所以我鼓励学生早点上手，自己探索怎么用它才最适合自己。

I think there's just new interactions and new modalities for us so I would just encourage students to adopt early and just to also pioneer their own path for how they would like to use it.

说实话，我觉得我当初太早进湿实验室了，就像前面说的，我根本不喜欢移液。

For me personally, I always actually felt like I got into the wet lab a little bit too early and like we mentioned earlier, I did not enjoy pipetting.

这事儿太典型了，没人爱干移液。

That's the theme here, nobody likes pipetting.

对，确实有太多枯燥的体力活，所以我希望以后AI模型能连上物理设备，把学习曲线变得有趣些，让学生能跟模型一起学，把时间都花在真正有意思的地方。

Yeah there's a lot of very intense manual tasks involved and so I hope that when our AI models can connect with physical devices that yeah, we can just make a lot of the learning curve more fun for students so that they can kind of learn with the models and then kind of like maximize their time with like the really interesting interactions spaces.

我最近在带一个学生，喜欢帮学生想项目，他把Codex连到了一个温室里，用来拍照片、分析植物生长情况。

So I've been working with a student, I like to help students come up with projects and one of them is we've taken codex and he's connected it to a greenhouse and basically using it to get photos back and to look at it and to evaluate it.

看他把AI技术和传统温室结合起来，慢慢培养出同时驾驭这两者的能力，真的挺有意思的。

And I think it's been fun to see how he's been taking both, you know, AI technology and then something traditional like a greenhouse and combining them two and basically building up the skill set of learning how to use the two of them.

你跟同行、实验室负责人或者研究人员聊天时，对他们有什么建议？

When you talk to your peers, you talk to people who are running labs or running experiments or researchers, what advice do you have for them?

但我发现，很多人一听到就说：这听着不错，可我没时间。

Because the problem I see is that a lot of them go, that's great, I just don't have the time.

但说到底，我们真正想做的，就是帮他们省时间。

But ultimately what we're trying to do is save them time.

那你有没有什么简短的建议，或者用什么办法能激励他们一下？

So do you have any kind of quick advice that you give them or any ways you try to maybe inspire them?

我认识的大多数人，我觉得在学术圈里，用AI主要就两种方式。

Most people that I know, I think in academia use AI in I think two main ways that I've seen.

一种是拿AI聊聊现有的论文，确认自己有没有理解对，或者顺便查查事实。

One is to kind of talk to AI about an existing piece of research paper or something, and just kind of make sure that you're understanding things the right way or kind of fact checking.

这正是我最喜欢用AI的地方，因为你可以问再傻的问题，也不会觉得丢脸。

And this is personally what I really like to use AI for, because you can ask really dumb questions and you don't feel any judgment.

学东西的时候，用它真的特别舒服。

It's actually just really wonderful for learning.

我觉得大家也经常用它来分析实验结果。

And I think people use it a lot for analyzing experimental results.

这其实又回到了我之前提到的统计学问题——有时候你根本不知道该怎么分析数据，或者你的数据可能涉及化学，甚至某个冷门的蛋白质生物学领域，跨学科的东西太多了。

And I think this comes back to the statistics piece that I learned where I mentioned before, where sometimes you don't know what the right way to analyze your data is, or there's just kind of so many different interdisciplinary fields that your data might touch on something in chemistry or something in like a random niche field of like protein biology.

特别棒的是，模型能帮你整合各种数据分析方法，自动探索这些不同的路径。

And the really nice thing is that a model can kind of like pull those different ways of data analysis and for you and kind of explore all of these different paths.

我觉得这两种方式都挺轻松的，适合先试试水。

I feel like both of those are pretty low lift ways to try things out.

你完全可以扔个PDF给AI，说：‘帮我看看这篇论文’，然后像聊天一样自然地交流。

So you could just kind of like throw a a PDF file at AI and just be like, hey, help me understand this paper and just have a natural conversation.

或者你直接打开Codecs，在自己电脑上做数据分析。

Or you can boot up Codecs and do some data analysis directly on your laptop.

是啊，我觉得关键是要让使用AI一开始就不像在干活。

Yeah, I would say that you'd have to start with making sure it doesn't feel like work right away.

所以如果你在推AI应用时，不妨先从一个爱好项目或兴趣项目入手。

So maybe it'll be easier when you're focusing on AI adoption to just like work on like a hobby project or a passion project.

就拿我来说，我其实是在搞创意写作的时候开始接触文献综述的——那完全跟我们日常研究没关系，虽然生物学本身就很需要创意，但我当时是通过另一种媒介去探索，也正是那时候，我才慢慢解锁了各种提问技巧和数据获取方式。

For me, for example, I actually started working on more literature synthesis tasks when I was doing creative writing projects which are kind of like just something that was not at all related to our day to day even though biology is a very creative space, I was just exploring that through like a different media and I think that's actually when I started unlocking like a lot of different ways to either prompt the model or to actually access different data sources.

所以我觉得这让我积累了很多模式识别的能力，因为第一次肯定做不好，这事儿本来就很难，而且这个领域发展太快了，每周甚至每月都有新突破，随时可能改变我们跟模型或AI互动的方式。

So I think that just gave me a lot of like pattern matching abilities for when I was trying to apply it because we're not gonna get it right in the first try and it is really hard and I feel like the progress and pace of this field moves so fast that every like week or month there is like a new like pretty exciting development that might change how we engage with models or AI systems.

所以我觉得最重要的是先动手干起来，另一个关键点是协作。

So I think it's just important to get started somewhere and I think another theme is the collaboration element.

我觉得如果有同事跟你做同样的日常工作，他给你推荐，效果会更好。

I feel like it's more powerful when you have a recommendation from either somebody on your direct team who is doing the same day to day tasks as you.

我们团队里经常发生这种事：有人突然说，嘿，我用Codabs连上了三个以前根本连不起来的内部数据库，简直太神奇了——模型的潜力深得超乎想象，很多东西你不去试，根本不知道它能干啥。

That happens a lot on our team as well where somebody will say, oh I got codabs to like now touch these three different like internal like databases that we weren't able to connect before and I don't even like the latent space, the latent capabilities are just so vast that there's a lot that we just don't know until again we can do it.

所以我觉得多跟朋友、实验室伙伴、团队同事聊聊，特别容易激发灵感，也能帮你更快上手。

So I think just having conversations with your friends, your lab mates, your teammates will I think spark a lot of those creative juices and then help you with your own adoption.

十年后的科学会是什么样子？

What does science look like ten years from now?

我们团队刚成立时就定了特别宏大的目标，其中一个就是希望能真正推动，哪怕只是协助，攻克某种疾病。

I think when we started this team, we do have like really just ambitious targets and one of those is like I think we do want to make meaningful strides towards or even if like assist with curing a disease.

现在有太多罕见病、孤儿病，根本得不到应有的关注和资源，因为这领域实在太难了——比如临床研究，要把成果真正送到病人手里、推向市场，难度大得吓人。

And I think there's just so many rare like orphan diseases that doesn't really have the attention and the resources that it warrants because it's just such a difficult field to actually like, for example, clinical research is so difficult to actually bring that to patients and to market.