本集简介
双语字幕
仅展示文本字幕,不包含中文音频;想边听边看,请使用 Bayt 播客 App。
将会发生巨大的变化。
There is going to be dramatic change.
每个人电脑上都会有一个专属的个人助手,能做的事情远超三个月前、六个月前的水平。
Everyone is going to get their own little personal assistant on their computer that allows you to do more than whatever was possible three months ago, six months ago.
这是蒂博ault。
This is Thibault.
他负责OpenAI的ChatGPT Codex和API,这些技术支撑着你每天使用的大部分AI功能。
He runs ChatGPT Codex and the APIs at OpenAI, the tech powering most of the AI you already use every day.
他刚刚告诉我,再过几个月,那些完全不用AI的人,也能享受到过去两年里那些先行者才获得的同样好处。
And he just told me that in a few months, people who don't use AI at all will get the same benefits as the ones who've spent the last two years figuring it out.
现在,你得主动去提问,还得动脑筋怎么问才有效。
Today, you have to actively prompt and be creative about what you're asking.
但将来就不用这样了。
In the future, that's not the case anymore.
那到底是什么会让人们变得如此出色呢?
What's gonna make people brilliant, I guess?
如果你是个有好点子的创意型人,现在正是大胆尝试的好时候。
If you're a creative person with good ideas, it's a great time to experiment.
你能给我展示一个每个人都该立刻开始用的工作流程吗?
Can you show me a workflow everyone should be deploying today?
我觉得这将会带来一个变化,那就是
One thing that I think this will mean is that
我的观众里有一部分是开发者,但更多是知识型工作者。
My audience is some of them are developers, but a lot of them are knowledge workers.
他们对人工智能特别感兴趣。
And they're very excited about AI.
我们在软件工程领域已经看到了一些变化,对吧?
And what we've been seeing with software engineering, right?
谷歌说,他们现在75%的代码都是AI写的。
Google says 75% of their code is now AI written.
所以AI已经彻底改变了这个行业。
So it has been really transformed by AI.
你提到,这种转变在未来六个月会扩展到所有知识型工作。
And you said something that this transformation will extend to all knowledge work over the next six months.
那你认为,对知识工作者来说,具体会有什么变化呢?
So what do you think is gonna be changing for people who are knowledge workers?
我觉得真正改变的不是人本身,而是技术成熟了,现在智能代理能在长时间任务中稳定运行。
I think it's not that much that people are going to change as the technology has matured and now agents are reliable over, you know, long horizons.
它们能熟练使用各种工具,比如操作电脑、浏览网页这些都不在话下。
That's very capable of using many different tools, you know, computer use being one of them, browser use.
我们已经接入了一百多个插件,覆盖了你日常用到的每一个小工具。
We have added over like a 100 different plugins that can tap into, you know, every little tool that you already use in your life.
GPT-55里的智能代理在这方面表现得极其可靠。
And the agent in g p t five five is like extremely reliable at it.
所以不是人要变,而是技术已经准备好了,每个人都能从智能代理中获益。
So it's more that the technology has matured and is ready, and so everyone will be able to get benefits from an agent.
以前你得有点技术背景,还得自己动手折腾才行。
Whereas before, you have to be a little bit technical and, you know, get in there.
如果不可靠,用五分钟就出问题,那你只能去手动调整设置。
If it wasn't reliable, if it was, like, struggling five minutes in, you're just gonna kinda maybe have to go into the configuration.
而这显然需要一定的技术背景。
And, obviously, that requires, a technical background.
但现在完全不一样了。
That's not the case anymore.
所以我觉得这项技术会迅速普及开来。
So that's why I think it's going to be become, like, very widespread.
你认为变化最大的会是什么?
What do you think is gonna change the most?
比如一个市场人员,有了智能代理后,他们的日常工作会有什么不同?
And say, like, you take a marketer, what's gonna change with their day to day with agents?
你要先想想,哪些事情可以自动化。
So you're going to look at what can you automate.
比如说,我以前花一个小时做市场调研。
Oh, maybe I spent, like, one hour, you know, doing market research.
或者我花一小时,整理一大堆收件箱邮件和信息。
Or maybe I spend one hour, like, summarizing a whole bunch of inbound and emails.
也可能要花两个小时,一个个查看潜在客户。
Maybe I spend, like, you know, two hours going through, like, prospects.
以前半年前还算是比较高级的功能,比如设置定时任务,现在完全不一样了。
And so a lot of the concepts that were, like, you know, advanced concepts, you know, maybe six months ago, such as, like, running something on on a cron schedule.
现在直接在应用里说一句,每12小时运行一次,做点市场调研,发我个PDF就行。
It's, like, nowadays, it's just, like, in the app, you can just say, like, you know, run this every 12 hours, you know, do some market research, send me a PDF.
然后会通过邮件发给你吗?
And it's gonna send it to you over email?
对。
Yes.
你可以选择邮件接收,也可以直接在应用里查看。
And then you can send it to you over email or you can, you know, just consume it in the app.
甚至还能打印出来。
Maybe you can print it.
几周前我做过一件事,每天把Slack上的所有新闻汇总一下,然后打印出来。
I did this, a couple of weeks ago where I would summarize, all the news from Slack and then print it on my printer physically, like, every day.
每天早上就和咖啡一起看。
So I would have it with the coffee.
哇,真厉害。
Oh, wow.
这做法挺有高管范儿的。
This is, very executive way.
有点儿过时了吧?
It's getting, like, know, maybe old school.
对吧?
Right?
就是想拿份自己的‘报纸’而已。
Just, you know, kind of want my newspaper.
这种事以后肯定会变得很普遍。
This kind of thing is just to be kinda is is going to be mainstream.
你知道我最近被什么启发了吗?
You know what I was really inspired by?
我看了你的一期播客,你和格雷格聊到早上起床喝咖啡时,有个仪表盘会弹出来,问你‘这个能批准吗?’
I was watching one of your podcasts where you were talking with Greg about having this dashboard, basically, when you wake up in the morning, you have your coffee, and it just tells you, like, can you approve this?
我就想自己也搞一个。
And I'm gonna run it.
你只需要看看AI代理在你睡觉时干了啥,然后点个头批准,它就能继续往下做。
Can you approve So you just basically read through whatever your AI agent did while you were sleeping and you're approving and it's continuing its work.
你觉得这种事什么时候能变成现实?
When do you think it's gonna be a reality?
没那么远了。
It's not that far off.
技术其实早就有了。
The technology already exists.
关键是怎么把它包装好。
It's about packaging it.
对。
Mhmm.
比如上周我们推出了自动审核功能,做这项研究的时候,我真的被震撼到了。
For example, last week we released auto review, which is the concept it definitely blew my mind when we did the research on it.
主代理负责执行操作,另一个代理则负责核查所有操作,确保它们不会对你造成任何潜在危害,尤其是在低风险场景下。
So you have the main agent doing actions, and then you have a second agent that is verifying all the actions of the first agent and verifying that they are not doing anything that could be harmful to you and, like, in our low risk.
这项创新主要来自我们的安全团队和对齐团队,核心是对齐研究。
And so that's an innovation that came out of our safety team and and and alignment team, primarily like alignment research.
它让你能够更长时间地让代理自主运行,甚至处理一些敏感数据,对吧。
And it allows you to run an agent much more, like, you know, for longer periods of time autonomously, even handling, you know, quite sensitive data Mhmm.
而不用担心代理不小心把你的个人信息发给陌生人。
Without risking, for example, you know, your agent, like, sending an email to a stranger with some of your personal information in there.
现在有了这些安全机制,系统更可靠了,你也就更愿意让它接触生活中更多方面了。
Now that you have these pieces to make it more secure, more safe, and you're willing to give access to a lot of things in your life.
我们该怎么组织自己的数据?
How do we organize our data?
想象一下,我们都要准备上线自己的第一个智能代理了。
Imagine we're all getting ready to deploy our first agent.
对吧?
Right?
没错。
Right.
关键是,智能代理需要高质量的数据。
And the thing is, agent needs good data.
我以前主要用的是自己的说话风格和个人资料,但后来意识到,如果我的代理能了解我创作内容时的策略,效果会好得多。
And sometimes, you know, I've been working with mostly my tone of voice like personal dossier, but then I realized that my agent would really benefit from knowing my strategy when it creates content for me.
那你对刚开始的人有什么建议?
So what would you say to someone who's just starting out?
他们该怎么整理这些数据?
How do they organize this data?
必须创建哪些关键文件?
What are the must have files that they have to create?
我把所有东西都存在电脑本地的文件里。
I organize everything on, like, local files on my computer.
我建了一个文件夹。
So I have a folder.
我的笔记都放里面。
I have my notes in there.
我尽量保持得整整齐齐,还让我的助手帮我整理。
I keep it, like, fairly tidy and, you know, I use my agent to organize it.
那旅行的时候怎么办?
What what happens in travel?
这正是我想问的。
That that's my question.
很多人都是用本地电脑处理的。
People work with local computers.
我出门会带一台笔记本电脑。
I have a laptop that I travel with.
我有两台麦金塔电脑,一台在工作室,一台在办公室。
I have two Macs in one in my studio, one in my office.
是的。
Yeah.
接下来的三个月,所有这些都会迁移到云端。
So what we're going to see over the next three months is all of this is going to come to the cloud.
对。
Mhmm.
你就不用再操心自己笔记本上的本地文件了。
And you're not going to have to manage, like, local files on, you know, just your laptop.
如果你出门旅行,用手机登录,那 suddenly 就不是同一个助手了。
And then, you know, if you happen to travel and you're on your phone, oh, now suddenly it's not the same agent.
是的。
Yeah.
这挺烦人的,因为你得在生活里同时维护两个独立的系统,还得把它们对上。
And that's that's annoying because you have, like, you know, two separate entities that you need to sort of, like, you know, map two different things in your life.
所以很多东西都会迁移到云端。
So a lot of this is going to come to to the cloud.
它会自动管理自己的内存。
It's going to manage its own memory.
它还会帮你管理那些存放在别处的文件。
It's going to manage you know, help you manage, like, files that are just, like, hosted somewhere.
到现在还没一个好办法,这其实挺让人意外的。
There's no good solution yet, which is, you know, sort of surprising.
是的。
Yeah.
我目前把所有东西都丢到Google Drive里,然后挂载一个Google Drive文件夹。
I'm I'm just putting everything to my Google Drive, and I'm connecting a Google Drive folder.
对。
Yeah.
这样确实可行。
That works.
这样就行。
That works.
对于智能代理来说,每个人应该准备哪些文件呢?
What are the files that everyone should come up with for their agents?
如果我们讨论的是知识工作者的效率问题的话。
If we're talking about productivity for knowledge workers.
对。
Yes.
语气风格是个不错的例子。
Tone of voice is a good one.
关于这一点,可能你想不到,我反而不建议你去详细解释自己的语气风格。
And for that, maybe surprisingly, like, I don't recommend, you know, actually trying to explain your tone of voice.
我建议直接提供一些例子,比如你过去发过的通讯稿,或者录音片段、你发给朋友的消息,涵盖不同场合——无论是工作还是私人交流。
I recommend, just including examples of, you know, like past, you know, newsletters maybe that you published or, you know, snippets from, you know, like recordings, or, you know, like messages that you sent, you know, to friends and and and, in different, like, maybe professional and personal settings.
我有很多关于各个项目的笔记。
I have a lot of, notes around, like, projects that I have.
每个项目都有自己的文件夹,里面放着各种文件。
There's, Like, each project has its own folder with various files in it.
对。
Yeah.
还有联系人列表。
And then contacts.
但你没必要把所有东西都存在文件里。
But you don't have to maintain everything in files.
你也可以直接用你平时已经在用的那些效率工具。
You can also just rely on, you know, like, all the productivity apps that you already have.
然后像Codex这样的工具,就能自动帮你找到需要的信息。
And then, you know, Codex, for example, will be able to go and pull the right information.
你怎么判断是该在项目里继续做,还是该让Codex去开发一个新工具或应用呢?
How do you know when you stay inside a project or you want to Codex to actually build something, like a tool or an app?
有时候我觉得,某个新场景确实需要一个专门的应用,就会新建一个文件夹,然后在里面随便试试看。
Sometimes I feel like, you know, I need, like, a dedicated app for for for like a new use case, and then I go and like I create a new folder, and then I just like experiment in there.
但现在,我越来越觉得,那些只能做一件事的静态应用用不着了,我更依赖我的智能助手来帮我搞定一切。
But nowadays, more and more, I feel like I need less of static apps that just can do one thing, and I rely just more on my agent to do everything for me.
如果有人开始用智能助手,有人不用,一年或三年后,他们的工作效率会有多大差别?
And if somebody starts using agents and somebody doesn't, what's the difference in their productivity in say one or three years?
一到三年?那太远了。
One to three years, that's like so far off.
到那时候,技术本身早就天翻地覆了。
The technology itself, like will change so much by then.
我们能看到的是,那些愿意适应、愿意探索的人,效率会高得多,他们能完成那些以前总想着拖一拖的事。
What we're going to see is that, you know, like, people who are willing to adapt and discover these things are, you know, going to be quite a bit more productive, and it's going to enable to do, like, you know, all these things that you sort of, like, know, maybe put off.
每个人都会有自己的个人助手,帮着报税、设置邮件过滤器、提醒你联系家人,帮你做各种琐事。
Everyone is going to, you know, get their own, like, little personal assistant on their computer, know, doing all of the, know, maybe filing your taxes, setting up, email filters, telling you and helping you, like, get more in touch with, like, you know, your loved ones.
这种变化会把很多人更紧密地连在一起。
That is something that will connect a lot of people.
当我开始更多地使用AI时,我感受到了一种责任上的纠结。
When I started using more AI, I felt this responsibility dilemma.
你刚提到报税,对吧?
You just mentioned like taxes, right?
没错。
Right.
是的。
Yeah.
AI完全可以帮我报税。
AI can totally do taxes for me.
我有个自动化工具,每个月都会告诉我该交多少税,以及可以采用哪些税务策略。
And I have this automation that tells me every month how much tax I'm gonna owe and what tax strategies I could be deploying.
我真得对这些事负责吗?
I want to be responsible for that?
我不确定。
I don't know.
像这类事情,作为创业者,我完全意识到,如果你现在才刚开始,效率会高得多。
Like, when it comes to these tasks, also as an entrepreneur, I totally realize if you're just starting out today, you're much more productive.
但如果你已经是创业者了,我不会因为有个AI代理能搞定就直接把团队里的人辞了。
But if you're an existing entrepreneur, I wouldn't really fire anyone on my team because an AI agent can run it.
我只会让那个人去用这个AI代理帮忙。
I would just ask that person to use that AI agent.
你觉得这事会发展到什么地步?
Where where do you see this go?
因为我发现,当我开始用这么多代理的时候,总觉得自己得盯着它们。
Because I feel like when I started running so many agents, I feel responsible that I need to check on them.
我得核对它们的输出结果。
I need to verify their output.
我干脆就让别人去管了。
I'll just let somebody else handle it.
这挺有意思的。
That's interesting.
归根结底,还是得由人来负责。
So at the end of the day, humans remain responsible.
我们看待这个问题的方式,其实是用来增强你自己的能力,对吧。
And then, you know, the way that we think about it is actually just really about augmenting your own capabilities Mhmm.
给你一个工具,让你能完成三个月前、六个月前根本做不到的事,还能自动处理那些枯燥乏味的工作部分。
And giving you a tool that allows you to do more than whatever was possible three months ago, six months ago, and take a lot of like the boring parts of the work as well, and just do that automatically for you.
在编程方面,我们觉得,如果你产出了一段代码,那你就要对它负责,这是当然的。
It is important as we see this for coding, for us, it's like if you produce a piece of code, you know, you are responsible for Yeah.
如果出问题了,不是代理的错。
If it breaks, it's not it's not the agent's fault.
是你自己的责任。
It's it's your fault.
代码审查也是一样。
The same for code review.
你没法把理解力外包出去。
And like you cannot outsource, like, understanding.
归根结底,人还是要掌控全局,得真正理解整个系统是怎么运作的。
At the end of the day, humans, you know, are still remain in control and remain the people who need to understand the entire system and how things work.
所以如果你从提升个人效率的角度来看,显然还是离不开人。
And so if you view it under the lens of like, you know, augmenting your own productivity as a human and say, you know, obviously, you know, you need humans.
毕竟,我们做这些是为了让自己的生活变得更好。
After all, we're building for ourselves to better our own lives.
对。
Yeah.
我感觉自己效率确实提高了不少,但后来发现,因为想优化的事情太多,脑子都快炸了。
It's just I feel like I've become much more productive first, but then I'm like, because I want to optimize so many things, my brain just started exploding.
这里有个陷阱,就是别太依赖它去做所有事,觉得自己什么都能干,但实际上你的能力还没跟上技术的发展。
There's a there's a pitfall of like, no, maybe, you know, falling too much into, you know, using using it for everything and feeling like maybe you can do things, but you're a little bit too early on the capability curve.
是的。
Mhmm.
也许未来模型出来之后,这些事就能做到了。
And so it's like maybe things, you know, will be possible with future models.
这挺好。
That's good.
这意味着你在不断尝试,摸索出它在哪些地方可靠,哪些地方还需要改进,也许这些功能三个月、六个月后就能实现。
That means that you're pushing and you're sort of, like, discovering, you know, where where it's reliable, where maybe it needs a little bit more work, and maybe this will be something that you can do in three months, six months.
明白了。
Okay.
咱们聊聊 vibe coding 这回事吧。
Let's talk about vibe coding in general.
我之前在播客里提过,我们在 Linguature 里大量使用 vibe coding 来开发各种应用,我们有个懂技术的人,编程能力很强,但同时也有我们这些不懂技术的人参与。
So I shared before the podcast that in Linguature, we're actually vibe coding a lot of different apps, and we have a technical person who can actually write code and he's really good at it, but then it's also us who can't.
当我们用 vibe coding 做出来的东西,用起来特别顺手,但技术同事一看就说:‘我本来会换种结构来写,因为如果以后要扩展的话——比如我们有个应用,里面收录了 300 个最常用的英语单词。’
And when we vibe code something, it works great, but then the technical person comes in and says, Oh, I would have structured it in a different way because if we want to grow it Like, for example, we have this app that has 300 most useful words in English.
对吧?
Right?
这个应用用 vibe coding 做得特别好,完全能用。
The way It was vibe coded in a perfect way, so it's usable.
但要是想扩展到一千个常用词,现在的架构就有点撑不住了。
But if we want to extend it and make it like a thousand useful words, then the architecture is a little bit off.
你现在会怎么去进行氛围编码呢?
What would you say, like, how would you approach vibe coding these days?
如果你打算打造一个可能成长为大公司的产品,是该找技术伙伴合作,还是继续自己搞氛围编码?
If you're thinking of building something like it's gonna be a big company, should you work with a technical person or still try and vibe code it?
如果只是想跟几个人分享一下,试试看,纯粹为了乐趣去开发,那完全可以直接自己动手,借助智能助手就行,这完全没问题。
If the goal is to share it with, you know, a couple of people and, like, try something out and, you know, build it for for the joy of it, you know, definitely like just cutting it by yourself with the help of an agent, it's like, that's perfectly fine.
对。
Mhmm.
但如果目标是做大做广,让成千上万甚至更多人用上,那还是得让技术背景的人参与进来,这非常有帮助。
If the goal is to scale it and, you know, make something that is successful, you know, for like, you know, hundreds of thousands of people or more, Having that technical person like in the mix is still useful.
我相信未来这种情况会慢慢消失,那时候智能助手能真正理解长期维护的需求,懂得合理的架构设计,全程陪你把产品从零做到大规模成功。
I do expect that, you know, over time to go away, where, you know, you will have agents which are able to like, you know, understand the long term maintenance aspects, the right structure and be able to handhold you all the way to scaling to a very large and successful product.
你觉得这个转变大概需要多久?
What do you think is the timeline?
毕竟技术变化实在太快了。
Because it's been changing so fast.
未来六到九个月内,代码的长期可维护性会有明显提升。
We will see significant improvements towards like long term maintainability of code, over the next six to nine months.
到时候就完全不需要找懂技术的人帮忙了。
In order to like not need the help of like someone more technical at all.
这听起来还是挺遥远的。
That feels like quite a, quite a, quite a far off.
对。
Yeah.
你对软件工程的未来怎么看?
What do you think in general about the future of software engineering?
我刚主持了一个关于深度学习和AI的小组讨论。
I was just moderating a panel around it with, you know, deep learning AI.
我们六位嘉宾都达成共识:你依然需要技术知识,但同时也得成为通才。
And we're talking about like how we all had a consensus of six people on the panel that you just you still need technical knowledge, but you also need to become a generalist.
现在人人都在用 vibe coding 做应用,作为创作者,十年前我根本不敢想自己能开发一个应用,因为觉得太贵了。
And because everyone is vibe coding apps, like, as a creator, I would have never thought of building an app ten years ago because I knew how expensive it was.
现在我会自己开发一个应用,但一旦看到成千上万的人在用,遇到各种bug,我还是得找技术人员帮忙。
Now I would build an app, but then I will see thousands of people using it, I will experience some bugs, and I will need a technical person.
对。
Yes.
照这么发展,我们需要的软件工程师会更多,还是会更少?
Are we gonna need more software engineers or less with this?
是的,我觉得我们会看到基础设施和应用程序的数量迎来爆发式增长。
Yeah, think we're going to see an explosion, in the amount of, you know, infrastructure and apps that we built.
现在太棒了,只要你有个想法,就能动手做出来,哪怕只是个雏形也行。
And nowadays it's, it's just, it's awesome because if you just have an idea, you can, you can build it or at least in some shape you can build it.
如果你是个有创意的人,想法不错,现在正是大胆尝试的好时候。
If you're a creative person, you know, with good ideas, it's a great time to experiment.
你可以快速原型、不断迭代,最终做出一个可能真的能成功的东西。
And then you can quickly prototype, iterate, and get to something that you feel, you know, like might be successful either.
但要界定到底需要多少软件,世界上对软件的需求到底有多大,这事儿还真说不准。
And then it's hard to, like, put a limit and, you know, how much software is enough, how much demand is there for more software in the world.
我们还能解决多少新的问题呢?
Like, many more problems can we solve?
感觉我们才刚刚起步。
It feels like we're at the very beginning
总觉得新的问题一直在冒出来,等着我们去解决。
of Feels like we keep coming up with problems we can solve.
没错。
Exactly.
对吧?
Right?
只要这种趋势持续下去,技术还能不断进步,
So it's like, you know, as long as that continues to be through and, you know, we are able to continue technological, like, advancement.
对。
Mhmm.
我觉得对技术人才的需求会越来越大。
It feels like, you know, there will be demand for a lot of technical folks.
是的
Yeah.
你能给我展示一个每个人今天都应该采用的工作流程,来最大程度提升效率吗?
Can you show me a workflow that everyone should be deploying today to maximize their productivity?
你能想到什么例子吗?
Can you think of something?
这因人而异。
It really varies.
要看具体情况。
It depends.
比如,我平时会用它来了解外界对Codex的看法。
Like, you know, one of the things that I do is I use it to stay in tune with what the world thinks, for example, about Codex.
对。
Mhmm.
所以我每天都会收到一份重点摘要。
And so I receive like a daily summary of the highlights.
给我来一个这样的功能,比如说
Something for me, So something like
我们新开个对话吧。
let's do a new chat.
举个例子,我正在参加一个关于AI和Codex的播客,你可以帮我从邮件里找出相关的内容,再根据我对优先级的设定,自动生成回复草稿。
For example, you know, I am on a podcast talking about AI and Codex, find relevant, emails in my inbox, and, prepare draft replies based on what you know about my priorities, from memory.
不过你目前还没法记住这些信息。
So you haven't, you haven't built, memories yet.
但如果你
But if you
如果你经常用Codex,它就会慢慢学会,虽然现在还不会
use, if you use Codex a lot, it will learn about But it's gonna new learn
ChatGPT也有记忆功能,对吧?
memory from ChatGPT as well.
是这样吗?
Right?
是的。
Yeah.
另一个功能是,去浏览一些热门的网络资讯,给我梳理一下 Codex 过去两周都更新了什么。
Another one is, you know, go through popular internet sources and give me a breakdown of what Codex shipped over the last two weeks.
然后我们再回头处理邮件。
And we can go back to email.
我的 Gmail 邮箱特别乱,可以设置一些规则来自动整理。
I have a messy inbox in for Gmail, set up filters for it to keep
你知道吗,我记得大概是 2019 年,我花了好几天时间整理邮件和文件夹。
You know, remember, I think it was like 2019 or I spent a couple of days just organizing my emails and folders.
结果发现,这纯粹是浪费时间。
And then I realized such a waste of time.
我后来才明白,花这么多时间在上面根本不值得。
I so much
真希望有个时间规划工具能帮我
wanted time plan a
旅行啊,你说我们去哪儿好呢?
trip, you know, to like where should we go?
得看我日程安排才行,对吧?
Based on my calendar availability, right?
是的。
Yeah.
或者类似的功能吧。
Or something.
对,对,对,比如
Yeah, yeah, yeah, For
举个例子,能连上我的日历。
example, can connect to my calendar.
规划旅行时,能根据我的日程安排来推荐。
A trip plan, and use my calendar availability to your planet, for example.
然后它就能这样,你看,它已经把你的邮件都处理过了。
And then so it's going to go like here, it's you can see it's gone through your email.
这个呢,是个讲AI和Codex的播客。
So this one is like, I'm a podcast to talk about AI and Codex.
它已经翻过你的邮箱,找到了几个跟今天相关的活跃讨论。
It's already gone through your email, found several live threads related to today's, thing.
对。
Mhmm.
我查了下你的日程。
You know, I went through the calendar.
它还在继续处理你的邮件。
It's all still going through, emails.
所以它找到了一堆相关信息。
So it's found like a bunch of things.
看来你这边有不少安排,你知道的。
Apparently, you have like, you know
而我这边用的是Codex,就是今天这个。
And I with Codex, which is today.
然后,你看,这儿还有另一个。
And then, you know, here you have another one.
所以你看,咱们现在同时在处理好多事情。
So, you know, you can see, like, we're doing, a lot of things, you know, in parallel.
你可以直接运行一些小型的智能任务。
You can just like run, you know, little agentic threads.
比如你正在做演示文稿,把内容丢进谷歌幻灯片,它就会自动帮你生成一份完整的幻灯片。
Here, you could say like, I am making a presentation, put this in a Google slide and it will go and it will just prepare like, you know, a Google slide for all of the content.
编辑是在应用里完成,还是得退出去用谷歌幻灯片?
The editing is gonna be within the app or I will have to leave and go to Google Slides.
应用里其实也内置了浏览器。
So there is an in app browser as well.
是的。
Mhmm.
我不知道你有没有设置好你的宠物。
I don't know if you've set up your pet.
还没有。
No.
还没呢。
Not yet.
我们还上线了一个挺可爱的小功能。
So we also shipped, like, this cute little feature.
哦,我刚在你的系统里看到它了。
Oh, I just saw it on your ex.
现在你可以把这东西跨平台移动了。
So now, you know, you can move this across Uh-huh.
你还能看到哪些线程还在运行中。
And you can see, like, you know, which which threads it's still working on.
比如,它现在还在总结 Codex 的更新内容。
So for example, like, it's still summarizing Codex updates.
它正在设置一些邮件过滤规则。
It's currently setting up, some some email filters.
我不想动你的收件箱,我只是把这一个停掉而已。
I don't wanna make changes to your inbox, I'm just gonna, you know, stop this one.
不过,嗯,这算是个简短的演示。
But, yeah, this is sort of like a a brief
你能给我讲讲这个仪表盘吗?我刚才看到。
Can you talk to me about this dashboard that I
你现在已经安排了去克里特岛的行程。
was You have a trip to Crete plant now.
真不错。
Nice.
它已经同步到我的日历里了。
So it went through my calendar.
明白了。
Okay.
今天早上,我根据自己的需求问了ChatGPT,该开发什么应用。
So this morning, I actually asked my ChatGPT based on my queries, what app should I build?
它给我提了五个点子。
And it gave me five ideas.
其中一个就是内容再利用。
And this was one of the ideas where it's content repurposing.
你觉得这个想法怎么样?
And can you rate rate it?
其实我当时用的提示词特别简单。
Like, my my so it's it was basically a very basic prompt.
我给了它一个链接,指向我的 MD 文件。
I gave it to a link to my MD files.
这不错。
Oh, that's good.
我开发了这个应用,本来是想能直接语音输入,让它自动生成邮件简报,但好像没成功。
I created this app, but the only thing I wanted to be able to dictate into the app so it can convert it into an actual email newsletter, but I don't think it worked.
明白了。
Okay.
那我们来看一下。
So let's have a look.
这个就是内容再利用引擎,从所有那些帖子来的。
So this was content repurposing engine, you know, from all the the post.
明白了。
Okay.
然后它就全都给你生成好了。
And then it just built all of that for you.
不过它生成的东西有点太花哨了。
Well, it built something too fancy.
我觉得咱们还是简单点好。
I'm like, let's let's just make it very simple.
你当时说,好啊。
You were like, okay.
你这有点太
This is like you're a little
有点太雄心勃勃了。
being being a little bit too ambitious here.
是的。
Yeah.
对。
Yeah.
然后它就生成了。
And then it has.
哦,我明白了。
Oh, I see.
明白了。
I see.
A/B测试。
AB testing.
这都快成邮件草稿了。
It's like already a newsletter draft.
是的
Yeah.
而且说实话,我喜欢它没等我提醒就自己行动了。
And, actually, I like that it didn't wait for my prompt.
它直接就干了。
It just went ahead.
做了好多事。
It just did a lot
这些东西。
of these things.
它自己就琢磨着,这大概就是你需要的。
Like, by itself, it was like, you know, this is probably what you need.
是的
Yeah.
然后唯一的问题是
And then the only thing
不太管用的是,比如说
that doesn't work is, like,
把语音直接输入到应用里。
voice voice into the app.
是的。
Yeah.
这问题靠另一个提示能解决吗?
Is it something that another prompt is gonna fix?
还是
Or
可以。
Yes.
没错。
Yes.
所以我想让你跟那个‘是的’说一下。
So I want you to speak to the yeah.
这里可以语音输入。
We have voice dictation here.
对吧?
Right?
所以
So
不对。
No.
我是说,这个应用。
I mean, the app.
对。
Yes.
对。
Yes.
对。
Yes.
那我们就用起来吧。
So let's use it.
我们来试试嵌套功能。
Let's do a nesting.
好的。
Okay.
所以这个应用里,语音输入功能还没跑起来。
So in the app, we have we don't have dictation, working.
你能把它和 OpenAI 的语音转文字 API 集成起来吗?
So can you integrate it with the OpenAI, speech to text API?
这需要一个 API 密钥,不过现在你可以先随便填一个凑合着用。
For that, you will need an API key, but you can just make it up for now.
你可以去网上查最新的文档,看看怎么把 OpenAI 的语音转文字 API 接到这个应用里。
And, you can go through the latest documentation online to figure out, you know, how to integrate the OpenAI speech to text, API into this app.
这个提示可真够专业的。
That was a very sophisticated prompt.
我怎么知道我需要这个提示呢?
How do I know that I needed that prompt?
因为我根本不会想到 API 密钥这回事。
Because I would never think about API keys.
所以有时候你还是需要一个懂点技术的人,不过现在看起来还得有技术背景才能搞定的事,以后都会变成主流。
Oh, that's why sometimes you still need someone who's like a little bit technical, but a lot of these things that right now maybe are still a little bit in the realm of like, oh, you need more of a technical background are gonna come and become mainstream.
所以现在你可能还需要我,但三个月后,你就自己能搞定这一切了。
So, you know, maybe you need you need me today, but, know, in three months, it's like, will be able to set all of this up.
它会直接连上你的 OpenAI 账号。
It will just connect it to your OpenAI account.
对。
Mhmm.
你也不用再单独去设置什么 API 密钥了。
And, you know, you don't need to set up, like, a separate API key.
它能自动连上所有东西。
And it's gonna be able to just connect to everything.
展开剩余字幕(还有 196 条)
好的。
Okay.
在它运行的时候,你能聊聊那些特别复杂的使用场景吗?就是那种工作流程很精细,但能极大提升日常效率的案例。
While it's building, can you talk about, like, the most sophisticated use case or maybe, like, the the use cases that are very sophisticated in terms of workflow, but that change productivity tremendously for for people day to day.
让人一听就受启发的那种。
Something that everyone could get inspired by.
比如说,你可以让系统连接很多不同的插件。
So you can say, like, you know, go through, you can you can tag a lot of different plugins.
像 Gmail、日历,还有你用的文档工具之类的。
So you can say Gmail, calendar, if you have, like, docs.
对。
Yeah.
然后当我的首席助理。
And and be my chief of staff.
帮我梳理一天的日程,总结出重点事项,提前做好准备。
And give me a breakdown of my day, a summary of what is important and prepare me.
然后你只要在早上运行这个,你就可以比刚才给的例子更具体一些。
And then you just run this at the start of the day, and then you can be more specific than the the example prompt that I I gave here.
比如,我看到有些人提的要求还挺复杂的,想知道怎么做好充分的准备。
Like, some some people that I've seen, like, you know, have, fairly sophisticated, you know, requests for, like, an how to be prepared.
对。
Yeah.
还有,我是不是在浪费时间?
And, like, where am I wasting my time?
我是该雇个人来帮我处理那些我发现自己花大量时间在做的事情,
Should I be hiring someone to fix something where you see I'm spending most
还是该去开发一个能帮我优化流程的应用程序呢?
of my
还是去搭建一个能提升效率的工具?
hours or building an app that's gonna like optimization.
你最好跟 Codex 好好聊聊,一起 brainstorm 一下,看看怎么把流程理顺。
You should really discuss, you know, with with Codex then and start to brainstorm, you know, about how you can streamline things.
你的日历上已经排到这儿了。
Your calendar says it this far.
这事儿影响挺大的。
It's with high consequence.
所以看起来你目前做得还不错。
So it seems like you're doing
是的。
good.
明白了。
Okay.
对。
Yeah.
因为我只把重要的事安排进日历。
Because I only put important stuff in the calendar.
日常琐事就不用写了。
Don't put, like, day to day.
好的。
Okay.
不错。
Cool.
你个人是怎么用的?按什么顺序?分别用来处理哪些任务?
What do you personally use and in what order for for which tasks?
所以我用它来写代码比较多。
So I use it for coding a lot.
我拿它来帮咱们规划不同内容的发布策略,比如按什么顺序出,还有怎么构建叙事线。
I use it to help strategize for the different pieces that we're planning to put out, you know, in what order, like the narrative to to to construct.
所以它就像个特别棒的思维搭档。
So it's like it's a great thought partner.
我也用它来整理一天里各种零散的笔记。
Use it to, like, organize all my own notes throughout the day.
对。
Mhmm.
所以我不再用单独的笔记应用了,现在直接在Codex里记笔记。
So instead of taking notes in, a separate notes app, now I just take them, like, straight into Codex.
哦,这挺有意思的。
Oh, interesting.
能慢慢积累记忆。
Builds up memories.
你只要想说什么,就新开一个对话就行。
Like, you just start a new chat for anything you wanna say.
我就说:嘿。
And I'm just like, hey.
我刚才在想这个。
This is what I'm thinking.
是的。
Yeah.
到了晚上,它是自动帮你总结,还是得你主动提醒它?
And then at the end of the day, does it automatically summarize it or you have to prompt it?
我特意提醒它,把所有内容都整理到一个文档里。
I prompted to like, again, you know, dump everything into a doc.
而且它还能慢慢积累记忆。
But also it's like builds up memories.
我经常用它来跟进那些没回复的重要邮件或Slack消息。
I use it to often stay on top of unanswered, like important unanswered emails or Slack messages.
你有没有注意到,AI现在更多是帮人写代码,还是更多用在提升效率上?两者比例怎么样?
Are you seeing this shift from more code work being done with AI to more productivity stuff, or has it been equal?
挺有意思的是,你想想,工程师和程序员一整天到底在忙些什么?
So it is interesting to think about, you know, what is what does an engineer, what does a programmer do all day?
其实现在,他们大部分时间根本不是在写代码。
And, like, already there, the majority of the time is not spent on coding, actually.
我们观察到,随着时间推移,这种变化一直在发生,甚至在我们的技术用户群体里也是如此。
So we saw we saw a shift over time, you know, even including in our technical user base.
在Codex上完成的任务里,超过一半都是非技术性的工作。
Over way over, like, 50% of the tasks done on Codex are, like, nontechnical tasks
这么快?
Already?
是的。
Yes.
那有多久了?
And it's been what?
你用这个产品大概一年了吧。
A year since you
这个应用才上线三个月。
The app has only been three months.
我们三个月前重新发布了这个应用。
We relaunched the app three months.
Codex 命令行工具是去年推出的,那时候主要给程序员用。
The Codex CLI, which was definitely for programmers, that was a year ago.
那大家现在主要用它来做什么?
So what are people using it for?
你看到的是什么?
What do you see?
主要有哪些使用场景?
What are the use cases?
很多人用它来帮忙规划行程、整理事务,或者思考旅行安排。
There's a lot of, like, help me plan, help me organize, thinking through trips.
现在有了电脑和浏览器的支持,它还能帮你下单DoorDash、代你购物之类的。
Now with computer use and browsers, it can, you know, do things such as, like, order on DoorDash for you, like, you know, do your shopping.
需要花很长时间吗?
Does it take long?
我还没在浏览网页上试过,不过有些智能浏览器确实会……
Haven't tried it on browse yet, but sometimes the agentic browsers are like
不用。
No.
我们目前是市面上响应最快的电脑操作实现。
We have the fastest computer use implementation out there.
对
Mhmm.
出乎意料地快。
It's it's it's surprisingly quick.
你会看到它,点几下就完成了。
So, you know, you will see it, like, you know, click around.
比如说,我可以
So for example, I can,
你能去帮我下载一下我的领英数据分析吗?
can we ask you to go and download my LinkedIn analytics?
好的。
Okay.
那我们来添加一下电脑操作吧。
So let's do let's do add computer.
行。
Yeah.
然后去 LinkedIn 下载我的数据分析。
And then go to LinkedIn and download my analytics.
我们之前说好的格式,就是表格文件。
And then the format we said, a spreadsheet.
Excel 表格。
Excel spreadsheet.
对。
Yeah.
再新建一个表格。
And create a spreadsheet.
好了。
There we go.
咱们来看看它会怎么操作。
So let's let's see what it does.
它会使用桌面端的控制界面。
It'll use the desktop controls.
这可能需要五到
This might take five to
十分钟。
ten minutes.
它和ChatGPT的不同之处在于,它能直接使用你的电脑。
So so it's the difference from ChatGPT is that it's it can use your computer.
对。
Yes.
它能访问你的电脑。
It can use your computer.
它是个完整的智能代理,能运行并访问你的浏览器,还能在你的应用程序里点来点去。
It's a full agent that works and has access to your browser, can click around on your apps as well.
你可以在这儿看看。
And you can look at it here.
它已经正在访问了。
So it's already it's visiting.
是的
Yeah.
它已经打开我的领英页面了。
It is already on my LinkedIn page.
对
Yes.
我们来看看它接下来会做什么。
So let's see what it does over time.
我觉得它
I think it's
你看,它正在这里操作呢。
Here, you can see it's like it's navigating here.
哦,它在保存了。
Oh, it's saving.
一些东西。
Something.
它已经在保存内容了。
It is already saving stuff.
这太厉害了。
That's amazing.
在等他完成的时候,我想问你件事。
While we wait wait for him to finish, I wanted to ask you something.
你博士读了两周就退学了,对吧?
You dropped out of your PhD, right, after two weeks?
就上了两周。
After two weeks.
如果现在有人正在读博士、硕士或者本科,但特别讨厌这段经历,你还会建议他们去创业吗?
That's To start a company, would you recommend to someone who's doing their PhD or bachelor's or master's now and feel that they're just hating it?
现在放弃是不是个好时机,还是
Is it a good time to quit or
那是很久以前的事了。
This was a long time ago.
这都快十五年前的事了。
This was almost fifteen years ago.
所以我不确定该怎么说,但我的感觉是……
So I'm not sure if It seems like I feel like
这挺相关的。
It's relevant.
但有意思的是,这十五年来,教育的意义已经变了。
But but it's interesting how the meaning of education has changed in these fifteen years.
你看到OpenAI和DeepMind这些地方的创新,觉得怎么样?
What what do you think from, like, seeing all the innovation inside OpenAI, DeepMind?
现在的人还该不该去追求那种特别深入的学术训练,还是干脆直接去动手做点东西?
Should people be still pursuing, like, very deep education or should they just go out and start building?
我当初退学的原因是,读了没多久就意识到,把自己四年的生命绑定在一个特定领域上,并不是我真正想走的路。
The reason I dropped out of my PhD is, I just actually after starting it, I felt like that was committing four years of my life to that specific subject was actually not what I wanted to do.
对。
Mhmm.
我当时还有很多想法,想尝试的东西。
And I had more ideas and more things that I wanted to try.
对。
Mhmm.
所以我觉得创业环境特别适合我。
And so I felt like a startup startup environment was like right for me.
是的。
Yeah.
我觉得很重要的一点是,一次又一次地,就我自己来说,我慢慢明白,听从直觉真的很重要。
I think it's very important to, like, time and time again, like, you know, in my own life, you know, I've realized like it's important to follow your instincts.
去做那些让你充满干劲的事。
And do things that give you energy.
对你来说,那是正确的决定吗?
Was it the right decision for you?
当然了。
Oh, absolutely.
我想是的。
I guess it was.
确实有
Had such
一种
a
你
You
特别
know, great
我创办了广告公司,做了一年,之后去了谷歌工作。
I built the ad company for like a year and then I went on, worked for Google.
一开始是在做谷歌地图和一些广告相关的项目,后来我加入了DeepMind,那正是它最辉煌的时期,一个接一个的重大挑战不断出现。
Initially, was, like, working on Google Maps and different parts ads, and then I ended up working at the DeepMind, which was during the the the golden ages of DeepMind, like, when there was a grand challenge after grand challenge.
再后来,事情就变得没那么有意思了,那时候大家都在拼命扩大规模,而OpenAI已经证明了Transformer模型可以通过规模提升效果。
And then it became, like, a little bit less interesting And around the time where, you know, everyone was scaling and like OpenAI had proven that you could scale transformers Mhmm.
然后做出一些特别吸引人的东西。
And create something like very compelling.
这大概就是通往通用人工智能的道路。
And this was probably the road to to building AGI.
那时候我就想,算了,还是去上学,去OpenAI工作吧。
That's when, you know, I was like, just, okay, let me go to school work for OpenAI.
一次又一次,我都在挑战自己,去尝试新的东西。
And, you know, time and time again, it's sort of like, you know, challenging myself to pursue new things.
在我们讨论这些成果之前,还有一个最后的问题。
So before we look at what has created, one final question.
根据你看到的这些创新,你认为三到五年后我们的生活会变成什么样?
What do you think is going to happen to our lives in three to five years from all the innovation that you've seen?
你觉得会像很多人说的那样发生巨大变化,还是他们在夸大其词?
Do you think we're going to see this dramatic change that a lot of people are talking about or are they exaggerating?
我觉得一定会发生巨大的变化。
I think there is going to be dramatic change.
我觉得对我们所有人,尤其是对我个人来说,把好处真正普及到每个人身上非常重要。
I think it's important for us and for me personally to really bring the benefits to everyone.
我认为这意味着,不管你有没有主动使用这项技术,你都能享受到它带来的好处。
And one thing that I think this will mean is that you will have the benefits, whether you're actively engaging with this technology or not.
现在你得主动去提问,还得动脑筋想想该怎么问才好。
Today you have to actively prompt and sort of like, you be creative about what you, what you're asking.
说白了,你能得到多少好处,很大程度上取决于你提问的水平。
And, you know, it's almost like the benefits that you get are definitely proportional to like, you know, how good your prompting is.
但将来就不会是这样了。
You know, in the future that will not be true.
对我来说,它会像找一位高明的裁缝一样——他一眼看你,就懂你,知道什么样的布料最能衬托你,让你焕发光彩,完全贴合你的风格。
It's like, it will, to me, you know, be very similar to going to a nice tailor sort of like, you know, instantly looks at you and gets you and is, you know, this is the kind of cloth that would, you know, make you shine and that is, like, you know, just, like, right for you.
那到时候取决于什么呢?
So what will it depend on then?
如果不再靠你提问,那人们怎么才能发挥出最大潜力呢?
If it's not your prompting, what's gonna make people brilliant, I guess?
是的。
Yes.
你会更像从朋友那里得到好的建议,或者像我们现在这样自然地聊天一样受益。
You know, it will be much more like benefiting from, you know, good advice from a friend or, like, you know, having a natural conversation like the conversation that we have right now.
所以,关键是能问出对的问题,而不是刻意去引导或提示。
So, like, the the ability to ask the right questions, not to necessarily prompt them.
你会以真实的自己参与对话,然后在恰当的时候获得恰当的帮助。
It will it will it will be about engaging, you know, as your authentic self in a conversation and then getting the appropriate help in the right at the right time.
对。
Mhmm.
你会受益于这种无处不在的智能,它默默支持着整个社会。
And, you know, having benefiting from this, you know, ambient intelligence that is just like supporting the society.
明白了。
Okay.
它已经生成了一个表格。
It has created a spreadsheet.
哦,成了。
Oh, there we go.
所以它已经把所有分析数据都下载下来了。
So it has downloaded all the analytics from
哇,太棒了。
Oh, that's amazing.
嗯,没关系的。
Well, well, it's okay.
我们本来需要更多,不过这样也行。
We we needed but this is okay.
比如,如果我让它做更具体一点的操作,它只截了一张图,但我们还可以
Like, if I ask it to do more specific because it did like one screenshot, but we can
你还想让它做什么?
What else would you like it?
我们可以再运行一次。
We could run it again.
我需要更多数据。
I need more data.
你明白吧?
You know?
比如每条帖子的曝光量。
For example, impressions, per post.
好的。
Okay.
我现在用Codex特别懒,全都靠语音操作。
So I am very lazy now with Codex, and I just do everything through voice.
所以我需要更多每条帖子的曝光数据。
So I need more data impressions per post.
是的。
Yeah.
我也是。
Same same.
我所有操作都用Whisper流程来处理。
I just use the whisper flow across all my yeah.
好的。
Okay.
咱们结束今天的简报吧。
Let us end this daily brief.
已经完成了。
It did it.
是的。
Yeah.
哦,这越来越好了。
Oh, this is getting better.
看到了吗?
See?
然后你可以继续深入,慢慢往更高级的方向发展。
And then you can continue to go and, you know, you can, you can get more advanced.
如果你已经有自己喜欢的工作流程,那就可以用技能创建器来实现。
If you, there's a workflow then that you like, then, you know, you can, you can use, the skill creator.
这本身就是一个独立的技能。
This is a skill by itself.
我觉得你可以创建一个技能,把这套流程固定下来,每天直接运行就行。
I think you can say, like, create a skill to capture, this workflow, such that, I can run it every day.
然后它会为你生成一个专属的技能。
And then it will create a bespoke skill for you.
所以我们能直接安装它。
So we can install it.
当然可以。
Yes.
明白了。
Okay.
这太有意思了。
This is fascinating.
非常感谢你。
Thank you so much.
我今晚回家后得好好干点活了。
I feel like I have some work to do when I come back home tonight.
这真的太有意思了。
This is really fascinating.
谢谢你开发了这个功能。
Thank you for building it.
当然。
Of course.
是的。
Yeah.
能帮上忙我很高兴。
It was a pleasure.
谢谢。
Thank you.
谢谢。
Thank you.
我还写一份通讯,深入探讨我使用的AI工具、职业发展策略,以及那些在三十分钟播客里讲不完的内容。
I also write a newsletter where I go deeper on AI tools that I use, career strategies and things I can't fit into a thirty minute podcast.
完全免费。
It's free.
链接在简介里。
Link is in the description.
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕,帮助你打破语言障碍,轻松听懂全球优质播客。