工作的未来（如何成为AI优先）

目前最顶尖的AI模型比85%的人类更聪明。

The top AI model is now smarter than 85% of humans.

到2026年，它将比99.9%的人类更聪明。

By the 2026, it will be smarter than 99.9% of humans.

而你却以为自己还能保住工作。

And you think you will still have a job.

这是大卫·帕特森在X上发布的一条帖子，埃隆·马斯克实际上对此做出了大致正确的回应。

That's a post from David Patterson on X, and Elon Musk actually replied to this with roughly correct.

现在让我稍微让你的大脑超负荷一下，因为Fiverr的首席执行官米迦·考夫曼上个月给他的团队发了一封邮件。

Now let me overload your brain for a moment because the CEO of Fiverr, Micah Kaufman, sent out an email last month to his team.

这是邮件的第一部分。

So here's the first part of it.

如果你愿意，可以自己读一读。

You can read it if you'd like.

但我主要想强调的是这里的三个关键点，也就是一个令人不快的真相。

But the main things that I wanted to focus on were the three key points here, which is unpleasant truth.

AI正在席卷你的工作、他的工作以及所有工作。

AI is coming for your job and his job and every job.

简单任务将不复存在。

Easy tasks will no longer exist.

困难的任务将成为新的简单任务，而不可能的任务将变得困难。

Hard tasks will become the new easy and impossible tasks will become hard.

对着枕头大喊，重新振作起来，为未来做好准备，我们将在本视频中学习如何做到。

Scream into a pillow, pick yourself back up and become future proof, as we're going to learn in this video.

此外，像Shopify的CEO这样的其他首席执行官也发邮件告诉公司，每个人都应该以AI为先。

And on top of that, other CEOs like Shopify, the CEO, sent out an email telling his company that everyone should become AI first.

我的意思是，就连这篇文章中提到的Duolingo也在转向AI优先，用人工智能取代外包人员。

I mean, even in this article, Duolingo is going AI first and replacing contractors with artificial intelligence.

这吓坏了很多很多人。

And so that scares a lot of people.

对吧？

Right?

很多人不知道该做什么。

Many people don't know what to do.

他们不知道该学习哪些技能。

They don't know what skills to learn.

他们不知道哪些职业和工作会继续存在。

They don't know what careers are going to be around, what jobs are going to be around.

虽然我对这个问题在本视频中有许多想法，但我主要想分享的是如何做到以AI为先，以及你现在可以做些什么来为未来做好准备。

And while I have many thoughts about this in this specific video, I want to give my main thoughts how to become AI first and what you can do right now to prepare yourself for what's coming.

我想先说一句，我真心认为，现在是活着的最了不起的时刻之一。

And I just want to say to start this out, I truly believe that this is one of the most incredible times to be alive.

你此刻拥有如此多的机会。

You have so much opportunity right now.

你只需要先弄清楚那是什么，然后抓住它。

You just have to, one, find out what it is and then take advantage of it.

到视频结束时，我相信你会对你能做什么有更清晰的认识。

And by the end of this video, I feel like you'll have some clarity on what you can do.

所以第一个解决方案是成为AI优先型的人，因为使用AI的人分为三类。

So the first solution here is to become AI first because there are three types of people who use AI.

一类是几年前试过一次ChatGPT，觉得没什么特别的人。

There's people who tried ChatGPT once a few years ago and thought it wasn't special.

一类是用各种AI工具进行网络搜索、摘要和一些简单任务的人，这些任务不用AI可能多花十秒钟就能完成。

There's people who use various AI tools for internet searches, summaries, and simple tasks that would take maybe ten more seconds to do without AI.

还有一类是使用足够多，已经看清了AI潜力，并在所有能用的地方都使用它的人。

And then there's people who have used it enough to see the light and are using it in every place that they can.

我对AI有很多看法。

Now I have a lot of opinions on AI.

我不认为AI让人变笨。

I don't think that it makes people stupider.

我不认为它剥夺了你的创造力。

I don't think it takes away your creativity.

但我想让这个视频更实用，所以这些观点留到下一次再说。

But I want this to be a practical video, so I'll save that for another.

但大多数人不明白的是，AI的输出取决于你的技能和想象力。

But what most people don't understand is that the output of AI is up to your skill and imagination.

并不是AI不如你。

It's not that AI isn't as good as you.

而是你还不够擅长使用AI，无法让AI超越你。

It's that you're not good enough at AI to make AI better than you.

大多数人还认为这仅仅是把问题输入ChatGPT，但事实完全不是这样。

Most people think it's still just about typing questions into ChatGPT when that's not the case at all.

如果你给AI非常具体的指令，无论这些指令是长是短，它都能很好地完成。

If you give an AI extremely specific instructions on what to do, and these can be long or short, it will do those quite well.

比如我最近在Substack上发布了一篇关于如何为数字产品创建着陆页的文章，详细讲解了如何撰写文案，并附上了针对该任务生成提示词的指导。

Like I recently posted on Substack how to create a landing page for a digital product and going through how to write copy for that, and I included instructions to create a prompt around that to write copy.

但其中涉及太多不同的要素和环节，如果你只是简单地问AI：‘你能帮我写一个有说服力的着陆页吗？’然后就指望它直接输出结果，却不根据你已有的知识给予具体引导，那你就得不到最佳效果。

But there's so many different things and moving pieces there where if you ask the AI, hey, can you help me write a persuasive landing page and then just expect it to spit stuff out out of the box without guiding it in a specific direction because you already have the knowledge to do so, then you're not going to get the best results.

如果你能理解一点：如何找到最适合AI的指令，因为撰写着陆页有多种不同的方法，你才能获得最佳结果。

You're going to get the best results if you understand, one, how to find the best instructions for AI because there's many different ways to write a landing page.

你不知道AI选择了哪一个，因此你还需要让它分析你的写作风格，以便它能模仿你来写作。

You don't know which one the AI is choosing, and then you have to have it do a voice analysis of how you write so it can write like you.

你得向它提供你的产品信息。

You have to give it your offer information.

你得懂得如何设计一个产品方案。

You have to understand how to create an offer.

你必须把这些小部分都整合起来，才能让输出结果与你原本会创造的内容一致。

You have to put all of these little pieces together for the output to be what you would have created in the first place.

所以，最终还是你在创造这个东西。

So it's still you creating the thing.

你只是做了更少的工作，因为你创建、复用并不断优化的提示词，现在就像你的小员工一样。

You're just doing a lot less work because your prompts, the prompts that you create and reuse and refine over time are like your little employees now.

当然，让AI为你完成特定任务，只是整个拼图中的一小部分。

And of course, having AI do specific tasks for you is only one piece of the puzzle.

虽然目前还没达到理想状态，但很多人觉得，即使我用了大量详细指令，AI给我的回应依然不够好。

And while it's not there yet, a lot of people think like, oh, AI doesn't give me that good of a response even if I use a bunch of detailed instructions.

还没到那一步，但迟早会的。

It's not there yet, but it's going to be.

你可能会想，既然还没到那一步，我为什么要先学AI呢？

You may think like, oh, why should I become AI first if it's not there yet?

关键就在这里：如果你能真正理解AI，等它真正成熟时，你将远远超越那些没学过它的人。

And that's the exact point is if you get to the point of understanding AI, then by the time it does get there, you will run laps around anyone else who didn't learn it.

所以我给你的最大建议是：练习如何使用AI，试着让自己从工作中解脱出来。

So my biggest piece of advice here to practice, to learn how to use AI is to try to automate yourself out of work.

这会很有道理的，我向你保证。

And this will make sense, I promise.

但我想先引用一句话。

But I wanna start with a quote.

成为AI原生一代，不是关于为用户构建功能。

Being AI native isn't about building features for your users.

而是一种运营公司的模式。

It's an operating model for how to run your company.

这是一种工作方式、思维方式，是你和你的公司赖以生存的方式。

It's how you work, how you think, how you and your company breathe.

这是一种重新架构、重新连接，是从‘如何扩大人力规模’转向‘如何借助机器扩大决策、创造力和行动力’的哲学变革。

It's a rearchitecture, a rewiring, a philosophical shift from how do we scale humans to how do we scale decisions, creativity, and action with machines.

促使我制作这段视频的，是一些长期积累起来的想法，其中一个就是‘自我主导的职业生涯’这一概念。

Now what sparked this video is a culmination of ideas that have just been building up over time, and one of those ideas is the notion of the self directed career.

在工业化之前，自由个体大多是工匠和农民。

So before industrialization, free individuals were mostly artisans and farmers.

自由人的标志是，他们应当根据自己的兴趣行事，并在一生中从事多种工作。

The mark of a free person was that they were meant to act on their own interest and do many things throughout their life.

当时约80%的自由劳动者是自雇人士，而今天这一比例仅为10%。

Around 80% of free workers were self employed versus a mere 10% today.

而奴隶则被要求终生重复做同一件事。

Now slaves, on the other hand, were expected to perform one task for the rest of their lives.

那是一种非常重复性的工作。

It was very repetitive.

这非常机械。

It was very mechanical.

这些工作现在将由机器来完成。

It's work that machines are going to do now.

但在工业时代，当机器开始出现时，我们却陷入了电子表格、算法和工厂中，从事这种重复性工作。

But during the industrial age when machines started to come in, we just found ourselves trapped in Excel sheets and algorithms and factories doing this repetitive work.

那么问题来了，这场人工智能革命是否会进一步剥夺我们按照自身兴趣行动的自主权和自由？

So the question then is will this AI revolution further remove us from autonomy and freedom to act on our interests?

很可能大多数人会从自动驾驶的生活转向自动补全的生活，从被分配工作转变为工作被替我们完成。

And it's absolutely possible that most people will shift from autopilot living to autocomplete living, from having work assigned to them to having work done for them.

但对于那些重视创造性工作并希望掌控自己选择、维护自身能动性的人来说，还有另一种选择。

But for those who value creative work and taking control of the choices they make, their agency, there's another option.

通常，在技术转型期间，人类会通过发展更高层次的技能和知识来适应变化。

Normally, during technological transitions, humans adapt by developing higher level skills and forms of knowledge.

当技能因自动化而不可避免地贬值时，我们会向上抽象一层。

We abstract up a layer when skills inevitably become less valuable through automation.

现在，随着人工智能的发展，这种情况再次发生了。

Now with AI, this is kind of happening again.

我们必须超越所学的大多数技能，向上抽象一层。

We have to abstract up a layer beyond most of the skills that we've learned.

我们必须从体力劳动转向脑力劳动。

We have to transition from labor to mind.

尽管人工智能现在能够像我们一样思考和执行，但我们必须思考如何在那些本可替我们思考的系统中思考自己的思考方式。

While AI can now think and execute as well as us, we must think about how we think in relation to systems that could do the thinking for us.

因此，在继续之前，我希望你始终记住这个问题：我什么时候该利用人工智能，什么时候该自己动手？

So as we continue, I want you to keep this question in mind: When should I leverage AI and when should I do it myself?

这引出了第二个理念：先成为人工智能。

And that leads us to the second idea of becoming AI first.

这位名叫Signull的人是我新近喜爱的匿名网络思想家之一。

So this guy, Signull, is one of my new favorite anonymous Internet thinkers.

他有很多优质的信息。

He just has a lot of good, information.

现在这是个付费帖子，所以我得花钱。

Now this is a paid post, so, I had I had to pay.

我订阅了他。

I subscribed to him.

希望你别介意，哥们，我展示一下这个，我只是想给大家看几个例子。

I hope you don't mind, man, if I show this, but I I just wanna show a few examples here.

对吧？

Right?

所以他 basically 说，大多数公司已经落后了，不是即将落后，而是已经落后了，因为它们没有做到 AI 优先。

So he just pretty much says most companies are falling behind, not falling behind, already behind because they aren't AI first.

它们没有跟上我们正在经历的这场技术浪潮。

They aren't hopping on this technological wave that we're going through.

但我想给你看几个‘AI 优先’的例子，因为我觉得他讲得非常通俗易懂。

But I wanna show you a few examples of what it means to be AI first because I think this is he just put it in a very easy to understand way.

在产品或构建产品方面，非 AI 原生公司反馈周期慢、依赖人工调研、产品路线图基于主观意见；而 AI 原生公司则能在几分钟内总结所有用户访谈、根据功能需求聚类生成路线图选项、在发布前模拟用户行为。

So in terms of product or building a product, non AI native has slow feedback loops, manual research, roadmap debates based on opinion, and then AI native is summarize all user interviews in minutes, generate roadmap options based on feature request clustering, simulate user behavior before launching anything.

所以，即使作为一个创作者、一个个体经营者，想象一下当你在开发自己的产品时，如果能够完成所有这些工作，你就不用再花时间亲自去做这些事了。

So even as a creator, as a one person business, imagine when you're developing your own products, if you were able to do all of this because then you don't have to spend any of the time actually doing that.

你可以专注于更重要的事情。

You get to focus on the more important things.

这里的示例提示包括：总结最近50次用户访谈，按频率和强度对主要痛点进行聚类；根据用户反馈和使用数据，提出三个信号最强、噪音最低的产品方向；撰写一份新用户引导流程的产品规格文档，包含边缘情况和反方观点。

So prompts here as examples would be summarize the last 50 user interviews and cluster the main pain points by frequency and intensity, suggest three product bets with highest signal to noise, given the user feedback and usage data, write a product spec for a new onboarding flow, include edge cases and counter arguments.

现在说说社交媒体方面。

Now something on social media.

在社交媒体上，比如说：

In terms of social media is like, okay.

帮我分析我最新帖子下的这50条评论，告诉我人们最大的痛点、兴趣或问题是什么。

Digest these 50 comments from my latest post and tell me what people's biggest pain points or interests or questions were from that.

你可以让这些分析结果直接发送给你，帮助你更清楚地了解应该创作什么样的内容。

And you can have that delivered to you to give you more insight on the type of content that you should create.

它能给你带来更多洞察，因为你还可以围绕这些信息开发产品，并用更相关的数据来指导你的业务方向。

And it gives you a lot more insight than that because you can also build products around that and you can steer your business with more relevant information.

现在还有工程、设计，但有一件事是营销。

Now there's engineering, there's design, but one thing, marketing.

所以非AI原生的内容是手动制作、猜测什么能引起共鸣、活动节奏缓慢；而AI原生的内容则是大规模生成与测试、根据心理特征而非仅仅人口统计特征调整信息、通过创意优化实现实时反馈循环。

So not AI native, manual content, guessing what resonates, slow campaigns, AI native, generate and test content at massive scale, tune message to psychographics, not just demographics, live feedback loops with creative optimization.

所以示例提示词包括：为这篇博客文章生成10个标题变体，针对五种不同的创始人类型进行优化，按预期参与度排序，然后用我们的品牌语气撰写一封面向曾从定价页面离开的早期采用者的发布邮件。

So example prompts are generate 10 variations of this blog post headline, tuned to five different founder archetypes ranked by project ranked by projected engagement, and then write a launch email in our brand tone optimized for early adopters who have bounced from the pricing page.

这简直太疯狂了。

So this is, like, crazy.

对吧？

Right?

这太惊人了。

This is incredible.

而你只需用一句话、短短一句话，就能替代数小时的工作。

And you do it with a sentence, a single sentence rather than hours of work.

最后我想谈的是客服。

Now the last one I wanna go over is support.

非AI原生的方式是人工客服反复回答相同的问题。

So not AI native is human triage support reps re answering the same questions.

AI原生的方式是用24/7的LLM自动解决一级问题，为人工升级汇总复杂对话，并根据工单量动态生成帮助中心内容。

AI native is auto solve tier one issues with twenty four seven LLMs, summarize complex threads for human escalations, generate help center content dynamically based on ticket volume.

我觉得支持领域中很多这类事情已经在发生了，因为我们已经看到了24/7的LLM处理一级问题。

I feel like a lot of this is already happening in the support space because we see these tier one issues with twenty four seven LLMs.

真正酷的是，由于我运营着一家软件初创公司Kortex，想象一下，只需将用户遇到问题的对话总结成一篇网站文档，供任何人随时查阅，或者让24/7的LLM直接推送给有类似问题的用户。

What is really cool, and since I run a software startup, Kortex, is that imagine just being able to summarize a thread with a user that was experiencing a problem and turn it into some kind of documentation article for the website that people can go and look up at any time or that the 24 LLM twenty four seven LLM can just feed to a person that has a similar problem.

那么现在的问题是，如何起步，才能在那个时刻来临时做好准备？

So the question now is how do you start so that you're ready when that point hits?

我的建议依然是：练习自动化，让自己从重复工作中解脱出来。

And my advice again is to practice automating yourself out of work.

那么，我们该怎么做呢？

So how do we do that?

当你着手任何任务时，就这么做。

When you go to do any task, do this.

把整个流程详细地写下来，就像你在教别人如何做你的工作一样，包括你的思考过程或创意过程。

Write down the entire process in detail like you were teaching someone how to do your job, including thought processes or creative processes.

我们将在未来的视频中专门讨论这一点，我会谈到数字产品的未来，以及我如何看待信息类产品的发展，比如课程、辅导、自由职业等，这些创作者长期以来一直在做的事情即将发生巨大变化。

We're actually going to talk about that specifically in a future video where I'm going to talk about the future of digital products, where I see information products going like courses, coaching, freelancing, other things like that, what creators have done for so long that are going to see a drastic shift.

下一步是假设流程中的任何任务或部分都可以通过提示来完成。

The next step is to assume that any task or piece of that process can be done with a prompt.

然后尝试将这个任务或流程的某个部分转化为一个提示。

Then you attempt to turn that task or piece of that process into a prompt.

接着进行测试，记录它表现不佳的地方，并不断优化提示，直到它至少达到90%的准确度，然后将这个提示安全地保存起来。

Then you test it, note where it doesn't do well, and refine the prompt until it's at least 90% of the way there, and then you store that prompt somewhere safe.

如果你不知道如何完成这个任务，可以向AI询问完成该任务的详细步骤。

And if you don't know how to do that task, you can ask AI for detailed instructions of how to do that task.

不过，这很大程度上取决于你试图完成的具体任务，比如创建内容或制作YouTube视频。

Now, this really depends on the task that you're trying to accomplish for something like creating content or a YouTube video.

很多时候，你只会得到一些通用建议，因为每个创作者都有自己独特的方法。

A lot of the times you're just going to get general advice because every creator has their own way of doing those things.

这是一个高度依赖个人经验的领域。

It's a highly relative domain.

对吧？

Right?

除非人们分享出自己具体的流程，否则AI在这里很难有太好的表现。

That's where AI isn't going to thrive too much unless people give out their own specific processes.

所以如果阿里·阿卜杜勒、亚历克斯·霍尔莫齐和我直接去问AI，没有任何方向地问：嘿，我该怎么制作我的下一个YouTube视频？

So if Ali Abdul and Alex Hormozi and I just started asking AI without any direction like, hey, how do I create my next YouTube video?

我想讲的主题是这个，等等等等。

Here's the topic I wanna talk about, so on and so forth.

如果我们只是自己动手做，结果会比这样问AI好得多。

We wouldn't get as good results compared to if we were just going to do it ourselves.

所以你需要更具体一些。

So you need to get more specific.

你需要给AI明确的指令。

You need to give specific instructions to AI.

举个例子，我可以去 Ali Abdaal 的频道。

So as an example, I can go to Ali Abdaal's channel.

我可以选取他最受欢迎的一条 YouTube 视频。

I can take one of his highest performing YouTube videos.

我可以把这条视频交给 AI，让它分析视频的精确结构。

I can give it to AI and tell it to break down the exact structure of the video.

哪些部分有效？

What worked?

他在这里使用了哪些心理模式？

What are the psychological patterns he's using here?

整个视频的结构是怎样的？

What's the structure of the entire video?

请逐行详细拆解，说明每一处为什么有效。

Just break it down in extreme detail, every single line, why it works.

现在我有了详细的指令，可以将其转化为 AI 提示，用我自己的想法来复现这种模式。

Now I have detailed instructions that I can turn into an AI prompt to recreate that with my own ideas.

而且每一个YouTube视频都是不同的。

And every YouTube video is different.

所以如果你把一个表现不佳的视频喂给AI，并要求做同样的分析，那么你的视频很可能也不会表现得很好。

So if you were to feed it a video that didn't perform well and ask for the same thing, then your video probably isn't going to perform well.

这完全取决于你谈论的主题、你想制作的视频类型等等。

And it completely depends on the topics that you talk about, the types of videos that you wanna create, so on and so forth.

所以当人们说所有这些内容创作者的东西都会被自动化取代时，这并不是事情的运作方式。

So when people say like, oh, all of this content creator stuff is just gonna be automated out of existence, that's not how this stuff works.

另一种寻找具体且有用信息来喂给AI以完成任务的方法，就是通过书籍或PDF文件。

Another way to find specific and useful information to feed AI in order to do a task is just from a book or a PDF.

如果你想为着陆页、邮件或其他内容撰写文案，你可以上传《突破性广告》或《绝佳线索》或其他任何文案书籍，然后说：嘿，总结一下这本书。

If you want to write, copywriting for a landing page or emails or whatever it may be, you can upload the book Breakthrough Advertising or Great Leads or any other copywriting book and just say, hey, summarize this.

给我一个撰写文案的精确、非常详细的步骤，这样你就能获得一套指令，可以喂给AI，转化为提示词来生成这些文案。

Give me the exact, very detailed way of writing copywriting, and then you have instructions that you can feed AI, turn into a prompt to write that copy.

现在，我生活中最有价值的东西之一，是一个能生成提示词的优秀提示词。

Now, one of the most valuable things that I have in my life right now is a very good prompt that creates prompts.

这实际上是我在我发布于Substack的文章中教授的内容的一部分。

So this is actually a portion of what I taught in that article that I posted on Substack.

但我真正做的是，用它来撰写落地页，教人如何写落地页文案，或者将文案写作转化为我可以反复使用的提示。

But here's what I did really to write landing, to teach how to write landing page copy or to turn copywriting into a prompt that I can use over and over again.

现在，我本可以更详细一些，但我使用了Gemini 2.5 Pro，只是说：研究《突破性广告》这本书，给我一份关于如何撰写有说服力文案的极其详细的指南。

Now, I could have gotten more detailed with this, but what I did is I used Gemini 2.5 Pro, and I just said research the book Breakthrough Advertising and give me an extremely detailed guide on how to write persuasive copywriting.

它确实做到了。

And it did just that.

它为我提供了一份关于有说服力文案写作每个步骤的精彩指南。

It gave me an incredible guide to for every step of persuasive copywriting.

对吧？

Right?

这之所以如此出色，是因为通常要花很长时间才能详细掌握这些内容。

And the reason this is so incredible is because this would usually take so long to learn in detail.

要写出优秀的文案，需要很长时间。

It takes a long time to be able to write good copy.

如果你不想学习如何写文案，而是想专注于自己的技艺，比如写作、开发产品或设计东西，无论你做什么，但你知道自己需要了解营销、销售或其他类似的事情。

And if you don't want to learn how to write copy, you want to focus on your craft, you want to write, you want to build products, you wanna design stuff, whatever it may be, but you know that you need to learn marketing or sales or other things like that.

你可以直接创建一些提示词，这些提示词就像是为你做事的‘小员工’，能完成90%的工作，这已经远胜于完全不会做。

You can just create prompts, which are little employees that do those things for you that are 90% of the way there, which is still much better than 0% of the way there.

如果你在没有任何这些知识的情况下，为自己的产品写一个着陆页，它的表现和转化率根本无法与使用这种方法相比。

If you were to go and write a landing page for a product that you have without any of this knowledge, it's not going to perform or convert anywhere near as well as if you were to do this.

于是我让它输出了一套操作指南。

So I had it spit out instructions.

然后，因为我喜欢把提示词分成两个阶段，我说：假设你要用以上所有原则为我撰写着陆页文案。

And then from that, since I like to structure my prompts in two different phases, I said, imagine you were going to write landing page copy for me using all of these principles.

为了写出一个出色的着陆页，你需要从我这里收集哪些信息？

What is all of the information you'd need to gather from me in order to write an incredible landing page?

于是，基于这些内容，它列出了所有针对我的具体所需信息。

And so what it did from all of this is it just wrote down all of the context specific to me that it would need.

比如产品或服务的名称、描述、核心问题、期望达成的结果与收益、独特机制、独特卖点、目标受众等等。

So the name of the product or service, a description, the core problem, the desired outcome and benefit, the unique mechanism, the unique selling proposition, target audience, so on and so forth.

有了这些，就好像我继续说下去，它就能自动生成一个着陆页。

So with all of this, it's like I could continue talking here and like somehow get it to spit out a landing page.

但把所有这些信息整理成文档，方便日后参考，并转化为一个元提示，要简单得多。

But it's so much easier to take all of this information, turn these into documents that I can reference later, and turn them into a meta prompt.

所以，我会去写‘出色的AI提示’，并告诉它我想做什么。

So what I would do here is I go to write incredible AI prompts, and I tell it what I want to do.

对吧？

Right?

这就是我拥有用于创建提示的提示的地方。

So this is where I have the prompt that creates prompts.

所以，如果我想把某件事转化为提示，我总会来这里，而且我经常这么做，但我还是给你看看它做了什么。

So if I want to turn something into a prompt, I always go here and I do this very often, but I'll show you what it did.

所以，我想创建一个AI提示，用于生成一篇500到1000字的数字信息产品着陆页。

So, I want to create an AI prompt that generates a 500 to 1,000 word landing page for a digital information product.

请将你的提示分为两个阶段。

Structure your prompt in two phases.

第一阶段：信息收集，根据用户所需问题获取所有必要信息。

Phase one, context gathering, acquire all information you need according to the information inside questions needed from user.

这其实就是所有这些内容，也就是第二个输出结果。

What that is is all of this, this second output.

所以，它需要的所有背景信息都是我提供的，比如我问过：嘿，你需要我提供什么？

So all of the context that it needs from me where I asked, like, hey, what do you need from me?

所以我在这里提供了这些内容，因为我已经把它们整理成文档，并做了链接以便它能引用。

So I gave that here because I turned it into a document, then I linked it so it can reference it.

然后我只说：每次向用户提两到三个问题，以避免让他们感到压力。

And then I just said, ask the user two to three questions at a time to prevent overwhelm.

第二阶段：撰写实际文案。

And then in phase two, write the actual copy.

所以，运用《如何撰写文案》中的原则，也就是这个问题的第一个回答。

So use the principles from how to write copy, which is the first answer to this.

对吧？

Right?

如何撰写文案就在这里。

How to write copy is here.

而且内容非常长。

And then it's very long.

所以我将它保存为文档，并将它导入到这里。

So I save this as a document, and I fed it here.

然后它说，好的。

And then it says, okay.

接着它生成了一个提示，但用的是代码块格式。

And then it created a prompt, but it did it in a code block.

所以我告诉它：嘿，用Markdown格式来做。

And so I just told it, hey, do this in markdown format.

然后它将其分成了两个阶段。

And then it broke it down in two phases.

首先，它会对我进行访谈。

So first, it's gonna interview me.

它需要什么？

What does it need?

我的产品是什么？

What's my product?

我的目标受众是谁？

What's my target audience?

所有这些内容。

All of that stuff.

然后第二阶段，它会利用所有这些信息。

Then phase two, it takes all of that.

并根据如何撰写文案，也就是提示中的访谈部分。

And according to how to write copy, which it gives the or this is the interview section of the prompt.

然后第二阶段，撰写落地页文案。

And then phase two, write the landing page copy.

所以你要这么做。

So here's what you do.

引人入胜的标题、开篇导语、问题与欲望的强化、然后是好处和证据，等等等等。

Compelling headline, opening lead, problem desire intensification, so on, benefits and proof, so on and so forth.

所以，如果我想写一个着陆页，现在我就有了一个可以完成这项任务的提示。

So if I want to write a landing page, I now have a prompt to do so.

我所需要做的就是发送这个提示，提供我的信息，然后它就会生成一份着陆页的初稿，这份初稿比那些毫无经验的人所能写出的要好得多。

And all I need to do is send that prompt, give it my information, and then it will spit out a first draft of a landing page that's better than what most people can create when they haven't learned anything.

这就是我如何实现以AI为先的方式。

So that's how I would become AI first.

首先，要非常熟悉这个过程：弄清楚如何做某件事，写下详细的步骤，将其转化为一个元提示，这样你就可以反复运行它，并随着时间推移不断优化提示，让它变得越来越好。

It's just get really used to that process is one, figure out how to do something, write down the detailed instructions, turn that into a meta prompt so that you can run it over and over again and refine the prompt over time so it continues to get better.

很快，你就会拥有一整套提示库，它们能为你完成大部分工作流程。

And then soon enough, you have this library of prompts that does most of your workflow for you.

但当然，成为以AI为先的人并不是你唯一需要做的事情。

But of course, becoming AI first isn't the only thing that you need to do.

这只是一个你现在就可以练习和培养的技能。

That's just a skill that you can practice developing right now.

如果你真的想深入了解这一点，我制作了一个免费的迷你课程，涵盖了我们刚才所讨论的多个不同例子。

If you do wanna go deeper into that, I have a free mini course that I put out that goes over multiple different examples of kind of what we just did.

但你可以去参加这个课程，以便更深入地理解这个过程。

But you can go and take that course so you can just understand the process a bit more.

现在进入下一部分。

Now onto the next section.

你必须明白，你很可能需要改变自己的生活，而这是一件非常好的事。

You have to understand that you're probably going to have to change your life, and that's a very good thing.

在本视频的这一部分，我无权告诉人们该如何生活、该做什么。

With this section of the video, I'm not in any position to tell people how to live, what they should do.

我并不真正了解你的处境。

I don't really understand your situation.

我不了解大多数人的处境。

I don't understand most people's situations.

我还没有孩子需要抚养。

I don't have kids to feed yet.

我并没有同时打两份工。

I'm not working two jobs.

我只是一个28岁的小伙子，喜欢散步、写作和外出吃晚餐。

I'm just a 28 year old dude who likes to go on walks and write and eat dinner out.

虽然我认为你肯定能做些事情，你仍然可以学习，你仍然有一些时间，可以通过改变生活中的某些习惯来获得更多的能量，或弥补你所缺乏的东西。

Now while I think there definitely is something you can do, you can still learn, you still have some amount of time, and you can make certain habitual changes in your life that gives you more energy or gives you whatever it is that you lack.

随着时间的推移，你会逐渐进入一个更好的状态。

And slowly over time, you'll be able to get into a better position.

但对于那些生活特别忙碌的人，我不知道该如何帮助你们。

But to those people specifically who have these very busy lives, I don't know how to help you.

所以下一节内容其实并不适合你们。

So this next section isn't really for you.

但对于大多数至少有几小时空闲时间、经济上相对稳定、生活总体舒适的人来说，问题出在哪里呢？

But for the majority of people who have at least a few hours to spare and are at least somewhat financially stable and generally have a comfortable life, what's the issue?

你们到底在抱怨什么？

Why are you complaining?

尽管这一点现在被夸大了，却依然引发争议，普通的朝九晚五工作简直糟透了。

Even though it's overblown at this point and somehow still sparks controversy, the average nine to five jobs fucking suck.

我不是在说那些能在谷歌的睡眠舱里小憩的百分之零点零一的人。

And I'm not talking about the point 01% who can nap in sleep pods at Google.

几十年来，我们 collectively 一直讨厌工作。

We've collectively hated jobs for decades now.

进化会解决问题，而现在你生活中最痛苦的问题之一——上班——正在被解决，你却感到愤怒？

Evolution solves problems, and now that one of the most painful problems in your life is being solved, working a job, you are mad?

你根本没有资格抱怨。

You don't really have any room to complain here.

你正面临一生中最好的机会之一，却依然陷入最陈腐的舒适陷阱，扮演受害者，试图紧抓已经过时的生活方式，而所有美好的事物都不是永恒的。

You're presented with one of the greatest opportunities of a lifetime, and you're still falling into the most cliched trap of comfort and playing victim and trying to hold onto your old way of life when all good things aren't permanent.

你仍然没有意识到，如果你从事的工作能被机器取代，那么你的生活很可能缺乏新意、意义、满足感、挑战性、复杂性、持续的成长与学习。

You still don't realize that if you work a job that a machine can replace, your life probably lacks novelty and meaning and fulfillment and challenge and complexity and continuous growth and learning.

这本身就是一个强烈的信号，提示你该做些新的事情了，这可不是什么主观意见。

That alone is a massive signal to do something new, and this isn't an opinion.

这些并不是关于生活是否具有意义或满足感的主观看法。

Those things aren't opinion about things being meaningful or fulfilling.

这是有文献记录的心理学模式。

That's documented psychological patterns.

首先，你满足基本需求，然后才会追求自我实现的需求。

First, you fulfill your basic needs, and then you pursue your actualization needs.

为了保持某种心流状态，你需要随着时间推移逐步增加你能应对的挑战，因为这要求你发展出与挑战相匹配的技能。

And in order to stay in some kind of a flow state, you need to gradually increase the challenge that you can take on over time because that demands that you develop skills to match that challenge.

但如果你被困在一种重复性的工作中，一旦过了某个阶段就再无挑战，日复一日地重复相同的日子，这并不会让你感到满足。

But if you get stuck in this repetitive line of work where there is no more challenge after a specific point, you're just repeating the same day, day in and day out, that's not fulfilling.

这并不令人愉快。

That's not enjoyable.

你什么也学不到。

You're not learning anything.

你没有作为个体成长或进化，尤其是当你每天八小时、占你生命三分之一的时间都花在这里，另外三分之一时间在睡觉，剩下的三分之一时间则在刷手机，因为你没有更好的事情可做——你的生活已经定型了。

You're not evolving or growing as an individual, especially if you're spending eight hours a day, a third of your life there, and you're spending another third of your life sleeping, and then you're spending the other third of your life scrolling because you don't have anything better to dedicate your time to because your life is just set.

你每天都在做重复的工作。

You're doing the repetitive work every day.

你不再需要去追求或探索其他东西了。

There's nothing more that you need to go after or pursue.

所以对于那些现在已经深信不疑的人，你们需要开始创造一些东西了。

So for those who are amply convinced now, you need to start building things.

我想在这里谈的第一点是，剩下的就是精通与意义，因为在几年内，你要么被裁员，要么保住工作但最终仍会被裁，要么不断学习新技能以胜任公司里的新岗位，要么就得自己谋生。

And the first point I wanna cover here is that what's left is mastery and meaning because you have a few years before you are either let go or you keep your job and are eventually let go or you upskill endlessly to take on new roles in the company you work at or you have to fend for yourself.

人工智能不仅仅针对公司和程序员。

AI isn't just coming for companies and programmers.

它正在向每一个有头脑的人逼近。

It's coming for everyone that has a brain.

我相信在遥远的未来，我们会生活在一个不必为金钱担忧的世界里。

Now I'm sure somewhere in the far future, we'll live in a world where we don't have to worry about money.

但就目前而言，我们确实需要担心。

But for now, we do.

我们将经历一个适应期，许多人会因此失去工作。

And we're gonna go through an acclimation period where people are going to lose jobs.

我们目前没有任何措施来帮助他们，比如全民基本收入，不管你是否支持这一点。

We don't have anything in place to help them out, like UBI, whether you agree with that or not.

这对很多人来说将会是痛苦的。

And that's going to be painful for a lot of people.

所以我试图向你们展示，除了工作之外，还有其他赚钱的方式。

So I'm trying to show you that there is a way to do something other than your job in order to make money.

而这个AI主导的世界留给我们的，是一个以精通和意义为核心的经济。

And what this AI first world leaves us with is a mastery and meaning economy.

换句话说，就是发现并追求你一生的事业——你本该一直从事的那件事。

In other words, discovering and pursuing your life's work, You know, the thing you could have and probably should have been doing all along.

我不知道我们未来是否会在VR中、在火星上，或在星际空间中做这些事，但目前，我们是在互联网上做。

Now, don't know if we're going to be doing this in VR or on Mars or somewhere in intergalactic space, but for now, we do it on the Internet.

你选择一件你真正关心的事，深入学习、研究，并以AI优先的思维无情地掌握它，然后毫不羞愧地在公众面前分享你所知道的、你所做的，以及你为什么这么做，因为当今世界唯一的真正保障，就是一份无法被忽视的作品集。

You choose something you deeply care about, you study, research, and master it ruthlessly paired with an AI first mindset, and you shamelessly share what you know, what you do, and why you do it in public because the only real safety net in today's world is a body of work that's impossible to ignore.

这就引出了第二个观点，即注意力是唯一的差异化因素。

And that leads to the second point here, which is attention is the only differentiator.

随着世界充斥着越来越多的AI，信任、注意力和有效信息变得愈发稀缺。

As the world is filled with more AI, trust, attention, and signal become more scarce.

是的，死寂的互联网正在蔓延。

Yes, the dead Internet is growing.

正在产生大量的内容。

There is a lot of content being pumped out.

互联网上有很多机器人。

There's a lot of bots on the Internet.

任何人都可以使用AI，然后说：嘿。

Anyone can go into AI and be like, hey.

帮我写一条推文，然后他们就可以发布；或者制作一个YouTube视频，用AI生成配音，配上屏幕上的画面或一些背景素材，然后发布出去。

Write me a tweet, and then they can post it, or create a YouTube video and have the voice over in AI and someone on the screen or some kind of b roll and they can post it.

但从前面那个YouTube标题生成的例子来看，这并没有太大意义，因为他们不会得到好的结果。

But from the example before with the YouTube title generation, that doesn't mean much because they're not going to get good results.

他们无法灵活调整和迭代，从而真正提升这些内容的竞争力。

They're not gonna be able to pivot iterate and make those things better to actually compete.

这仅仅意味着市场的成熟度将继续快速提升。

It simply means that the level of market sophistication will continue to increase rapidly.

人们会厌倦千篇一律的内容，而这种模板每个月都会变化。

People will get bored of the cookie cutter, and the cookie cutter will change every month.

人们会逐渐失去对大多数内容的信任，而你仍需真正了解自己在做什么，才能让AI创造出独特且引人入胜的内容。

People will lose trust in most content, and that doesn't account for the fact that you still have to know what you're doing in order to have AI create something unique and compelling.

我的建议是，用AI来处理你作为个体经营者不想做的那些事情。

Now my advice is to use AI to do the things you don't want to do as a one person business.

第二点是，不要把那些你非常在意的事情完全交给AI掌控。

And the second thing is don't give AI complete control over the things you deeply care about.

不要让它替你完成工作。

Don't have it do the work for you.

我个人喜欢写作。

I personally like writing.

这就是我的专长，不一定是语法层面，或者写一些非常诗意的东西，而是把我想要表达的想法以我期望的方式写下来、表达出来。

That's my craft, not necessarily like the grammar aspect of it or writing something that's super poetic, but putting ideas on paper in the way that I want them to be put across or to be articulated.

当AI接管了整个过程时，我就感觉不到那是我的作品了。

And when I have AI take over that entire process, it doesn't feel mine.

我感觉不到自己创造了任何东西。

It doesn't feel like I created anything.

我感觉不到自己把在乎的东西交给了别人。

It doesn't feel like I've given something that I care about to someone else.

对吧？

Right?

正是在这里，意义才体现出来。

That's where the meaning comes into play.

但对于所有其他方面，这正是为什么对许多作家或创作者来说，现在是活着的绝佳时代——你可以专注于自己的专长，人们会看到你独特的内容并信任它。

But for all of the other aspects, this is why this is an incredible time to be alive for many writers or creatives in general is because you can focus on your craft, and people will see your unique content and trust that thing.

但现在，你可以处理一个人企业或任何企业生存所需的其他所有事情，比如营销、销售、客服、设计，无论你的企业规模多大，都取决于这一点。

But now you can do everything else that a one person business or a business in general needs to survive, like the marketing, the sales, the support, the design, however big your business grows, it depends on that.

但你可以让AI处理这些事情，让你的作品被更多人看到，从而维持你想要在人生中追求的事业。

But you can have AI handle those things so that your writing gets seen by more people so that you can sustain what you want to do with your life.

这并不虚假。

That's not inauthentic.

这是大规模的真实表达。

That's authenticity at scale.

这和雇人帮你做是一样的，因为无论哪种方式，你都没有亲自做那些工作。

It's the same thing as hiring employees to do it for you because either way, you're not doing that work.

但这对你实现使命来说是必要的。

It's just necessary for you to achieve your mission.

那这意味着什么？

Now what does that mean?

这意味着你就是那个细分领域。

It means you are the niche.

你就是你的独特之处。

You are the differentiator.

你的专长、经验，以及通过一生中所处理的每一条信息所形成的独特世界观。

Your mastery, experience, and way of looking at the world through a perception forged by every bit of information you've processed over the entirety of your life.

当人工智能制作的90%产品都千篇一律时，实际上没什么改变。

When AI makes 90% of the products the same, nothing really changes.

人们依然会购买他们认识、信任并关心的人和品牌的产品。

People continue to buy from people and brands that they know, trust, and care about.

这与其说是构建一个销售漏斗，不如说是打造一个让人们可以探索的世界。

It's less about building a sales funnel and more about building a world that people can explore.

我也写过一篇关于这个主题的文章。

I wrote an article about that as well.

这篇文章叫《如何打造一个世界：两小时内容生态系统2.0》。

It's called how to build a world, the two hour content ecosystem two point o.

所以我会在描述中留下链接，供你去查看。

So I'll leave a link to that in the description to check it out.

现在，这里第三个也是最后一个观点， arguably 最重要的，是‘一千个铁杆粉丝’正变得越来越相关，因为大多数人并不想出名。

And now the third and final point here, and arguably the most important, is that 1,000 true fans is increasingly relevant because most people don't want to be famous.

即使你有这个目标，也请等到你真的能靠它谋生时再实现。

And even if you do, save that goal until you're actually making a living.

这正是大多数人想要的，而你并不需要数百万粉丝才能接近这个目标。

That's what most people want, and you don't need millions of followers to get anywhere close to that.

你只需要大约一千名忠实粉丝。

You need about a thousand true fans.

说实话，甚至不需要那么多，因为如果你擅长自己所做的事情，你可以每名客户收费5000美元。

And honestly, it's even less than that because if you're good at what you do, you can charge $5,000 per client for a service.

你可以为付费通讯收取10美元。

You can charge $10 for a paid newsletter.

是的，我现在对Substack很着迷。

And, yes, I'm on the Substack kick now.

也许我以后会拍一个视频，谈谈我对Substack的看法以及我认为它的发展方向。

Maybe I'll make a video on that in the future about my thoughts on Substack and where I think it's going.

但就目前而言，它看起来很有前景。

But for now, it seems promising.

你可以为任何类型的产品定价在50到150美元之间，然后还可以推出衍生产品，比如书籍或软件，因为任何人都能编写一个解决买家特定问题的应用程序，让他们再次购买。

You can charge anywhere from $50 to a $150 for any type of product, and then you can create spin off products like a book or software since anyone will be able to code up an app that solves a specific problem for buyers to buy again.

我认为，大多数人每月5000美元就能过得很好。

Now, I would argue that most people can live just fine off of $5,000 a month.

如果你有家庭，责任更重，那每月需要1万到1万5千美元，这只需要三个客户。

And if you have a family and more responsibilities than $10,000 $15,000 a month, that's three clients.

每个月三个客户就够了。

That's three clients a month.

我知道，如果你一直在在线商业领域，这听起来像是个老生常谈。

I know if you've been in the online business space, this is like a cliche thing.

人们总说，搞定一个客户或三个客户，就能替代你的收入，但事实确实如此。

It's like, oh, land one client or land three clients and you replace your income, but it's true.

如果你有一个10美元的订阅服务——我并不推荐新手这么做——请看我上一个视频，那是基于创作者建议的问答内容。

And if you have something like a $10 subscription, which I really don't recommend as a beginner, watch my last video, which was a q and a based on creator tips.

但如果500人购买了这个订阅，那就是每月5000美元的持续收入。

But if 500 people buy that subscription, then that's $5,000 a month recurring.

要达到这个目标需要花更多时间，因为你得卖出更多产品，但这只是说明你能做到什么的一个例子。

It takes a lot more time to get there because you have to sell a lot more, but it's just an example of what you can do.

在拥有四十亿到五十亿用户的互联网上，找到五百个用户购买你的产品，并不是什么难事，你完全可以分到自己的一块蛋糕。

500 people on an Internet that has four to 5,000,000,000 people there isn't that much task for you can take a piece of your pie.

如果在一年内，你逐步建立起一个产品组合，包括高价产品、低价订阅服务和中间价位的产品，那么你每个月只需要一个人购买高价产品。

And now if over the course of a year, you build out an offer stack where you have a high ticket product, you have a low ticket subscription, and you have a product in between, then you need, like, one person to buy the high ticket a month.

你需要十个人购买订阅服务。

You need 10 people to buy the subscription.

你需要十个人购买实际的产品。

You need 10 people to buy the actual product.

考虑到社交媒体上注意力的流动方式、人们如何关注新账号，以及那些局外人看不到的自然波动，实际上有足够的注意力可供分配。

And considering how attention flows on social media, how people go on to follow new people, and other natural fluctuations that those outside of the game can't see, there is really more than enough attention to go around.

这里的关键是，你要把这件事当作你工作的反面来看待。

The key here is that you treat it as the opposite as your job.

这是一种不断进化的劳动。

It's work that evolves.

这是一份要求持续学习的工作。

It's work that demands continuous learning.

这份工作有一个前提，那就是你至少要把自己的生活理顺一些。

It's work with a prerequisite that you have at least some of your life together.

一旦你停滞不前，熵就会增加，而这是一件好事。

The minute you stagnate is the minute entropy increases, and that's a wonderful thing.

保持不变是美好生活的敌人。

Staying the same is the enemy of a good life.

感谢你观看这个视频。

Thank you for watching this video.

请查看描述中的链接，了解我们的笔记和AI软件Kortex，它集成了所有模型和25个以上的工作流程。

Check out links in the description to Kortex, which is our note taking and AI software with all models and 25 plus workflows in one place.

你可以订阅我的Substack，每周免费收到两封信，或者选择包含更多附加福利的高级版本。

You can subscribe to my Substack for two free weekly letters, or you can check out the premium version of that that has some more bonuses.

本视频中提到的其他资源链接也都在描述中。

And other links to the resources I mentioned throughout this video are also in the description.