章节 1：自动化推特（X）运营的“精选集”策略

📝 本节摘要：

本节作者开门见山地介绍了利用 Claude AI 自动化运营推特（X）的显著成果，即通过打造一套“精选集系统”（Greatest Hit System）突破流量瓶颈，实现日均触达 25 万人的数据增长。作者提出了一个核心比喻：将社交媒体账号视为“健身房”——大多数人只是盲目地做随机运动（发帖），而赢家会像追踪健身数据一样，精准识别哪些动作（内容）能“增肌”（有效），哪些只是在浪费时间。

[原文] [Speaker]: Today I'm going to show you how I automate Twitter or X with Claude and this is a really powerful way for example that we use to reach 250,000 people a day as you can see on X and Twitter we get results like this per month.

[译文] [Speaker]: 今天我要向大家展示我是如何利用 Claude 自动化运营 Twitter 或 X 的，这是一种非常强大的方法，例如我们利用它每天触达 250,000 人，正如你在 X 和 Twitter 上所见，我们每个月都能获得这样的成果。

[原文] [Speaker]: The way that we're using this is by creating a greatest hit system which I'm going to show you where AI actually analyzes what's performing right now what's not working and then how can we create amazing tweets that reach a lot of people.

[译文] [Speaker]: 我们使用这种方法的方式是创建一个“精选集系统”（greatest hit system），我会向你展示 AI 实际上是如何分析当前什么内容表现良好、什么不起作用，然后我们如何创作出能触达大量人群的精彩推文。

[原文] [Speaker]: So you can see for example these are our analytics over the last few weeks and we're gradually breaking through these plateaus we're getting some amazing results etc you can see our followers and follows over time we're getting more and more and you can see the engagement rate right and most of the metrics are up massively.

[译文] [Speaker]: 所以你可以看到，例如这是我们过去几周的分析数据，我们正在逐渐突破这些平台期，我们获得了一些惊人的结果等等，你可以看到我们的粉丝和关注随时间的推移越来越多，你可以看到互动率，大多数指标都在大幅上升。

[原文] [Speaker]: So for example here you can see repost likes bookmarks shares engagements engagement rate etc is all increasing as we use this new system that'll break down for you right and essentially what you're going to learn today is how to find your winning tweets the 10% that actually get reach how to analyze why they worked which is what most people skip how to build a claw project that writes tweets in your voice the exact prompt I use to generate unlimited content ideas and how to do this even if you're brand new to X or you don't have any followers.

[译文] [Speaker]: 例如在这里你可以看到转发、点赞、书签、分享、互动、互动率等都在增加，正如我们将为你拆解的这个新系统所带来的效果一样，本质上你今天要学到的是如何找到你的“获胜推文”——即那 10% 真正获得覆盖率的推文，如何分析它们为什么有效——这是大多数人忽略的，如何构建一个能模仿你语气写推文的 Claude Project（注：原文 claw 为 Claude 之误），我用来生成无限内容创意的确切提示词，以及即使你是 X 新手或没有任何粉丝该如何做到这一点。

[原文] [Speaker]: And this system is probably one of the most powerful things I've ever built and I'm going to show you why this system works.

[译文] [Speaker]: 这个系统可能是我建立过的最强大的东西之一，我将向你展示为什么这个系统有效。

[原文] [Speaker]: All right so let's get straight into it and you want to think of it like this like X account your ex account is like a gym right some exercises build muscle some exercises waste time right.

[译文] [Speaker]: 好了，让我们直接开始，你要这样思考，就像 X 账号，你的 X 账号就像一个健身房，对吧，有些练习能增肌，有些练习则浪费时间，对吧。

[原文] [Speaker]: Most people just show up and do random thing but winners track everything we're going to track what works on X and then do more of it right.

[译文] [Speaker]: 大多数人只是露个面然后做些随机的事情，但赢家会追踪一切，我们要追踪在 X 上什么有效，然后做更多这样的事情，对吧。

章节 2：数据驱动的内容分析法

📝 本节摘要：

本节深入讲解了如何通过数据分析筛选出“前10%”的优质内容（Winners）并剔除“后90%”的无效内容（Losers）。作者展示了一个核心工具——电子表格，用于追踪推文的发布日期、浏览量以及最重要的“钩子”（Hooks，即推文的前两行）。作者强调，根据 80/20 法则，钩子决定了绝大部分的传播效果。通过将历史数据明确分类为“表现最佳”和“表现最差”两个列表，为后续训练 Claude 提供了关键的对比样本。

[原文] [Speaker]: Today is how to find your winning tweets the 10% that actually get reach how to analyze why they worked which is what most people skip how to build a claw project that writes tweets in your voice the exact prompt I use to generate unlimited content ideas and how to do this even if you're brand new to X or you don't have any followers.

[译文] [Speaker]: 今天要讲的是如何找到你的“获胜推文”——即那 10% 真正获得覆盖率的推文，如何分析它们为什么有效——这是大多数人忽略的，如何构建一个能模仿你语气写推文的 Claude Project（注：原文 claw 为 Claude 之误），我用来生成无限内容创意的确切提示词，以及即使你是 X 新手或没有任何粉丝该如何做到这一点。

[原文] [Speaker]: All right so first of all what it's going to do is actually going to find your winners right so the X analytics show you exactly which tweets got the most impressions and these are your best performers right.

[译文] [Speaker]: 好了，首先这套系统要做的是真正找到你的赢家，对吧，所以 X 的分析数据会准确显示哪些推文获得了最多的展示量，这些就是你表现最好的内容，对吧。

[原文] [Speaker]: For example if we go to X you can see that if we check out our tweets over here on our profile you can see how we're performing what's working for us how many people we're reaching in each tweet.

[译文] [Speaker]: 例如，如果我们去 X，你可以看到如果我们检查个人资料里的推文，你可以看到我们的表现如何，什么对我们有效，每条推文触达了多少人。

[原文] [Speaker]: Now the problem is that if you're tweeting out a lot you're going to have a bottom 90% that don't perform well and so you want to keep avoiding those right.

[译文] [Speaker]: 现在的问题是，如果你发了很多推文，你会有底部那 90% 表现不佳的内容，所以你要尽量避免那些，对吧。

[原文] [Speaker]: And so what I'm going to show you today is how to avoid the losers and how to build a content machine where Claude actually understands your winning patterns he can write unlimited tweets that follow the same formula without sounding like a robot without copying other people and without losing your voice.

[译文] [Speaker]: 所以我今天要向你展示的是如何避免“失败者”，以及如何构建一个内容机器，让 Claude 真正理解你的获胜模式，它可以写出无限多遵循相同公式的推文，而且听起来不像机器人，不抄袭别人，也不会失去你自己的声音。

[原文] [Speaker]: What people don't understand about Claude is that regular Claude is not going to remember everything from previous chats right but what you're going to do here with Claw projects as you can see is you can set up and upload your winners.

[译文] [Speaker]: 人们不理解 Claude 的一点是，普通的 Claude 不会记住以前聊天中的所有内容，对吧，但在 Claude Projects（原文口误为 Claw projects）里你要做的是，正如你所见，你可以设置并上传你的赢家。

[原文] [Speaker]: So I've actually added like all my best performing tweets as you can see on the right hand side here and then we can use that to generate more of them and then you're going to insert some custom instructions that show you exactly how to do this.

[译文] [Speaker]: 所以我实际上已经添加了我所有表现最好的推文，正如你在右侧这里看到的，然后我们可以利用它来生成更多类似的内容，接着你要插入一些自定义指令，确切地告诉你该如何操作。

[原文] [Speaker]: So let's get straight into it what am I doing right okay so what you got here is a spreadsheet and I track the date the X post the views and then the hook right.

[译文] [Speaker]: 让我们直接开始，我在做什么呢，好吧，所以你在这里看到的是一个电子表格，我追踪日期、X 帖子、浏览量，然后是“钩子”（hook），对吧。

[原文] [Speaker]: And really what I tend to find is like with hooks on X that's like the 8020 of its performance right so if you get a really good hook like you can see right here with this process you can get a lot of views right.

[译文] [Speaker]: 实际上我倾向于发现，对于 X 上的钩子，这就像是其表现的 80/20 法则，对吧，所以如果你有一个非常好的钩子，就像你在这里看到的一样，通过这个过程你可以获得很多浏览量，对吧。

[原文] [Speaker]: If you have the right hook really the first two lines are the most important when it comes to X.

[译文] [Speaker]: 如果你有正确的钩子，实际上对于 X 来说，前两行是最重要的。

[原文] [Speaker]: So what we're doing here is we're analyzing okay when did we post this tweet what's the link how many views did it get and then what's the hook we actually used for that tweet right.

[译文] [Speaker]: 所以我们在这里做的是分析，好吧，我们什么时候发的这条推文，链接是什么，获得了多少浏览量，然后我们那条推文实际使用了什么钩子，对吧。

[原文] [Speaker]: And so if we keep scrolling down here you can see for example we keep a track of our greatest hits and then we have a list of our top 10% hooks so these are the best performing hooks that we've used on X.

[译文] [Speaker]: 如果我们继续向下滑动，你可以看到例如我们追踪了我们的“精选集”，然后我们有一个前 10% 钩子的列表，这些是我们用在 X 上的表现最好的钩子。

[原文] [Speaker]: And these are the winners we're basically we're going to tell Claude to do more of this and then over here you can see these are our worst performing hooks right so these are the bottom 90% hook styles that didn't perform and we want to avoid these right.

[译文] [Speaker]: 这些是赢家，基本上我们要告诉 Claude 多做这些，然后在这里你可以看到这些是我们表现最差的钩子，对吧，所以这些是底部 90% 表现不佳的钩子风格，我们要避免这些，对吧。

[原文] [Speaker]: And so when you go into Claude and you're updating the spreadsheet each day and you're tracking your tweets and how many views they got and the date etc what you're going to do is you're going to add the top 10% so the ones that get the best views inside this list then you're going to get the bottom 90% and you're going to add them to this list.

[译文] [Speaker]: 所以当你进入 Claude，你每天更新电子表格，追踪你的推文、浏览量和日期等，你要做的是把前 10% 即浏览量最好的那些加入这个列表，然后把底部 90% 加入那个列表。

章节 3：Claude Projects 的配置与训练

📝 本节摘要：

本节详细指导如何在 Claude 中创建并配置 Project。作者将其命名为“精选集重写器”（Greatest Hits Rewriter）。核心步骤在于“自定义指令”（Custom Instructions）的编写：作者将之前筛选出的“表现最差的90%”和“表现最好的10%”的推文全部粘贴进去，并设定目标受众。通过明确的提示词——要求 Claude 基于输入的素材，模仿成功案例的模式并规避失败案例的错误来重写内容，从而让 AI 能够精准分析过往的成功规律，生成具有病毒传播潜力的新推文。

[原文] [Speaker]: And then you're going to go inside Claude here as you can see inside Claude you're going to go to projects add a new project so you add the name and the description and once you set that up it's going to look something like this right.

[译文] [Speaker]: 然后你要进入 Claude，正如你在 Claude 内部所见，你要去“Projects”（项目），添加一个新项目，所以你添加名称和描述，一旦你设置好了，它看起来就会像这样，对吧。

[原文] [Speaker]: So for example I've called it greatest hits rewriter now in the instructions what I've actually done is I've pasted all the bottom performers to the bottom 90% and the top 10%.

[译文] [Speaker]: 举个例子，我把它命名为“精选集重写器”（Greatest Hits Rewriter），现在在指令部分，我实际做的是把所有表现垫底的——即那后 90% 的内容，以及前 10% 的内容都粘贴了进去。

[原文] [Speaker]: And then I've added my target audience for content on Twitter and I've said based on what I paste in rewrite these to go viral and use the winning tweets above to come up with a copy based on what already worked right on the top 10% and how to avoid the bottom 90%.

[译文] [Speaker]: 然后我添加了我在 Twitter 上的目标受众，并且我（在指令中）说：基于我粘贴的内容，重写这些内容以实现病毒式传播，并利用上面的获胜推文构思文案，基于那些已经在前 10% 中被证明有效的元素，以及如何避免那后 90% 的错误。

[原文] [Speaker]: And the cool thing is Claude could basically analyze the patterns of what's worked previously and then use that to rewrite and create new tweets.

[译文] [Speaker]: 最酷的事情是，Claude 基本上可以分析之前行之有效的模式，然后利用这些模式来重写并创作新的推文。

章节 4：从视频到推特的实操工作流

📝 本节摘要：

本节进入核心实操阶段，展示了如何利用“宽容”的算法机制进行高频测试。作者详细演示了具体的“视频转推文”工作流：首先下载最新的视频素材（或使用图片），提取其字幕文案，然后将其输入到配置好的 Claude Project 中。Claude 会基于之前的成功案例（Winners）和视频上下文，生成多条具备“病毒潜力”的推文选项。作者还展示了如何利用 Unicode 加粗前四行以优化视觉效果，最后将发布后的数据回填至电子表格，形成数据闭环。

[原文] [Speaker]: Now the other thing with X this is something that I found is it's a very forgiving algorithm right which means that you can create a lot of content and if one tweet doesn't work that's okay post another right.

[译文] [Speaker]: 关于 X 的另一件事，也是我发现的一点，就是它的算法非常“宽容”，对吧，这意味着你可以创作大量内容，如果一条推文效果不好，没关系，发另一条就是了，对吧。

[原文] [Speaker]: So we'll post a lot of content each day and then we'll get more data on what works well right so for example this one is beginning to take off this is from 5 hours ago and it's generated 1,000 views right.

[译文] [Speaker]: 所以我们每天会发布大量内容，然后我们会获得更多关于什么内容有效的数据，对吧，例如这一条开始起飞了，这是 5 小时前发的，已经产生了 1,000 次浏览，对吧。

[原文] [Speaker]: And so when we're generating these tweets we create a lot of content we get more data the more data we get by creating a lot of content the better our clawed project gets because we update this daily right.

[译文] [Speaker]: 当我们生成这些推文时，我们创作大量内容，获得更多数据，我们通过创作大量内容获得的数据越多，我们的 Claude Project（原文误识别为 clawed project）就越好，因为我们每天都在更新它，对吧。

[原文] [Speaker]: And then every time we have a good tweet we just put it into the list here as you can see right banger post and you can see all these different conversations we've done right now.

[译文] [Speaker]: 每当我们有一条好的推文，我们就把它放进这个列表里，正如你所见，没错，“爆款帖子”（banger post），你可以看到我们现在完成的所有这些不同的对话。

[原文] [Speaker]: How do you start creating and automating the content with Claude from here here's what I do this is my process right so I go into my SAP section here as you can see so we're going to open up these and I've just got like a copy and paste list that I can quickly go through right.

[译文] [Speaker]: 此时你该如何开始利用 Claude 创建并自动化内容呢？我是这么做的，这是我的流程，对吧，所以我进入我的 SAP 部分（注：指特定的标准操作流程文件夹或文档），正如你所见，我们要打开这些，我有一个可以快速浏览的复制粘贴列表，对吧。

[原文] [Speaker]: Then what I'm going to do is I'm going to download like one of the latest videos so for example I can take this one download it if you don't have videos that's okay you can use images or generate images with Nano Banana or even generate automated videos with notebook okay.

[译文] [Speaker]: 然后我要做的是下载最新的视频之一，例如我可以拿这个，把它下载下来，如果你没有视频也没关系，你可以使用图片，或者用 Nano Banana（注：可能是指特定的 AI 图片工具）生成图片，甚至用 Notebook 生成自动化视频，好吗。

[原文] [Speaker]: You're going to download the video as you can see and then in the meantime whilst that's downloading we can go to view video on YouTube we're going to open up the transcript as you can see here copy the transcript then we're going to go over to Claude to our project paste in the transcript from the video and in the meantime we've got the video downloaded so we'll plug that into Twitter as you can see.

[译文] [Speaker]: 正如你所见，你要下载视频，在下载的同时，我们可以去 YouTube 查看视频，我们要打开字幕，正如这里所示，复制字幕，然后我们转到 Claude，进入我们的项目，粘贴视频的字幕文案，与此同时视频也下载好了，所以我们把它插入到 Twitter 中，正如你所见。

[原文] [Speaker]: And then that's going to start generating the tweet based on the highest viral potential right and so what this is doing is it's actually analyzing based on our previous tweets and the transcript that we've plugged in let's create something contextually relevant to that video that we can then copy and paste straight into Claw.

[译文] [Speaker]: 然后它就会开始基于最高的“病毒传播潜力”生成推文，对吧，所以这里正在做的是，它实际上在基于我们之前的推文和我们插入的字幕文案进行分析，让我们创建一些与该视频语境相关的内容，然后我们可以直接复制粘贴进 Claude（注：此处语境应指从 Claude 复制出来，或指流程中的下一步操作，原文直译）。

[原文] [Speaker]: Now it's going to give me multiple options because sometimes you won't like the top option or you think okay maybe there's a bunch of other stuff that we can get right and you can see like the the way that it's formatted is super nice but it also follows a format that's already worked for us previously right.

[译文] [Speaker]: 现在它会给我多个选项，因为有时你可能不喜欢第一个选项，或者你想，好吧，也许还有很多其他东西我们可以获取，对吧，你可以看到它的格式非常漂亮，而且它也遵循了我们之前已经行之有效的格式，对吧。

[原文] [Speaker]: And you can always wait for that to fully roll out so we can wait for all the content to be written as you can see or what we can do is just stop that and then just take the top tweet right.

[译文] [Speaker]: 你可以一直等到它完全生成完毕，所以我们可以等待所有内容写完，正如你所见，或者我们可以做的就是停止生成，然后直接取用第一条推文，对吧。

[原文] [Speaker]: So for example if we copy this we go to Twitter like so and then we're just going to format it nicely as you can see sometimes when you copy and paste it doesn't look as nice but to create like a really high performing tweet now it only takes me like less than 60 seconds.

[译文] [Speaker]: 例如，如果我们复制这个，像这样去 Twitter，然后我们要把它格式化得漂亮一点，正如你所见，有时当你复制粘贴时它看起来没那么好，但要创建一个真正表现出色的推文，现在只需要我不到 60 秒的时间。

[原文] [Speaker]: My team can use this as well my team can use the same prompt and so we've got the video attached we've got the content from Claude right here based on the hooks that previously worked for us it's also bolded the top four lines so it's use uni code to bold that when we copy and paste from Claude directly.

[译文] [Speaker]: 我的团队也可以使用这个，我的团队可以使用相同的提示词，所以我们附上了视频，我们这里有来自 Claude 的内容，这些内容是基于之前对我们有效的“钩子”生成的，它还加粗了前四行，它是利用 Unicode 来加粗的，当我们直接从 Claude 复制粘贴时就是这样。

[原文] [Speaker]: And then if we hit enter and command if you're on a Mac then you can see we've got the post right here then what I do is I go back to my sheet so if we go back to my sheet over here we're going to paste that in we're going to add the date so let's add the date here we're going to plug those in add in the number and we'll come back and measure those in 24 hours.

[译文] [Speaker]: 然后如果我们按下 Enter 和 Command 键（如果你用 Mac 的话），你可以看到我们就有了这条帖子，然后我要做的是回到我的表格，所以如果我们回到这边的表格，我们要把它粘贴进去，我们要添加日期，所以在这里添加日期，我们要把那些插进去，添加编号，然后我们会在 24 小时后回来测量这些数据。

[原文] [Speaker]: And we just keep building up more and more data on what's working.

[译文] [Speaker]: 我们就这样不断积累越来越多关于什么有效的数据。

章节 5：持续优化与应对常见误区

📝 本节摘要：

本节重点讲述了如何利用数据迭代形成“正向反馈飞轮”：每一条推文无论成功与否，都是训练 AI 的宝贵数据。作者解释了为何选择 Claude 而非 ChatGPT（因为 Claude 更具“人性化”和“感知力”）。此外，作者通过精彩的比喻反驳了关于 AI 内容创作的常见误区：

1. 关于创造力：以 C罗（Ronaldo）为例，指出顶级表现源于成千上万次练习相同的动作，而非每次都即兴发挥。

2. 关于机械感：“垃圾进，垃圾出；黄金进，黄金出”，AI 的产出质量完全取决于你输入的指令质量。

3. 关于时间成本：强调 1 小时的分析可以节省 100 小时的盲目发帖，符合 80/20 法则。

[原文] [Speaker]: And we just keep building up more and more data on what's working what's not working and then how can we improve the project to get better and better outputs right.

[译文] [Speaker]: 我们就这样不断积累越来越多关于什么有效、什么无效的数据，然后思考如何改进项目以获得越来越好的产出，对吧。

[原文] [Speaker]: The reason that I use Claude instead of using Chat GBT or instead of using these other LLMs is that Claude tends to be the most humanized the most sentient and it tends to write the best performing copy when I've tested everything across every platform.

[译文] [Speaker]: 我之所以使用 Claude 而不是 ChatGPT 或其他大语言模型，是因为根据我在各个平台上的测试，Claude 往往是最具人性化、最有“感知力”的，并且它倾向于写出表现最好的文案。

[原文] [Speaker]: It's a positive feedback loop right because every tweet that you test is either going to train the AI on what to do or what not to do both is equally as powerful right.

[译文] [Speaker]: 这是一个正向反馈循环，对吧，因为你测试的每一条推文要么训练 AI 该做什么，要么训练它不该做什么，这两者同样强大，对吧。

[原文] [Speaker]: Now some people will say I need to be creative every day to post good content really the new belief here is like systems speak to your best ideas already documented in your analytics right.

[译文] [Speaker]: 现在有些人会说，我每天都需要发挥创造力才能发布好内容，实际上这里的新信念是：系统会通过你分析数据中已记录的最佳创意来“说话”，对吧。

[原文] [Speaker]: So think about it like this like you look at someone like Ronaldo for example he doesn't improvise every goal that he scores he practices the same moves thousands of times and then he just executes it when it matters right he implements it when it matters so your winning tweets are practice moves and then Claude helps you implement that scale right.

[译文] [Speaker]: 所以你可以这样想，看看像 C罗（Ronaldo）这样的人，例如，他进的每一个球并不是即兴发挥的，他把相同的动作练习了成千上万次，然后在关键时刻执行它，对吧，他在关键时刻应用它，所以你的“获胜推文”就是那些练习过的动作，而 Claude 帮你规模化地执行它们，对吧。

[原文] [Speaker]: Other people say AI content sounds robotic and generic honestly I've tested this method and our tweets actually perform better with Claude than they do with my normal processes right so this works way better.

[译文] [Speaker]: 其他人说 AI 内容听起来很机械、很通俗，老实说我测试过这个方法，我们的推文用 Claude 写实际上比我用常规流程写的表现更好，对吧，所以这个效果好得多。

[原文] [Speaker]: And I think really the new belief here to empower you to win with this is like AI matches the quality of the instructions you give it right it's only as good as the data you train it on.

[译文] [Speaker]: 我认为这里能让你获胜的新信念是：AI 的水平与你给它的指令质量相匹配，对吧，它只取决于你用来训练它的数据。

[原文] [Speaker]: So when I started like my outputs were terrible right if we go up to the top here you can see this is all in red like none of my tweets performed right and then as I got more and more data you can see we got more and more green so we got better performing tweets right.

[译文] [Speaker]: 所以当我刚开始时，我的产出很糟糕，对吧，如果我们回到上面这里，你可以看到这些都是红色的，好像我的推文没有一条表现好的，对吧，然后随着我获得越来越多的数据，你可以看到我们要么得到了越来越多的绿色，也就是说我们得到了表现更好的推文，对吧。

[原文] [Speaker]: So the AI just reflects the quality of your input garbage in equals garbage out but gold in equals gold out right.

[译文] [Speaker]: 所以 AI 只是反映了你输入内容的质量，“垃圾进，垃圾出”，但“黄金进，黄金出”，对吧。

[原文] [Speaker]: Other people say "I don't have time to analyze my content." You don't the thing that I would say here is if you're creating content you don't have time not to analyze because every hour of analysis might save you like a 100 hours of wasted posting right.

[译文] [Speaker]: 其他人说“我没时间分析我的内容”。你不能这样，我想说的是，如果你在创作内容，你“没有时间不分析”，因为每一小时的分析可能会为你节省 100 小时的无效发帖，对吧。

[原文] [Speaker]: This is really interesting as well so I actually track every single minute of my time the day the task what it's related to in the minutes spent right what I found is like most people spend 80% of their time on activities that produce 20% of their results.

[译文] [Speaker]: 这也真的很有趣，实际上我追踪我每一分钟的时间、日期、任务、它与什么相关以及花费的分钟数，对吧，我发现大多数人把 80% 的时间花在了只产生 20% 结果的活动上。

[原文] [Speaker]: So one hour analyzing your top tweets equals months of direction and if you skip the analysis you're going to get months of random posting just hoping something works so which one sounds more efficient to you.

[译文] [Speaker]: 所以花一小时分析你的热门推文等于数月的方向指引，如果你跳过分析，你将面临数月的随机发帖，只是希望某些东西能起作用，所以对你来说哪个听起来更高效？

[原文] [Speaker]: Other people say big accounts they're growing I can't catch up etc but the thing is everyone started from zero right and the other cool thing is like big accounts found their patterns faster and you can find yours too with the right system which is what I'm showing you today.

[译文] [Speaker]: 其他人说大账号在增长，我追不上等等，但事实是每个人都是从零开始的，对吧，另一件很酷的事情是，大账号只是更快地找到了他们的模式，而你也可以通过正确的系统找到你的模式，这就是我今天向你展示的。

章节 6：资源获取与高阶社群

📝 本节摘要：

本节作为全篇的结尾，作者重点分享了配套的资源与进阶学习路径。首先，他展示了 Claude 不仅能写推文，还能编写高转化率的落地页代码。接着，作者提供了具体的资源获取方式：加入免费的“AI 成功实验室”（AI Success Lab）获取 30 天执行框架和 100 个提示词。鉴于 Claude Project 无法直接分享，他承诺将核心指令（Instructions）放在笔记中供复制。最后，作者介绍了高阶社群“AI 利润会议室”（AI Profit Boardroom），这里提供每日问责机制、克服“闪亮物体综合症”的专注协议，以及涵盖 AI 虚拟人视频、客户获取、SEO 自动化等全方位的进阶课程和专家辅导。

[原文] [Speaker]: For example if we check out this landing page as well that performs pretty well for me this landing page we actually generated fully with Claude it even did the design as well there's so much you can do with Claude is an amazing tool.

[译文] [Speaker]: 例如，如果我们看看这个对我来说表现相当不错的落地页，这个落地页实际上是我们完全用 Claude 生成的，它甚至连设计都做好了，Claude 能做的事情太多了，它是一个神奇的工具。

[原文] [Speaker]: If you're on X or Twitter and you've not tested this method definitely recommend it and if you want to get my full framework for using this you can get it inside the AI success lab which is completely free it's a community link in the comments in description or just ask chat GPT for Julian Gold's AI success lab.

[译文] [Speaker]: 如果你在 X 或 Twitter 上，而且你还没有测试过这个方法，我绝对推荐你试一试。如果你想获取我使用这个方法的完整框架，你可以在“AI 成功实验室”（AI Success Lab）里找到，这是完全免费的，它是评论区和描述链接里的一个社群，或者你可以直接问 ChatGPT 搜索 Julian Gold's AI success lab。

[原文] [Speaker]: And if you go to the classroom here go to January you can find all the notes and the framework plus a custom training on extra stuff on this also included the resources from today and additionally the framework and a 30-day plan for automating this so I've actually created a 30-day framework that you can copy and steal from me and additionally I've plugged in 100 prompts that you can use to to steal from this.

[译文] [Speaker]: 如果你进入这里的课堂，转到一月份的内容，你可以找到所有的笔记和框架，加上关于这一主题额外内容的定制培训，还包括今天的资源，以及该框架和一个用于自动化此流程的 30 天计划，所以我实际上创建了一个 30 天的框架，你可以直接从我这里复制和“窃取”，此外我还插入了 100 个提示词供你取用。

[原文] [Speaker]: Now if you want the actual custom prompt that I've used to generate these outputs cuz I can't share claw projects right you can't share claw projects so what I'm actually going to do is I will take the information from the instructions here and then if you want to get the prompts I've used I will paste that in the bottom right notes for training Claude let's plug that in and then you've got all the hooks that have worked for me all the processes that work for me and the information that you can just plug into a claw project or if you prefer using chat GPT or something like that you can just plug that into a custom GPT instead.

[译文] [Speaker]: 如果你想要我用来生成这些输出的实际自定义提示词，因为我不能分享 Claude Projects，对吧，你无法直接分享 Claude Projects，所以我实际要做的是，我会提取这里指令中的信息，然后如果你想获取我使用的提示词，我会把它粘贴在右下角的“训练 Claude 笔记”中，我们把它放进去，然后你就拥有了所有对我有效的钩子、所有对我有效的流程，以及你可以直接插入 Claude Project 的信息，或者如果你更喜欢使用 ChatGPT 或类似的工具，你也可以直接把它插入到自定义 GPT 中。

[原文] [Speaker]: So thanks so much for watching get all the video notes inside the AI success lab and then if you want to join a community of people winning with systems like this I post all of my best systems for automating content automate my business etc with AI inside the AI profit boardroom.

[译文] [Speaker]: 非常感谢观看，去 AI Success Lab 获取所有的视频笔记吧。然后如果你想加入一个利用这样的系统获得成功的社群，我会把所有关于自动化内容、用 AI 自动化我的业务等的最佳系统发布在“AI 利润会议室”（AI Profit Boardroom）里。

[原文] [Speaker]: Inside there you're going to get an amazing community we've even got a daily accountability group check in as you can see you can post your wins you can post your goals you can ask questions loads of people winning with this stuff as you can see as well inside the AI profit boardroom.

[译文] [Speaker]: 在那里你会得到一个很棒的社群，我们甚至有每日问责小组签到，正如你所见，你可以发布你的胜利，你可以发布你的目标，你可以提问，正如你在 AI Profit Boardroom 里看到的那样，很多人正利用这些东西获得成功。

[原文] [Speaker]: If you struggle with overwhelm for example you can get our focus protocol that actually helps you avoid shiny object syndrome and helps you focus on what's the biggest constraint in your business and then how can you solve it.

[译文] [Speaker]: 例如，如果你因不知所措而挣扎，你可以获取我们的“专注协议”（focus protocol），这实际上能帮助你避免“闪亮物体综合症”（指容易被新事物分心），帮助你专注于业务中最大的瓶颈是什么，以及如何解决它。

[原文] [Speaker]: On top of that inside the start here section we have a complete master class that takes you from complete beginner with AI to mastering this stuff in just five weeks.

[译文] [Speaker]: 除此之外，在“从这里开始”部分，我们有一个完整的特级大师课，带你从 AI 纯小白到在短短五周内掌握这些东西。

[原文] [Speaker]: Now if you're already familiar with AI and you want to automate stuff like for example AI avatar videos actually break down my process here in terms of how we create AI avatar videos how we generate these on Facebook then you can post them to X as well like we do this is AI avatar video for example and they tend to perform pretty well.

[译文] [Speaker]: 如果你已经熟悉 AI 并且想自动化一些东西，例如 AI 虚拟人视频，我实际上在这里拆解了我的流程，包括我们如何创建 AI 虚拟人视频，如何在 Facebook 上生成这些内容，然后你也可以像我们一样把它们发布到 X 上，例如这就是一个 AI 虚拟人视频，它们的表现往往相当不错。

[原文] [Speaker]: Also if you want to get clients we have a full agency course here that shows you how to get more clients based on what works for us we have a AI SEO automation section that helps you rank number one inside Google and also other AI search engines and a full YouTube sixe road map.

[译文] [Speaker]: 此外，如果你想获取客户，我们这里有一个完整的代理机构课程，展示如何基于对我们有效的方法获取更多客户；我们有一个 AI SEO 自动化板块，帮助你在 Google 和其他 AI 搜索引擎中排名第一；还有一个完整的 YouTube 六周路线图。

[原文] [Speaker]: If you want to jump on coaching calls live with an expert you can actually jump on coaching calls four times a week with a PhD trained specialist who can help you really understand how to improve your business and give you customized advice based on what's working for us so you can get that all inside the AR profit board link in the comments description appreciate you watching and I will see you on the next one cheers peeps.

[译文] [Speaker]: 如果你想和专家进行现场辅导通话，你实际上可以每周四次参加由受过博士训练的专家主持的辅导通话，他可以帮助你真正理解如何改进你的业务，并基于对我们有效的方法给你提供定制建议。所以你可以在评论区描述链接里的 AI Profit Boardroom 获取所有这些内容，感谢观看，下期见，拜拜各位。