← 返回任务列表

5 Essential AI Skills for Instructional Designers

298 段 · 1 位说话人 · 原片 23:53
M1
M10:00

大家好,我的朋友们。

Hello, my friends.

我今天想告诉你,artificial intelligence 并不是 instructional design 的未来,因为它已经来了。

I'm here to tell you that artificial intelligence is not the future of instructional design, and that's because it's already here.

所以,如果你不去培养这些 AI 技能,你可能就会被落在后面。

So if you don't build these AI skills, you might be getting left behind.

我不希望这种事发生。

I don't want that to happen.

所以呢,我们会来看一看你需要掌握的五个关键 AI 技能,帮助你做出更好的项目、工作得更快,也让你的职业发展更能应对未来。

So yeah, we're going to take a look at the 5 essential AI skills you need to create better projects, work faster, and future-proof your career.

那我们会先深入讲这些技能。

So we'll dive into the skills first.

我也准备了很多其他资源、工具和例子要给你看。

I also have a lot of other resources, tools, examples to show you.

所以,就算你已经熟悉其中一些技能,这个视频对你来说也应该会很有价值。

So this should be a valuable video for you, even if you're already familiar with some of these skills.

第一个技能是写出有效的 prompts,也就是大家常说的 prompt engineering。

So the first one is writing effective prompts, also commonly referred to as prompt engineering.

第二个技能是熟练掌握各种工具。

The second one is mastering the tools.

第三个技能是教别人了解 artificial intelligence,或者更广泛地说,能够清楚地沟通 artificial intelligence 相关内容。

The third one is teaching people about artificial intelligence or communicating about artificial intelligence more generally.

第四个技能是读写 JavaScript。

The fourth one is reading and writing JavaScript.

第五个技能是创建由 AI 驱动的 automations。

And the fifth skill is creating AI-powered automations.

所以,如果你能扎实掌握所有这些技能,我觉得在接下来相当长一段时间里,你都会是 instructional design 团队里很抢手的人。

So if you get a solid handle on all of these skills, you are going to be in demand on instructional design teams for a while, I would say.

那我们先来看第一个,prompt engineering。

So let's take a look at the first one, prompt engineering.

本质上,如果你的 prompts 写得好,就能得到很好的结果。

So essentially, if you write good prompts, you'll get great results.

如果 prompts 写得不好,你就会浪费很多时间。

And if you write bad prompts, you're going to waste a lot of time.

如果你对这些内容完全是新手,prompting 基本上就是你写下并提交给正在使用的 AI 工具的那段文字。

If you are brand new to this stuff, prompting is essentially just writing the text that you submit to the AI tool that you're working with.

我们马上会看一些例子,不过你提出需求的方式,会对你得到的结果产生很大影响。

So we're about to take a look at some examples, but the way that you ask for what you want will have a big impact on the results that you get.

我也要提一下,prompting 的最佳做法取决于你使用的工具。

I'll also mention that prompting best practices depend on the tool you use.

比如你用 ChatGPT 这样的工具时,prompts 会和你用 Midjourney 之类的工具生成图片时很不一样。

So if you use a tool like ChatGPT, the prompts are going to be quite different than if you were generating images with a tool like Midjourney or something like that.

所以我等会儿会分享更多例子,但总体来说,你要知道,什么样的 prompt 算好,确实取决于你用的是什么工具。

So I'll share some more examples in a bit, but generally know that, yeah, it depends on the tool you're using, what makes a good prompt.

那么,在我们进入例子之前,先来看一下我推荐的一般 prompt 格式。

So yeah, before we dive into the examples, let's take a look at the general prompting format I would recommend.

这是针对 large language models 的。

So this is for large language models.

我也想说,如果你对这里的一些术语不熟,或者不知道 LLMs 是什么,那就去看一下我们的 Intro to AI 视频。

I'll also say if you are unfamiliar with some of these terms or you don't know what LLMs are, go ahead and watch our Intro to AI video.

我是在几周前发布的。

So I posted it a couple of weeks ago.

如果你打算更深入地学习 AI,我觉得那个视频是必看的。

That one's a must-watch if you're going to start diving deeper into AI, I'd say.

所以,下面就是我在使用 ChatGPT、Claude,或者其他这些 LLMs 时,有时会用的一种 prompt 类型。

So here's the type of prompt I use sometimes when working with ChatGPT or Claude or any of these other LLMs.

首先,我会告诉 AI 它的角色是什么。

First, I'll tell the AI what its role is.

然后我会告诉它任务是什么,接着我会提供一些相关信息或上下文,我也会给你展示我是怎么做的。

Then I'll tell it what its task is, and then I'll share some relevant information or context, and I'll show you how I do this.

然后在这个基础上,你甚至还可以提供一些示例。

And then from there, you can even provide examples.

所以你可以给一个 one-shot 示例,也可以给一个 multi-shot 示例。

So you could do a one-shot example, you could do a multi-shot example.

同样,这个我们在 Intro to AI 视频里讲过,不过我们现在先来看一下,当我们只是写一个很笼统的请求或者任务时,得到的输出会有什么不同。

Again, we talk about that in the Intro to AI video, but let's just see the difference in output we get when we just write a general request or a task.

为 instructional designers 写一个关于 artificial intelligence 的五分钟 eLearning 课程大纲。

Write an outline for a 5-minute eLearning course about artificial intelligence for instructional designers.

好,这是它给出的这个大纲。

So here's the outline.

它给了我们一个介绍部分。

It gives us an introduction.

什么是 AI?

What is AI?

它还给了一些常见应用,比如,哦,interactive experiences,personalized learning。

It gives us some general applications like, oh, interactive experiences, personalized learning.

然后它讲到一些 ethical considerations,最后还有一个小回顾。

It goes into some ethical considerations and then a little recap.

所以也不算差。

So not terrible.

同样,如果你想看它推荐了什么,可以暂停一下;不过这就是我们用的 prompt,而且这是给 ChatGPT-4o 用的。

Again, you could pause it if you want to see what it's recommending, but this is the prompt we used and this is for ChatGPT-4o.

现在,如果我们把这个 prompt 稍微扩展一下,给它一个角色、任务和背景,结果看起来就会有点不一样。

Now, if we build that prompt out a little bit and we give it a role, task, and context, things look a little bit different.

所以注意看这个角色部分,我写的是:你是一位 world-class instructional designer,同时也是 artificial intelligence 方面的专家,因为我们可以把角色设成任何我们想要的样子,对吧?

So notice for the role, I said, you are a world-class instructional designer who is also an expert in artificial intelligence, because we can make the role whatever we want, right?

也就是说,AI 足够聪明,会尽量去配合我们告诉它的角色和工作内容。

Like the AI is smart enough to, yeah, try to accommodate whatever we tell it that its role and job is.

然后在任务部分,我就给了它刚才这边用过的同一个任务,接着又给了它一些背景信息。

So then for the task, I just gave it that same task I used over here, and then I gave it some context.

我说,这些 instructional designers 可能以前听说过 artificial intelligence,但他们不知道该怎么用它。

And I said, the instructional designers may have heard of artificial intelligence before, but they don't know how to use it.

来帮他们在工作中节省时间。

To save them time on the job.

实际应用是最重要的。

Practical application is essential.

所以,对,我们当然希望这里的内容保持实用。

So yeah, of course we want to keep things practical over here.

我们不是为了培养技能而培养技能。

We're not building skills just to build them.

我们是要用这些技能看到真实的成果。

We're using them to see actual results.

所以我给了 AI 这些背景信息。

So I gave the AI that context.

我大概多花了,嗯,三十到六十秒来写这个 prompt,但在我看来,输出结果好很多。

It took me about, what, an extra 30 to 60 seconds to write this prompt, but the output is way better in my opinion.

所以它的组织结构更清楚。

So it's organized more nicely.

首先,它告诉我们学习目标会是什么,而且我觉得这些目标相当不错。

First of all, it tells us what our learning objectives are going to be, and I feel like these are pretty good.

识别三个关键的 AI 工具或应用,应用至少一个 AI-powered workflow 来节省时间,并认识到 AI 如何增强创造力。

Identify 3 key AI tools or applications, apply at least 1 AI-powered workflow to save time, and recognize how AI can enhance creativity.

所以说实话,这些目标都挺扎实的。

So those are some pretty solid objectives, to be honest.

这就是那个 world-class instructional designer 在发挥作用。

That's that world-class instructional designer doing their work.

因为在这边,我们有这些吗?

Because over here, do we even have that?

提供对 AI 应用的基础性理解。

Provide a foundational understanding of AI applications.

这就比我们这边这些学习目标弱多了。

That's way weaker than these learning objectives we have over here.

然后等我们进入大纲部分的时候,看起来也好很多。

And then when we get into the outline, it seems a lot better as well.

所以它给了我们一种抓住注意力的方法。

So it gives us a way to capture attention.

它也给了我们一个互动元素。

It gives us an interactive element.

它建议做一个 poll。

It suggests a poll.

这个我们在这边肯定没有得到。

We definitely didn't get that over here.

这边没有任何 poll 的建议。

There was no poll suggestion.

然后,是的,我们接着进入一些关键应用,还有非常具体的内容创作工具。

And then, yeah, we get into some key applications, very specific content creation tools.

这里还有一个你可以用的例子,media generation tools。

And here's an example you can use, media generation tools.

它提到了其中一些工具,并给了例子。

It mentions some of them, gives examples.

所以这真的很酷。

So that's really cool.

然后它还给了我们一个完整的实用 workflow activity,用来展示 instructional designer Jane 是怎么把 AI 整合进她的项目里的。

And then it gives us a whole practical workflow activity to show how Jane, the instructional designer, integrates AI into her project.

所以再说一遍,我们这里比较的是这个 prompt 和我们一开始用的那个初始任务 prompt 之间的差异;这个稍微更长一点,但我会说,输出结果差别相当大。

So again, what we're comparing here is the difference in this prompt, a little bit bigger than this initial task one that we used, but the output differs pretty drastically, I would say.

然后,如果你对这个格式感到好奇,它其实就像 Markdown 一样。

So, and then if you're wondering about this format, it's just like Markdown.

所以这只是用来表示一个 H2 HTML element。

So this is just for like an H2 HTML element.

这有点让人困惑。

That's confusing.

基本上,我就是用两个井号,然后把我想给那一段用的标题写在后面。

I basically, I just use two number signs and put whatever the heading I want is for that paragraph.

所以你可以稍微玩一玩这个。

So play around with this a little bit.

它可以写成任何你想要的内容。

It can say whatever you want.

不一定非得是 role、task、context。

It doesn't have to be role, task, context.

有时候我会写 guidelines 或者 instructions。

Sometimes I do guidelines or instructions.

等一下我也会给你看一个这样的例子。

I'll show you an example of that in a little bit as well.

不过我们先来讲一些 prompting tips。

But let's get into some prompting tips.

所以在你写完 prompt 之后,其实可以问 AI,在开始之前它有没有任何问题,或者需不需要进一步澄清。

So after writing your prompt, you can actually ask the AI if it has any questions or needs any clarification before it gets started.

Sabrina 跟我讲过她是怎么做这个的。

And Sabrina talked to me about how she does this.

是的,这是一种让你的 prompt 更扎实的方法。

And yeah, it's a way for you to make your prompts more robust.

也就是说,你先尽量在 prompt 里给它所有信息,然后你说,嘿,你有什么问题吗?

So you try to give it all the information in your prompt and then you say, hey, do you have any questions?

在继续之前,你还需要我提供什么吗?

Do you need anything from me before continuing?

然后它就会告诉你,它的知识里还有哪些缺口,如果想给出更好的回答,它希望补充哪些信息。

And then it will let you know whatever its gaps in its knowledge are that it would like to have to give a better response.

然后你其实可以把这些内容加回到你的初始 prompt 里,可以说是把这个 prompt 维护起来。

And then you can actually add that to your initial prompt and maintain that prompt, so to speak.

这个我们后面还会再深入讲。

Again, we'll dive deeper into that.

另外,我也建议你去读一读你所用工具的 prompting guide。

And then also I would just suggest reading prompting guides for your tools.

比如你在用 ChatGPT,那你就可以去搜 ChatGPT prompting guide。

So if you're using ChatGPT, you could look up ChatGPT prompting guide.

我觉得我们这里放的就是这个。

I think that's what we have right here.

那如果你在用 Midjourney,也对 Midjourney 做同样的事。

So if you're using Midjourney, do the same thing for Midjourney.

总之要去做这些研究,因为,你要确保自己遵循的是 best practices,并且尽可能得到最好的输出结果,因为这会决定它到底能不能用到实际生产里。

But do that research because, yeah, you want to make sure you're following best practices and getting the best possible output you can, because that's going to make the difference between whether you can use it in production or not.

我是说,我不知道你有没有见过,但我确实看到过一些社交媒体帖子,类似于,哦,看这个大公司在用的 AI artwork 多糟糕,而且一看就很明显是 AI 做的。真正擅长 prompt engineering 的人,会让别人很难分辨它到底是不是 AI 生成的。

I mean, I don't know about you, but I've actually seen some social media posts where it's like, oh, look at this terrible AI artwork that some big corporation is using, and it's really obviously AI. People who are really good with prompt engineering make it really hard to tell whether or not it's generated from AI.

所以你要掌握这种能力。

So you want that power.

这里有一个例子,就是我刚才说的,把 prompt 进一步搭建起来。

And here is an example of what I was saying to build out the prompt.

这是我在前一个例子里用的同一个 prompt,但我问它,在你开始之前,你有什么问题要问我吗,或者有没有需要我澄清的地方?

It was the same one I used in the previous example, but I asked it, do you have any questions for me or do you need any clarification before you begin?

然后它就告诉我,有,我想知道目标受众具体是谁。

And then it tells me, yeah, I'd like to know the target audience specificity.

还有,我应该使用什么语气和风格?

Yeah, what tone and style should I use?

你希望互动性达到什么程度?

What level of interactivity are you looking for?

我们是否要推荐特定的软件?

Are we suggesting specific software?

所以这些都是我可以加到最初那个 prompt 里的内容,尤其是如果这是一个我想反复复用的 prompt。

So that's all stuff I could add to that initial prompt, especially if this is a prompt that I'd be wanting to reuse over and over.

如果它包含所有这些信息,那会很方便。

It would be nice for it to have all of this information.

不断迭代你的 prompts,也是一个很好的 prompt engineering 做法。

Iterating on your prompts is also a good prompt engineering practice.

你不希望和 AI 聊了一整段,把这么多有价值的信息都告诉它了,结果等到要做一个新项目或者别的事情时,又从零开始开一个新 chat。

You don't want to have a whole chat with an AI and you give it all this valuable information, and then when it's time to do a new project or something, you just start a new chat from scratch.

如果你能管理并迭代你的 prompts,从长远来看会给你省下很多时间。

If you can manage and iterate on your prompts, it's going to save you a lot of time in the long run.

所以也有一些工具可以帮你做这件事。

So there are also tools to help out with this.

我这里就提几个。

So I'll just mention a couple.

这个叫 PromptHub,level up your prompt management。

This one is called PromptHub, level up your prompt management.

如果你是在一个团队里工作,而且大家都想一起贡献 prompts,那它特别好用。

And it is especially good if you're working with a team that all wants to contribute to the prompts.

你可以在不同的 AI models 上测试你的 prompts,看看它们效果怎么样。

So you get to try out your prompts on different AI models and see how they work.

基本上,你可以把它们跑几百次甚至几千次,然后看看哪些 models 表现更好。

You get to basically run them hundreds or thousands of times and see which models perform better.

当然,你也可以看出哪个 prompt 表现更好。

And then of course you could see which prompt performs better too.

所以还挺酷的。

So pretty cool.

可以去看看这个。

Check this one out.

Promptmetheus 是另一个工具。

Promptmetheus is another one.

他们把它称作 prompt engineering development environment。

They're calling it a prompt engineering development environment.

所以对,你可以比较自己不同的 prompts,测试它们,优化它们,还可以协作。

So yeah, you get to compare your different prompts, test them, optimize them, collaborate.

所以还挺酷的。

So pretty cool.

现在有很多这样的工具不断冒出来。

A bunch of tools like this are popping up.

它们有点贵,但如果你是为了你的团队在做这类事情,那绝对是一个很值得拥有的工具。

They're a little bit expensive, but if you're doing something like this for your team, definitely a good tool to have.

好。

Okay.

我想我本来想给你看——哦对,我想给你看一个我一直在反复迭代的 prompt。

I think I wanted to show you— oh yeah, I wanted to show you a prompt I've been iterating upon.

这就是我们做的 Become an Instructional Designer custom GPT。

So this is the Become an Instructional Designer custom GPT that we built.

我上周做了一个关于它的视频。

I created a video about this last week.

它本质上是用来帮助人们成为 instructional designer 的。

It's essentially to help people become an instructional designer.

它是基于我们的流程训练的。

It's trained on our process.

这里就是我是怎么搭建它的。

And here's how I built this.

我最近收到了很多关于这个的问题。

I've been getting a lot of questions about this.

这个 prompt 是我自己一直在迭代、并且随着我们收到这个 custom AI bot 的反馈,不断尝试改得更好的。

And this is a prompt that I'm personally iterating on and trying to make better as we get feedback on this custom AI bot.

所以我会告诉它,你是一位经验丰富的 instructional design coach,专门帮助人们成为 instructional designer,并且你代表 Devlin Peck 这个品牌。

So I tell it, you're an experienced instructional design coach who specializes in helping people become instructional designers, and you represent the Devlin Peck brand.

然后我们给它这些指令。

And then we give it these instructions.

提问来评估用户目前处在什么阶段,评估他们现在的知识和技能。

Ask questions to assess where the user's at, assess their current knowledge and skill.

以后我会更深入地讲这个,但你可以看到,我们也会告诉它我们的方法。

And I'll dive into this deeper at a future point, but you can see we also tell it our approach.

所以这就是我前面说的意思。

So this is what I meant earlier.

它不一定只能是 context,比如我们在某个地方提到了我的方法,那现在这个 AI bot 就知道该去哪里进一步了解我们的方法。

It doesn't have to just be context, like if we mention my approach somewhere, well, now the AI bot knows where to learn more about our approach.

还有我们的 guidelines 也是一样。

And then our guidelines too.

当我们发现这个 bot 有哪些地方可以做得更好时,我就会继续往这里加内容。

This is what I'm adding to as we're finding stuff that the bot could be doing better.

所以对,我会做一个关于这个的视频。

So yeah, I'll do a video about this.

我觉得创建一个像这样的 custom GPT,也是一个很棒的 portfolio project。

I think that's a great portfolio project too, creating a custom GPT like this.

所以如果你对这种类型的东西感兴趣,告诉我。

So yeah, let me know if you're interested in that type of thing.

不过我们先继续讲下一个技能。

But let's move on to the next skill here.

也就是总体上掌握这些工具。

So mastering the tools in general.

如果你擅长 prompting,那你就能把这项技能用在多个工具上,但你真正需要建立的是你自己的个人 AI toolkit。

So if you're good at prompting, you're going to be able to use that skill across multiple tools, but you really want to build out your own personal AI toolkit.

你需要知道自己有哪些工具可以用,以及你实际上如何使用它们来节省时间,并在工作中产生影响。

You want to know which tools you have available to you and how you're actually using them to save time and make an impact on the job.

所以首先,要了解哪些工具在哪些任务上表现最好,对吧?

So first of all, learn which tools excel at which tasks, right?

希望你现在已经知道 ChatGPT 了,但可能还有——我是说,我会告诉你我最常用的前三个。

Hopefully you know about ChatGPT by now, but there's probably— I mean, I'll tell you my top three.

其实我下一页幻灯片就会讲这个,不过你可能还是应该先了解一下,现在 AI 到底能做哪些事情。

I actually have a slide for that next, but you should probably know everything that AI is capable of right now.

所以,如果你还不清楚现在都有哪些工具,就去看一下那个 Intro to AI 视频。

So watch that Intro to AI video if you don't know about all the tools out there.

另外,也要弄清楚你所在的组织到底能使用哪些工具,因为有些公司正在搭建内部 AI 工具,只有员工可以用;也有些公司不允许你使用未经批准的工具。

Also discover which tools your organization actually has access to, because some orgs are building out internal AI tools that only their employees can use, and some don't let you use unapproved tools.

所以这一点一定要特别注意。

So definitely be sensitive to that.

不要把敏感的客户数据,或者公司的机密信息,放进未经批准的工具里,比如 ChatGPT,或者其他工具。

Don't be putting sensitive customer data or company confidential information into unapproved tools like ChatGPT or any others.

所以,是的,如果你要使用公司数据,一定要知道你们公司对 AI 的立场是什么。

So yeah, make sure if you're using company data, you know what your company stance is on AI.

其实我对这一点也挺好奇的。

And I'm actually quite curious about that too.

所以如果你方便的话,可以留个评论,告诉我你们公司现在在这方面是什么情况。

So if you can drop a comment, let me know where your company is at with it.

他们是在搭建内部工具吗?

Are they building internal tools?

他们是在批准某些外部工具,还是只是说,不要用这个?

Are they approving certain external tools or are they just saying, don't use this?

因为我听到的答案真的是各种各样,所以能做一个小小的非正式调查也挺好的。

Because I've been hearing answers across the board, but it'd be nice to do a little informal poll.

然后总体来说,我会说,就是要多练习使用这些工具。

And then overall, I'd say just get the practice in with these tools.

这是唯一能让你更熟练使用它们,并弄清楚怎么产出更高质量内容的方法。

That's the only way to get better with them and figure out how to create higher quality output.

但这真的很值得,尤其是如果你在做客户项目的话。

But it really does pay off, especially if you're doing client work.

我们有些客户对这些 AI 生成的素材反馈非常好,因为项目在这么短的时间内,看起来定制化程度高了很多。

We've had clients give really good feedback to some of these AI-generated assets because the project feels so much more custom on such a quicker timeline.

所以,是的,作为一个人,你也能做出比原本高很多的制作水准,而不一定要依赖一个大团队。

So you can get, yeah, as one person, you're able to get a lot higher production value than you would otherwise if you were relying on a big team.

但这确实要求你掌握一些相关的方法和知识。

But it does require you to have some know-how.

你不会被 AI 取代,但你需要知道怎么使用 AI。

You're not going to get replaced by AI, but you need to know how to use AI.

那么这里也说一下我最推荐的三个工具。

So here are my top 3 tools also.

首先是 ChatGPT。

First of all, ChatGPT.

我觉得原因在于 custom GPTs,你可以为自己、你的团队,或者你的受众创建这些 custom GPTs。

I think the reason for this is the custom GPTs, being able to create them for yourself or your team or your audience.

是的,而且它现在看起来也是在 UX 上投入最多的一个。

Yeah, and it just seems to be getting the most UX investment right now.

它就像一个通用型 LLM,基本上什么都挺擅长。

It's like a general LLM that's good at pretty much everything.

所以它绝对应该在你的工具箱里。而且每个月花十九美元来创建 custom GPTs 之类的功能,对你来说可能是值得的,或者可以让你有更多使用额度。

So that should definitely be in your toolkit and paying the $19 a month to create the custom GPTs and stuff might be worth it for you or to get the more usage out of it.

Freepik Pikaso 也是我现在最喜欢的前三个工具之一,因为它里面集成的东西非常多,而且真的很容易上手。

Freepik Pikaso, I'm really liking as one of my top 3 tools because it does have so much built into it and it is really easy to use.

所以,是的,可以去看看他们的 AI image generator。

So yeah, check out their AI image generator.

再说一次,它比去 Discord 上用 Midjourney 简单太多了。

It's again, way easier than going on Discord and using Midjourney.

所以可以去试试看。

So check it out.

他们的 AI video generator 现在还不算真正的专业级质量,但我相信他们肯定在继续改进。

Their AI video generator, not really professional quality right now, but I'm sure they're working on it.

而且他们这里还有各种各样的东西,比如 icon generators、characters、music,我听了其中几个。

And they have all sorts of stuff here, icon generators, characters, the music, I listened to a few of those.

这些声音其实真的挺不错的。

The voices are actually really good.

所以我喜欢这类工具,因为它们基本都挺好用,而且都集中在一个地方,这样你就不需要订五个不同的服务了。

So I like tools like this because they're all pretty good and it's all in one place, so you don't need to have 5 different subscriptions.

所以 Freepik Pikaso 肯定算一个。

So Freepik Pikaso for sure.

然后是 NotebookLM,它在自己擅长的事情上真的太出色了。

And then NotebookLM, it's just so excellent at what it does.

同样,我在介绍视频里讲过它,所以这里我不想讲得太深。

Again, I talked about that one in the intro video, so I don't want to go too deep into it.

不过对,你可以上传 PDF,对吧?

But yeah, you get to upload PDFs, right?

也可以是任何你需要处理的数据。

It can be whatever data you need to work with.

它可以给你很不错的总结,然后你还可以向它提问。

It can give you nice summaries and then you can ask it questions.

它会几乎立刻给你回答,并且把来源高亮标出来。

And it'll give you near-instant responses with the source highlighted.

所以对,如果你在处理技术文档,或者总体上就是大量内容,而且你需要能够引用这些内容,那它真的很有价值。有点像是在搭建你自己的个人 SME,或者说 subject matter expert。

So yeah, that's really valuable if you're working with technical documents or just a lot of content in general and you need to be able to reference it, kind of like building out your own personal SME, so to speak, or subject matter expert.

所以这是一个很棒的工具。

So that's a great one.

所以这些就是我排前三的选择。

So these would be my top 3.

如果你不知道从哪里开始,就从这里开始。

So if you don't know where to start, start here.

如果你想继续深入,就去看 Intro to AI 那个视频。

If you want to go further, watch the Intro to AI video.

我在那里讲了很多这类工具。

I talk about a lot of them there.

好。

Okay.

关于 AI 的沟通,或者教别人了解 AI。

Communicating about AI or teaching people about AI.

我觉得这是一个必备技能,因为这毕竟还是一项新技术。

I think this is an essential skill just because this is still a new technology.

不是每个人对它的理解都在同一个水平上。

Not everyone's on the same page about it.

而且,如果你要使用这些工具,并且尝试在项目里落地这些东西,你就需要能够和 stakeholders 讨论这件事。

And yeah, you might need, if you're going to be using these tools and trying to implement this stuff in projects, you're going to need to be able to talk to stakeholders about this.

所以在我讲得太超前之前,这件事的第一部分,是先建立你自己的 AI 知识,对吧?

So before I get ahead of myself, the first part of this is building up your AI knowledge, right?

你需要知道 AI 相关的术语。

You need to know the AI terms.

你需要了解它的能力、限制,以及伦理方面的顾虑。

You need to know the capabilities, the constraints, the ethical concerns.

所以对,不能低估这一点。如果你还没看过那个介绍视频,就去看一下;但如果你已经看过了,我会说你已经走在一条挺不错的路上了。

So yeah, can't underestimate, watch that intro if you haven't already, but if you have watched it, you're on a pretty good path, I would say.

如果我现在说的这些术语你都懂,那你已经做得挺好了。

If you know all these terms I'm saying, you're doing pretty good.

不过对,还是要持续跟进 AI 相关知识,因为你需要去教育客户、subject matter experts、经理、学习者,对吧?

But yeah, keep up to date on the AI knowledge because you're going to need to educate clients, subject matter experts, managers, learners, right?

他们可能会问你,你做出来的东西是怎么搭建的,或者是什么技术让它成为可能。

They might have questions about how you built what you built or what technology is making it possible.

然后对,如果你想使用某个特定工具,你可能还需要在组织里争取 business stakeholders 的支持。

And then yeah, you might need to get buy-in from business stakeholders at your organization if you want to use a specific tool.

所以对,能够自在地谈论这些事情,会很有帮助。

So yeah, getting comfortable talking about this stuff is going to help.

哦,我刚才已经把我的 bot 关掉了。

Oh, I already closed my bot.

如果你回到我展示我的 bot 的那一段,其实我当时问它,你能不能扮演一个公司内部客户,我想在一个项目里使用 generative AI,然后你来问我相关问题,我们必须回答这些问题。

If you go back to where I showed off my bot, I actually asked it, can you role-play as an internal corporate client and I want to use generative AI on a project and you ask me questions about it and we have to answer them.

所以你可以用我做的那个 custom GPT 来练习这类内容。如果你还没看上周那个关于如何用 AI 成为 ID 的视频,也可以去看一下。

So you can use that custom GPT I built to practice stuff like this and watch last week's video if you haven't already about using AI to become an ID.

不过是的,你越能自在地去沟通 AI、教别人理解 AI,你就越适合去参与这些对话。

But yeah, the more comfortable you are communicating about AI and teaching people about AI, the better suited you're going to be for those conversations.

然后,在你的 portfolio 里展示你的 AI 知识和技能。

And then demonstrating your AI knowledge and skills on your portfolio.

我觉得这很有用,尤其是如果你在考虑自己的职业发展,向招聘经理和潜在客户展示你正在使用这项技术,真的帮你节省时间,并且做出更好的学习体验。

I think this is good, especially if you're thinking about your career, showing hiring managers and potential clients that you're using this technology to actually save you time and make better learning experiences.

我觉得这会显得非常加分,尤其是在短期内,或者接下来几年里,因为大多数 instructional designer 可能并不会真正深入去学习这些技能。

I think that is going to look really good, especially in the near term or the next few years when most instructional designers probably are not actually going to dive into these skills.

如果你正在看这个视频,你其实已经走在前面了,因为你在研究这些东西,并且保持跟进。

If you're watching this, you're already ahead of the curve because you're researching this stuff and staying up to date on it.

但有很多 ID 可能还在埋头做事,他们不会去 LinkedIn,也不会看 YouTube 视频,或者 Google 要怎么提升自己的技能。

But there are a lot of IDs who are probably chugging away and they're not going on LinkedIn or watching YouTube videos or Googling how they're going to build their skills.

他们可能只是专注于自己的工作,专注于眼前的需求。

They're probably doing their job and focusing on their immediate needs.

所以这些变化可能会在他们没注意的时候突然到来。

So some of this stuff might sneak up on them.

如果在这第一波浪潮里,随着这些技能变得越来越有价值,而且我们看到,好的,这些东西会长期存在,公司也在使用它们。

If in this early wave, as these skills are becoming more valuable and we're seeing, okay, these are here to stay, companies are using them.

那是的,把这些展示在你的 portfolio 里,会带来很大的帮助。

Yeah, showing that off on your portfolio is going to go pretty far.

同时也能帮助整个行业了解你正在做什么。

And just helping educate the industry on what you're doing.

如果你在做一些很酷的项目之类的内容,把它们突出展示在你的 portfolio 里,也会帮助整个行业和你一起提升。

If you're working on cool projects and stuff like that, highlighting that in your portfolio is going to help bring the whole industry up with you.

好。

Okay.

持续跟进 AI 的发展。

Staying up to date with AI developments.

是的,我们最近做了很多 AI 相关内容。

Yeah, we've been doing a lot of AI content lately.

我们和 Luke Hobson 做了一场 AI upskilling 的直播活动。

So we did an AI upskilling live event with Luke Hobson.

所以他在里面分享了一些很值得关注的人。

So he shared some good people to follow in there.
M1
M117:25

最后,是的,加入一些有 AI 爱好者的社群,这样你就可以和其他 instructional designer 聊 AI,了解我们是怎么使用 AI 的,也看看我们在哪些项目里用了 AI。

And then finally, yeah, join communities with AI enthusiasts so that you can talk to other instructional designers about AI and learn how we're using AI and looking at projects that we're using AI in.
M1
M117:47

太好了。

Awesome.

好,那现在我们进入一些 scripting 或 coding 技能。

All right, so now we're getting into some scripting or coding skills.

这是第四项技能,就是 JavaScript,学会写一点代码。

So this is the fourth skill, just JavaScript, learn to write some code.

当你擅长 JavaScript 的时候,你就可以把 AI 整合进学习体验里。

So when you're good with JavaScript, you can integrate AI into learning experiences.

比如说,我们很多人会用 Articulate Storyline 360 或者 Rise 360 来做开发。

So, like for example, a lot of us do development in Articulate Storyline 360 or Rise 360.

你可以把类似 ChatGPT 的功能加入到那样的学习体验中,不过这确实需要一些 JavaScript 方面的知识。

You can incorporate ChatGPT-like functionality into learning experiences like that, but it does require some JavaScript know-how.
M1
M118:28

但总体来说,一旦你学会了 JavaScript 的基本语法,你就可以学着使用 APIs。

But overall, once you learn the basic syntax of JavaScript, you can learn to work with APIs.

你可以学习怎么解析从这些 APIs 返回的数据。

You can learn how to parse the data that you get back from those APIs.

你可以学习怎么和 servers 配合工作。

You can learn how to work with servers.

如果你刚接触这些东西,这听起来可能有点复杂,确实也有一点复杂,但现在有 AI 帮我们,一切都变得更容易了。

This might sound a little complex if you're new to this stuff, and it is a little complex, but everything is easier now with AI to help us.

所以我会在下一张幻灯片上分享一些资源。

So I'll share some resources on the next slide.

另外,如果你想要更精细的输出,JavaScript 也可以让你对正在使用的模型进行 fine-tune。

But also JavaScript makes it possible to fine-tune the models that you're working with if you want more refined outputs.

就今天早上,我们还在一个 Bootcamp AI workshop 上讨论 fine-tuning,因为有人想做一个项目,就是——对,她想根据用户的需求,生成非常定制化的 wedding cakes。

So just this morning we were on a Bootcamp AI workshop and we were talking about fine-tuning because someone wants to do a project where— Yeah, she wants to be able to generate really custom wedding cakes based on what the user is looking for.

然后她发现,现在市面上的那些模型没办法做到她想要的那种具体程度,因为她本身是专业的 cake decorator。

And she's finding that the models that are out there aren't able to be as specific as she wants to because she's a professional cake decorator.

Creator。

Creator.

所以对,我们就在聊,给她用一批高质量数据来 fine-tune 一个模型,这些数据会由她来描述,然后再用这个模型生成更好、更定制化的蛋糕图片。

So yeah, we're talking about her fine-tuning a model on a bunch of high-quality data that she's going to describe and then use that to generate better, more custom cake images.

所以对,fine-tuning 能做的事情真的很酷。

So yeah, really cool work you can do with fine-tuning.

我觉得如果你们偏技术,或者对这个足够感兴趣,这会是一个非常有价值的技能。

I think that's a really valuable skill if you all are technical or interested enough in doing that.

我个人对它真的很感兴趣。

I'm personally really interested in it.

所以等你们当中有足够多人感兴趣之后,我肯定会深入做一些这方面的内容。

So once enough of you all are interested, I'm going to definitely dive into some content around that.

这里有一些 JavaScript 的资源。

Here are some resources for JavaScript.

我大概是在 2017 年左右学的时候,用的是 freeCodeCamp,后来又发现了 Scrimba,也特别有帮助。

When I learned back in, like, 2017 or so, I used freeCodeCamp and then I later found out about Scrimba, which was really helpful.

这两个资源我都很喜欢。

I love both of those resources.

也有一些非常棒的 YouTube 频道,可以用来学习 JavaScript 技能。

There are some really great YouTube channels to learn JavaScript skills.

AI tools,就像我刚才说的,现在让写代码变得容易多了。

AI tools, like I said, this makes coding so much easier.

到目前这个阶段,只要你把它们当成一个懂代码的同事,跟它们聊你想做什么,把你现有的代码给它们看,它们基本上可以帮你完成大部分工作。

They can do most of the work for you at this point if you just treat them like a coworker who does know code and talk to them about what you're trying to do and show them the code that you have.

这样能走得非常远。

It goes really far.

后面那两个工具,GitHub Copilot 和 Cursor,基本上都可以一边陪你写代码,一边像代码的自动纠错一样帮你。

Those latter two tools, GitHub Copilot and Cursor, those both will write the code along with you and serve as autocorrect for your code, essentially.

所以这两个工具也挺酷的。

So those are pretty cool.

然后我之前提过这个想法,在那个 AI Haunted House live event 上,好像大家还挺感兴趣的。

And then I floated this idea before and it seemed to get decent interest at that AI Haunted House live event.

不过对,如果你们想要的话,我可以做一个 JavaScript playlist。

But yeah, if y'all want, I can make a JavaScript playlist.

我有时候也会想,外面已经有这么多 JavaScript 内容了,我为什么还要做这个呢?不过对,也许做一个专门面向 ID 的 JavaScript playlist 会有帮助,这样我可以把内容放到更合适的语境里,也让它和我们作为 ID 所做的工作更相关。

I question sometimes why I should do that when there's so much JavaScript content out there, but yeah, maybe it would be helpful to do an ID-specific JavaScript playlist just so I can help contextualize things better and make it more relevant to the work that we do as IDs.

所以如果你想要 JavaScript 教程,请留言告诉我。

So please leave a comment if you want JavaScript tutorials.

也许如果这个视频在一个月内拿到,比如说,二百五十个赞,那我接下来就做 JavaScript playlist。

Maybe if this video gets, let's say, 250 likes within a month, that's the next thing I'll do is a JavaScript playlist.

所以如果你们真的这么想要,就给这个视频点赞,我会把它算作投给 JavaScript playlist 的一票。

So if y'all want it that bad, like the video, I'll count that as a vote for JavaScript playlist.

好,接下来进入 automation,也就是第五个技能。

Okay, moving into automation, the fifth skill.

这部分真的很强大。

This is really powerful stuff here.

这也是我个人非常想在自己的业务里深入做的事情,就是搭建 AI automations,让重复性的任务变得更轻松。

This is personally also what I want to really get involved with for my own business, building AI automations to make repetitive tasks easier.

所以,automations 本来就已经很有用了。

So, automations are already helpful.

比如说,每当有人提交这个表单,就给他们发送这封邮件,并把他们加入 community,诸如此类。

Like, whenever someone signs up to this form, send them this email and add them to the community, for example.

但现在你可以把 AI 融入到里面。

But now you can incorporate AI into that.

所以,你基本上可以利用 ChatGPT 的能力,来决定工作流下一步该做什么,或者起草邮件,或者起草给别人的回复。

So, you can use the power of ChatGPT essentially to make decisions about what to do next in the workflow or to draft emails or draft responses to people.

对,你的自动化背后就有一个 large language model 的完整能力在支撑。

Yeah, you have the full power of a large language model behind your automations.

比如说,你可以上传一些教学内容,或者一些源材料,然后设置一个自动化流程,让 AI 根据这些内容生成脚本,生成旁白音频文件,生成可以配套使用的字幕文件,然后把所有这些都上传到一个 drive 里。

So for example, you could upload some instructional content or some source material, and then you could have an automation set up where AI generates a script from that content, generates narration audio files, generates subtitle files that can go along with it, and uploads all of that to a drive.

它还可以生成一个开发指南,或者说基本上是一个 storyboard,建议在旁白播放的时候应该搭配展示哪些视觉画面。

It can generate a development guide or a storyboard essentially that suggests which visuals to show with the narration as it plays.

然后它可以给你发一封邮件,告诉你它的任务已经完成了。

And then it can send you an email to say that its task is complete.

所以很明显,这里面还会有更多步骤。

So obviously there would be more steps in here.

你可能会想加一些检查节点,让你在流程继续之前,真的可以先审核一下内容,不过你可以用像——哦对,我刚才说的是像 Zapier 和 n8n 这样的工具。

You would probably want to add in checks where you can actually review things before it continues, but you can use a tool like— oh yeah, I said tools like Zapier and n8n.

这里就是 n8n。

Here's n8n.

所以这是我目前最关注的、用来搭建这类东西的首选工具。

So this is my top tool I'm keeping an eye on for building out this stuff.

它看起来就是这样。

And this is what it looks like.

所以它像这种可视化的工作流,对,你可以把 AI 的完整能力带进这些流程里,而且它也可以 self-host。

So it's like a visual workflow like this, and yeah, you can bring in the full power of AI into this stuff, and it's self-hostable too.

所以如果公司希望所有数据都保留在内部,对,他们就可以把它托管在自己的服务器上,而不是基本上依赖 cloud。

So if companies want to be able to have all of their data in-house, yeah, they can host this on their own servers instead of relying on the cloud essentially.

所以真的很酷。

So really cool.

如果你对自动化这块感兴趣,一定去看看这个工具。

Definitely check out this tool if you're interested in the automation stuff.

不过对,这就是这五项技能。

But yeah, these are the 5 skills.

我觉得接下来的步骤,肯定是如果你还没看过 Intro to AI 那个视频,就先去看,然后也许再回来看一遍这个视频,这样很多东西会更好理解。

I'd say the next steps from here are definitely watch the Intro to AI vid if you haven't already, then maybe rewatch this one so things make more sense.

如果你想挑战一下自己,对,就从这些 AI 技能里选一个,开始练习。

And if you're up for a challenge, yeah, pick one of these AI skills and start practicing.

我下周会发布一个视频,讲如果是我,我会怎么从零开始搭建自己的 portfolio。

I have a video going live next week about how I'd build my portfolio from scratch.

很明显,AI 会是其中很重要的一部分,我们也会深入讲一些 AI 项目。

And obviously AI is going to be a big part of that, and we're going to get into some AI projects.

所以如果你们想要一点作业,就在下一个视频下面留言,告诉我你一直在练习哪项 AI 技能,以及进展怎么样,因为对,我们越早开始练这些东西越好。

So if y'all want some homework, comment on that next video with whatever AI skill you've been working on and how it goes, because yeah, the sooner we start practicing with this stuff, the better.

然后对,如果你喜欢这个内容就订阅,如果你想看更多 AI 内容也订阅。

And yeah, subscribe if you like this, subscribe if you want more AI content.

就像我说的,下周我们会深入讲 AI portfolio 和项目相关的内容,我对那个视频真的很期待。

Like I said, next week we're going to dive into AI portfolio and project content, and I'm really excited about that one.

所以对,谢谢你看到最后,我们下周见。

So yeah, thanks for making it to the end and I will see you next week.
已剔除 3 处广告(点击展开查看)
M1
M117:18广告 · 已剔除

I'll try to grab that, but yeah, check out that whole webinar if you want, if you want to learn about how to stay up to date with this stuff.

M1
M117:37广告 · 已剔除

So yeah, we're doing a lot of good AI work over in my instructional design community, which you are more than welcome to join.

It's completely free and you can go to devlinpeck.com/id to join.

M1
M118:19广告 · 已剔除

So check out the AI Haunted House live event that we did on Halloween if you want to see an example of this in practice and learn how I did it.