I Watched Dan Koe Break Down His AI Workflow OMG

Transcript

00:00

If you've been on the internet, you've seen Danco. He's got millions of followers and I've seen him everywhere.

So, I reached out to him and I was like, "How do you do it?" And he told me that his secret is LLM's, AI, and prompts. So, in this episode, he shows his exact

00:17

playbook for coming up with ideas, creating the content, and most importantly, content that is going to go viral. Now, why I find this interesting is right now there's an unfair advantage to create X accounts, to create YouTube channels that get distribution.

If you

00:34

can figure out distribution, you could create a bunch of startups for those audiences. So, if you stick around to this end of this episode, you will understand how to create content like Danco.

And, you know, people generally charge thousands of dollars for this.

00:50

It's free. All I ask is for a like and a comment on this episode.

I can't wait to see what you build. Have a creative day.

01:06

>> What's up, Dan? Great to have you.

By the end of this episode, what are people going to take out of it? Glad to be here, dude.

Thank you for having me on. Uh, I have a few things to share, but the outcome as a whole will be a big picture overview of my content

01:22

ecosystem. So before we just started, you were mentioning like uh you're curious about how my output is so high when I personally don't think it's that high.

It doesn't take too much. So uh I think it just comes down to having a

01:38

really smart system and then being able to augment that with AI in a unique way, which we're going to learn. But the other thing that we have to learn in order to do that well is just how to use LLMs in a way to get them to do exactly

01:54

what you want. So I'm not a coder.

I don't use too many of the fancy tools. Mostly just claw chat GPT.

So how are we going to use the base tools that most people have access to and understand to create something incredible?

02:10

>> I love it. I only have one ask from you.

You know, I just ask that you don't hold back any sauce. You give the prompts.

You show how it's really done. Because what would be success for me is if thousands of people consume this episode

02:26

and understand how to use LLMs to be like you. Because when I see Dan Co, I see millions of followers.

I see you everywhere. I see content here.

I see content there. And I'm like, how does this guy do it?

Dan, can you commit to it? >> Absolutely.

Yep.

02:42

All right, let's get into it. >> Cool.

Uh, so the first thing there, just to touch on, uh, millions of followers, having so much content on so many different platforms. The the thing about that is all of my content is a same is the same across all different platforms.

02:59

So the base of it all comes down to two things. It comes down to my newsletter, which I write every week, sometimes twice a week, but once a week is like baseline.

And then two is two to three social media posts a day and those are

03:15

usually written. And so I write Twitter first because it's like the 28 280 character limit.

And then since that's like the limiting factor out of all the platforms that can pretty easily transfer over to any other platform. So

03:30

that gets posted to threads by Instagram. I turn that into an image and post that onto LinkedIn onto Instagram.

Sometimes I use them as a real script. So I pull up my camera, look at my phone, read my best tweet, recite it to

03:47

the camera, maybe add a bit on there. Um, and then I have reals, Tik Toks, shorts, etc.

And the good thing about that is since the ideas themselves are almost always validated ideas or have

04:04

the potential to do very well, if I can nail the first idea that goes out on Twitter, then I'm practically nailing all the other content that goes out on all other platforms because I'm not much of a trend follower. Sometimes I am, but

04:19

most of the time I believe that the algorithms are based off of human psychology. And if you understand human psychology and can use AI to help you understand that more, your content will see a notable increase in engagement and other things of that nature.

So, uh,

04:37

first I'll just give an overview in this canvas style so you understand my weekly process. So, every week I plan out my newsletter and the idea for the newsletter.

04:53

It's based on a a few different things. It could be a top performing tweet of mine that I had before because I know that that tweet has already done well.

So, I'm just going to take on it take it expand it into a newsletter that turns into a YouTube video podcast. Good to go.

05:11

uh or it can be based on a video that is already doing well on YouTube. So a good way to generate ideas that help you grow is to go to YouTube, look at the accounts in your niche or the people the kind of like tribe you want to join

05:28

online. Go to their account, go to videos, filter by most popular, and then just write down 10 of the videos that you think you could recreate.

Now, the secret there is you're not watching the videos. You're not trying to steal

05:43

anything from the video itself. You're trying to take the topic, the angle, and write your own perspective on it.

Because if other people are watching those videos since they have a lot of views, then they're going to just by the

05:58

nature of YouTube and people binge watching things, they're going to be recommended your video. They're going to watch that.

And if they hear you say the same things as the other person, then you're not going to stand out or be unique. So that's what this process is.

So that's usually the two ways that I

06:14

come up with newsletter ideas. And I keep a document where I tend to I just have a list of ideas that I can pull from.

I pick the one that I think is going to do the best that week. Now what I do for that is I of course have a

06:30

place where I write the newsletter just in a note or document. But I tend to find a few things that I've either talked about before relating to that.

So previous newsletters, previous tweets, whatever they may be. And then Sam

06:46

Alman, the co-founder of OpenAI just said that it is the era of the idea guy. And he is not wrong.

I think that right now is an incredible time to be building a startup. And if you listen to this podcast, chances are you think so, too.

Now, I think that you can look at trends

07:04

uh to basically figure out uh what are the startup ideas you should be building. So, that's exactly why I built ideabser.com.

Every single day, you're going to get a free startup idea in your inbox, and it's all backed by high

07:19

quality data trends. How we do it, people always ask.

We use AI agents to go and search what are people looking for and what are they screaming for in terms of products that you should be building and then we hand it on a you know silver platter for you to go check

07:35

out. Um we do have a few paid plans that you know take it to the next level uh give you more ideas give you more AI agents and more almost like a chat GBT for ideas with it but you can start for free ideabrowser.com and if you're listening to this I highly recommend it.

07:53

A lot of the times I like to weave in other concepts from specific books or videos. So this is where AI really comes into play because I'm talking about uh survival and how to reinvent yourself

08:08

in this newsletter. And when I say survival, I'm talking about psychological survival, something that I find really unique and interesting.

And I've previously watched three videos on that. They're each by a channel called actualize.org and they're like three hours long.

So,

08:26

it's a lot of learning when I'm on walks and other things of that nature that allows me to write in-depth newsletters, but I'm usually not taking notes during this time. And this is where LLMs can come into play because they can usually

08:42

pull from YouTube videos or PDFs allowing you to talk to them and pull up the relevant points that you want to talk about in a newsletter. So the research process there is thinking about the newsletter and the topic I want to write, talking with AI a bit and just

08:59

getting specific key points and summaries and uh other unique perspectives and maybe throwing in a few things about the topic that I'm talking about just so I have a chat that I can refer to that

09:18

turns 6 hours of videos into a thousand words, right? So then I can write about it myself.

>> Uh now the other thing here is if I bake in something I've written previously. So this is on reinventing yourself.

It's kind of my thoughts on that. So you can

09:34

think of this newsletter as like combining these two. So when I take the summary, paste it into another chat or link it to another chat, then I can say, okay, what are the similarities here?

Is there anything that I'm missing? I'm kind of just in this process of ideiation with the LLM until

09:53

I feel like I have enough firepower to start piecing together an outline and then to start writing it. Now, I don't have the LLM write for me, like the actual newsletter.

Um, but with what we're going to talk about,

10:09

you could technically do that. And if we have time, I'll actually show you how to do that specifically.

But uh that's the research process. Nothing special here.

Just getting information from the getting the ideas so that I can put them

10:25

into the newsletter. Otherwise, you just get stuck staring at a blank screen.

>> So I have a question on this. So what your thesis is basically you use X Twitter as a litmus test on ideas.

10:41

>> Yes. And then once you have an idea that gets validated, then you're like, okay, this might have legs to extend this to other mediums.

Then you

10:56

look to YouTube as a encyclopedia of knowledge. you find popular videos um or or videos that just are high quality that you know might be you know

11:11

might be it might be hours long and then you use looks like you use Gemini in the first example to summarize it. You can use you know really anything you want, right?

Yeah, I I tend to go with Gemini 2.5, especially when there's Oh, this one's

11:28

only 9 minutes, but for like the three to six hour videos, I'll use Gemini just because the large context window, >> right? So, you get smart about the videos and then from that you you extend it, right?

And you're you're not using

11:44

the LMS to at this point write it. But um I guess my question is cuz someone someone listening this is going to be like okay but you never talked about the hardest part which is your great ideas

11:59

on Twitter. You know what I mean?

>> So >> how do you think about that? >> So it's kind of circular, right?

The ideas for Twitter can come from two different places. It can come from the newsletter itself.

So I'll show over

12:15

here in a second, but they can come from the newsletter itself and that's a good way to generate ideas. But the mistake that people make there is they'll have a really good newsletter and they'll think, "Okay, I'm supposed I'm supposed to pull ideas from this and they'll read a certain sentence or paragraph and just

12:32

try to copy paste that directly and post it on Twitter with a maybe a few changes when in reality what you're looking for is just the idea that you can turn into a high performing tweet. So,

12:49

it's kind of the same thing there where the first step is learning how to write high performing tweets. And there's courses on there.

There's courses on this. There's YouTube videos on this.

The best thing I can recommend is actually what I'm going to show you, but

13:07

you need to kind of immerse yourself in good writing and ideas. I think the best thing you can do to write better social content in general is to just keep a swipe file of really good ideas that you'd like to emulate under your own brand.

I think of brand as the ideas

13:24

that you associate yourself with. And so you're not necessarily copying other people's ideas or taking other people's ideas.

You're synthesizing a bunch of different ideas that are technically what compose your mind and your

13:39

worldview and you're bringing them together in a very articulate way under your own brand and then as people consume those over time you gain that authority in their head. So uh that's one way to come up with

13:56

ideas is just keep a swipe file. You have to become like a ruthless notetaker and just consistently save ideas until it's until you've read so many of them and written so many of them that it's become second nature to you on how to just structure an idea.

Like you can give me a word and then if you give me a

14:14

minute to think about it, I could probably come up with a really good post about that thing. >> And do you use any tools for that?

Um, I use, if we go to Twitter,

14:30

so I've used both Tweet Hunter X and SuperX. It It feels like >> a lot of people or no company has really mastered this yet, just the Twitter sidebar in general, where when you go to a certain person's profile, this is the

14:46

Super X extension, and it just shows their top performing tweets. So if you write down five to 10 accounts that you really like uh and you kind of want to I think you refine your writing style by

15:01

reading the writing you aspire to have. And so if you have the accounts that talk like you want to talk and you consistently visit their highest performing tweets with SuperX or Tweet HunterX and save all of these down and try to exchange either the structure or

15:18

the idea and practice posting. What I mean by that is right here there's a clear structure.

Go on more walks, walk for no reason, walk to solve a problem, etc., etc. You can take this structure or bring another idea into it.

So rather

15:34

than walks it could be code code more code every day something of that nature. And then the more you or that's exchanging the taking the structure and exchanging the idea from it.

But you can also take an idea like go on more walks

15:52

and exchange the structure of it. So then you look at another structure where this is just a single sentence uh I don't know platitude of sorts this one at least and you can just take the idea

16:08

of walks try to write it as a single sentence now you have two tweets from one idea. Um, and the more you do that and practice it, then really any idea has the potential to do very well cuz you can just pop it into a pre-existing structure or you can take a structure

16:26

and plug in one of the ideas that you're trying to include in your brand. So, uh, one thing there that I'll talk on and then we'll get into actually doing all of that.

Uh what the after I write the

16:41

newsletter, I have three prompts that I like to run it through. So this first one is just a YouTube title generator.

And I'll show you how this one's structured. And if you do this, Greg, I can uh put all of these in a Google doc

16:57

so people can copy paste them and read through them and dissect how they're structured. But for this prompt specifically, I uh wrote a prompt.

I took my best like 15 titles on YouTube

17:15

and I whatever I plug into it, the newsletter I plug into it, it pulls out the key points from that newsletter and tries to spin them into uh 20 to 30 YouTube titles that are similar and follow the

17:31

same psychological patterns and principles as my best performing titles. And so it spits these out here and then I tend to read through these, cross out ones that I don't think would work, highlight ones that I think would potentially come up with a better one

17:46

from my own mind, test that out on YouTube. If it doesn't work after 2 weeks, I'll try changing the title.

Usually that'll get it to a decent place in terms of views. And the reason I create YouTube titles for this is because

18:01

my newsletter is a practically an outline for my YouTube videos, right? I have it pulled up as I'm looking at the screen, I mean at the camera.

I'll look down, read a sentence, recite it to the camera. Sometimes I'll riff.

Sometimes I'll read straight from it. uh if I'm

18:18

looking down and reading at the newsletter, my editor just adds B-roll or text screens over my face so people can't see me looking down. But that's the general structure of it and people quite like it.

The objection there usually is

18:34

uh well, if you're posting the same thing on two platforms, won't people notice that or won't they get bored? Right?

>> And most of the time, I've seen the opposite. Like there are people that read it and then when they watch or listen to it, they get something entirely new because it's a different

18:50

medium. Other people prefer to watch or listen rather than read.

And I I just feel like my philosophy behind it is I'd rather produce one amazing thing a week

19:05

and just put all of my attention into that and put it out across all platforms rather than try to create something new for each platform and each of those things not be as good. And there's probably marginal difference between the

19:20

two paths, but I've really gotten into this flow of the weekly flow of newsletter, YouTube video, social posts across all platforms, and it's worked out pretty well. People don't get angry, and it's seems like it's working.

19:38

But the next thing I do is I'll take the newsletter or the YouTube video and I'll plug them into a deep post generator prompt and just a content ideas

19:53

generator. And these are both kind of the same thing.

For the deep post generator, I wrote a prompt that breaks it breaks down the newsletter or whatever you feed it. It could be a PDF, a YouTube video, anything.

And

20:10

I came up with this prompt. It's still kind of fuzzy how I did it because it was like a huge conversation, but I broke down I asked it to break down a bunch of these social posts that I really liked.

Uh the way they were structured, the way they were written,

20:25

and the key things that made them do really well. And after that conversation, it said, "Okay, well, a lot of these have paradoxes and counterintuitive truths.

A lot of them have a transformation arc. A lot of them have core problems.

Uh, a lot of them

20:42

have examples or they handle objections or they have action steps or aspirational statements." And so I was like, "Okay, take each one of these and uh break down the newsletter, whatever

20:57

you're fed, into like five compelling post ideas, like the most impactful ideas that could be turned into posts. And then I want you to ideulate three core paradoxes, a bunch of key quotes, transformation arc, core problems, key

21:13

examples." So with this, I'm not asking it to write tweets for me because the LLM isn't going to be too good at doing that unless you guide it to in a very specific way. But instead here is I have it give me the pieces of ideas like the

21:31

the deconstruction of high-erforming content and ideas and just give me a bunch of those. And so I can read through here and when one thing sparks an idea for me, boom, I can go and start writing a tweet on that.

Or I

21:47

can combine a bunch of these together since technically each of these like all of these paradoxes, quotes, problems are all related. They're all about one idea.

So I just have a bunch of building blocks for good content in here.

22:07

The next prompt that I use is just a general content idea generation. And I have it give me 60 ideas.

And these are usually based around the tweets that have done really well for me or the social post in general. I found that people really like when I give harsh

22:23

life advice and I like that too. Uh counterintuitive truths which is very similar to the paradoxes.

core problems or pain points. Again, very similar key insights, wisdom, big ideas.

So, I can just read through here and I'm not technically I'm not trying

22:40

to copy paste these as a tweet. I'm just trying to generate a starting point for an idea so that I can have those ideas faster, right?

because I feel like good ideas are the fuel for content, but usually good ideas come from sifting through hours of YouTube

22:58

videos on a walk or reading a good book or uh read finding a really good Substack article and having that generate ideas or just sitting and thinking with your thoughts until something emerges. And those are great ways to come up with ideas, but this is

23:14

just a way to kind of force that to an extent and still maintain the quality without having the LLM actually write it for you. Now, so we just went over like in general how I go over or how I create

23:29

my newsletters and content each week. That's just the writing ideiation process.

How I actually structure that on a day-to-day basis is quite simple where I wake up, do a little morning routine, shower, walk, whatever it may

23:45

be, and then I sit down at my computer and the next two hours is just dedicated to writing. So in that writing, I have to finish one section of my newsletter and I have to finish three posts, three social posts.

And then within that time,

24:01

you usually have time within that two hours still to paste those social posts to all platforms or at least schedule them. And so that's kind of my routine every day.

It's just 2 hours of writing. And that's the base of all of my social

24:16

media, all of my newsletter. Uh, one day out of the week, I record my YouTube video and pass that off to the editor.

And then boom, my content is done. And the rest is history.

You got millions, millions of followers,

24:32

millions of likes. Well, that that was the other thing that I found relatively quick is once I had a post that started to

24:49

generate more followers than other ones did, I knew that I just had to start creating spin-offs of that post because then it's like I almost had a predictable way to increase the follower growth. So, it's like the the key part here is one, you

25:07

need to experiment until you kind of strike gold and know what brings in followers. Then you need to make that a consistent part of your process and consistently create spin-offs of that while with your other time.

And let's say that's 30% of the time. That's like

25:22

one post a day out of the three. The other two posts are for continuing the experimentation on until you find the next one that is like you striking gold.

Then you make that a part of your process. Now you have two posts a day that are just consistently most like they they're not all going to bring in a

25:38

ton of followers, but you're getting closer and closer to just consistently increasing followers over time by doing this. And if you keep that balance of your core ideas that are consistently going to bring in followers with a

25:53

healthy dose of experimentation until you find more and then you just cycle out the core ideas. That's really the entire growth process on any social platform in my opinion.

>> So there's there's one thing that's missing from this which is visual assets

26:12

like media. So on X, I've noticed that like if I tweet just words versus I tweet words with an image, the words with image, you know, call it performs 40% better.

26:28

Um, and so how do you think about that? How do you think, you know, okay, you do this process to come up with ideas to write the content, but what do you do about the image?

I don't do anything. But but that's just

26:44

me. Um because I've ever since I started out, I kind of put the constraint on myself of like I want to do this with just writing.

I I didn't start YouTube until I think 2 years after I was on

26:59

Twitter. And ju just by creating that constraint, it made me be more creative and I think it made me refine my writing a lot more.

I've occasionally incorporated visuals in my Instagram posts uh and newsletters

27:14

for like drawing out specific concepts or having certain infographics or I even did animations on Instagram at one point and those things definitely helped but uh the animations eventually died down because a lot more people started doing them. And then same with the visuals.

27:32

Like I noticed they were a good pattern interrupt when I would post those after not posting them for a while and they would get a lot more engagement. But then if I tried to make that a consistent thing, they would just balance out as most of the posts did.

27:47

And so then I'm like, I should just stick with writing because the ideas in in my head, the thing that I feel built my audience the most, there's no way to know this, but I

28:02

think it came from my focus on idea density and novel perspectives. So, in my YouTube videos especially, I try to bake in at least one thing that I think will like blow people's minds as an exaggeration.

It's

28:20

like the thing that I found so interesting that I feel like not too many people know about. If I can find a way to incorporate that into a newsletter that has a pretty broad and validated angle, so people go into the video because it's already validated.

28:37

they like that topic, they'll just watch it to watch it. And then I have a good hook, a good lead, and then I lead them into like that mind-blowing insight that makes them think a bit deeper and learn something new and have something new to explore that gives them this unique

28:55

connection with me that not too many other people do. >> Yeah.

I mean, >> does that make sense? >> It does.

cuz I think like if you have really good optimized ideas, you don't need visuals,

29:13

it's just so hard to come up with good ideas. I mean, you're sharing the sauce around how to come up with good ideas, but it it's by the way, it's not just good ideas.

It's good ideas that are optimized for feeds, >> right? because >> you need people to reshare it to get

29:30

seen um or like it or comment on it. You you need them to stop scrolling to to you know.

So for a lot of people the images or the videos are a bit of a crutch like you can take a not so good idea crutch

29:47

it to a really nice visual and likely you you'll get more seen. I think I do think that if you added images or and or videos to your ex posts, you would see not that you need,

30:04

you know, another 5 million impressions a month, but I think the way I would I mean I think it's fun to explore like Sora 2 just came out. >> Um, you know, what are some short films, cinematic films that you can, you know, you could be creating.

That's one. The

30:20

other thing is, you know, Nano Banana, Gemini Flash 2.5, >> you know. Okay, I'll show you something real quick.

Actually, go to my my ex. So, I had a um post

30:36

that I posted that got like no like scroll, keep scrolling. It's like a It's a lot.

It's a lot down, but I had a post um Yeah, I'll let you know when that basically didn't go anywhere and then I

30:52

reposted it with an image. There it is.

Stop. It went up.

Went up. There it is.

The cringe mountain one. H.

So, I posted this. Everything I ever got in life was because I climbed cringe mountain.

You sweat, you shake, you look stupid. And

31:08

then one day it breaks open and suddenly you're the only one with the view. It went nowhere without the, you know, Yeah.

basically nowhere without the image, but once the image com combined with, you know, maybe it wasn't a great idea, maybe it was a pretty good idea with, I think, a really

31:25

great image. It it got 148,000 impressions.

And by the way, I hear you listening to this and you're like, "Yeah, but you have 500,000 followers or whatever." The followers don't really matter anymore, realistically. >> You know, it's actually easier to get

31:40

impressions if you have less followers. It's, you know what I mean?

>> Yeah. >> Like I can post something that is >> just like a throwaway tweet and it will go nowhere.

It'll get like a hundred likes compared to a typical 2 to 3,000.

31:57

And yeah, it's just the the audience is definitely a force amplifier, but only if you're good at that thing. Like if if most people who haven't practiced writing content for a long amount of time were given a

32:13

million followers, I do not think that that would help them in the slightest. >> Yeah.

Cool. >> But yeah, let's actually show a cool way to do this with uh Claude or in in creating a prompt to help you write

32:30

tweets or to help you, I guess, learn how to write those high performing posts. So, one thing I love to do is there's a two-step process here.

Anything that I like and want to

32:45

incorporate into my work, I will ask Claude or Chat GPT to break it down, teach me how to do it, I will take that as if it were a an SOP or just a set of instructions. Then I will take a prompt

33:01

that helps me create prompts, send that, paste the instructions into it and say, I want to turn this into a prompt that interviews me for the exact thing you need in order to execute on this as best as possible. So as an example here is I

33:21

said break down the structure of this post so that I can recreate it from scratch. break down why it works, the psychological patterns involved, what context is needed from me, and anything else I would need to understand how to recreate it.

Then I just pasted I believe this is one of my tweets that

33:39

did pretty good. Uh how to know you were doing something meaningful.

You feel like you don't make any progress for weeks, months, or years. Then the growth hits you all at once.

When you create your own path, results aren't predictable, and that keeps many people from sticking with it. So, you find a

33:54

good post. Well, first you write this, tell it to break it down.

You find a good post that you want to understand why it did so well or how it was written. But this can also be applied to something like a landing page or if you're trying to write a book and you

34:10

read someone else's book and you're like, damn, that paragraph was good. I want to understand how I can take my own idea and replicate the structure of this so my writing flows the same amount.

Um, you can do that for writing YouTube scripts if you're able to break down the

34:26

transcript. There's just so many different things that you can do.

Uh, I actually saw when Culie first launched, they had a landing page that I really liked and I copy pasted the landing page, wrote exactly this, told it to

34:43

break down every single part of the landing page copy and it called it paradigm shift marketing and just taught me exactly how to replicate it. So when I could take that and turn it into a prompt and then plug my own ideas for my

34:59

own product into it, then it could rewrite that same style of copy but with my product, my features, my problem that I'm solving, unique mechanism, etc. So with here uh we're breaking down a tweet

35:15

and it breaks down the core structure, hook statement, pain and struggle, the payoff, the insight and warning, why it works. And just by doing this alone, you can learn so much about writing content.

you you like this alone is a master class in

35:33

writing one piece of content and then you can continue conversing with it and that helps you write better content on your own. Then it gives you a formula additional elements to make this work.

And then what I did here is I did the

35:49

same thing do the same thing with this and I took one of Zack Pogra's tweet and or Instagram posts and it was just the best way to get your spark back is burning everything down. You have to reset your life.

Very similar to the

36:04

things I like to read and write. And so I like this.

I want it to break it down. I told it to.

It has It broke down the structure, provocative thesis, purge list, the promise, the elite few, the permission,

36:19

teaches me why it works, writing techniques, context needed, what makes it work, and then I did the same thing for one last tweet. So do the same thing with this again.

People in third People in third world countries

36:36

spend their days talking, playing, dancing, and making love. People in first world countries aspire to spend their days talking, plans, playing, dancing, and making love.

I think about this often. This was just one I found and I'm like, that's a good idea.

I'll throw it in here, too. So, the process of this is pretty much break down three

36:52

tweets that you really like and then combine all of these into a singular guide. So now the anatomy of viral philosophical posts a complete guide breaks down the three archetypes which were the tweets

37:10

and then core psychological patterns technical structure power techniques. >> This is insane.

This is absolutely >> manipulation. It's it's really cool.

And like well imagine doing this for more than just tweets like YouTube scripts and other things like the this is where

37:28

it starts to all tie together. Once you use this process a few you use this process a few times you realize that you can use it for almost anything.

And then if you rewind this podcast to looking at the canvas that I was showing you, you

37:44

can now see that each chat in there was one of these prompts that I created. So, if I want to turn my newsletter into tweets and I use this there, then it like that's so much better than me

38:00

pasting my newsletter into an LLM and being like, "Hey, what are 10 of the best ideas from this? Can you write a tweet?

Have write 10 tweets for me." And then it writes like the worst tweets you've ever read and still includes hashtags. So this is like how you give

38:16

the inspecific instructions so that it can write like you or write how you want it to and then you can paste those directly if you'd like or you can use those as ideas to like as a first draft a starting point.

38:33

Now the last important thing here is I said from that guide what is everything you would need from me in order to start writing tweets. Everything I need to know from you.

Core identity, audience psychology,

38:49

philosophical stance, voice parameters, specific insights, transformation, narrative, emotional territory, so on and so forth. A lot of stuff.

So, we're making this very comprehensive because the context is king. the the more context you give to the AI, the

39:05

more it can get close to putting out what you actually want it to put out or putting out things that will blow your mind and that you feel confident in replicating week by week in your content. So now we have two things here.

We have a description of the context we

39:20

need and we have instructions for how to actually write posts. So what I did next is I took these two things.

I personally pasted them into just a separate note and then start a new chat.

39:37

And what I did here is I have this special this thing is a lifesaver. I use it all the time.

It's just a prompt that helps you create better prompts. So, uh I can share this in some way with you guys after, but it just it it's so good.

39:55

You just have to try it. makes very comprehensive prompts.

Um, maybe you can pause, read through this if you'd like, but I send that and then I try to use the most powerful. I know 4.5 just came out, but 4 point I use 4.1 extended

40:11

thinking on this because I just like to be as advanced as I can when I'm creating a specific prompt. And what I did or said, what is the topic or role of the prompt you want to create?

I said I want to create a prompt that helps me

40:26

ideulate social posts based on the guide from writing posts below. Structure this prompt in two phases.

The first is context gathering. So interview me for the ideas that I want to write about.

And phase two is postw writing. So write three variations of each type of post using the ideas I gave you.

Here's the

40:44

additional context. So inside of here on this one, I didn't paste the context that it needed from me, but I could have.

Right? In phase one, I could have said, "Use the uh guide on what context you need from

40:59

me in order to create interview questions that you can ask me." And then phase two, use the tweet writing guide to take everything I gave you and turn them and turn them into tweets. And so what it did after that, sometimes this prompt will ask clarifying questions um

41:17

if it has them to make the prompt better. But then it starts to spit out the prompt and it has phase one context gathering interview.

Oh, I guess I did paste it in there, >> right? >> So it's somewhere in the thing.

But the

41:33

when you send this prompt for writing tweets, it goes through it asks you what your domain and expertise is, audience painoint, unique observation, transformation, vision, personal connection, uh what tone, specific examples, and then once it has enough information,

41:50

it'll go into phase two, which is to generate nine posts using uh yeah, three variations of the patient observer post, which was one of the examples. uh three variations of the dramatic

42:05

profit post, three variations of the quiet devastator posts, post structure requirements, constraints, output format, and then I just had it rewrite it as markdown without the code block so that I could paste that into my notes

42:22

because not a coder, don't use cursor, any of that, but I just keep things in notes. But, uh, Greg, do you have questions on this?

>> What's going through my mind is like it feels like

42:37

I've been doing content like a Neanderl, you know? That's that's what it feels like cuz I just I just >> brute force my way into it, you know?

I just I have an idea and I d, you know, and I, you know, I start typing and then

42:53

I just throw it out there. Um, it works for me.

Um, and sometimes, you know, do what works type thing. >> Oh, yeah.

>> Um, that being said, what's going through my mind is we we have a lot of

43:08

accounts, you know, cuz I run a holding company and >> I find that some of my, you know, I love my team, but some of my team haven't been creating content. Like, I've been writing, you know, I started a blog 20 years ago, you know what I mean?

I've been writing for so many years that it's

43:24

just kind of uh it's it's inst it's it's my instincts almost now. Um so this is really helpful for managing multiple accounts with multiple people and you really don't

43:41

have an excuse not to be posting something every single day if you have this. >> Yeah.

I mean, another way to think of it, too, is if you're a ghostriter, social media manager, uh you're managing

43:57

clients, you can kind of send them a prompt that gets all of their you send it to them rather than like a onboarding questionnaire that pulls all of their details from them, gets examples of their best content,

44:13

etc., etc., and then you just create prompts for like tweets, tweet writing for them. you create another prompt for YouTube script writing or landing page generation all based around their voice, their profile, etc.

And one last example

44:29

that I'll share here is you can do a very similar thing for almost anything where imagine I know how to create offers. I've like been in the marketing space for so long.

But a lot of people just starting, they have this superpower

44:45

where they can start learning by doing faster, right? You're it's not that you're avoiding reading a book like a hund00 million offers by Alex Hermoszi.

You're able to get the information you need to act and have AI act for you as a

45:02

first draft so you get the first iteration out of the way as well. So you can fail faster on creating your offer so that you can refine faster.

And in my eyes, doing things this way isn't like outsourcing your agency or your cognition to AI. It's actually getting

45:19

you closer to being high agency and uh learning through doing like right the the best way to learn is by building projects, failing, iterating, etc. So

45:35

the example here and how to turn this into a prompt to create an offer and start selling and failing on whatever product it is, a SAS startup, a physical product, a digital product, whatever. Uh I can just ask, give me a detailed guide

45:50

on how to create offers like Alex Horoszi. The reason we do this rather than, hey, here's my product, help me create an offer from it is because I want to narrow the context from the beginning, right?

I don't want AI. I I don't want the LLM to pull just general

46:06

context that it thinks is best for creating an offer or writing copy. I want it to pull from the expert that I know has skin in the game here.

So, it gives me a detailed guide with the value

46:22

equation, building a grand slam offer, pricing, 10x value test, naming your offer, etc., etc. And I think when I was doing this I cut myself off.

But what you can do here is

46:38

then I can say okay take this and act as if you are guiding me through creating an offer. What are the exact steps?

make it extremely detailed and then I can do the same thing I did here or sorry then I can ask next okay what's

46:55

all the context you need from me uh in order to take my product and turn it into a very good offer then it'll spit all that out now you have two things you have the set of instructions and you have the context that you need so the next thing you do is you paste

47:12

the prompt that helps you create great prompts it's going to say what do you want to create I say, "I want to create a two-phase prompt. Here's the information that you need to interview me for." Then phase two, here's exactly how to turn this into an offer.

So, I

47:28

want you to create my offer blueprint for me. And then you, it'll output a prompt.

You copy paste this. You send it.

You answer the questions. It spits out an offer blueprint that is probably 10 times better than something

47:45

you would have done on your own just then, right? And it would probably be on par with what you would be able to do after two to three months of actually learning about it.

And I believe that is it.

48:03

Well, you delivered, dude. I I you know that's why I wanted to have you >> you actually overd delivered to be honest >> nice >> and because you know I think uh my mind is just racing with how much more I can

48:19

do with content and and I you know what are other audiences I could be spinning up using this method. Um if you can uh yes let's include that Google doc or or that doc with some of this in the show notes.

Um, and I appreciate that

48:38

generosity. Dan Co, I hope you come back on the pod.

I'll include links to where you can follow Dan on YouTube, on X, his newsletter. Uh, what else, Dan?

>> That's it. You'll, if you're curious,

48:54

you'll find stuff that branches off of that. >> Amazing.

Thanks, Dan. I really appreciate it.

And uh dude, takes a lot for me to get to get my to get my mind buzzing, but you did. You did it.

>> Good. Yeah, I appreciate it, man.

And

49:10

I'm glad that it did provide value. That's one thing I'm always like hopeful for because you never really know.

I've used this so much that it's like second nature to me. So, I'm very glad.

I mean, I've used LLMs with content, like how can you make this more clear or you know

49:29

that that kind of stuff, but I I this the way you approach content creation is a very methodical way of approaching content creation. It's like very surgical.

And um by the way, I think this is the best

49:45

way to learn how to create content. you you might actually like everyone should set this up.

>> Maybe over time you you you're like, you know what, I'm just gonna now I understand what works and what doesn't work. I'm just going to shoot from the hip like Greg.

Um and that might be what

50:02

you what might, you know, end up being great for you. Or you might be like, hey, uh I'm going to double down on this and I'm going to make 20 accounts and that's what this is going to look like, you And I think I'm really interested in that

50:18

because um I just think that we're in this window where there's a huge o arbitrage opportunity to earn attention. So you know any any way to help that you know I'm looking I'm looking for ways looking for ways to optimize.

50:34

>> Yep. Agreed.

Danco Startup Ideas podcast thanks for coming on. Thank you, Matt.

Summary

Transcript