Nicolas Cole: Build A Digital Brain So AI Repeats You, Not Slop
A ghostwriter turned SaaS founder explains why AI can't invent your point of view, it can only repeat what you've already written down.
Posted
yesterday
Duration
Format
Interview
educational
Views
6.3K
258 likes
57 · 43
Big Idea
The argument in one line.
AI can't invent a point of view, so in the AI content era the real leverage comes from building a personal library of your own experiences, opinions, and approved language for AI to repeat, not invent.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
A creator, founder, or ghostwriter using AI to write content who wants a system for feeding it real material instead of guessing at your voice.
Someone building a personal brand or company blog who keeps getting told their AI-assisted writing sounds generic and wants to know why.
A SaaS founder trying to differentiate a product through methodology rather than features.
SKIP IF…
You're looking for prompt engineering tricks or specific AI tools, this is a content strategy framework, not a tutorial.
You only want quick viral hooks with no interest in building a long-term content library.
TL;DR
The full version, fast.
Nicolas Cole, a ghostwriter turned Typeshare founder, argues AI didn't remove the value from writing, it just exposed which writing was already worthless. Most text is commodity content, the same sentences anyone could type, and readers who spot 'AI slop' are really spotting commodity language. The fix is building what Cole calls a digital brain: your life facts, lessons, opinions, and previously approved sentences, so AI repeats and recombines material that's actually yours instead of guessing. Ownership works out to association built through volume and personality, and voice reduces to three adjustable settings: word choice, sentence length, and sentence structure. The long-term asset isn't any single post, it's the growing library both you and AI keep pulling from.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Greg introduces Cole, reads the Brex sponsor spot, and frames the episode as a masterclass on the value of everything you write.
03:36 – 05:40
02 · POV is the Moat
Cole argues SaaS still has a moat when the product carries a specific method, not just features that copycats can ship in a month.
05:40 – 11:24
03 · Language As Open Source
The alphabet, most words, and most sentences belong to no one, so what readers call 'AI slop' is really just commodity language.
11:24 – 14:26
04 · Value Is Relative To The Reader
A lesson that bores the writer to repeat is still the most useful thing a beginner can receive, so value depends on where the reader stands.
14:26 – 17:43
05 · Approved Language
Cole's decade of ghostwriting distilled into four steps: collect, approve, recombine, repeat, the same job he says a personal AI model should do.
17:43 – 22:09
06 · The Value of AI in Writing
AI should automate repeating commodity ideas you've already worked out, freeing time to think through ideas you haven't written yet.
22:09 – 24:23
07 · Commodity Ideas
Two opposing commodity opinions about coffee and saving money show that the value isn't the idea itself, it's which side you're willing to claim.
24:23 – 27:58
08 · How To Set Up The System
Cole plugs Typeshare but says the library can live anywhere, what matters is that AI has access to it and you've done the thinking first.
27:58 – 33:56
09 · Ownership IS Association
Ownership isn't legal, it's association built through volume and personality, illustrated by Cole's own repeated World of Warcraft opening line.
33:56 – 39:53
10 · Build your Personal Data Set
Everyone prompts the same public data, so your own formative experiences and life facts are the one dataset AI doesn't already have.
39:53 – 45:17
11 · Branded Content vs Founder-Led Content
Branded accounts underperform named humans, but branded pages can still repeat a founder's commodity ideas, or be built around several named employees.
45:17 – 46:35
12 · What is Original Content?
Typing a sentence doesn't make it yours, original thinking is the rare result of crystallizing a point of view over years, not volume alone.
46:35 – 48:58
13 · Writing Versus Short-Form Video
Writing versus video is a false choice, most short-form is scripted anyway, so it's writing in a different format.
48:58 – 49:50
14 · Three Types of Hooks
Adding a credibility detail or a 'just launched' timestamp to a generic hook moves it up a tier, at the cost of expiring it.
49:50 – 53:27
15 · Timely Content vs Timeless Content
Every piece sits on a spectrum from a 24-hour meme to a ten-year essay, match your expectations to which one you're making.
53:27 – 56:32
16 · Voice is 3 dialed settings
Voice unbundles into word choice, sentence length, and sentence structure, dials you can adjust deliberately instead of rewriting a whole draft.
56:32 – 1:00:49
17 · Finding your Voice
Most people who dislike their voice can't describe what makes it theirs, it's usually fear of publishing wearing a voice complaint.
1:00:49 – 1:07:30
18 · The Company Brain As The Moat
Companies are starting to pay a deep thinker, or pay millions to acquire data outright, to build the library their marketing machine repeats from.
1:07:30 – 1:09:34
19 · Closing Thoughts
Cole points people to Ship 30 for 30 and X, and Greg closes by calling the episode timeless rather than timely.
Atomic Insights
Lines worth screenshotting.
AI slop is really just commodity language, the same sentence structure thousands of other people have already used.
The ghostwriter's job was never inventing new ideas, it was collecting a client's own words and recombining them, which is exactly what a personal AI model should do.
Ownership of an idea has no legal basis in most writing, it's just association built through repeated volume and personal details.
A single sentence repeated deliberately across hundreds of pieces can become someone's most recognized idea.
Less credible writers with more personality routinely outperform more credible writers with none, credibility alone is not the winning card.
Your own formative experiences and mistakes are the one dataset every competitor's AI model is missing, because you haven't written it down yet.
A branded corporate account will always underperform a specific named human telling the exact same commodity idea.
Typing a sentence doesn't make it yours, most individual sentences are open-source groupings of words nobody owns.
Short-form video is rarely spontaneous, it's usually a script, so writing and video are the same underlying skill in different formats.
Adding a 'just launched' phrase or a specific year to a hook boosts its short-term reach and simultaneously expires it for reuse.
Voice breaks into three adjustable dials, word choice, sentence length, and sentence structure, and changing roughly 1% of a draft can make it sound like the author again.
Nearly all professional ghostwriting uses the same authoritative default voice, the personal difference is usually a tiny fraction of the total wording.
Companies are starting to pay millions of dollars just to acquire and mine another company's written data.
Paying one person to think deeply about an industry full time can out-produce a scattered content team, because AI can slice one deep library a thousand ways.
Takeaway
Build A Digital Brain, Let AI Repeat It
WHAT TO LEARN
The AI era didn't remove your moat, it moved it: from typing words to owning the point of view and approved language AI repeats back at scale.
02POV is the Moat
A SaaS product built around commodity features has no real moat, but one built around a specific method forces competitors to copy the person behind it, not just the UI.
Every time you explain your product's logic, you reveal the point of view that makes it hard to copy quickly.
03Language As Open Source
Most sentences people flag as 'AI slop' aren't badly written, they're just commodity language that thousands of other people have already said the same way.
Writing 'you should do this' repeats what the whole topic already says, writing 'I did this, I think this' gives the reader your specific view of the world.
04Value Is Relative To The Reader
A lesson you've repeated a thousand times and find boring to write is still the single most useful thing a beginner in your field can read.
Commodity content isn't worthless, it's calibrated to wherever the reader currently stands, so it keeps paying off long after it stops teaching you anything new.
05Approved Language
A ghostwriter doesn't invent a client's ideas, they collect what the client already said in books, posts, and meetings, then recombine those exact words into new pieces.
The same four steps, collect, approve, recombine, repeat, describe what a personal AI model should do with your own writing instead of guessing at your voice.
The bigger your library of your own approved words gets, the more new combinations AI can build from material that's actually yours.
06The Value of AI in Writing
AI is most useful for automating the low-leverage work of repeating ideas you've already fully worked out, which frees your time for ideas you haven't thought through yet.
The real bottleneck isn't prompting skill, it's whether you've captured your own high-quality ideas somewhere AI can actually pull them from.
07Commodity Ideas
Two commodity opinions, like skipping coffee to get rich versus coffee not being the problem, aren't unique, what matters is which one you're willing to put your name on.
AI writes generic advice not because it's a bad tool, but because it doesn't know which side of a commodity debate you actually agree with.
08How To Set Up The System
Where your notes live, a markdown file, Obsidian, or a paid tool, matters less than whether AI actually has access to them and you did the thinking first.
Skipping the step of deciding what you actually believe and asking AI to figure it out for you is the shortcut that becomes the longer road.
09Ownership IS Association
Ownership of an idea isn't legal, it's association, the more you write about a topic, the more people connect that topic with your name by default.
One writer repeated a single line about playing competitive World of Warcraft as a teenager across hundreds of pieces until it became his most recognized idea.
Personality details, even mundane ones like moving across the country twice, differentiate you more reliably than credentials do, since less-credible personal writing regularly outperforms more-credible corporate writing.
10Build your Personal Data Set
Everyone prompting the same AI models pulls from the same public data, so your own formative experiences, mistakes, and choices are the only dataset that's actually unmade.
You already run a personal language model at every coffee meeting when you repeat the same life stories, writing is just the scalable version of that same habit.
The goal isn't finding one dramatic life event, it's writing down the mundane details you take for granted, because other people recognize themselves in them.
11Branded Content vs Founder-Led Content
A branded account with no name behind it will always underperform a specific person telling a specific story, even when both are repeating the same commodity idea.
Branded content doesn't need a single voice, it can be several named employees, each attaching their own personality details to the company's commodity ideas.
12What is Original Content?
Typing a sentence doesn't make you its owner, 'I woke up today and made coffee' is nobody's original thought no matter who typed it.
Original thinking isn't a single insight you stumble into early, it's the result of narrowing your focus and crystallizing your view on one topic over years.
13Writing Versus Short-Form Video
Asking whether writing, video, or podcast wins is the wrong question, a multi-seven-figure content business can run almost entirely on writing with minimal short-form video.
Short-form video is rarely spontaneous, it's usually a script read off a teleprompter or clipped from longer writing, so it's writing wearing a different format.
14Three Types of Hooks
Adding one credibility detail, like a specific revenue number, to a generic how-to hook moves it from purely commodity content into personality content.
Naming the current year or a feature that 'just launched' can boost a hook's short-term pull, but it also expires the piece, you can't reuse it in three years.
15Timely Content vs Timeless Content
Every piece of content sits somewhere on a spectrum from a 24-hour meme to a ten-year essay, decide which one you're making before you start writing it.
A timeless piece will never out-perform a meme in its first 24 hours, judging it by that number just measures the wrong kind of value.
16Voice is 3 dialed settings
Voice breaks down into three adjustable dials, word choice, sentence length, and sentence structure, so you can diagnose exactly what's off instead of rewriting the whole piece.
One ghostwriting client rejected a draft as not sounding like him, and changing about one percent of the wording, two added citations, fixed it completely.
17Finding your Voice
People who say they dislike their own voice usually can't describe what makes their writing theirs, it's often fear of publishing wearing a voice complaint.
Roughly 98% of most professionally ghostwritten writing uses the same authoritative, concise default voice, the personal 2% is what people actually mean by 'sounding like them.'
18The Company Brain As The Moat
Companies are starting to pay millions just to acquire and mine another company's data, the same logic applies to hiring one person to build your internal content library.
Paying one skilled writer to think deeply about your industry full-time can out-produce an entire scattered content team, because robots can slice one deep library a thousand ways.
Most teams publish far less content than they're capable of, the real constraint usually isn't output capacity, it's the lack of a documented point of view to pull from.
Glossary
Terms worth knowing.
Approved language
Words and phrases a person has already said or written themselves, which a ghostwriter or AI model can recombine into new content without inventing new thinking.
Digital brain / personal language model
A person's collected life facts, lessons, opinions, and approved language, stored somewhere AI can pull from so it repeats the writer's actual material instead of guessing.
Commodity content
Writing built from shared, widely repeated ideas and phrasing that isn't attributable to any one person, useful for beginners but low in long-term leverage.
Personality content
Writing that includes specific, identifiable details about the writer's life or opinions, which builds stronger association with a topic than credentials alone.
Original content
Writing that results from years of crystallizing a specific point of view on a narrow topic, rare and not achievable just by writing a lot quickly.
Unmade data set
A person's own formative experiences, mistakes, and life facts, the one category of material current AI models have no access to until the person writes it down.
Company brain
An internal content library built by one or a few deep thinkers inside a company, which AI and teams can then remix across every marketing channel.
“Just because you wrote it doesn't mean you own it.”
contrarian one-liner on originality→ TikTok hook↗ Tweet quote
07:00
“The key to losing weight is eating healthy. That's a commodity sentence.”
clearest possible example of what 'AI slop' actually is→ newsletter pull-quote↗ Tweet quote
1:06:20
“That's why people are starting to say the company brain is the moat.”
names the episode's closing idea in one line→ TikTok hook↗ Tweet quote
16:40
“Value is always relative to where the audience is.”
reframes when commodity content is actually worth writing→ newsletter pull-quote↗ Tweet quote
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
metaphoranalogystory
Nicholas Cole's made over $30 million writing on the internet. So when he talks about content, I listen. And this episode honestly felt less like an interview and more like just a free masterclass.
He breaks down why most people's content gets ignored in the AI era, why AI is making generic writing worthless, and how to build what he calls a digital brain. a library of your stories and your opinions and your lessons and ideas that gets you customers 24 -7, that builds you trust and gives AI something real to work with.
If you write online, you sell online, or you want ideas to compound, this is honestly worth thousands of dollars, but it's free for you. The people that stick around to the end of this are not going to be writing AI slop. And I think the people that use AI to actually create interesting content, I mean, of course, those are the people that have the highest probability of success.
Enjoy the episode. I learned a lot. I think you will too.
Today's episode is brought to you by Brex. My company's been on Brex for a year and a half, and I started because I kept hearing companies like Vercel, OpenAI, Anthropic were using Brex, and I figured if they're using it, why shouldn't I? It's been a game changer.
The thing that got me is how smooth it is. It's got high limit cards, it's got banking, it's got AI that handles the back office busy work like expense reports, which I don't want to do on its own. It's really just built for this agentic world.
If you're building something new, it's time to get Brex. Check it out at brex .com slash solutions slash startups.
Link in the description.
Nicholas Cole, thank God you're here, man. By the end of this episode, what are people going to learn? By the end of this episode, you will better understand the short, medium, and long -term value of everything you create and write and market on the internet before you create it.
That is the goal. Before you invest the time, you will understand why you're doing what you're doing in the first place. Yeah, and I think that matters a lot because all of us are vibe coding software.
We know there's a lot of opportunity to build apps, ship SaaS, things like that. But it's hard to get customers. And what I love about you is through writing, you understand how to get seen, how to create stuff that goes viral.
And there's sort of a bit of art to it. And what I hope to learn, selfishly, by the end of this episode, is to get some of that art from you, to learn about that. So that the products that I create, I could accompany it with content writing.
that tells stories that hopefully goes viral or that just resonates with people. And Cole, can you commit that to me? Will I learn that?
Yeah, that's what we're going to talk about. So something I see people talk about a lot these days is that in the AI era, SaaS no longer has a moat. And I actually think that that's not true.
I think that SaaS has a moat. when you have a point of view. That was a nice little rhyme there.
But what I mean by that is if you're just building commodity SaaS, meaning you're just providing features that accomplish tasks to people, that's fundamentally a commodity. SaaS with a point of view means that you're building some sort of product or platform that enables users to do things in a certain way. Like they're executing a certain method.
So for example, like we have a SaaS platform called TypeShare and the goal of TypeShare is not, well, there's a gazillion scheduling apps and there's a gazillion writing apps and our writing app does writing and scheduling too, right? Like that is a commodity. The goal of TypeShare is to quote unquote automate or make tasks easier for people who want to execute.
a different methodology for writing on the internet. So there's a point of view that drives the product. And if someone says, well, why would that create a moat?
For example, if we say inside TypeShare, well, your goal is we're going to help you build your personal language model. And our point of view is that there are certain types of content, commodity content, personality content, original content. Each of these different types of content creates different leverage inside of your personal language model.
Even as I'm articulating that, imagine you're trying to build a SaaS platform that does everything our SaaS platform does. Every time you build a feature, it's like you have to turn back and look at me and go, oh, but wait, what does that actually mean? What's the point of a personal language model?
What does commodity content mean? So you will forever be one step, two steps, three steps, five steps, ten steps behind because you constantly have to look at the person who's creating the point of view or creating the methodology. to go, wait, they're the one doing the thinking.
I'm just going to copy them, which means you're always behind. And so I think it's funny when people say there's no moat in SaaS because it's like, well, yeah, there's no moat in anything if you don't have a point of view. But there's a ton of moat in SaaS and product and writing and marketing if you do have a point of view.
And so that's kind of what inspired this convo is like, so how do you create a point of view in the first place? So where do we start? So an idea I've been thinking about a lot is the fact that the English language or any alphabet, any language is really just an open source language for communication.
Meaning no one owns the individual letters of the alphabet. No one owns words like the vast majority of words, right? Like you don't own the word tomato.
I don't own the word catastrophe. We also don't own even our names. There are a lot of people in the world named Greg.
There are a lot of people in the world named Cole. And if someone says, my name is Cole or my name is Greg and you write, my name is Greg, right? You're not plagiarizing them and they're not plagiarizing you.
And as you start to like tease that idea out, you realize that the vast majority of words and groupings of words and thoughts and ideas and sentences are shared. They are not actually attributable to any one individual.
And I think this is what people are wrestling with when they read writing that they think was written by AI, is when you identify, quote unquote, AI slop, what you're really identifying is commodity content, meaning you're reading something, but It feels the way that it does because it's not attributable to any one individual.
It's like reading health and wellness content. Someone writes, the key to losing weight is eating healthy. That's a commodity sentence.
There's a gazillion people who have said that exact grouping of words. And we don't look at the millions and millions of people who have repeated that same grouping of words and go, oh, everyone's plagiarizing each other. We just go, that's a commodity sentence.
as you start to tease this out, it raises a really interesting question in the age of AI. And the age of AI is the creation of infinite text -based content. Anyone can write anything on any topic at any time with any amount of volume and speed they want, which raises the question, so what is worth writing in the first place?
And I think it's interesting, even like your content is a really good example of this, where The simplest delineation is when you're writing and you say the difference between I think, I did this, I perspective versus you should. You should do this.
Because when you're talking to the reader through the lens of you, inherently the content is going to lean more commodity. Because you're kind of just like repeating all the things that everybody knows everyone should do. But when you're writing from a place of I, like I just did this, or I think this, or you share all the time, it's like, I think this would be a great startup idea.
When someone consumes that content, that is personality content. That is, I am consuming what Greg thinks about the world, not what everyone in this topic area typically says about this topic in the world. You know what I mean?
That's why all content isn't created equal because sometimes I'll see people go insanely viral on X, let's say, and I look at the content and I'm like, I've read this before. It's a Wikipedia article essentially summarized or it's just a take that's been said before and they plopped it into ChatGVT and asked to create an essay on it.
It could still get 5 ,000 likes or 10 ,000 likes, but at the end of the day, It's like, what game are you trying to play? When I'm creating content, I think about two things.
I think about, how do I drive as much value for the person consuming the content? Constantly as I'm writing, I'm thinking about the person in the other seat and thinking about, how can I help them? That frame of mind is super helpful because...
That means actually you're not going to write things like, you should do X, Y, Z, you should do X, Y, Z. Because that's kind of coming from a place of like, I'm better than you and I know everything, therefore you should do this. And I don't think that's what I want to avoid.
And then the second thing is creating affinity. So if I add value to people's lives, it's sort of an effective number one. it generates some affinity to, hey, this guy helped me out a little bit.
I got an idea from him. I never thought of this topic in that way. For example, what you were talking about with the English language is open sourced.
I use English every day of my life. I never thought about that concept before. And for the rest of time, I'm going to be associating that with you, Cole.
People undervalue how valuable it is to have a new idea and ship that value to someone and it actually affects their day -to -day, especially within the context of a business context because that means that they're going to be making more money, they're going to be being more productive, and that's going to put food... food on their plates, feed their family, get them a job promotion.
There's so much good to that. And that's what I want to unpack today is that piece for people. Yeah, so, okay, a ton of thoughts.
First, value is always relative to where the audience is, meaning something that you could find really valuable because maybe you're a more advanced entrepreneur. A beginner would not find as valuable because they're not far enough along yet. And in reverse, something that's very basic to you that you've known and you've learned that lesson a thousand times is not valuable to you anymore because you already know it.
But to a beginner, that's like the most helpful thing you could give them is step one. So value is always relative to where the other person is. What prompted this whole mode of thinking for me is especially in the past two or three years, and as I've seen AI start to accelerate, I've found myself living in a constant state of frustration where I know that these models can write a lot of things just as well as I can.
My frustration is knowing that that solution exists, and yet I still find myself spending time writing and rewriting things. to tell beginners that I know would be helpful to them, but are really boring for me to repeat. And so I've been writing nonfiction online for 15 years.
I've been writing about how to write on the internet for over a decade. I'm not exaggerating, like probably written, I don't know, north of 50 million words on the topic, like thousands and thousands and thousands of pieces. And especially in the past few years, this frustration for me is like, I know that there are still a tremendous amount of people in the world who need help taking the first step and going, how do I start writing on the internet?
What's my niche? How do I name and claim what makes me different? Ideas that are very simple to me that are very novel to beginners.
And yet, every time I write about them for the 999th time, I feel like I'm wasting my finite time on earth because I'm just reswizzling the same groupings of words over and over and over again. And so that's what originally prompted this way of thinking.
And it made me realize that commodity content doesn't have zero value. Commodity content does have value. And for a lot of people, particularly beginners in any industry, commodity content is actually the most helpful thing you could give them.
The problem is that as you write more and more commodity content about a specific topic or in a niche, you will feel what I'm feeling. You probably already do, where you're like, I find myself just repeating the same ideas that I've already thought through. I'm not actually thinking through anything new.
I'm just repeating things I already know. And that is very low leverage. And so as you begin to build your content library, It, you have to sort of examine why am I writing the things that I'm writing and each new thing that I write, am I, what is the additive value of the thing that I just created to my, what I'll call digital brain, my personal language model.
And in an ideal world, like, so just a quick little backstory. I've been ghostwriting for a decade. I used to do this framework with clients.
where when I would be working with a client who was really busy, maybe they wrote a book, maybe they would have a podcast, more likely maybe they would just record internal meetings. And I would pull from those assets, pull from their book or pull from blog posts or pull from things that they had created or said out loud.
And I would take those words and phrases and I would assemble them into content. And that's essentially what a ghostwriter does. The misunderstanding is that the ghostwriter isn't really creating new ideas.
The ghostwriter is not really like putting net new thinking in the client's mouth. The ghostwriter is essentially taking what I like to call approved language. The client said it.
And so why can't I take a sentence from chapter one and a sentence from chapter two and two sentences from chapter three and two sentences from chapter four of their book and stitch them all together and create something new? The client said all of those words in those groupings of words. And so I did the manual version of what AI can and should do here soon, which is the larger your library grows, the more ideas and approved language you have, the more net new combinations you can create.
And I think that fundamental idea is something that the masses do not understand about the leverage you get. in crystallizing your thinking and building a library of content on the internet. Because the AI is not actually writing.
AI is taking approved language from things that you've written and creating net new combinations. And if you play that out, the benefit of doing that is now I can continue to help beginners, telling them the ideas I know they need to hear, but I don't need to take my finite time on earth. rewriting the same thing over and over again.
Which, what does that mean? That opens up more time for me to do what? To think about the next thing I haven't thought about.
And I think this is the hope, or this is the utopian hope of AI, is that AI will automate the low leverage stuff, which opens up time for higher leverage things. And so in the case of writing, what that means is AI will help me repeat the same commodity ideas that I've already written 10 ,000 times, which opens up time for me to think about non -commodity ideas, personality ideas, or ideally original ideas, which is more additive to the long -term value of my digital brain, which imagine the societal implications of more people thinking about things they haven't thought about before.
rather than just churning out blog posts of things that they've already thought about and just need to repeat. To me, that creates two bottlenecks, and I'm curious your thoughts on this. The first is, how do you create a second brain that people use Obsidian or whatever second brain software you want to use so that you're properly capturing those ideas?
Those ideas are in a safe place. And then the second bottleneck is, okay, I've got these high quality ideas. Actually, some of them might be high quality, some might be low quality.
But I've got these ideas in this bank. How do I essentially plug it into an LLM such that I don't get slopped? There's a bunch of people who are just able to mangle their LLMs such that they understand how to prompt it in a way or have agents run in a way such that it's eloquent.
So I think to me those are the two bottlenecks. Am I missing something? Yeah, I mean, I think that there's certainly outlier use cases and I'm not against the sort of person who is not a writer, is not a proficient writer, doesn't want to be a proficient writer, but knows a lot about an industry, wants to brain dump, into an LLM and then have the LLM take that information and just write it the way that the LLM would, that's really no different than how ghostwriters work with clients.
And inherently in hiring a ghostwriter, you get subjectivity, right? Ghostwriters are human beings. They all write in different ways.
You're going to get a subjective output because you're working with a specific ghostwriter. An LLM is no different. And I'm not against that way of working.
I think that The fundamental mistake and the point of view that I'm really leaning into for the foreseeable future is that the highest leverage thing you can do is the writing. The robots should do everything else.
And I think where people get lost and why slop happens is they might agree or understand that idea. Then they want to skip the step of writing and go, well, my LLM knows enough and it can just do it for me. And you don't realize that in doing that, you're giving up a lot of leverage.
I don't want the robots writing for me. What I want is the robots repeating me. What does that mean?
So, for example, you post on a bunch of different social platforms, right?
I know that I'm like, okay, I have this giant library of short form text. Giant. I've tweeted 10 ,000 times, whatever it is, right?
And I want to keep publishing two short form pieces a day. And I also want to put out short form on Instagram threads and YouTube community posts and some of these other platforms. It is extremely low leverage for me to go through my library and find the highest performing posts and repost them on multiple platforms.
People got around this with a social media manager. Cheaper solution, you get around that with a VA. And even cheaper solution, now AI and agents can essentially do that.
But the step further and where I see long -term value in this idea of building a digital brain or a personal language model is that it is not difficult for AI to understand. Cole, you wrote this short form piece about... five benefits of journaling.
And you wrote this other short form post on seven benefits of journaling. And I can take some from the first list and some from the second list and create a new third list that is just repeating a different combination of those two posts. And this is where every time you sit down to write, I think one of the most helpful things to ask yourself is what is the short, medium, and long -term value of the thing that I'm creating?
And I'll give you an example of – so what is the value of you writing commodity content? If commodity content is a commodity, what's the point? The point is that there is borderline infinite commodity content inside every topic.
The question is not how do you write about every type of commodity content in your topic. The question is which types of commodity content do you agree with? And which types do you disagree with?
So I'll give you a simple example. In the world of personal finance, two pieces of commodity content are, if you want to build wealth, you should skip your morning coffee. Don't waste that $3.
And you should just buy the S &P 500. And 30 years from now, you'll be a millionaire. Can I guess the second?
Yeah, yeah. Because I'm just curious. Is the other one, if you only rented?
and didn't pay a mortgage, you would actually outperform in general? I feel like I see that one all the time. Yeah.
So these are examples of commodity ideas. These are not unique to any one individual. One person says, save your coffee, buy the S &P, you'll get rich.
Another person says, saving your coffee money isn't going to make you rich, you're focused on the wrong thing. Both of those are commodity ideas. The question and why you should write about it is which one do you agree with?
Because those are two different ways of looking at the world. And so where people run into, oh, AI is creating slop for me, is they don't realize how many of those micro decisions they're expecting AI to just understand about them. So AI might, like, let's say their topic is personal finance.
AI might write something and use the example, well, you should save your... $3 of coffee money and just invest in the S &P 500. And the person reads that and goes, oh, this is slop.
But they don't realize that the reason AI wrote that is because AI doesn't know which perspective you agree with. So what is the long -term value of your library? Well, on the commodity content side, the long -term value is you actually yourself having clarity over which points of view do you agree with?
Which points of view do you disagree with? It's not the ideas themselves that are unique. It's which ones are tied to your identity.
And the more of those that you crystallize, the more long -term leverage you get out of working with AI. Because AI knows, oh, I'm never going to use that example because that's not how you see the world. How do you actually set this up?
If someone listening to this is like, I agree. I agree with Cole and what he's saying. It makes a lot of sense.
How do I actually set this up? Well, shameless plug, this is the point of view. This is what we're building inside of TypeShare.
So if you want SaaS that does that, there you go. But the reality is there's a lot of different ways that you can accomplish this. And fundamentally, it actually doesn't really matter where the library of content lives.
What matters is that AI has access to it. And more importantly, Like you can't skip the step of understanding what you're doing in the first place.
I fundamentally believe that the reason people are struggling with writing an AI slop and writing with AI is because deep down, everyone wants to skip the step of thinking and just have the robots do the thinking for them. And in doing so, it's like you're trying to find a shortcut and the shortcut is the longer road in disguise.
And so. Like an example that's very visceral for me is when I was writing on Quora back in 2013, 2014, I wrote a handful of answers about me playing World of Warcraft professionally as a teenager. And I would always like in each answer, I sort of referenced it in different ways.
I was like, oh, when I was a teenager, I was a really competitive World of Warcraft player. And then another piece, I would say the same thing, just like a different combination of words. And one day I wrote a piece and the first sentence.
was when I was 17 years old, I was one of the highest ranked World of Warcraft players in North America. And that Quora answer went viral. And so me, myself, manually looked at that and when that first sentence and that grouping of words, that way of talking about that experience, my hypothesis is clearly that resonated with people.
So what did I do? I wrote hundreds. of other pieces over the next five, 10 years, repeating that exact sentence somewhere.
Maybe it was the first sentence, maybe it was a hook somewhere in it, maybe it was before I told a story, but I used that grouping of words hundreds of times. That is the goal of what AI can and should help you leverage, is that as you get clarity over these ideas and groupings of words and points of view and perspectives, once you say it really well once, or twice, or two different variations, and you like both, AI can then go, yeah, yeah, you've said that.
I got it from here. You want AI doing repeating. You don't want AI doing net new writing, thinking, I'm assembling groupings of words and groupings of sentences that you've never done before.
That's why then it gives you something and you're like, this isn't me. Because it's not you.
Yeah, and people sniff it out too. And it actually hurts you more than it helps you. Because then they associate you with, in my opinion, then they associate you with someone who isn't adding value.
And it's like, there's nothing worse. I don't think there's anything worse that you can do than steal someone's time. And I think that's why people are so...
hate slop so much. It's because they consume this piece of content and they're like, wow, that was sloppy. I hated that.
I think this is a second really fascinating question in the age of AI, which is, so what is ownership? If language is open source, no one owns the alphabet, no one owns 99 .9 % of the words, and most people don't even own 99 .9%. percent of the groupings of words.
If language is open source, what is ownership? And ownership, and obviously there's like a legal component. It's like you can trademark names and like there is a legal component.
But generally speaking, ownership is really just a different word for association. You associate ideas with certain people. Now, you can achieve association in lots of different ways.
The easiest way to achieve association is to write about a topic a lot. Like, people associate AI startup ideas with you. And we can just say, at a fundamental level, for no other reason than you create more about that topic than the vast majority of people.
Or I'll give you another example, two examples. of your coffee, drinking coffee and wasting money, which is when I hear that concept, I think of Ramit Sethi, I Will Teach You To Be Rich, who always says, keep drinking your coffee, even if it costs $6, live your rich life. And that's against the common idea of save, put in the S &P.
And the other side of that coin is Dave Ramsey. who's like, save, invest, don't go in debt. And so both ends of the spectrum, same core idea.
Exactly. And I mean, so to build on that, Dave Ramsey is known for that meme where he's taking the live call and he's like, so you spent all your money on what? And then it's a funny answer.
And he's like, and you owe how much money to who? And then it's a funny answer. What's actually happening is that that's a format, that's a structure that people associate with him.
So ownership is really just association, generally speaking. And the easiest way to establish association is through volume. You write a lot about a topic, people associate that topic with you.
Very basic. The next layer up of association is personality.
It is, okay, so it's not just that you write about this topic, but the audience knows all of these specific things about you. So play it out. It's like you have a personal finance expert and you know nothing about them.
They just write a lot about personal finance. And then there's another personal finance expert who is maybe just as qualified or less qualified. But they share a lot about their personal life, their personality, how they see the world.
She's a personal finance expert. She's a single mom. She got divorced three years ago.
She has two kids. She's had to work multiple jobs to put them both through school. Really hard worker.
Started a blog. Blog took off. You know all these things about that person.
And something that I think a lot of people forget is that less credible, more personality content typically outperforms more credible commodity content. Take the extreme.
There are not as successful personal finance people or finance people with significant personality content that get more distribution and have bigger audiences than Ray Dalio. It was like, a multi -billionaire, right?
So the credibility of the person is actually not like it's important. You want to learn or hear from someone who has done something and, you know, knows about the thing. But I think people forget that credibility itself is not the winning card.
The winning card is actually the personality of the person and how much they're willing to let you into their world and how many things about them you associate with that topic. And that idea should be really liberating for a lot of people. I'm sure a lot of your listeners, people that build products or are first -time founders or second -time founders, they want to build organic audiences.
You actually don't need to be coming up with game -changing, radically new, original thinking. All you need to do is just go, yeah, there's a gazillion startup founders, but I am a startup founder, and this is what makes me unique. This is how I see the world.
This is where I live. This is how many siblings I have. This is where I went to school.
This is what I studied. This is what my hobbies are. And all of those little data points actually now in the age of infinite text -based commodity content, all of those little data points are differentiators.
And more importantly, every time you write about one of those personality details, that personality detail should make its way into your digital brain, your personal language model. Just like With me, when I was 17 years old, I was one of the highest ranked World of Warcraft players in North America.
There's a reason why I repeated that 500 times. And so then association is really just I associate these life details, life events, ways of seeing the world with you. I'm sure you're familiar with Gary Vee repeated this, has been repeating this for 15 years, which is I want to buy the New York Jets.
Over and over and over and over and over again. What is that? That's a personality detail.
And you associate that detail with him. And I think he lives in New Jersey.
So it's kind of funny, right? I know he works, maybe he has a place in Manhattan, but I think he grew up and still lives in New Jersey. So it is interesting that you can have someone who lives in another place and his dream is to buy the New York Jets.
Tens of millions of people associate Gary Vee, who's not necessarily a sports person. I know he has a sports business now, but made his wealth in wine and social agencies. All of a sudden people connect him with the New York Jets.
I think something really interesting you said is basically like, hey, you're going to have... Your competition, basically, is going to just plop ideas to create commoditized content. And there's just going to be more and more of that.
Whereas if you could do an audit of interesting tidbits about you, your brand, how you think about things, point of views as you call it, throw it in a markdown file or use a second brain software like Obsidian or something.
And it's going to just stand out by virtue of everyone else is using the same tool and the same methodology to crank out content. So it's sort of like fish where the fish are and you're going to be able to fish with Dynamite if you actually have these stories and stuff like that.
To sharpen the precision of the language, it's that everyone is using the same LLMs, meaning they are all creating from the same data set. Your personality content, your library of content is the unmade data set. You can't expect AI to write a paragraph inside of an essay or a newsletter or something that you're writing about.
what you learned as a kid going through an experience that you've never articulated. AI doesn't know that. That is an unmade data set.
And so the way I think about personality content and that aspect of a personal language model is that you want to spend your time building the unmade data set. And the way that you do that is by writing about how you became you. What are your formative experiences?
What lessons did you learn? What mistakes did you make? All of these things seem very...
mundane to ourselves, but that's because we lived it. We experienced it. We take it for granted.
But you articulating a visceral mistake, especially in a story, instantly differentiates that piece of writing from everybody else writing about the same fundamental commodity idea, right? And so that's where you want your time to be spent is you want to be building these net new additive data points, the unmade data set about how you became you.
What do you say to people who are thinking to themselves, well, I lead an uninteresting life and I don't have anything interesting or specific or unique about me? Yeah, I mean, A, that's incorrect. And B, I don't know any other way to say it.
It's just wrong. And B, it's like you have to get outside of yourself and realize that Even the most mundane experiences are things that other people resonate with and see themselves in.
Like a really simple example is I moved across the country twice in the past four years. That's not like the most unique and different and captivating data point ever, but it is an experience. And I bet if I was at a dinner party and I said, yeah, I moved coast to coast twice in the past four years, I might.
be talking with someone else and they're like, oh my gosh, I did the same thing six years ago. Wasn't it crazy? And then we're telling each other stories and all of a sudden we're like friends at a dinner party.
We fail to realize that we do all of these things in conversation every single day of our lives. And the build on idea, and I've been thinking about this for a decade because I was trying to explain this to ghostwriting clients back in 2016, 2017, which is in real life, We walk around as a walking, talking, personal language model already.
And what happens is we go have a coffee meeting. Think about the first time we met up for coffee in Miami. And you go have a coffee meeting or meet up with someone.
And what do you do? You start sharing things from your personal language model. You're like, here's where I'm from.
Here's how many siblings I have. I moved to this city two years ago. Here's what I'm into.
Oh, you're into the same thing. We're just doing the unscalable version of that. Writing on the internet is the idea that instead of repeating the same life facts and stories and perspectives over and over again manually at one meeting and then another dinner party and then another grouping with friends and then a picnic one afternoon, instead of doing that manually, What you can do is you can write about it once and then publish it on the internet and have that same coffee meeting sharing that same perspective or insight or life fact or story with borderline infinite people.
And then with AI, that piece now gets added into your library. And now you can go, I don't want to tell that story anymore. I've already told that story 10 times.
I've written about it 10 times. We've got 10 different variations. I just want you to repeat it for me.
And that is the... fundamental leverage of writing with AI for the next decade, foreseeable future. So that works with personal brand.
If you're trying to build a personal brand, you want to get out there, you want to be Gary Vee, that works. How does this work with branded content?
First of all, I think that for a long time, I remember my first job out of college was working at an ad agency. I was repeating this as a 23 -year -old. I was trying to tell my boss and my boss's boss, nobody wants to hear from the branded Twitter page.
They want to hear from the CEO. They want to hear from a human being. And that only proves more and more true.
So I think that regardless, the human being personal branding decision is continuing to accelerate and branded content just always underperforms. But also, There's no reason why your branded content can't repeat all of your chosen commodity ideas.
Like if we create, so let's say TypeShare, our SaaS platform has an X account and a LinkedIn page. There's absolutely no reason why those social pages can't just repeat all of my commodity ideas. Now, it wouldn't make sense for TypeShare to be like, when I was 17 years old, I was one of the highest ranked World of Warcraft players in North America.
Like that doesn't make sense because that's a personality data point. But there's no reason why TypeShare can't tweet, here are the five benefits of writing online. It's a commodity and it's the same thing I would have said.
So I actually think that you get double leverage when you have the human being, the founder or an executive team or whoever, building their own personal language model that then the branded content is like, great, we're just going to ignore the personality details and we're just going to repeat the commodity ideas, which is what brands do anyways.
You're just choosing which perspectives do we agree with. Yeah, and I think you can still, there's so many brands. There are a lot of brands like Taco Bell comes top of mind, which it feels like it's personality driven.
So just because it's commoditized content doesn't mean it can't be human centric. I think the future of branded content is like, you know, basically not just one human behind it, but like five, you know, and you can even like sign off, oh, it's Tiffany who did this or Jeremy who did this, who wrote this and different stories around the brand.
So for example, like, you know, Taco Bell, let's just use that. Like you can have, instead of just. Taco Bell writing commoditized content around, or just content around, oh hey, we dropped a new crunchy burrito today and here's what it looks like.
It's more around Tiffany, who works at the Tulsa, Oklahoma store, has been coming to Taco Bell since she's been 13. In fact, her parents met at a Taco Bell. And then you tell that story.
Think about – OK. So that's a fantastic example. Think about how with a little bit of creativity, if you take – or Taco Bell takes that approach, what's really happening is you're opening the door for personality content.
And personality content – that's what I was saying where lots of like high personality content often outperforms more credible commodity content. And so I actually think in that example, Tiffany, the taco czar would outperform the other taco company brand that's just only doing the commodity version. And this is why I think this lens is so helpful is because it forces you to raise the question like, what are we actually creating here?
What is the value? And if you're only creating commodity content with no personality component, you have to fundamentally realize it's like you are creating the lowest leverage, lowest value form of content in your topic.
That's not to say it has zero value, but it has the lowest value. A little creativity means you can open the second tranche, which is, well, now we give ourselves permission to create some personality content. And the ultimate goal, the ultimate goal is to spend enough time crystallizing your thinking through writing.
on a specific topic where you ultimately create original content. And one of the things that I've been writing about recently is just because you wrote it doesn't mean you own it. I think that a lot of times people think because my fingers touched the keys, my fingers typed that grouping of words, that means that grouping of words is mine.
And that's not true. If you write the sentence, I woke up today and made coffee. That is not attributable to you.
That is an open source grouping of words. No one owns that. And there's this faulty belief that people have where they think if they're the ones who wrote it, they're the ones who did the work, that means they own it and that means they created something original.
That is not original thinking. You didn't think a single new thought. You didn't say anything that hasn't been said before.
And so original thinking is not something that's like, oh, I started writing online and three days later I had a game changing insight. Original thinking comes as the result of crystallizing the way that you see the world in a specific topic or a small number of topics for a prolonged period of time. And one of the best examples of original thinking, like original does not mean I wrote a hundred thousand word book on it.
Great example of original thinking was Pythagoras saying the earth is round, not flat. That's six words, like fundamentally six words that completely changed the course of humanity that we still associate with that person. And you only arrive at those six words after years and years and years of crystallizing thinking in a topic.
And so if you kind of loop all the way back to the beginning here, that's why there's an irony when people say, I want to write about everything. It's like you don't realize that in doing that, you deprive yourself of ever being able to go deep enough where you end up saying something original. All you're doing is you're just typing and repeating commodity ideas all day long, thinking they're yours and thinking that you have ownership or originality because you were the one who wrote the words.
When in reality, those groupings of words have existed forever. How do you see writing versus short form video? In what sense?
So we talked a lot about content in general, on capturing ideas, writing ideas. And when I hear that, I think of tweets, I think of blog posts, I think of email newsletters, I think of copy on websites. But on the other hand, we have all the social feeds just hungry for short form video.
So my question to you is, how do you think about writing as it relates to short form video? First of all, I don't, one of the most common questions people ask is which is better, which is better writing or video, which is better podcast or long form YouTube.
The reality is that the answer is yes. It's all of them. They all work.
Me and Dickie, my business partner create essentially zero short form video. The only short form video we publish is we have an internal editor and we record a long form and he clips out some random short form and posts them. It is very low, I'll even say low value short form.
We're not intentionally investing in that. And we run a multi seven figure business. It's all driven through writing.
So yes, there's always a trend or some other form of content that is working or maybe working a little bit better, but that doesn't mean that the other ones aren't working. The second idea is video is no different than writing.
Everyone uses teleprompters now. When you do shorts, do you ever script them? I actually, do I script them?
When I do shorts, it's like cutouts of long form content. Or it's, I do like animation explainers that are scripted. So they are scripted?
Yeah, those are scripted. Either it's cuts or scripted. Yeah.
So more and more short form video is scripted and a script is what? A script is writing and it's the same fundamental idea. To make this really tangible, like walking through the value of the thing that you're creating.
So pretend I create a short form video and the first sentence is, here's how to start a newsletter on Substack. That is commodity content. That's not to say that it's not valuable, but that is fundamentally commodity content because I am just explaining how to do something that a lot of other people can and probably have explained how to do.
If we change the first sentence and we say, I run a $400 ,000 a year. annual recurring revenue, paid newsletter on Substack. Here's how to start a Substack.
Now we've moved up a tranche. That has a personality detail in it. It's a little like credibility too, but that's really a personality detail.
Like I did this, I'm doing this, right? If we add in the variable, which is sort of the last variable we haven't talked about, which is timely versus timeless, everything that you create sits on a spectrum. Timely content has very high short -term value and very low long -term value.
Timeless content has very low short -term value and very high long -term value. And so, for example, I saw you tweet yesterday or two days ago like an AI image meme where you were like, I'm not waiting for the foldable phone. I'm waiting for the desktop that's foldable, right?
So that's a great example of hyper -timely content. where you capitalize for 24 hours, but then the value of that content falls off a cliff. You can't repost that a month from now or three months from now or a year from now, right?
So this raises a very interesting question of, all right, so everything that you create, like how much value are you actually extracting and over what time horizon? You created a meme, you capitalized for 24 hours, great. But like, That's all the juice you're going to get out of it.
You're not going to get any long -term value out of that. In contrast, I wrote this essay about AI, took me two weeks to write, spent a lot of time on it. It got some short -term value, but I'm going to reference and use that essay for like the next 10 years.
And so when you're creating something, you want to ask, well, where on the spectrum am I sitting? And are my expectations aligned with the thing that I'm creating? You can't go build something or write something or create something that you know is intended to be very timeless content and expect it to get the same short -term distribution as a meme that capitalizes on an Apple announcement in 24 hours, right?
There are different expectations. So if we tweak the first sentence of that short -form video and we go, Substack just launched this new feature. I run a $400 ,000 a year paid newsletter.
Here's how to start a Substack. in 2026. That word, this just happened, or in 2026, fundamentally changes the short, medium, and long -term value of that content.
Because I can't repost that in 2030, right? Maybe I could use pieces of it, or I got to rerecord it, or I got to rewrite half of it. And so when you put all these variables together, you start to realize everything you create in some way is is adding something to your digital brain, your knowledge of the world, your personal language model.
The question is, how much did it add? How additive was it? If you spend 20 minutes and you're just repeating the same commodity idea that you've repeated a hundred times, maybe you added like one point of, I said it slightly differently.
Okay, cool. We have another like grouping of words, approved language. Whereas if you sit down and write a net new essay about a topic you've never written about before and you crystallize a whole point of view around it, the additive value of that to your digital brain is exponentially higher than the first one.
And so that's my hope with AI is that the human is doing the writing, the human is doing the thinking, the human is doing the choosing of how do I see the world, what ideas do I agree with, do I disagree with, what are my personality details, etc. But the robots are very effective at repeating and remixing and doing all the things that social media managers and even ghostwriters and VAs have done for a very long time.
You still got to give the robots direction around writing in a style that isn't generic sounding AI. So now you have touched on the, I think the final piece, and it is the hardest one for people to understand, which is voice. So what is voice?
What is style? Voice is a bundled term. When we say someone has a voice, someone's content has voice, we think that it's this intangible quality that can't be replicated.
The reality is voice is really a bundled term made up of a couple different data points. It's like, Word choice dictates voice, right?
If I say something is really cool, that gives you one perspective of who I am. If I say something is exquisite, that gives you a different perspective of who I am. And if I say something is fucking dope, that gives you a different perspective of who I am.
So individual word choice, like literally a single word in a sentence can change your perspective of what the voice of that sentence or person is. Another one is sentence length and sentence structure. If someone writes a paragraph with no periods and like seven different sentences broken up by semicolons, you have a very different perspective of that person rather than someone who writes in short, terse sentences with very simple third grade reading level vocabulary.
And what's important about this is Your takeaway should be that voice is something that you intentionally create. It's not something that just naturally happens.
And when you become proficient enough at writing, what's funny, I ran into the same thing, ghostwriting for clients. I will never forget. I had a ghostwriting client who I sent him a piece and he sent it back and he was furious.
And he said, this doesn't sound like me. And I looked at it and I went back and listened to the call and I was like, all right, he didn't really tell me what it was. So I have to guess what he's thinking.
Like, why doesn't this sound like him? And it occurred to me, oh, I was using the wrong voice. I made this piece lean too much into opinion.
And this guy actually spoke referencing like a lot of studies and a lot of facts. I didn't rewrite the piece. All I did was go back in and add in like.
two or three sentences that just said, according to the New York Times or according to Harvard Business Review, I changed 1 % of it. And I sent it back to him and he was like elated. And he was like, oh my gosh, this is so much better.
It sounds just like me. And so the irony of writing with AI is oftentimes AI will give you an output and people will look at one single variable. They'll look at one individual word choice and go, the whole thing is shit.
And the reality is it's like, no, if you were proficient, you understood the qualities that made your voice, your voice, which most people don't know. All you would have to do is just go, I want to trade that one adjective for a different adjective. And all of a sudden the sentence sounds like me.
What do you think of people who don't like their voice, but who know which creators voices they like? So they might say, you know, I don't really like my voice, but I really like how. Greg Eisenberg writes his essays on X.
I truly believe that 99 % of the time when someone says that, they can neither articulate what makes their voice their voice, what it is they don't like, and they can't articulate what makes your voice your voice. And usually when people say that, it's more of a mask of they're just afraid to write, period. And so it's a lot easier to point to and make this like a voice issue.
The reality, I have ghostwritten for north of 300 different founders, executives, investors. The vast majority of the time, I used the exact same voice in the writing. It was authoritative.
It was concise. It maybe had like a couple little vocabulary choices in there specific to that person, but like. 98 % of the thing was just an authoritative thought leadership style voice.
And almost every one of them said, this sounds like me. And so if you go, I'm not joking. And so if you go back to like language is open source, most groupings of words are commodities.
The reality is that the voice component in writing is a minority component. It's a very small piece. It's not like, this entire piece of writing is an original voice.
Very rarely is that the case. It happens, but very rarely. Instead, a piece is 80 % commodity groupings of words with maybe some subjective swapping of adjectives and things.
And I think this is one of those, it's like you can't possibly discover the qualities that you want or like about your voice without writing. How are you going to discover, oh, I actually prefer these types of adjectives over these types of adjectives? I prefer this rhythm in my writing opposed to this rhythm in my writing.
And so if you're not willing to do that, how in the world should you expect AI to have any inkling of any of those data points or choices? That's the unmade data set. Yeah, I mean, the truth is, you know, like deep down, you know, like it's the same.
It's the same concept around you know what kind of food you like. You know if you're a pizza or a pasta person. Of course you know the types of words that you connect with.
You know your voice. It might be a little deep down. You have to be a little self -reflective on it, but it's there.
To be clear, this is not exclusive to the everyday person. I think oftentimes people think, you know, oh, I'm the everyday person. I haven't achieved this massive outcome.
I bet this is so much easier for people who have achieved all these things, right? They make up all of these problems. The reality is I have ghostwritten for some of the most, like, you can't accomplish things any more accomplished than what they've accomplished in the world of business.
And even they, the core problem was They get nervous staring at a blank page and they're nervous to put themselves out there. The majority of my value as a ghostwriter was not actually the writing.
It was helping the client and giving them permission to put themselves out there. And it was comical to me. Ghostwriting for so many people where I would have a client be like, I'm really afraid of putting my perspective out there.
And their perspective is like, the key to marketing is knowing your customer. I'm like, there's nothing. There's nothing to be afraid of.
You are literally repeating a commodity idea that most of the planet agrees with. And so a lot of these are self -inflicted issues. And the whole key is getting out of your own way and hitting publish so that you can discover what these variables are.
What makes your personality? What makes your writing style? What makes your voice?
The other piece of that is Generally speaking, people are underpublishing. The volume of content that they're putting out is lower than they can actually do.
I was having this conversation with our agency actually and LCA literally works with the biggest companies on the planet building their new products. In the AI world, as you can imagine, that team has just an incredible data set and an incredible knowledge base around what's working and what's not.
There's so much content that they could be putting out. They put out some content here and there, but I think one of the things we were talking about is how do we ship a minimum of five pieces of content a day? And what does that look like?
What systems do we need in place? And we could use AI in that process, but what systems do we need in place for that to happen? Because if we're not posting five pieces of content per day, I don't believe we're giving it a fair shot.
I would agree. And I would say that it's important for people to remember that in many ways, that is the baseline expectation now. And that is just the infrastructure.
The question on top of that is, what is the input? If you're posting five times a day, where are you pulling the ideas from? It goes back to, well, you need a digital brain.
You need a personal language model. Where is it all coming from? The same way, I know you're aware of the trend.
You see OpenAI acquired that podcast, TBPN, and there's a huge move to these large companies realizing they need a media component. I think that It might be idealistic, but I think that some of the best marketing money that any Fortune 500 company could spend would be plucking really thoughtful journalists and nonfiction writers in relevant or tangentially relevant fields and being like, you have one job.
We're going to pay you a fat salary, and your only job is to think deeply about our industry and write long -form content about it. That's it. Because what's really happening is it's not that like that individual piece of writing is the silver bullet and that's going to drive all the sales and they make it all super promotional.
What's really happening is you have a very smart person who's paid to think deeply about a topic, hopefully over time creating net new original thinking and ideas that get associated with the company. And that library of content feeds the infrastructure. It feeds your whole supply chain.
It's like, just to make it visceral, Greg, it's like, imagine you were like, cool, I'll pay you a million bucks a year. I want you to do nothing but think deeply about the future of AI and technology and the implications it's going to have on product building. And I just want you to build like a world -class long -form content library that all of our different departments can pull from.
The robots do the repeating. We can pluck short form. We could turn it into automated videos, like all the infrastructure.
I just want you to create the highest quality input. Now imagine if Salesforce did that with like a Pulitzer Prize winning journalist or Slack did that. I think that the leverage you get out of that is so, it's so hard for people to wrap their heads around and they're so focused on the minutia of like, I post five times a day and what time of day and how do we go viral and can we pay for retweets and like all, none of it matters.
None of it matters. What matters is you have the infrastructure and you have a very valuable content library that you pull from and leverage for a long period of time. That's why people are starting to say the company brain is the moat.
If I hired you, Cole, and you're adding to this company brain and it's getting more and more valuable and I can use robots to go and slice and dice it a million ways. I know it sounds crazy to pay someone a million dollars, but it's not that crazy if the value from that million dollars is thousands of pieces of content. It's improving workflows.
It's helping with the product. It's pretty insane, right? And that's why you're starting to see actually companies spend millions of dollars, tens of millions of dollars, and I think it'll be hundreds of millions of dollars on data licensing or buying data.
I saw Micro One buying data, Spirit Airlines data, for I think millions of dollars. And why? Because they're going to use that data, they're going to mine that data.
And I think that's where everything is going. That's exactly it. It's just people understand this idea more in technology.
They go, oh, we acquired a company that just filed for bankruptcy and we bought all their data. And everyone's like, Yep, I instantly understand the value.
The reality is language and writing and libraries of content are data. And you are buying groupings of words, you are buying ideas, you are buying crystallized thinking that you can leverage the same way you leverage data. Yeah, it's funny because I bet the Spirit Airlines people are probably like, ha ha ha, I can't believe they bought our data for millions of dollars, idiots.
And the Micro One people are like, Ha ha ha ha. I can't believe they sold us the data for so cheap.
Yeah. Well, it's, it's always a question of, it's actually not really the value of those assets in and of themselves. The question is, how do you leverage it?
You know, like if you're, if you're a company and you buy some writer's entire catalog and library of content, and then you don't have the operational infrastructure. to scale it and squeeze all the juice out of it, then that was a horrible decision because you're not able to do what you should do with it. But if you have a really smart person as your input for your entire content marketing machine, like literally one person doing that for Slack or Salesforce or Airbnb or whoever, one smart person could build the entire digital brain library.
of content that their entire marketing machine uses and reswizzles and remixes and posts and chops and cuts and pace. To me, that's the future. I mean, we're out of time, but that's like a whole other startup idea that someone needs to do, which is just do company brains as a service.
Come on. It's a home run. If I was excited at all about B2B, I would go do it, but I'm not.
You've been so generous with your ideas, the whole playbook. You open sourced it. I really appreciate it.
Where could people get to know you better, find you, and get more of your ideas? Always pumped to chat, my man. You can find me on X.
I still want to call it Twitter. You know, I'm an old head. You can find me on X.
And if anyone wants to build their digital brain and library of content, you should come join us for the next Ship 30 for 30 cohort. We're going to do that in October. Should be a lot of fun.
Cool. Yeah, I'll include links for all that in the show notes in the description. Cole, you've been on the pod probably, what, three, four times?
Is it three? Three or four? Yeah.
I love having you on. I hope this episode is well loved. It's one of those things where you're talking about timeliness, timely versus timeless.
This is a timeless episode. People who ingest these ideas, 2026 now, 2029, 2032, these are ideas that are always going to be so, so valuable. It's sort of, in some ways, an eat your vegetables episode.
Versus drinking a smoothie, there's just so much nutrients in this. And I do believe that people who actually, if you've made it here, people actually who've made it here, they are going to understand how to have strong points of view, how to use AI to create some of this content, how to think about how it all plays together.
And they're lucky. Well, thanks, man. Yeah, that's my goal.
As timeless as possible. I'll catch you later, Cole. Thanks, everyone.
The Hook
The bait, then the rug-pull.
Greg Isenberg opens by calling this 'less like an interview and more like a free masterclass,' then hands the mic to Nicolas Cole, a fifteen-year nonfiction writer and ghostwriter who has turned his own writing into a repeatable system he says AI can extend without ever producing slop.
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
Jonathan Courtney walks through his four-step Promoter Blueprint, then shows live how he used Claude and Claude Code to build a $450K webinar campaign in about an hour.
Startup Ideas host Greg Isenberg gets engineer Ras Mic to open his actual agents.md file and walk through the isolate-build-prove-ship loop he runs on real pull requests.
Greg Isenberg and Remy put the invite-only Instinct agent through real errands, a haircut, a restaurant table, and a Bali visa, and find out it keeps a copy of your email even after you disconnect it.
A two-person breakdown of OpenAI's top-tier model that skips the game demos and goes straight to code audits, nine money-making agent prompts, and a Raspberry Pi speaker built and shipped in about 30 minutes.
01:02
01:58
02:57
04:27
05:20
07:23
08:56
09:06
11:20
12:55
13:48
15:39
16:29
17:06
18:48
19:48
20:49
21:42
22:55
24:00
25:43
26:37
27:31
29:20
30:27
31:23
32:26
33:39
35:36
35:55
37:02
38:23
39:19
40:49
42:08
43:02
43:56
44:50
45:56
46:22
46:35
47:25
49:24
50:18
51:11
52:05
52:59
54:36
55:22
56:08
57:49
58:40
59:31
60:23
61:59
63:12
64:42
65:49
66:56
68:12
68:56
44:45
28:24
31:29
27:30
22:52
38:46