Jan 20, 2025 · 55m · startup-ideas

I Discovered The Perfect AI Writing System (Life-Changing)

Nicolas Cole · 36m spoken Greg Isenberg · 9m spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

Digital writer Nicolas Cole joins host Greg Isenberg to demonstrate a practical, step-by-step framework for using conversational AI as a collaborative writing assistant to transform unpolished seed ideas into high-performing long-form content.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Greg holds 19.9% of the talking time here. How this is scored →

Greg as informed peer 5.3 Guest teaching 5.3 Guest disagreement 1.1 Greg pushing back 1.3
05100:0015:0030:0045:004:54–7:25 · Greg as informed peer 4/10 Brainstorming the Seed Idea: Reed Hastings and Blockbuster Greg introduces a raw seed idea from his notes about Reed Hastings banning the word Blockbuster at Netflix, framing it around psychological guilt rather than late fees. Cole enthusiastically accepts the seed as an ideal starting point for the writing demonstration.7:26–12:14 · Greg as informed peer 5/10 Deconstructing Ideas and Defining Content Goals Cole teaches the foundational framework for seed ideas, noting that writers must decide if a concept is the central theme or a sub-point. Greg contributes his specific audience niche expertise around startup ideas, while candidly admitting his usual writing process lacks strategic goal-setting.12:16–18:39 · Greg as informed peer 6/10 Conversational Prompting and the Editor-in-Chief Dynamic Cole demonstrates conversational AI prompting and introduces the editor-in-chief metaphor to describe high-leverage taste. Greg demonstrates his own domain expertise by selecting the word 'banning' and breaking down why conflict-driven headlines perform virally on X.18:40–28:14 · Greg as informed peer 5/10 Reverse-Engineering Viral Hooks with Mini Few-Shot Models Cole breaks down how to construct a few-shot mini-model using Greg's past viral tweet hook, firmly dismissing common writer objections about stealing format. When Greg asks if users should build custom GPTs, Cole bluntly dismisses the idea as premature complexity.28:20–35:02 · Greg as informed peer 4/10 Structuring Long-Form Content and Subhead Specificity Cole explains why 90% of an article's value rests in headline and subhead specificity rather than stylistic prose, contrasting vague subheads with tangible promises. Greg validates the concept, acknowledging that his intuitive 'cowboy' writing style will benefit significantly from this structured approach.35:03–49:08 · Greg as informed peer 6/10 Applying the 10 Magical Ways and Finalizing the Outline Cole guides Greg through his 10 Magical Ways framework to map out distinct subheads and actionable lesson formats. Greg exhibits sharp editorial judgment by quickly flagging and rejecting an ambiguous 'me too trap' phrasing generated by ChatGPT.49:10–53:22 · Greg as informed peer 7/10 The Meta-Value of Frameworks and Educational Leverage Cole argues that paying for educational frameworks provides extreme leverage when prompting AI. Greg pushes back slightly to broaden the claim to free educational media like podcasts, then cites Mr. Beast's leaked playbook on consultant hiring to reinforce the meta-learning theme.4:54–7:25 · Guest teaching 2/10 Brainstorming the Seed Idea: Reed Hastings and Blockbuster Greg introduces a raw seed idea from his notes about Reed Hastings banning the word Blockbuster at Netflix, framing it around psychological guilt rather than late fees. Cole enthusiastically accepts the seed as an ideal starting point for the writing demonstration.7:26–12:14 · Guest teaching 5/10 Deconstructing Ideas and Defining Content Goals Cole teaches the foundational framework for seed ideas, noting that writers must decide if a concept is the central theme or a sub-point. Greg contributes his specific audience niche expertise around startup ideas, while candidly admitting his usual writing process lacks strategic goal-setting.12:16–18:39 · Guest teaching 6/10 Conversational Prompting and the Editor-in-Chief Dynamic Cole demonstrates conversational AI prompting and introduces the editor-in-chief metaphor to describe high-leverage taste. Greg demonstrates his own domain expertise by selecting the word 'banning' and breaking down why conflict-driven headlines perform virally on X.18:40–28:14 · Guest teaching 7/10 Reverse-Engineering Viral Hooks with Mini Few-Shot Models Cole breaks down how to construct a few-shot mini-model using Greg's past viral tweet hook, firmly dismissing common writer objections about stealing format. When Greg asks if users should build custom GPTs, Cole bluntly dismisses the idea as premature complexity.28:20–35:02 · Guest teaching 6/10 Structuring Long-Form Content and Subhead Specificity Cole explains why 90% of an article's value rests in headline and subhead specificity rather than stylistic prose, contrasting vague subheads with tangible promises. Greg validates the concept, acknowledging that his intuitive 'cowboy' writing style will benefit significantly from this structured approach.35:03–49:08 · Guest teaching 7/10 Applying the 10 Magical Ways and Finalizing the Outline Cole guides Greg through his 10 Magical Ways framework to map out distinct subheads and actionable lesson formats. Greg exhibits sharp editorial judgment by quickly flagging and rejecting an ambiguous 'me too trap' phrasing generated by ChatGPT.49:10–53:22 · Guest teaching 4/10 The Meta-Value of Frameworks and Educational Leverage Cole argues that paying for educational frameworks provides extreme leverage when prompting AI. Greg pushes back slightly to broaden the claim to free educational media like podcasts, then cites Mr. Beast's leaked playbook on consultant hiring to reinforce the meta-learning theme.4:54–7:25 · Guest disagreement 0/10 Brainstorming the Seed Idea: Reed Hastings and Blockbuster Greg introduces a raw seed idea from his notes about Reed Hastings banning the word Blockbuster at Netflix, framing it around psychological guilt rather than late fees. Cole enthusiastically accepts the seed as an ideal starting point for the writing demonstration.7:26–12:14 · Guest disagreement 1/10 Deconstructing Ideas and Defining Content Goals Cole teaches the foundational framework for seed ideas, noting that writers must decide if a concept is the central theme or a sub-point. Greg contributes his specific audience niche expertise around startup ideas, while candidly admitting his usual writing process lacks strategic goal-setting.12:16–18:39 · Guest disagreement 1/10 Conversational Prompting and the Editor-in-Chief Dynamic Cole demonstrates conversational AI prompting and introduces the editor-in-chief metaphor to describe high-leverage taste. Greg demonstrates his own domain expertise by selecting the word 'banning' and breaking down why conflict-driven headlines perform virally on X.18:40–28:14 · Guest disagreement 3/10 Reverse-Engineering Viral Hooks with Mini Few-Shot Models Cole breaks down how to construct a few-shot mini-model using Greg's past viral tweet hook, firmly dismissing common writer objections about stealing format. When Greg asks if users should build custom GPTs, Cole bluntly dismisses the idea as premature complexity.28:20–35:02 · Guest disagreement 1/10 Structuring Long-Form Content and Subhead Specificity Cole explains why 90% of an article's value rests in headline and subhead specificity rather than stylistic prose, contrasting vague subheads with tangible promises. Greg validates the concept, acknowledging that his intuitive 'cowboy' writing style will benefit significantly from this structured approach.35:03–49:08 · Guest disagreement 1/10 Applying the 10 Magical Ways and Finalizing the Outline Cole guides Greg through his 10 Magical Ways framework to map out distinct subheads and actionable lesson formats. Greg exhibits sharp editorial judgment by quickly flagging and rejecting an ambiguous 'me too trap' phrasing generated by ChatGPT.49:10–53:22 · Guest disagreement 1/10 The Meta-Value of Frameworks and Educational Leverage Cole argues that paying for educational frameworks provides extreme leverage when prompting AI. Greg pushes back slightly to broaden the claim to free educational media like podcasts, then cites Mr. Beast's leaked playbook on consultant hiring to reinforce the meta-learning theme.4:54–7:25 · Greg pushing back 0/10 Brainstorming the Seed Idea: Reed Hastings and Blockbuster Greg introduces a raw seed idea from his notes about Reed Hastings banning the word Blockbuster at Netflix, framing it around psychological guilt rather than late fees. Cole enthusiastically accepts the seed as an ideal starting point for the writing demonstration.7:26–12:14 · Greg pushing back 1/10 Deconstructing Ideas and Defining Content Goals Cole teaches the foundational framework for seed ideas, noting that writers must decide if a concept is the central theme or a sub-point. Greg contributes his specific audience niche expertise around startup ideas, while candidly admitting his usual writing process lacks strategic goal-setting.12:16–18:39 · Greg pushing back 1/10 Conversational Prompting and the Editor-in-Chief Dynamic Cole demonstrates conversational AI prompting and introduces the editor-in-chief metaphor to describe high-leverage taste. Greg demonstrates his own domain expertise by selecting the word 'banning' and breaking down why conflict-driven headlines perform virally on X.18:40–28:14 · Greg pushing back 2/10 Reverse-Engineering Viral Hooks with Mini Few-Shot Models Cole breaks down how to construct a few-shot mini-model using Greg's past viral tweet hook, firmly dismissing common writer objections about stealing format. When Greg asks if users should build custom GPTs, Cole bluntly dismisses the idea as premature complexity.28:20–35:02 · Greg pushing back 0/10 Structuring Long-Form Content and Subhead Specificity Cole explains why 90% of an article's value rests in headline and subhead specificity rather than stylistic prose, contrasting vague subheads with tangible promises. Greg validates the concept, acknowledging that his intuitive 'cowboy' writing style will benefit significantly from this structured approach.35:03–49:08 · Greg pushing back 2/10 Applying the 10 Magical Ways and Finalizing the Outline Cole guides Greg through his 10 Magical Ways framework to map out distinct subheads and actionable lesson formats. Greg exhibits sharp editorial judgment by quickly flagging and rejecting an ambiguous 'me too trap' phrasing generated by ChatGPT.49:10–53:22 · Greg pushing back 3/10 The Meta-Value of Frameworks and Educational Leverage Cole argues that paying for educational frameworks provides extreme leverage when prompting AI. Greg pushes back slightly to broaden the claim to free educational media like podcasts, then cites Mr. Beast's leaked playbook on consultant hiring to reinforce the meta-learning theme.

speaking balance: gold is Greg, purple is the guest (3 minute bins)

0:00 · Greg 5% · guest 95%0:00 · Greg 5% · guest 95%3:00 · Greg 28% · guest 72%3:00 · Greg 28% · guest 72%6:00 · Greg 40.5% · guest 59.5%6:00 · Greg 40.5% · guest 59.5%9:00 · Greg 36.9% · guest 63.1%9:00 · Greg 36.9% · guest 63.1%12:00 · Greg 0% · guest 100%12:00 · Greg 0% · guest 100%15:00 · Greg 6.5% · guest 93.5%15:00 · Greg 6.5% · guest 93.5%18:00 · Greg 31.9% · guest 68.1%18:00 · Greg 31.9% · guest 68.1%21:00 · Greg 10% · guest 90%21:00 · Greg 10% · guest 90%24:00 · Greg 0.3% · guest 99.7%24:00 · Greg 0.3% · guest 99.7%27:00 · Greg 7.6% · guest 92.4%27:00 · Greg 7.6% · guest 92.4%30:00 · Greg 0% · guest 100%30:00 · Greg 0% · guest 100%33:00 · Greg 39.9% · guest 60.1%33:00 · Greg 39.9% · guest 60.1%36:00 · Greg 20% · guest 80%36:00 · Greg 20% · guest 80%39:00 · Greg 31% · guest 69%39:00 · Greg 31% · guest 69%42:00 · Greg 16.3% · guest 83.7%42:00 · Greg 16.3% · guest 83.7%45:00 · Greg 0% · guest 100%45:00 · Greg 0% · guest 100%48:00 · Greg 6.7% · guest 93.3%48:00 · Greg 6.7% · guest 93.3%51:00 · Greg 62.4% · guest 37.6%51:00 · Greg 62.4% · guest 37.6%54:00 · Greg 82.2% · guest 17.8%54:00 · Greg 82.2% · guest 17.8%
Sharpest disagreement ▶ 27:24 Cole forcefully rejects custom GPTs as unnecessary distraction

Cole dismisses the trend of building custom GPTs by comparing it to an amateur driver trying to drift a Porsche 911 before learning basics.

Hardest push from Greg ▶ 51:15 Greg challenges Cole's paid-education framing

Greg pushes back against Cole's exclusive focus on paid courses by pointing out that podcasts and free content offer the exact same leverage.

Biggest teaching moment ▶ 23:15 Cole explains why writers fail by opening blank documents

Cole systematically explains that writing quality is determined by pre-writing architecture decisions, educating Greg and viewers on treating AI like an intern.

Greg holds their own ▶ 43:13 Greg catches bad AI output missed by Cole

Greg immediately spots that ChatGPT's suggestion 'ignore the me too trap' carries heavy unwanted cultural baggage, demonstrating superior editorial taste.

the scores for every segment, with the reasoning behind each
ChapterTopicGreg as informed peerGuest teachingGuest disagreementGreg pushing backWhy
Brainstorming the Seed Idea: Reed Hastings and Blockbuster 4200 Greg introduces a raw seed idea from his notes about Reed Hastings banning the word Blockbuster at Netflix, framing it around psychological guilt rather than late fees. Cole enthusiastically accepts the seed as an ideal starting point for the writing demonstration.
Deconstructing Ideas and Defining Content Goals 5511 Cole teaches the foundational framework for seed ideas, noting that writers must decide if a concept is the central theme or a sub-point. Greg contributes his specific audience niche expertise around startup ideas, while candidly admitting his usual writing process lacks strategic goal-setting.
Conversational Prompting and the Editor-in-Chief Dynamic 6611 Cole demonstrates conversational AI prompting and introduces the editor-in-chief metaphor to describe high-leverage taste. Greg demonstrates his own domain expertise by selecting the word 'banning' and breaking down why conflict-driven headlines perform virally on X.
Reverse-Engineering Viral Hooks with Mini Few-Shot Models 5732 Cole breaks down how to construct a few-shot mini-model using Greg's past viral tweet hook, firmly dismissing common writer objections about stealing format. When Greg asks if users should build custom GPTs, Cole bluntly dismisses the idea as premature complexity.
Structuring Long-Form Content and Subhead Specificity 4610 Cole explains why 90% of an article's value rests in headline and subhead specificity rather than stylistic prose, contrasting vague subheads with tangible promises. Greg validates the concept, acknowledging that his intuitive 'cowboy' writing style will benefit significantly from this structured approach.
Applying the 10 Magical Ways and Finalizing the Outline 6712 Cole guides Greg through his 10 Magical Ways framework to map out distinct subheads and actionable lesson formats. Greg exhibits sharp editorial judgment by quickly flagging and rejecting an ambiguous 'me too trap' phrasing generated by ChatGPT.
The Meta-Value of Frameworks and Educational Leverage 7413 Cole argues that paying for educational frameworks provides extreme leverage when prompting AI. Greg pushes back slightly to broaden the claim to free educational media like podcasts, then cites Mr. Beast's leaked playbook on consultant hiring to reinforce the meta-learning theme.

Statements from this episode (15)

Prediction Not checkable as stated
Cole: AI writing platforms will turn everyone into a writer
“And ChatGPT and Claude and all these AI writing platforms are going to do the same exact thing. They are going to make everyone a writer because it's reducing the barrier to entry.”
Nicolas Cole Jan 20, 2025 ▶ 2:46
Insight
Cole: AI will multiply output and income for knowledge workers
“So if you add AI into the equation, it's not that AI removes people's jobs. AI becomes the stress test of, do you know how to use this to double, triple, quadruple your output? If yes, you will make even more money than you were making before because you are m…”
Nicolas Cole Jan 20, 2025 ▶ 4:00
Assertion Contradicted
Isenberg: Reed Hastings banned the word 'Blockbuster' in early meetings
“The founder of Netflix had a strange rule in Netflix's early days. No one was allowed to say blockbuster in meetings.”
Greg Isenberg Jan 20, 2025 ▶ 5:29
Insight
Isenberg: Early Netflix competed against consumer guilt, not Blockbuster
“Netflix wasn't actually competing with Blockbuster. They were competing with guilt.”
Greg Isenberg Jan 20, 2025 ▶ 5:49
Insight
Cole: AI cannot automate tasks users do not already understand
“People expect AI to automate and do things that they don't understand how to do themselves. And so it is worth recognizing that, you know, AI is not a silver bullet. On some level, you do have to have a baseline understanding of how the thing is supposed to wo…”
Nicolas Cole Jan 20, 2025 ▶ 7:27
Insight
Cole: Struggling writers lack process, not talent
“The majority of people don't have a talent problem. They have a thing called, I don't want to practice and execute the basics, and I would rather, no offense, drink a bunch of coffee, close my eyes, and hope that something smart comes out when I hit the keyboa…”
Nicolas Cole Jan 20, 2025 ▶ 11:52
Insight
Cole: Drafting in a blank document without planning leads to failure
“The thing that most writers and most people who want to write don't realize is that many of the decisions get made before you start writing. So if you start, if you come up with an idea, and your first thought is I should open a Google Doc and I should start w…”
Nicolas Cole Jan 20, 2025 ▶ 15:10
Insight
Cole: AI elevates writers to taste-driven editors-in-chief
“Instead of me passing this along to my junior columnist, I'm going to feed this to ChatGPT, which means we just got to elevate ourselves up to editor in chief, where our primary value is taste.”
Nicolas Cole Jan 20, 2025 ▶ 17:18
Insight
Cole: Writing intuition is just subconscious pattern matching
“Every time you sit down to write, your brain is cycling through previous quote unquote templates that you have seen work. Either something you've read and were inspired by, or something that you've done already and you received a positive response from.”
Nicolas Cole Jan 20, 2025 ▶ 21:08
Prediction Not checkable as stated
Cole: Writers who do not learn AI orchestration will lose jobs
“Either you get on board with that and you become an orchestrator and editor in chief, or you don't have a job anymore.”
Nicolas Cole Jan 20, 2025 ▶ 26:59
Opinion
Cole: Base ChatGPT and Claude make custom GPTs unnecessary
“The ChatGPT and Claude are so powerful that you do not need much more than this for a long time.”
Nicolas Cole Jan 20, 2025 ▶ 27:47
Insight
Cole: 90% of an article's value is the headline and main points
“And this is something that took me a really long time to understand, is that 90% of the value of a piece is the headline and the quality of the three main points.”
Nicolas Cole Jan 20, 2025 ▶ 29:39
Insight
Cole: 99% of nonfiction writing maps to 10 tangible formats
“You could reverse engineer 99% of all nonfiction writing back into these 10 ways.”
Nicolas Cole Jan 20, 2025 ▶ 35:15
Insight
Cole: Writers should ideate first before refining with AI
“I think that you should always try yourself first and then weigh it against what ChatGPT comes up with and then improve the quality together. That is why I don't defer the writing to AI. I write with AI”
Nicolas Cole Jan 20, 2025 ▶ 39:57
Insight
Cole: Sentence-style subheadings retain readers better than clever titles
“I've always found that the best main points and subheads are actually more like sentences.”
Nicolas Cole Jan 20, 2025 ▶ 44:15
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.