3 questions founders actually ask, each with a straight answer and the resources worth your time.
As a founder, content is one of your cheapest growth channels, but writing consistently is exactly the kind of work that gets dropped when you're busy, AI removes the blank page and turns a 4-hour blog post into a 45-minute editing job. The catch: readers (and algorithms) now punish generic 'AI voice', so the winning setup is you supplying the ideas, stories, and opinions while AI drafts and polishes. Think of yourself as editor-in-chief and AI as your junior writer, then ruthlessly strip out the tells like 'delve', 'game-changer', and the endless 'It's not X, it's Y' constructions.
One of the world's most-read writers explains the 'editor-in-chief, AI as junior columnist' system, the clearest case for why founders should write with AI and how to keep the judgment human.
Open youtube.com →A concrete checklist of AI fingerprints (contrast framing, rule-of-three overuse, vague filler) with fixes and prompts to edit them out.
Open writewithai.substack.com →A working editor's step-by-step revision system for de-slopping AI drafts, including the actual prompt template his team uses, free and immediately usable.
Open louisbouchard.substack.com →The pattern top creators follow: capture your thinking once, a voice note, a weekly essay, a podcast, then use AI to 'atomize' that one idea into platform-native pieces: a newsletter, five LinkedIn posts, a thread, a YouTube script, short-video hooks. Dan Koe is the best public example: he spends a couple of hours each morning with ChatGPT and Claude acting as his research and strategy team, and that feeds millions of impressions a week across platforms. The key is that AI handles reformatting and volume, while the original idea and point of view stay 100% yours, that's what stops the engine producing spam.
Dan Koe screen-shares his actual daily AI content engine, research, ideation, drafting, repurposing, the single best over-the-shoulder look at a working founder content machine.
Open youtube.com →The 5-minute distilled version of the same workflow, good to skim before committing to the full video.
Open x.com →Koe in his own words: treating AI as a 'digital employee', building a reusable prompt library, and wiring it into a content and business workflow.
Open letters.thedankoe.com →A free, ready-to-clone implementation of Dan Koe's framework, six skills covering ideation, drafting, titles, and tweets, if you want to run the engine instead of just watching it.
Open github.com →You don't 'train' the model, you feed it evidence of how you write. Collect 10-20 of your best posts or emails, ask the AI to analyze them and produce a 'voice guide' (tone, sentence rhythm, pet phrases, words you'd never use), then save that guide as standing instructions in a Claude Project, custom GPT, or skill so every draft starts from your voice. From there it's iteration: every time output sounds off, tell it why and update the guide, within a few weeks the AI's first drafts need only light edits.
One of LinkedIn's biggest AI-writing practitioners shares his exact method for getting AI to mimic your voice, short, copyable, battle-tested on millions of impressions.
Open linkedin.com →Step-by-step: export your past LinkedIn posts, feed them to Claude, and build a reusable skill that writes new posts in your proven voice and patterns, free.
Open ruben.substack.com →A real operator's full setup, identity file, anti-AI style guide, and a custom LinkedIn skill, with his actual config files shared as free GitHub gists.
Open ranthebuilder.cloud →
A practical walkthrough of writing Claude Project instructions so the voice setup persists across every chat, aimed at non-technical business owners.
Open youtube.com →