Most people think newsletters are either fully automated garbage or handcrafted artisan content. The automated ones scrape headlines and spit out summaries. Nobody reads them. The handcrafted ones are great but don't scale past one person working 40 hours a week.
Neither works if you're trying to run multiple ventures and stay sane.
Last year I gave myself an exploration year. Work alone, test a bunch of business ideas, see what stuck. The constraint was bandwidth. How many ideas could I actually validate in 12 months while also writing a weekly newsletter that doesn't suck?
That constraint pushed me into AI. Not because I'm an AI maximalist. Because I wanted to see if I could create what basically amounts to a three-person team while working solo.
Turns out you can. But not the way most people think.
The stuff I build and break in practice becomes what I write about here. Good feedback loop. But there's still only 24 hours in a day, and at some point you have to choose between shipping content and shipping products.
Unless you build a system that compresses the gap between rough idea and published draft.
Here's how mine works.
Idea generation happens in two modes. First, I have a research assistant that pulls pulse content from Reddit and X once a week. Surfaces what's resonating in the AI operator space. Keeps me plugged in without doomscrolling for three hours.
Second is the more interesting part. Every week I block time for what I call an interview session. I use a prompt that asks me about what I've been working on, who I've talked to, what problems came up. It keeps asking follow-up questions, one at a time, pursuing different threads until I say stop.
I'll sit down for 20 minutes with coffee and answer the questions out loud using Whispr Flow. Then it summarizes everything into a structured list organized by content pillar.
Reading that list is usually when something clicks. One of those bullets becomes the seed for the next post.
Here's the prompt:
Knowing what you know about me and my content ambitions, I want you to interview me one question at a time.
Ask follow-up questions where it makes sense.
Interview me about what I've done over the past week: what I've worked on, projects, things I've learned, challenges, things that didn't go as planned, people I've met, what I've talked with them about, things I like, knowledge that I've gathered. Help me mine for content ideas.
You'll ask me one question at a time until I say that we can stop. When we stop, your next reply to me should be a structured list of content ideas organized according to our content pillar or content buckets. Once you have given me that list, you should be ready to start over questioning from the top the next time that I activate the conversation.
Do you understand?And here’s an example of what the output looks like:
Once I have the seed, I open Whispr Flow again and just talk. Challenge myself on the framing, the practical value, why this matters from my vantage point. What comes out is messy. A few pages of rambling, half-formed thoughts, no polish.
This is where it gets interesting.
I drop that raw dump into a Claude project I've been refining for months. My newsletter copywriter. It's got business context for Inside Track, the full ICP, example posts, detailed tone of voice instructions.
And here's the part that matters more than people realize: a long list of things to avoid.
Telling an AI what you don't want is as important as telling it what you do want. Maybe more important.
I don't want em dashes. I don't want "the kicker." I don't want strawman arguments or saggy conclusions that recap everything. I don't want correlative constructions like "it's not X, it's Y." I don't want rhetorical questions as filler.
Every week I tweak the prompt based on what lands and what doesn't. It's a living document that gets sharper over time. The AI learns to write more like me by learning what I definitely won't say.
What this actually buys me: I can go from a 10-minute voice note to a ready-to-edit draft in about 30 minutes. That's the difference between publishing once a week and publishing never.
But the AI isn't writing the newsletter. I am.
The AI is compressing the gap between rough idea and structured draft. Handling the grunt work of organizing, tightening, translating my rambling into something readable.
Without this system I'd be choosing between the newsletter and everything else. With it, I can do both.
If you're juggling multiple projects and trying to scale output without hiring a team, steal the prompts. The system works because it's tuned to my voice and workflow, but the structure transfers.
The businesses winning at content aren't choosing between AI and authenticity. They're compressing the time between having something worth saying and actually saying it. That's the game.
Reader Question: "I'm a consultant. How do I position AI as part of my service offering without sounding like every other person who just discovered ChatGPT?" – Maria T.
Don't sell AI. Sell the outcome AI enables you to deliver.
Everyone's pitching "AI consulting" or "AI strategy." That's noise. Instead, lead with the business problem. "I help law firms cut proposal turnaround from 5 days to 5 hours." "I help agencies onboard clients without burning 8 hours per kickoff."
Then, when they ask how, you casually mention the AI-powered systems you've built. The AI is your competitive advantage, not your marketing message.
Clients don't buy tools. They buy solutions to expensive problems.
If you have ideas you want me to write about, hit reply!
Until next week,
Martin
PS: Go connect with me on LinkedIn and say hi! It’s always fun to chat with readers
PS2: Want to kickstart you business’ AI transformation? Fill out this interest form and my consultancy will be in touch asap for a free discovery call.