18 Hours of AI That Bought Me Back 500 Hours This Year

I spent one weekend to build five AI processes and I'm breaking down the time math for you.

One weekend, I spent 18 hours building five Substack AI systems, and they are projected to buy back more than 500 hours this year.

Real quick: The AI Leverage Lab opens August 3rd, and this article is the clearest example of what it’s built to do, which is spend a little time now and get back an enormous amount later. Now, let me show you the weekend that changed my whole week, recurring!

I know how that sounds, so let me put the real math right up front before you decide I’m doing the inflated-number thing that everybody in this space does lol. One Substack article (like the one you’re reading right now, so meta am I right?!) used to take me about 6 hours start to finish. I run two Substack publications, this one and my lifestyle one, More Life, and I will be publishing somewhere between 12 and 16 articles a month across them both, so at 6 hours a pop that’s basically a part-time job just to get the words down. And that was the problem I sat down to solve, and the math at the bottom is going to walk you all the way to 500+ per year. Stay with me.

The Things Almost No One Does:

Here’s the thing almost nobody actually does: they will not spend time to save time. They’re so buried in the doing that they never stop long enough to build the thing that ends the doing. That’s the whole busy trap right there. That’s exactly why they stay busy and burned out forever and never scale, because scaling requires you to step out of the work for a minute and build the machine, and stepping out feels impossible when you’re drowning… so they keep swimming, and they never build the boat.

The ones who scale aren’t smarter than you… They just stopped long enough to build the thing, and you haven’t yet. That’s the whole advantage: spending the time up front when almost nobody else will.

I almost didn’t build any of this, actually. It was a whole weekend, give or take: six hours Friday, six Saturday, six Sunday, and that is not how I wanted to spend those three days. I did it anyway, and every week after has been (and will be) better for it, so I’m not going to pretend the up-front time feels good, because it doesn’t; it’s just the thing that’s worth doing. And side note… I made it fun… I went to Whole Foods, grabbed a few fun snacks, treats, and some kombucha, turned on my Brain Food playlist and just rocked it out!

Why you think AI can’t do this (and why you’re wrong):

Okay, so here’s where I want to push on you a little, because I think most entrepreneurs have already decided what AI can do for them, and they decided it after about 16 minutes of trying.

Here’s what usually happens… You open ChatGPT, you type “write me a post about X,” you get back something that sounds like a LinkedIn robot wrote it during a corporate training, and you close the tab. You conclude that AI writing sounds like nobody, that it can’t capture your voice, and that using it would make your brand generic, so you go back to doing everything by hand.

I get it, and honestly you’re right about that output; it is crazy generic.

But that’s just the front door of what AI is.

And listen, I wasn’t ahead of you on this. I didn’t think AI could run repeatable processes like this at all. I definitely didn’t know it could go out onto the web or into my actual computer and do things for me. For a long while, I just treated it like a chat box you asked questions to, same as you probably do right now.

What you did in that 16 minutes is Level 1 of my 4 Levels of AI Leverage framework:

Level 1: Prompting

You typed a request into a blank box and hoped. Level 1 is the rung almost everyone lives on, and it’s also the rung that produces the exact bad output that convinces people to quit or to use it endlessly until you finally get a good output. You asked a tool that knows nothing about you to sound like you, so of course it didn’t lol.

Level 2: Advisory

This is where AI knows your business well enough that you stop re-explaining yourself every single time you open it.

Level 3: Acceleration

This is where you’ve built repeatable processes, and the work moves at a completely different speed.

Level 4: Automation

And this is where the thing runs without you touching it.

What I built that weekend is Level 3. And the gap between the Level 1 experience that made you give up and the Level 3 system that gives me back 500+ hours a year is not talent, and it’s not some technical background I have that you don’t.

It’s setup time, and that’s a gap that keeps most entrepreneurs stuck in prompting and re-prompting.

I walk through all four levels and how to build in each one inside the AI Leverage Lab, because once you see the ladder you can’t unsee how much room there is above where you’re standing. It’s seriously so exciting!

First things first, why I stay the brain:

Real talk though, here’s what this was never about. It was never about cranking out more content (aka harnessing systems to do more instead of less). I built this to protect my integrity, my creativity, my brain, and my voice. Because here’s what freaks people out about AI, and they’re super right to be careful: they think leverage means handing over the thinking. I never hand over the thinking. I stay the brain at every single decision point in my personal brand, and the AI only does the grunt work in the gaps between my decisions. I hand over the labor, and I keep the judgment. Those are two completely different things, and the whole system lives on that line.

---

---

I stay the brain for two reasons that actually matter to me.

1. The first is that everything you put out there represents you, so the second I publish something I don’t fully stand behind, that’s my whole reputation on the line, and I want people following me for my brain, not for the AI’s.

2. The second is that I just started this Substack, so these articles are literally the first time I’m running this system live, and without it I’d have been stuck with two options I hated, either post way fewer articles right when I’m trying to warm up a brand new audience with real value, or work a whole lot more, and I wasn’t about to do either one.

Okay, so let me tell you what I actually built, because I want you to see the machine.

The AI x Substack System I Built:

Before I built any of it, I had AI go study the top Substack creators and hand me back a report on what’s actually working right now. Not what worked in 2022, but what’s working this month, and all other systems got built on top of that report.

I have these run in a very specific order, and they are all designed to keep my brain, voice, and integrity.

1. It starts with a Substack Advisor, loaded with Substack strategy that’s specialized to my business, so when I ask it something, the strategy I get back is mine, not the internet’s average take that already lives in every other creator’s head.

2. An article builder that interviews me instead of writing for me. I talk my ideas out loud using Wispr Flow (you can get a free month of Wispr Flow Pro right here), and it pulls the article out of my head.

3. A refiner that sharpens the draft, hunts for the spots where a smart reader would get confused, and checks every line against my actual voice.

4. A pre-publish audit that handles SEO, AEO, and GEO, so the piece gets found by search engines and by the AI search tools people use instead of Google now. This one replaced a $50/month SEO subscription that I cancelled.

5. A promoter that turns the one article into a stack of content pointing back to it.

If you want the actual build to these systems, I wrote up the complete teardown for paid subscribers: the exact tools I used, what's inside each process, the research prompt I started with, my real minute-by-minute workflow, and the mistakes I had to go back and fix:

---

---

Here’s the principle underneath all of it, and this is the part I want you to really understand even if you forget everything else. If you do something more than twice, it should be a process. Build it once, run it the same way forever, and stay the brain at every decision point so it never stops sounding like you. Fun fact: this exact article ran through the system you just read about… I mean, of course!

The real time-saving numbers:

The time invested: 18 hours, 1 weekend, 5 processes.

Each article used to take 6 hours and now takes 2 to 3, so I’m saving 3 to 4 hours every single time. Run that against 12 to 16 articles a month and the savings land somewhere between 430 and 770 hours a year, depending on how heavy the month is. A normal month sits around 575. I say “over 500” because it’s the round number I can defend on my slowest month without stretching, and I’d rather hand you the real range than the biggest number I could get away with.

A Note: While I spent 18 hours building the initial systems, all systems need to change, evolve, and be updated over time. I’ve already spent about 45 minutes refining some of the processes as I’ve worked through them in real time. Still overall a major time saver, but I want to remind you that once systems are set up, they’re not set it and forget it. If you evolve, your business changes, your processes change, you find hiccups or leaks in the process, it needs to be updated, period.

Where the hours saved actually go:

Here’s the deeper meaning behind this though, and it’s the reason I did any of this.

The hours don’t go back into more work; that was never the goal. They go into more life. The numbers were never the “prize” of productivity… the life the time buys back is the prize. That’s the More Life throughline in everything I teach: you leverage the tool so you can work less and live more, not so you can grind harder with the time you just freed up. If you save 500 hours and pour all 500 straight back into the business, you didn’t build leverage; you built a bigger jail cell.

So build the boat, spend the weekend, and be the person who stops swimming long enough to make everything after it easier. Ok?!

If you built one process this month, just one, what would it be? Tell me in the comments, and I’ll tell you where to start. ↓

---

Get on the AI Leverage Lab Waitlist ↓

The AI Leverage Lab is a live, 8-week cohort where we build your custom AI system together, one level at a time, so you walk out with AI that sounds like you and runs the busywork. This is the first live round, and I won’t always run them live. Get on the waitlist, and you’re first in line when doors open August 3rd.

---