This week's prompt will ask you to revisit past improvement discussions and report back to your GPT. Then get the next tweak, and report back in another week.
The Army calls these After Action Reports, and the military runs these after every mission. Talking through expectations and reality is a great method of iteration with your AI rather than starting every conversation from scratch.
Your Prompt
Here's an update on the process we discussed
What I changed: ______.
What actually improved: ______.
What still feels forced or ignored: ______.
What’s the smallest next adjustment to this same process that I should test next week?
Assume I’m optimizing for compounding capability, not quick wins.
Pushing Against The Grain
There’s a progress distortion field around AI right now.
You scroll LinkedIn and someone claims they’ve fully automated their firm.
You watch a YouTube breakdown of a founder’s “AI stack” that looks like it was designed by NASA.
A podcast guest casually mentions replacing three roles with workflows you’ve never heard of.
Meanwhile you… built one solid GPT. Or automated meeting summaries. Or tightened up a reporting process.
And it feels small. Useless.
Is there even a point in trying?
Your feed is not the market. Algorithms elevate ambition. When you surround yourself with operators you admire, it’s easy to forget how many businesses are still figuring out the basics you’ve already solved.
Step outside LinkedIn and the adoption curve changes. Most businesses still haven’t integrated AI into a core, repeatable workflow - especially across a whole team. Many haven’t tied it to a real business process at all. Of those experimenting, a lot are using it sporadically, once or twice a month. A loud minority are pushing hard. A tiny fraction are integrated deeply enough that it changes how decisions get made.
Doing anything thoughtful and repeatable still places you near the front of the curve.
The danger right now is confusing visibility with capability.
In powerlifting, we talk about "show muscles" and "go muscles".
“Show muscles” - biceps, abs, and quads - are out front. They make you look strong, build the Instagram following.
“Go muscles” are the triceps, back, and hamstrings. They're literally behind the scenes, doing the heavy lifting.
It doesn't feel like it, but implementing AI is a long game. New models will keep releasing. Interfaces will get cleaner. Use cases will become obvious in hindsight. But if you haven’t built the muscle of working with these systems, the obvious adoption won’t feel obvious when it arrives.
So get in the reps.
Not because today’s workflow is perfect. It won’t be.
Not because this version of the model is final. It isn’t.
Reps matter because you are training yourself while the tools improve.
Think about AI transcribers like Fireflies or Zoom AI Companion. They slide effortlessly into scheduled video calls. Clean workflow. Automatic summaries. Searchable history.
But if your company still runs on impromptu landline emergencies and drop ins, that same tool feels awkward. Clunky. Often forgotten.
The technology may be ready. Your workflow might not be.
Set your meeting policy before the AI transcribers arrive.
Structure and sync your company calendars before an assistant starts auto blocking your week.
Build your AP process on OCR technology and approvals before advanced reasoning shows up.
Small, steady experiments shape your operating system so that when the right tool appears, you’re structurally (and mentally) prepared to absorb it.
One clean experiment that saves two hours a month compounds.
One decision you improve with structured AI-first thinking compounds.
One tip shared within your team per week compounds.
Speed is seductive. It gets likes and views and speaking engagements.
Capacity is durable.
This isn’t a call to crawl forever. It’s a call to build rhythm.
Reading about AI is not training. Talking about AI is not training. Running one small experiment inside your business is training.
And when the breakthrough feature drops, the companies who have been training will recognize it… and the ones waiting for the perfect moment will start training.
When your strength is called upon, it's obvious who's been showing up consistently.
So put in the work.
Track what changes.
Stack the next rep.
Disruption does not reward panic. It rewards consistent adaptation.
Quiet compounding still works… even when the internet insists you’re too late.
