Getting Back to the Future

What’s Happening?

The Idempotent Trap:
Why Safe Routines Keep You Exactly Where You Are

I've been so far down the AI rabbit hole — coding, agents, development — that a word I'd never heard before has wormed its way into my daily vocabulary.

The word is idempotent.

Here's what it means: an action you can repeat as many times as you want, and the result never changes past the first time. Do it once, and you get an outcome. Do it ten times, and you get the exact same outcome.

Think of an elevator button. You walk up, you press it, it lights up. Smash it twenty more times, and nothing comes faster. The elevator doesn't care. The state is locked.

You actually want this in parts of your work. You call ten referral partners. Three call back. Not 100% — but reliable. Over time, you learn that ten calls buy you three conversations, and you build on that.

In software, that's idempotency. Repeat the action forever; the result never moves after the first time. Perfectly safe. Zero side effects.

It's also exactly how we can run their businesses — and our lives.

We optimize for idempotent acts because they feel safe.

So I've been exploring the opposite: AI as an asymmetric action. A question I started asking - What if you could spend 10% of your time on something that is the furthest thing from idempotent — something that could return 10x or 100x?

The Math of Stagnation

Look at your calendar. You check your email. You run the weekly sync. You nudge the ad to spend 2%. You post the daily graphic.

All idempotent. All status quo. All baseline protection.

Here's the brutal part: you cannot out-work a system built to stay the same.

If every action you take is idempotent, you will not grow 10% or 50% this year. Doing a safe thing 1,000 times does not create a 1,000x return. It creates a beautifully polished version of exactly where you already are.

Safe is comfortable. Safe feels like work. Safe is an illusion.

Asymmetric Layering

If you want exponential growth, you have to break the loop. You have to hunt for asymmetry.

An asymmetric bet is a different piece of code entirely. Massive uncapped upside. Strictly capped downside.

You want your losses to be idempotent. You want your wins to be explosive.

The idempotent loss: You cold-email a dream referral partner. They ignore you. Cost? Three minutes of typing. Your business doesn't shrink. Your bank account doesn't move. You hit a button, the system ignores it, and you're exactly where you started.

The asymmetric win: They reply. They sign. Your monthly recurring revenue doubles. One action. Infinite cascade.

You need the idempotent routines to keep the lights on. But if you never allocate 10% of your energy to swinging for asymmetric wins, you're just pedaling a perfectly optimized stationary bike.

The framework: stop optimizing for predictability. Protect your baseline — then spend your real currency, your focus, on the few things that can actually change the numerator.

Just a thought. Take it for what it means to you.

Ship a really raw piece of content. Pitch the impossible partner. Build the feature that might break. If it fails, you reset to baseline — you've still got 90% of your days running the normal routine. If it hits, everything changes.

That's why I've been quiet. For six months, I've gone deep on AI and bet on the big swing. I stole time from this newsletter and a few other things to make it happen. It's happening.

Find your big swing in 2026 — the asymmetric one — that builds a 2027 you can't yet imagine.

REAL ESTATE, INVESTING AND WEALTH
SOME THINGS THAT STOOD OUT?

ON BEING HUMAN
What’s Worth Sharing?

DOPAMEMES
And Other Happy Moments…

AI
and The Future of Work…

I’m on this chart, I guess, far right…

Borrow Smart University || National Institute of Financial Education