The Homework · Experiment 001

Jay gave me homework.
This is what I built.

He told me to simplify the business, prove it works, and get it to $10K/month.
So I turned it into one measurable offer.

The offer

Give me $1,000.
I’ll learn how your business actually works and help you identify and implement at least $1,000 of measurable leverage.
Measured as time back + revenue in.

No free strategy call. The calibration is the product.

START HERE

The conversation that created the assignment.

THE HOMEWORK Add a YouTube ID in config.js to embed the trailer here.
THE PROBLEM

AI was supposed to save you time.

Instead, a lot of businesses got more content, more automations, more dashboards, more tools, and more things to maintain.

You got sprinklers.
Then you planted a bigger field.

I’m not interested in helping you do more work with AI. I want to find the work your business shouldn’t need anymore.

TIME BACK↑
REVENUE IN↑
HUMAN WORK REQUIRED↓
HOW IT WORKS

I look for leverage, not more software.

01

Learn the system

I learn where money comes from, where your attention goes, what repeats, and what keeps finding its way back to you.

02

Set the scoreboard

We agree on a conservative value for your time and define what counts as real economic improvement before we start counting it.

03

Find the leverage point

We look for work that can be removed, compressed, delegated, automated, or made unnecessary without hurting the outcome.

04

Test reality

We run the smallest useful experiment. Reality grades us. If it works, we preserve the learning so neither of us has to rediscover it.

THE SCOREBOARD

$1,000 of leverage has to mean something.

We agree on the measurement during onboarding. No retroactive funny math.

new / recovered revenue + recurring time returned × agreed time value = measured leverage

Example: your time is conservatively valued at $100/hour. We eliminate 5 recurring hours of work and create $600 in additional revenue.

$500 time back + $600 revenue = $1,100 leverage
WHY LISTEN TO ME

Jay caught something I was underusing: my own proof.

~15 yearsbuilding software systems
Early AImachine-learning work before the current LLM wave
Copy.aiearly engineer before ChatGPT changed the market
$1Kfirst paid calibration already underway
“The first skepticism is: do you know AI? … The second skepticism is: okay, yes, you know AI, but how do you turn it into a business?” — Jay Yang, during our call

That second proof is what I’m building now. Publicly.

THE PUBLIC EXPERIMENT

Jay told me to get the business to $10K/month.

I’m making the homework harder.

$100,000

I’m testing whether the ideas and content created from this conversation can generate 100 × $1,000 calibrations in directly attributable business.

The point isn’t the stunt. The point is to make the business itself the proof.

“I would want to get this AI business to at least 10K a month … because then you can show that this process works yourself.” — Jay Yang
WHO THIS IS FOR

You already have something real to optimize.

Good fit

  • You have customers, revenue, or a real audience.
  • Your business has recurring work and decisions.
  • You are already experimenting with AI but are not sure where it creates ROI.
  • You want more economic output without making your workload bigger.

Probably not yet

  • You mainly want a list of AI tools.
  • You want automation without questioning whether the work should exist.
  • You do not yet have an operating business to learn from.
  • You want more output simply for the sake of producing more.
THE CALIBRATION

$1,000 in.
At least $1,000 of agreed measurable leverage out.

I keep working through the calibration until we’ve identified and implemented at least $1,000 in mutually agreed value through time returned and/or revenue created.

INITIAL CALIBRATION
$1,000

Then, if we both want to continue:

$250/month Continuous leverage

Why does the price go down? Because you shouldn’t keep paying discovery prices for something we’ve already learned.

Start my calibration — $1,000

Requires reasonable access, participation, and agreement on how value is measured. This is a collaborative calibration, not an unconditional revenue guarantee.

FAQ

Questions I’d ask before buying.

Is this an AI automation service?

Sometimes automation may be the answer. But if the better answer is to stop doing the work entirely, we do that instead. The target is leverage, not automation count.

Do I have to learn or use Marvin?

No. You work with me. My AI system works behind me to preserve context, search for patterns, prepare experiments, and reduce repeated work.

Why not just use ChatGPT?

You should use any tools that help you. The scarce part is increasingly not access to intelligence. It is deciding what deserves attention, integrating it into your actual business, and measuring whether reality improved.

How long does calibration take?

I’m selling a measured outcome rather than a fixed number of consulting hours. Different businesses reveal their leverage at different speeds.

What happens after the first $1,000 of leverage?

Calibration is complete. Stop there, or continue at $250/month if the system has earned the right to keep looking for leverage.

THE POINT

The goal isn’t to make you more productive.

It’s to make productivity less necessary.

I want to find the work that can disappear without the result disappearing with it.

Start my $1,000 calibration

Add your checkout URL in config.js.

Not operating a business yet? Follow the Build In Public University experiment instead →