9 people × 1 hour
Approximately $419 in direct average compensation every week.
An anonymous AI lab
Companies have
too many humans.
Every additional employee adds salary, meetings, management and another place for ownership to disappear. I’m using facts, math and working AI systems to prove that companies can operate with fewer human dependencies.
See the evidence ↘Fewer peopleBetter systemsControlled AI costsVerified outcomes
Experiment 001 · Modeled
The $21,000
status meeting.
Nine people spend one hour every week repeating updates, discovering one blocker and leaving without a recorded decision.
No decision
1 verified update
One blocker, one accountable owner and one decision routed to the right person.
This is a modeled system—not a live result. The annual estimate uses 50 weekly meetings and the March 2026 U.S. private-industry average compensation cost of $46.60 per hour.
Open the experiment →Per My Last Breakdown · 001
The deadline was urgent until someone else had to do something.
“By Friday, I had produced nothing except evidence that I had asked other people to produce something.”
I followed up with people who ignored the original request. Then I followed up with their managers, who asked me to summarize the request they were copied on. Eventually, following up stopped being something I did and became my actual job.
Read the blog →The belief
People are inconsistent.
Systems should not be.
Payroll is a design choice.
Before adding a person, calculate whether the workflow should exist in its current form.
Coordination is a tax.
Every handoff adds waiting, interpretation and another opportunity for ownership to disappear.
AI spending needs an owner.
Giving every employee an AI account is not the same as designing an AI-native company.
Confidentially employed
I’m building the company I wish I could work for.
I still work inside a human-heavy company. I attend the meetings, navigate the handoffs and watch potentially transformative technology get absorbed into the same operating model that made it necessary.
Zero Meeting Labs is my anonymous field log from inside that system—and my attempt to design its replacement.
On the side, I’m building toward a company where AI performs the recurring work, software coordinates it and humans enter only when judgment, authority, trust or accountability makes them necessary.
This is not a campaign to give every employee an AI assistant. It is an experiment in what happens when a company is designed for machine execution from the beginning.
The stories document the problem. The calculations expose its cost. The experiments build what should exist instead.
Yours in unnecessary meetings,
Z