AI Employee Training
Build Your
AI Employees
+ Consultants
One person, a whole team, plus a panel of advisors
No coding required. All you have to do is speak.
Sunday · Three-hour hands-on workshop
This is a training ground, not a lecture hall
What matters today is not how much you understand, but that you build your own AI employee right here.
Operating is the AI's job, not yours.
What in your current workflow is eating you alive?
- ✕Files scattered everywhere. Just finding them takes half the day.
- ✕Repetitive documents that you rebuild from scratch every time
- ✕Appointments and messages that slip through the cracks
- ✕Busywork consuming the time that should go to real work
- ✕Major decisions made in a vacuum, going in circles alone
Right now, who is doing all of this? (Raise your hand.)
Steel was not invented to reinforce wooden houses
When steel first appeared, people used it as nails to build wooden houses one floor taller.
Skyscrapers only became possible once someone designed directly in steel from the start.
99% of people use AI to patch their existing workflows.
The real question to ask: what tasks should have been handed to AI entirely from the beginning?
Do not use AI to do what software already does well.
You need two kinds of AI: Employees and Consultants
They work the way you work, replicate what you know, and get it done faster.
They apply an external perspective to challenge your blind spots and bring angles you lack, helping you think more clearly.
No team to manage? Start with consultants. Anyone can build those.
Here is how I use AI to get things done
You say one sentence to the director, and the right person gets assigned. You talk, the team works.
Mika is the one installed inside your laptop
You speak
Say one sentence, assign the task
Mika dispatches
Routes to the right employee
Employee takes over
The right person does the right thing
Result comes back
You confirm, and it's done
In the first half of today, we will get Mika running on your own laptop.
When today ends, you walk out with your own team
They are built from your habits and your language.
Hand over one thing you used to do yourself
- ①Assign it, do not do it yourself first
- ②Resist the urge to take over and operate it manually
- ③Watch it work
- ④If it gets something wrong, add a rule rather than rewriting it yourself
- ⑤Once it succeeds, save the instruction as a reusable template
Imperfect the first time is perfectly normal. That is what training looks like.
An AI employee works your way, but only once you surface what your way actually is
- ①Every rule you just added was you surfacing an unspoken judgment from inside your head.
- ②This is tacit knowledge distillation: taking the calls, the nuances, and the standards you know how to apply but cannot quite articulate, and turning them into rules an AI can follow.
- ③The more you distill, the more the AI employee resembles you. And no one can take that away. Those are your judgments, not the tool's.
Organizing is not the same as distilling
Collecting facts that were always expressible
A client in inclusive education: five years of tour logs, stakeholder interviews, route maps, wheelchair accessibility notes. Organizing these is useful, but the information was already there.
Surfacing the judgment buried inside decisions
She knows which advisor's words to trust and which ones to set aside. Distilling that judgment into a rule is what lets an AI help her make decisions.
Organizing collects what is already explicit. Distilling surfaces what is still tacit. Only the latter is something no one else can replicate.
Tacit Knowledge Distillation System
Every layer deeper you go unlocks automation for everything above it: exponential leverage
One AI employee is one node. Connect them and you have a workflow.
- ①Each individual employee handles one thing. The real power comes from chaining several nodes into a single pipeline.
- ②For example: Research Lead Feng gathers the data, Editor Chun shapes it into a draft, Secretary Bee schedules and sends it out. One sentence starts it. One pipeline runs it end to end.
- ③Today you will first build one solid employee. Connecting them into a workflow is the next step you take home.
Bring a decision you are sitting on right now
For example: should I take this project or not?
What all three flag is your real blind spot.
The hardest and most valuable ability:
knowing what a problem is worth
All you need to know: what problem, who will pay, and what it is worth right now.
Not something to remember. Something to use tomorrow.
Within the next 24 hours,
what is the first real task you will hand to your AI employee?
Write it down. Only then is it a commitment.
Ready to step out of the operator role?
Every month I host free online talks on exactly this: how to move from being the one who operates to the one who assigns and designs. From the employee framework and tacit knowledge distillation to workflow design, I walk you through building your own AI team step by step.
Two free sessions every month.
Join the LINE Community ↗