Degree 3 · Unit 3.1
From command to commission
When you give a new employee a piece of work, you do not simply say "write a report". You tell them who they are in this particular task, who the work is for, what material you already have, what to avoid, and how the two of you will know when the work is good. Then you read their draft and send it back with notes.
Working with an AI system is exactly that. And the thing people complain about most — that the results come back general and shallow — happens because they handed it a question and then expected the work of a consultant. The output never exceeds the context you put in. That is the first rule of this degree, and there is no exception to it.
A comparison
WEAK PROMPT
"Write me a policy for using artificial intelligence."
What you get back is a general text that would suit any organisation anywhere in the world — which is another way of saying it suits yours not at all.
PROFESSIONAL BRIEF
"You are an expert governance consultant. Our organisation is an engineering consultancy, 40 employees, handling confidential client drawings, and bound by contracts that forbid sharing client data with third parties. Write a two-page AI use policy, divided into: permitted, forbidden, requires approval. Executive language with no padding, and do not name any tools. Constraint: every clause must be verifiable as complied with. Standard: a new engineer understands it in ten minutes and works by it without asking. And if you lack necessary context, ask me before you write."
That last sentence — "if you lack necessary context, ask me before you write" — is one of the most useful things you can add to any commission. It turns the system from a blind producer into something closer to a colleague you are talking to.
The quality of the output is a direct reflection of the quality of the input. Four ingredients cover most cases.
fawzooz.ai
One example holds them all: "write a summary of the main challenges of solar power in desert regions, in a simplified academic style for university students, in a paragraph of no more than 150 words". Delete the context and what is left is a general request whose answer suits anyone — that is, no one.
The review: where quality is made
A beginner writes a commission and takes whatever comes back. Someone who has mastered this writes a commission and treats the answer as the place where the work starts. Three moves will double the quality of what you get. Ask for alternatives. Ask for criticism — "criticise your own answer the way a seasoned opponent would". And ask for narrowing — "delete every sentence that adds no information". Then close, every time, with the strongest question of the four: "what might be wrong in your answer?"
Do this
1 — On the tool. Take the last question you asked an AI system and rewrite it in the five layers. Compare the two outputs side by side.
2 — On the tool. Ask for an answer, then ask for its criticism, then ask for the corrected version. Record: how much better is the third than the first?
3 — In your field. Write one commission template for the task that recurs most for you, and save it in a file called "my templates". You will add to it in every unit after this.
Where to after this unit? You have mastered the single commission. The next unit widens it into a broader skill: what you put in, in what order, and what you withhold.
