How we invite AI to surface gaps, ambiguities, and unstated dependencies before it begins the work

_{—Part of CSS Primer for the AI Era — CSS Communication Toolkit}

AI does not naturally pause.
It does not naturally ask questions.
It does not naturally say, “I don’t have enough information.”

If something is missing, AI will:

• infer
• assume
• approximate
• generalize
• fill the gap with a pattern
• proceed confidently

So this next communication skill is simple:

We ask AI to identify what information is missing
so the work begins with clarity instead of assumption.

Missing information is not a problem —
it is an opportunity to define the work.

When we give AI an incomplete request,
it will not say:

“I need more context.”

“This is ambiguous.”

“This depends on something you didn’t specify.”

Instead, it will:

• guess the intention
• guess the structure
• guess the constraints
• guess the tone
• guess the behavior

A simple truth:

AI fills gaps
unless we ask it to expose them.

So, we ask.

—

Every gap falls into one of three categories:

1. Missing intention
   What we want, but have not fully described.
2. Missing structure
   What the system looks like, but AI cannot see.
3. Missing constraints
   What must remain stable, but we have not named.

When we ask AI to identify the missing pieces,
we prevent misalignment before it begins.

Let's see how.

We say:

“Identify what parts of the intention are unclear.”

“List what you need to know before proceeding.”

“Describe which aspects of the goal are underspecified.”

“Explain what you cannot infer about the purpose.”

We avoid:

“Is this enough information.”

We ask AI to surface the ambiguity,
not to judge completeness.

AI cannot see the system unless we show it.

We say:

“Identify what structural information is missing.”

“List the selectors, containers, or relationships you need to see.”

“Describe what parts of the layout you cannot determine.”

“Explain which elements you need to inspect before making changes.”

We avoid:

“Do you need more code.”

We ask AI to name the dependencies,
not to request the entire file.

Constraints are invisible unless we state them.

We say:

“Identify which constraints are missing.”

“List what boundaries you need clarified.”

“Describe what you assumed was flexible or fixed.”

“Explain what limitations you need before proceeding.”

We avoid:

“Is this safe to change.”

We ask AI to reveal the unknowns,
not to guess the risk.

We define the scope:

“Identify only the missing information about spacing.”

“Identify only the missing information about tone.”

“Identify only the missing structural details.”

“Identify only the missing constraints.”

Missing information becomes useful
when it is specific.

We explicitly say:

“Do not infer missing context.”

“Do not assume the intention.”

“Do not guess the structure.”

“Do not propose solutions yet.”

Then we add:

“Identify what you need to know before proceeding.”

This keeps the collaboration grounded in inquiry,
not improvisation.

Once AI identifies the gaps,
we can respond with clarity:

“Here is the missing selector.”

“Here is the intention you needed.”

“Here are the constraints you asked for.”

“Here is the structural context.”

The work becomes grounded,
not speculative.

A complete request looks like this:

“Identify what information is missing before you can refine the spacing.
Do not infer or assume anything.
List only the missing structural details.”

Or:

“Identify the missing constraints before making visual changes.
Do not propose solutions.
Reveal only what you need clarified.”

Or:

“Identify what parts of the intention are unclear.
Do not guess.
Describe the gaps only.”

This is how we turn ambiguity into alignment.

When we ask AI to identify missing information:

• the gaps become visible
• the assumptions disappear
• the intention becomes clearer
• the constraints become explicit
• the work becomes safer
• and the collaboration becomes more deliberate

A simple closing thought:

Missing information is not a flaw.
It is the beginning of clarity.
We ask AI to reveal the gaps,
and the work finds its shape.

• Suggested Next Reading: Closing Notes — The Quiet Skills That Make Collaboration Work

_{Tony de Araujo —New York | Lisbon}

• Top↑
• Back to Topic ↔
• Home←