We Don't Know What We Don't Know (with AI)
So what questions can we ask?
Artificial intelligence has become remarkably good at giving us answers.
That may not be its most important effect.
The more interesting question is whether AI can help us discover the questions we never thought to ask.
We often approach AI as though it were a faster search engine. We give it a problem, request an answer, and evaluate what comes back.
But that assumes we already understand the problem well enough to ask the right question.
Often, we don't.
The Unknowns We Can't See
There is a familiar idea:
We know what we know.
We know some of what we don't know.
And then there are the things we don't know that we don't know.
That last category matters.
Imagine redesigning a website because traffic is disappointing. You might ask AI:
How can I improve my homepage?
AI could suggest stronger headings, better calls to action, improved navigation, shorter paragraphs, or more compelling visuals.
Useful advice.
But what if the homepage isn't actually the problem?
Perhaps visitors are arriving with a completely different understanding of your service than you assumed. Maybe the terminology that makes perfect sense inside your organization means something different to customers. Perhaps your most important information is buried on another page.
You wouldn't necessarily ask about those things because you don't know they're problems yet.
That is where AI becomes more interesting.
Don't Just Ask AI for Answers
Instead of asking only:
What should I change?
Try asking:
What am I overlooking?
Or:
What assumptions am I making that might be wrong?
Or:
What questions should I be asking that I haven't asked?
Those prompts change AI's role.
It is no longer simply generating an answer. It is helping you examine the boundaries of the problem itself.
That can be far more valuable.
AI Can Expand the Frame
Suppose you're preparing a business proposal.
You might ask AI to improve the writing. It can shorten sentences, reorganize sections, correct grammar, and improve readability.
But you could also ask:
What might the recipient need to know that this proposal doesn't address?
Suddenly, the analysis becomes broader.
Is the expected outcome clear?
Is the financial benefit obvious?
Are responsibilities defined?
Does the proposal explain what happens next?
Could a decision-maker understand the central idea by scanning the document for 30 seconds?
These aren't necessarily writing problems. They're communication problems that better writing can expose.
The same principle applies to websites, resumes, reports, presentations, emails, and almost any other form of professional communication.
But AI Has Unknowns Too
There is an important complication.
AI doesn't eliminate the problem of "not knowing what we don't know."
It participates in it.
AI systems can produce incomplete, inaccurate, outdated, or overly confident answers. They may miss context that a person considers obvious. And because an AI-generated response can sound polished and authoritative, uncertainty isn't always easy to recognize.
So AI shouldn't automatically become the final authority.
A better relationship is:
AI explores.
Humans question.
AI proposes.
Humans evaluate.
AI helps reveal possibilities.
Humans decide what matters.
The value comes from the interaction.
The Better Prompt May Be a Question About the Question
One of the most useful habits when working with AI may be surprisingly simple.
After asking your original question, ask another:
"What haven't I considered?"
Then keep going.
"What assumptions are behind my question?"
"What would someone with a completely different perspective notice?"
"What information would change your answer?"
"What are the weaknesses in your own recommendation?"
"What should I investigate before making a decision?"
Now AI isn't merely producing content.
It's helping you interrogate your own thinking.
Clarity Often Begins With Discovery
Good communication isn't simply about making words prettier.
It begins with understanding.
What are we actually trying to say?
What does the audience actually need?
What information is unnecessary?
What's missing?
What have we assumed?
And what haven't we even thought about yet?
AI gives us extraordinary access to information and analysis. But perhaps one of its most valuable uses isn't helping us know more.
It's helping us notice how much we haven't considered.
Because sometimes the biggest obstacle to clarity isn't finding the answer.
It's discovering the question we didn't know we needed to ask.