The Art of Being Specific

Why vague questions get vague answers

The street-corner problem (and why it matters)

Imagine you’re standing on a street corner in a new neighborhood, hungry and a little stressed. You stop someone and ask, “Which way to the restaurant?”

They blink. “Which restaurant?”

Of course. There are dozens nearby. Your question made perfect sense in your head because you were picturing a specific place—maybe the Italian spot with the red awning you saw online. But none of that context actually made it into your words.

That’s the same trap people fall into with AI every day.

Only the words make it through

When you type a prompt into an AI tool, only the words make it through the doorway. Not your background. Not your tone. Not the earlier conversation you had with yourself while staring at the screen. Not your unspoken preferences.

So when you say:

• “Give me some ideas.”
• “Help me write something.”
• “What should I do?”

…you’ll often get an answer that’s technically fine but emotionally flat, overly broad, or weirdly irrelevant. Not because the AI is stubborn, but because it’s responding to the information it has—your words—and nothing else.

Vague input creates vague output. It’s not a moral failing. It’s physics.

How humans cheat (in the best way)

With a friend, you can ask, “Should I take that job?” and they immediately know what “that job” is.

They remember the last time you complained about your current boss. They know you value flexibility. They understand your anxieties without you having to list them. They can fill in the gaps because they have context—shared history, emotional cues, and a mental model of you.

AI doesn’t have any of that unless you provide it. Even in an ongoing chat, it only knows what’s in the conversation. There’s no secret backstory, no assumption that it “gets what you mean.”

That’s why talking to AI can feel strangely literal: it’s not being difficult; it’s being exact.

The specificity spectrum (from foggy to useful)

Specificity isn’t about writing longer prompts. It’s about including the details that change the answer.

Here’s a simple example:

• “Help me write an email.”
• “Help me write an email to my manager about time off.”
• “Write a brief, professional email requesting three days off next week, acknowledging the team is busy and offering to hand off my tasks.”

Notice what’s happening. Each version adds information that narrows the possibilities and makes the AI’s job easier.

When your prompt is specific, the AI doesn’t have to guess what matters. It can focus on delivering a result that fits your real situation.

If you’re not sure what to include, think in three buckets:

Context: What’s going on? Who’s involved?
Constraints: What are the limits—length, tone, rules, timing?
Purpose: What outcome are you trying to create?

A prompt with those three elements almost always produces a leap in quality.

The hidden reason specificity feels hard

Being specific can feel strangely vulnerable. It requires you to admit what you actually want.

“Help me with my resume” is safe.
“Help me rewrite my resume for a mid-level marketing role after a career break, and I’m worried I look unreliable” is honest.

But honesty is what creates accuracy.

Specificity also forces clarity. Many of us walk around with fuzzy goals: we want “something better,” “more time,” “a healthier routine,” “a career that fits.” When you try to translate that into a prompt, you bump into the truth: you’re not fully sure what you mean yet.

That’s normal. It’s also useful.

When you don’t know what you want (use the AI to find it)

Sometimes vague is the right starting point because you’re exploring. In those cases, treat the first AI response as a sketch—not the final painting.

You can refine with simple feedback like:

• “More casual.”
• “Shorter.”
• “Make it sound more confident, less salesy.”
• “Give me three options: direct, warm, and playful.”
• “This isn’t right—ask me 5 questions to clarify what I need.”

This is one of the best ways to use AI: not as a mind reader, but as a conversation partner that helps you discover your preferences. Often, the fastest path to what you want is seeing what you don’t want.

Specificity can be iterative. You don’t need perfection upfront—just forward motion.

A practical checklist for better prompts

If you want an easy upgrade, add one line to your prompt that answers:

• Audience: Who is this for?
• Tone: What should it sound like?
• Format: Bullet points, outline, script, email, caption?
• Success: What would a “good” answer help you do?

Even a few words can transform the result.

“Give me ideas” becomes:
“Give me five simple self-care ideas for a busy parent with 15 minutes a day. No buying products.”

Now the AI has traction.

Closing: “Not perfect—just enough”

Back on that street corner, you try again: “Italian place, small, red awning, something with Mario in the name.”

The stranger lights up. “Mario’s Trattoria. Two blocks down.”

Not perfect. But enough.

That’s the goal with AI, too. It can’t read your mind. It can only read your words. The gap between what you know and what you say is where communication breaks down—and where better prompts begin.

When the response isn’t quite right, it’s not failure. It’s a signal. It’s the conversation showing you what was missing.

And every vague question is a chance to learn what you actually meant.