Your First AI Conversation

You’ll have your first real conversation with an AI tool in the next 2 minutes. No local installation, no configuration. You may need a free account depending on the tool, but that only takes a minute. Just you, a browser, and a prompt.

Why this matters

Many job postings now mention AI skills or AI-assisted workflows. Hiring managers aren’t just asking “can you use ChatGPT?” — they’re asking “can you get useful results from AI tools?” That starts with knowing how to talk to one.

The difference between someone who gets garbage from AI and someone who gets gold? How they phrase the question. This lesson teaches you that skill with your hands on the keyboard, starting right now.

The mental model: AI as a new coworker

Imagine you just got a new coworker. They’re brilliant — encyclopedic knowledge, incredible work ethic, available 24/7. But they’re extremely literal. If you say “write something about our project,” they’ll produce a generic essay. If you say “write a 3-paragraph status update for the VP covering our Q1 progress, the timeline risk on the API migration, and our hiring plan” — you’ll get exactly what you need.

AI works the same way. Vague instructions produce vague output. Specific instructions produce useful output. You’re about to see this firsthand.

Step 1: Pick your tool

Open one of these in a new browser tab right now:

All three offer free tiers, though limits and availability vary by region. Pick whichever one you already have an account for, or whichever loads first. It doesn’t matter which one — the lesson works with any of them. If you don’t have an account for any of these, take a minute to sign up for a free one now.

Step 2: Paste this exact prompt

Copy the following prompt and paste it into the AI tool you just opened. Don’t modify it — paste it exactly as written:

Explain what a REST API is in 3 sentences. Use an analogy to a restaurant.

Hit Enter (or the send button) and read what comes back.

Step 3: Notice what happened

You just got a response. Before you move on, pay attention to three things:

1. Structure — The AI likely gave you close to 3 sentences, because you asked for 3. It usually follows this kind of constraint. If it didn’t, try re-sending with “exactly 3 sentences.”
2. Analogy — It used a restaurant analogy, because you told it to. The explanation probably compared a customer ordering food to a client making an API request.
3. Length — The response was concise. You didn’t get a 500-word essay because you specified a boundary.

This isn’t magic. The AI is following your instructions. The more specific your instructions, the more useful the output.

Step 4: Change one word

Now send this prompt — it’s identical except for one word:

Explain what a REST API is in 3 sentences. Use an analogy to a library.

Read the new response. The explanation structure is the same (3 sentences, same concept), but the analogy is completely different — probably something about borrowing books, a librarian looking up records, or a catalog system.

One word changed. The entire framing shifted. This is the core insight of prompt engineering: you have strong influence over the output. Every word in your prompt is a steering wheel.

Step 5: Now try being vague

Send this:

explain APIs

Read the response. Compare it to your first result. Notice:

• It’s probably much longer (no length constraint)
• It might use jargon you don’t understand (no audience specification)
• There’s no analogy (you didn’t ask for one)
• The structure is whatever the AI defaulted to (no format instruction)

The AI isn’t being unhelpful — you just didn’t tell it what you actually needed. Vague in, vague out.

The specificity principle

Here’s the pattern. On the left, what most people type. On the right, what actually works:

Notice the pattern in every specific prompt:

• What you want (email, script, explanation)
• Constraints (length, format, tone)
• Context (who it’s for, what you’re working with)
• An anchor (analogy, example, structure)

Practice: Write your own specific prompt

Now it’s your turn. Think about a topic from your own work or studies, and write a prompt that uses all four specificity levers: what you want, constraints (length, format), context (audience, purpose), and an anchor (analogy or structure).

Try it in your AI tool right now. Compare the result to what you’d get from a vague version of the same question. But first, practice the formula with this exercise:

Three more prompts to try

Now that you’ve seen how specificity shapes AI output, let’s explore three more prompt styles. Go back to your AI tool and try each of these. After each one, notice what the structure of the response looks like.

Prompt A: Structured list with ratings

List 5 ways AI is used in healthcare. For each, give one
specific real-world example and rate the maturity from 1-5
(1 = experimental, 5 = widely deployed).

What to notice: The AI follows your exact structure — numbered list, example for each, rating scale. You defined the output format, and it delivered.

Prompt B: Audience-aware explanation

I'm a complete beginner. Explain what a database is like
I'm 10 years old. Use bullet points.

What to notice: The language is simple, there’s no jargon, and the format uses bullet points. You set the audience and the format, and the AI adjusted.

Prompt C: Comparison table

Compare Python and JavaScript. Use a table with 4 columns:
Feature, Python, JavaScript, Best For.
Include at least 5 rows.

What to notice: You should get a formatted table with the columns and rows you requested. The AI can usually produce structured output — tables, lists, numbered steps — when you tell it to. If the format is off, try adding “Output only a markdown table with exactly 5 rows.”

Field-specific first prompts

Pick the prompt below that best matches your interests and try it in your AI tool. Notice how the same specificity formula (topic + constraints + analogy) adapts to different fields.

I'm a biology student. Explain what a BLAST sequence
alignment search does in 3 sentences. Use an analogy to
searching for a specific phrase in a library of books.

I'm a business student. Explain what a KPI dashboard is
in 3 sentences. Use an analogy to a car's instrument panel.
Include one example KPI for a retail store.

I'm a government developer. Explain what Section 508
accessibility compliance means in 3 sentences. Use an
analogy to building codes for physical buildings.

Key takeaways

• Specificity is everything — the more precise your prompt, the more useful the AI’s response
• One word changes the output — swapping “restaurant” for “library” completely reshapes an explanation while keeping the same structure
• AI follows the structure you provide — ask for bullet points, you get bullet points; ask for a table, you get a table; ask for 3 sentences, you get 3 sentences
• Vague in = vague out — “explain APIs” gets you a generic essay; a specific prompt with constraints, audience, and format gets you exactly what you need
• Every prompt has four levers: what you want, constraints (length/format), context (audience/situation), and an anchor (analogy/example)