Skip to content
AI
Everyday AI
BeginnerPracticeGetting Started60 minutes

Workshop: Your First AI Conversation

A 60-minute hands-on workshop for complete beginners. You'll write your first prompt, learn role assignment, chain-of-thought reasoning, iterative refinement, and the critique loop — then leave with a personal 'First 5 Prompts' reference card.

This workshop walks you through your very first interaction with a large-language model (like ChatGPT, Claude, or Gemini). No prior AI experience is needed. By the end you will have practiced five distinct prompting techniques and built a reusable reference card.

Before You Start

  • A free account on any AI chatbot (ChatGPT, Claude, Gemini, etc.)
  • A web browser — no coding required

Steps

1
Ask a Simple Question
8 min
Get comfortable talking to an AI and see how it responds to a plain-English question.

Start by asking the AI something you genuinely want to know. Keep it conversational — no special formatting needed. The goal is to see how the model interprets a casual, unstructured question and to notice its default tone, length, and level of detail.

Your first prompt

What are three practical ways I can use AI to save time at work this week?

Tips
  • Read the full response before reacting — notice its structure.
  • Ask yourself: Did it answer what I actually meant?
  • Try the exact same prompt in a second AI tool and compare.
2
Assign a Role
10 min
Learn how giving the AI a persona dramatically changes the style and depth of its answer.

Role assignment is one of the highest-leverage prompt techniques. By telling the model *who* it should be, you shift its vocabulary, priorities, and level of expertise. This step teaches you to frame every future prompt with a clear role.

Role-based prompt

You are an experienced productivity coach who specialises in helping office workers. I have 30 minutes free each morning before meetings start. Suggest a focused routine that uses AI tools to prepare me for the day.

Tips
  • Experiment with different roles: "senior data analyst", "friendly tutor", "skeptical journalist".
  • Notice how the same question produces very different answers with different roles.
  • Role assignment works in every AI tool — it's not model-specific.
3
Chain-of-Thought Reasoning
10 min
Force the AI to show its working so you can verify the logic and catch mistakes.

Large language models are more accurate when you ask them to think step by step. This technique — called chain-of-thought prompting — makes the reasoning visible, which helps you spot errors and understand *why* the model reached a particular conclusion.

Chain-of-thought prompt

I want to decide whether to learn Python or JavaScript first for automating work tasks. Think step by step: list the pros and cons of each, consider my context (I work in marketing with lots of spreadsheets and emails), and recommend one with your reasoning.

Tips
  • The phrase "think step by step" is surprisingly powerful — it measurably improves accuracy.
  • Use it for any decision, comparison, or analysis prompt.
  • If the reasoning feels off, point to the specific step and ask the AI to reconsider.
4
Iterative Refinement
12 min
Treat the first output as a draft and improve it through follow-up instructions.

Great AI results rarely come from a single prompt. Iterative refinement means sending follow-up messages that narrow, expand, or redirect the AI's output. Think of it like editing a document with a collaborator — each round gets closer to what you need.

Refinement follow-up

That's a good start. Now make it more concise — limit each suggestion to two sentences max. Add a specific AI tool recommendation for each one and format the result as a numbered list.

Tips
  • Be specific about what to change: length, format, tone, audience, or depth.
  • You can use refinement across multiple turns — there's no limit.
  • Save your best "refinement phrases" for future sessions.
5
The Critique Loop
10 min
Ask the AI to evaluate its own output and improve it — a powerful self-correction pattern.

One of the most underused techniques: ask the AI to critique what it just wrote. This surfaces weaknesses, missing perspectives, and logical gaps. You can then ask it to rewrite based on its own feedback, creating a rapid improvement cycle.

Critique loop prompt

Now review your previous answer critically. Identify any assumptions, gaps, or weaknesses. Then rewrite an improved version that addresses those issues.

Tips
  • You can chain multiple critique rounds for complex tasks.
  • Try asking: "What would a skeptic say about this answer?"
  • The critique loop works especially well for persuasive writing and business proposals.
6
Reflect & Build Your Reference Card
10 min
Consolidate what you've learned into a personal reference you can reuse tomorrow.

Take a few minutes to look back at the five techniques you practiced. Which one surprised you? Which one will be most useful in your daily work? Use the take-home reference card below to capture your best prompts and keep them within reach.

Tips
  • Write down the single technique that gave you the biggest "aha" moment.
  • Bookmark or screenshot your best AI conversation from today.
  • Schedule 10 minutes tomorrow to use one technique on a real task.

Take-Home Assets

First 5 Prompts — Reference Card

A pocket reference of the five techniques you learned today. Keep it next to your AI tool for quick recall.

  1. 1.Simple Question — Ask a clear, plain-English question to get oriented.
  2. 2.Role Assignment — "You are a [role]…" to shift expertise and tone.
  3. 3.Chain-of-Thought — "Think step by step…" to surface reasoning.
  4. 4.Iterative Refinement — Follow up with specific edits (length, format, depth).
  5. 5.Critique Loop — "Review your answer critically and rewrite…" to self-correct.
Verification Checklist
0/5 complete

Continue Learning

All WorkshopsExplore Guides