AI WORLD HQ
You've just seen another AI tool demo and felt that familiar itch to try it. Before you click "download," pause and ask: what problem am I actually trying to solve? Focusing on the need, not the shiny new tech, saves time and delivers real value.
What you need before you start
This is a thinking exercise, not a software install. You'll need:
- A free AI chat account (ChatGPT, Claude, or Gemini all work for the testing step)
- About 20 minutes of quiet time
- A pen and paper, or a blank note on your phone
No paid plan is required for the prototype test in Step 4.
Step 1 — Write the problem in your own words
Open a blank note on your phone or computer. Instead of writing "I need to use [some AI tool]," write what you're actually struggling with right now. Be specific. Vague problems lead to vague results.
For example, instead of "I want to use AI for writing," try: "I'm spending two hours every Sunday drafting a weekly update email to my team, and I keep putting it off."
💬 Example line you can copy into your note: "The problem I'm trying to solve is: [fill in your own situation here]."
You'll know it worked when you can read your sentence out loud and a friend would understand what's bugging you, without you having to mention any AI tool.
Step 2 — Ask "Why?" three times
Each "why" peels back another layer and gets you closer to the real cause. Take the problem from Step 1 and write it at the top of your note. Then answer "why?" underneath it, three times.
Example:
- Why am I spending two hours on the update email? → Because I have to gather updates from five different people first.
- Why does gathering take so long? → Because everyone sends their updates in different formats (Slack, email, a meeting).
- Why don't they use one format? → Because I never set up a shared template.
💬 Example line you can copy: "Why is this a problem? Because… Why does that happen? Because… And why does THAT happen? Because…"
You'll know it worked when you reach a cause you can actually change with an action, not a cause that's out of your control (like "because my boss is disorganized").
Step 3 — Decide if AI is really the right fit
Not every problem needs AI. Some need a better template, a new habit, or a simpler automation tool. Run through this quick checklist with your problem in mind:
- Is the task repetitive or data-heavy? (AI is good at spotting patterns.)
- Do I need help understanding or generating natural language (human-like text)? (That's exactly what large language models, or LLMs — the engine behind tools like ChatGPT — are built for.)
- Can I test a solution in under 30 minutes? (AI tools let you type a prompt, the instruction you give the AI, and get a result right away.)
- Is privacy a deal-breaker? (Many AI services store your data on their servers. If you handle sensitive info, check the privacy policy before pasting anything.)
If you answered "yes" to two or more, AI is worth a try. If most answers are "no," a plain template, a spreadsheet, or a manual fix is probably faster and cheaper.
You'll know it worked when you can say out loud either "yes, AI fits" or "no, I actually need a [template / spreadsheet / new habit] instead." No more guessing.
Step 4 — Build a tiny prototype with a real prompt
A prototype here just means a small, throwaway test, not a finished product. Open a free AI chat (ChatGPT, Claude, or Gemini — pick whichever you already have). Type a prompt based on the problem from Step 1.
For the Sunday update email example, your prototype prompt might be: 💬 "Here's a Slack message, an email, and meeting notes from my team. Combine them into a 5-bullet weekly update I can send on Monday morning. Keep the tone friendly and under 200 words."
Paste in one example of the messy inputs you actually deal with. Read the output. Does it save you time? Does it miss anything important?
💬 Another example prompt to try: "Look at these three customer emails and tell me which ones mention a refund request or the word 'urgent'. List them in order of urgency."
You'll know it worked when the AI's output is close enough that you'd actually use it with small edits, not when it's perfect. Perfection on the first try isn't the goal — saving 30 minutes is.
Step 5 — Decide your next move based on the test
Now step back and look at what happened. You have three honest options:
- It worked. Use the prompt as a starting point and refine it over the next week. Save it somewhere you'll find again.
- It almost worked. The AI missed one thing (maybe tone, maybe missing detail). Add that requirement to the prompt and test once more. This loop of small tweaks is how people get great results.
- It didn't work. Either the problem isn't a good fit for AI, or the prompt needs to be much more specific. Go back to Step 2 and try a different "why."
💬 Example line you can copy to log your decision: "Test result: [worked / almost / didn't]. Next step: [use it / refine the prompt / pick a different tool]."
You'll know it worked when you have a clear yes/no/maybe decision and a next action, instead of a vague feeling of "I should try AI sometime."
Common mistakes
Mistake: Jumping straight to a tool because of a demo or a friend's recommendation. Fix: Always start with Step 1. If you can't write the problem in one sentence, you're not ready to pick a tool yet.
Mistake: Expecting the AI to understand your situation from a vague prompt like "help me with emails." Fix: Paste in a real example of what you're working with. AI is much better when it has something concrete to react to.
Mistake: Treating the first output as the final answer and giving up if it's not perfect. Fix: Plan to test at least three slightly different prompts before you decide AI doesn't work for you.
Mistake: Skipping the privacy check before pasting in real customer data, internal notes, or anything with names and numbers. Fix: Before your first real test, read the AI tool's data policy. If the data is sensitive, use fake sample data instead.