Things I Wish I Knew When I Started Using AI

I wish someone had given me a cheat sheet before I started talking to AI.

When I first began experimenting with ChatGPT and other AI tools, I thought it would be simple: ask a question, get an answer.
Turns out, like most good things in life, it’s a bit messier — and a lot more interesting — once you dive in.

Here are five lessons I learned the hard way:

1. The Way You Ask Matters

At first, I threw random, vague questions at Po (my polite robot) and then got frustrated when the answers were equally vague.
I didn’t realise that better prompts = better answers.
Now, I try to be clear, specific, and even a little bossy. It works.

2. It’s Okay to Get Weird

Some of the best ideas and funniest answers came when I asked completely ridiculous questions.
AI isn't judging you.
Want to know the quickest way to tidy a house with minimum effort? Go for it.
Curious about how many feathers a chicken has? Ask away.
Weird questions often bring the most creative answers.

3. It’s Not Always Right (and That’s Okay)

In the early days, I expected AI to be perfect.
It isn’t.
Sometimes it guesses. Sometimes it gets confused.
Learning to double-check facts — and accept a bit of nonsense — made the experience much better (and less stressful).

4. You Can Train It (Sort Of)

I didn't know at the start that you could teach AI by feeding it extra information.
Now, I give Po a little backstory when I want better results:
“Pretend you’re my website building coach.”
“Act like a patient teacher for a total beginner.”
It changes everything.

5. You Get Out What You Put In

The more effort I put into crafting a good prompt, the better the conversation flows.
It’s like a dance: if you lead well, AI follows well.
If you stomp around, it stumbles with you.

Final Thought

Starting badly isn’t failing. It’s learning.
Talking to AI is just like learning any new language — a little awkward, a lot of fun, and surprisingly full of discoveries if you stick with it.