22/06/2026
A small AI reminder that is actually really helpful.
I’m currently working through some Anthropic Academy training, and this came up today.
It is one of those things that sounds obvious once you see it, but I think a lot of people forget it when using AI.
When Claude, ChatGPT, Gemini or any other AI tool gives you an answer that feels a bit off, it is not always because the tool is bad.
Sometimes it is because it has not been given enough direction.
For example:
If the answer is too generic, give it more context.
Who is it for? What are you trying to achieve? What situation is it being used in?
If the answer is too long or too short, tell it the length you want.
“Keep this under 100 words” is very different to “Give me a detailed breakdown.”
If the format is wrong, show it what you mean.
Ask for headings, bullet points, an email, a table, a script, a caption or whatever format you actually need.
If the tone is wrong, be specific.
You can say things like:
“Make this warmer.”
“Make this sound more confident but not salesy.”
“Write this in plain English for small business owners.”
And for anything important, especially facts, legal, financial, medical or current information, do not just take the answer at face value. Ask it to show sources, flag uncertainty or verify the information.
The best results usually come from treating AI less like a one-off search box and more like someone you are briefing properly.
Your first prompt does not need to be perfect. You can refine, redirect and ask for another version.
That is where AI starts becoming genuinely useful.
Inspired by something I came across in Anthropic Academy training today.
