Before you pay for an AI course, try this ten-minute test

Before paying for an AI course, try one small task you can judge yourself. A short email is a useful starting point: you already know what you want to say, so you can spot a missing question or an invented detail.

A ten-minute exercise

Use an AI assistant you already have access to. This fictional example does not require real names, contact details or payment information.

Write a friendly email to a community hall asking whether it is available for a neighbourhood meeting. Ask about the hire price, seating capacity and available dates. Keep it under 100 words. Do not invent a date, price, booking or person's name.

Read the reply as if you were about to send it. Does it ask all three questions? Has it made a booking you never requested? Does it sound like something you would write?

Then ask for one change: “Make it less formal and remove anything that sounds like a confirmed booking.” Compare the two versions. Which needs less editing?

Judge the result, not the impressive wording

Give yourself one point for each of these checks:

  • The draft includes all the questions you asked for.
  • It adds no unsupported facts or promises.
  • You can explain and edit every sentence.

This is an informal exercise, not a recognised assessment. Even three points would not prove that the assistant is reliable on unfamiliar subjects. The useful question is whether you can steer its answer and check the result.

Choose your next lesson

If the exercise felt useful, look for a course with examples and exercises you can actually try. For a free starting point, Spinnit's beginner prompting course has five lessons that can be read without signing up. The course is free; that does not mean every AI service you might use alongside it is free.

This blog is connected to Spinnit. Its course page was checked on 5 October 2026. The hall enquiry above is our own suggested exercise, not a report of a test we carried out or a promise of better results.

Comments

Post a Comment

Popular posts from this blog

Travelling with a South African SIM? Check this before buying data

Before a trip from South Africa to Kenya, Ghana or Nigeria, check what happens to your existing data bundle abroad. A bundle on your home SIM and a plan for using that SIM overseas are separate things to investigate. Start with your home operator Vodacom's international roaming terms say ordinary bundle offerings and promotional tariffs do not apply to its international roaming offerings. Its roaming products have their own conditions. That is a Vodacom-specific statement; check your operator's current terms rather than applying it to every South African SIM. Ask your operator which roaming option covers your destination, what activates a charge, when the allowance ends and what happens after it runs out. Include calls and SMS in those questions if you need them. Compare the whole trip Write two short lists before choosing: Your existing SIM abroad: the roaming activation requirements, eligible destination and network, charges, allowance and expiry. A destination SIM or ...
Read more

The cheap-looking data bundle that costs more when you use it

A big data allowance can look like a bargain until you ask when you can use it. If part of the bundle is restricted to overnight hours, adding it to your daytime allowance can give you a misleading price per GB. Count the data that fits your day Consider two fictional bundles. These are arithmetic examples, not current offers from any network. Bundle A costs R50: 2GB of general-use data and 3GB of night-only data. Bundle B costs R60: 3GB of general-use data. Dividing Bundle A's price by the whole 5GB gives R10 per GB. But if you use your phone during the day and none of that night allowance, you are paying R50 for the 2GB you actually use: R25 per GB. Bundle B works out at R20 per GB if you use its full allowance. In this example, it costs R10 more upfront but gives you an extra 1GB for daytime use and a lower cost per GB used. If you can use all of Bundle A's night allowance, the calculation changes. The point is to compare the data that fits your routine, rather than tr...
Read more