Five useful ways to use AI this week

1. Turn a list into a realistic plan

A scattered list often hides the real problem: there is more to do than the available time allows. An assistant can suggest an order and expose assumptions. Begin with invented chores or a list containing no private information. Give the time available and identify anything that must happen first.

Suppose you have 45 minutes. A workable draft might allocate 15 minutes to a desk, 10 to plants, 15 to a shopping list and five as a buffer. Adding up is only the first check. If your desk needs an hour, a mathematically correct plan is still unrealistic. Ask for a smaller version of that task rather than accepting an impossible schedule.

I have 45 minutes to clear my desk, water plants and write a shopping list. Leave five minutes spare. Suggest an order, allocate minutes and list your assumptions. Do not add extra tasks.

2. Compare information you already have

Paste the public details of two options and ask for a comparison using your priorities. For example, compare two community classes by timetable, travel time and total cost. Ask the assistant to mark missing information as unknown.

The useful output is a list of differences and questions to resolve. Check every number against the original information, including whether a price is per session or per course. Do not let a neat table turn an unstated cancellation policy into a fact. If you need current availability, check the organiser directly.

3. Make a short practice exercise

Choose a concept you have already encountered, such as fractions or the difference between a claim and evidence. Ask for one question at a time and attempt it before seeing the answer. This creates an opportunity to practise rather than just read a smooth explanation.

If the feedback seems strange, return to your textbook or another reliable source. A confident marking comment can be wrong. Try explaining the concept later without the assistant; that is a more useful check of your understanding than how familiar its explanation felt.

Help me practise distinguishing facts from opinions. Give me one everyday example at a time. Wait for my answer, then explain your classification. Avoid specialist knowledge.

4. Organise a rough draft without changing its meaning

Use your own non-confidential notes, such as ideas for a hobby project. Ask for headings and a sequence while keeping all the original points. The result can help you see repetition or a missing step.

Compare it with your original. Did the assistant turn a possibility into a commitment? Did it remove an exception? For example, “might borrow a venue” must not become “venue booked”. Keep the original alongside the organised version until you have checked it.

5. Generate options for something playful

Try names for a fictional gardening club, ideas for a birthday scavenger hunt, or different openings for a story. Give two or three preferences so the ideas are relevant. The goal is to produce material you can adapt, not to find an objectively best answer.

For a scavenger hunt, check that every clue fits the real space and the participants. Remove activities that need unsuitable equipment or permissions. If you intend to publish a design or name commercially, further checks may be needed; a generated suggestion is not proof of originality or availability.

Decide whether the experiment helped

After one exercise, ask whether it saved effort after checking, whether the result was usable, and whether you could have done it more easily yourself. A short list or calculator may be sufficient.

These are suggested practice activities, not measured promises of productivity. Start with one, adapt it to your circumstances and keep only the part that helps. For broader limitations and risks of generated information, see the NIST resource below.

Sources and further reading

The exercises and examples are Virel AI’s practical suggestions. These resources provide supporting guidance and broader context.

• NIST: Generative AI Profile