Practice · module 100 of 100
Practical test: create your own AI method
Combine a good prompt, a real practice conversation, and your own quality control.
After this lesson
You create a reusable method with a source, prompt, checked result, and clear boundary.
In view · How it connects
From a small trial to a working method
Choose a small task with permitted input and a visible end result.
Try your prompt and one follow-up question; keep track of what you actually changed.
Record source, prompt, checkpoints, boundaries, and an alternative route.
Part 1 of 3 · Goal and source
Your preference for less movement is on. Choose each part yourself.
Choose a feasible mini-assignment
Take a small task that you can assess yourself: a fictitious invitation, a summary, or an internal notification. First, note the desired result and which source facts are known. Then, create a prompt with task, context, format, and boundaries. Avoid making decisions with major consequences for this first test. The goal is to learn a workable routine, not to use as many tools or complicated steps as possible.
Execute and improve with purpose
Present your prompt to Ende when the AI connection is available. Read the answer and request one concrete change. Save the first and second versions if you want to compare the difference. Separately check factual accuracy, usability for the target audience, and the agreed format. If AI is not available, you can practise the method with written examples; honestly call this a local exercise and not a generated result.
Save a method card
Record which task you performed, which data you used, which prompt helped, how you checked the result, and when a human must decide. Also note the time spent on improving and correcting. A reusable card helps you repeat the same task later and notice when sources or agreements change. This practical test is self-study: your own notes are not independently assessed and the certificate of participation is not a recognised diploma.
New practice example
From explanation to a usable result.
Your final assignment shows a small AI method involving source, prompt, improvement, and demonstrable control.
Fictional practice situationA fictional book exchange is taking place this Saturday in the community centre. Participation is free. The time and a contact person have not yet been confirmed. You are creating a participant message.
View a possible solution
This is how it can be done
Save the source, your first prompt, and a short invitation. Create a second version for a poster. Check for free, Saturday, and community centre; leave time and contact unknown. Finally, write down who checks the final message before publication.
Why this works
You are demonstrating a repeatable approach, not just a single piece of nice text. The individual work is self-study; the fixed case and knowledge question remain the control for lesson completion.
Make it slightly more difficult
Use a different fictional activity next week and the same workflow sheet. Which fields will you fill in again and which checks remain useful?
Extra practice, at your own pace. You do not need to submit anything. This does not count towards an automatic assessment.
Your workplace · learning by doing
Make something. See what changes.
Checking the AI connection… You can already view the written example.
Present your assignment to Ende.
Read the answer, change one thing, and compare.
Previously held
New version
You remain in control
What do you want to improve in your next attempt?
On this device only. AI answers remain only in this page memory. This is self-review, not a grade.
Why this approach works
You are demonstrating a repeatable approach, not just a single piece of nice text. The individual work is self-study; the fixed case and knowledge question remain the control for lesson completion.
A step further
Use a different fictional activity next week and the same workflow sheet. Which fields will you fill in again and which checks remain useful?
A prompt to get you started
Help me create a method card for this fictitious task: preparing an online workshop invitation based on approved facts. Provide fillable fields for goal, permitted source, starting prompt, one follow-up question, fact check, format check, human approval point, and safe fallback. Do not fill in unknown facts and do not perform any external action.
Apply it
After your AI test, you only save the final result. A week later, you do not know which source or check you used.
Which approach best fits this situation?
- In the next execution, let the model guess the missing source.
- Also save the source, prompt, relevant follow-up question, checks, and human decision boundary in a workflow sheet.
- Consider a beautiful final result as sufficient proof of a repeatable method.
With JavaScript, you can check your approach and receive immediate feedback.
Test your insight
Which result best demonstrates that you have created a reusable AI method?
- A long answer that sounds professional.
- Only a list of favourite AI tools.
- A card with source, prompt, checkpoints, boundaries, and what you actually improved.
Extra practice
Fill in your own method card with fictitious facts. Save the card and one concrete improvement from your test.
Goal: a short workshop invitation. Source: Saturday 10–11 o'clock, online, free. Follow-up question: shorter while retaining all source facts. Fact check: find day, both hours, online and free. Boundary: do not invent a date or registration link; a human decides on publication. Fallback: write manually using the same checklist.
