Course Content
AI Driving Licence
Welcome to the training "AI Driving Licence" – your guide to the working life of the future!
0/1
Introduction
Part A – The Basics and Responsibility
Here we lay the theoretical foundation. We explain in detail what generative AI actually is and how it can elevate your productivity. We also look at the indispensable traffic rules – from information security and copyright to the EU AI Act – so that you can navigate safely and legally.
0/9
Introduction to Generative AI
Let AI Elevate Your Working Day
Navigating Safely in the AI Landscape
The Human at the Helm
The EU AI Act
AI-Driven Security Threats (Social Engineering)
Copyright and Intellectual Property in the AI Age
Sustainability and Environmental Impact
Summary of Part A
Part B – AI in Practice
In this part, we open the bonnet. We explain how the technology works in an understandable way, compare the major AI assistants, and teach you how to mentally and practically implement AI in your daily processes, with your critical thinking as a compass.
0/7
Understanding the Technology
Getting Started with Generative AI
Your New Digital Colleague
Getting Started with Your AI Assistant
From Words to Action
Critical Thinking
Summary of Part B
Part C – Prompts
Here we put our hands on the steering wheel. We dive deep into the craft of communicating with the machine, so-called ”prompting”. You will receive proven frameworks for text, methods for analysing complex documents multimodally, and techniques for directing fantastic AI images.
0/7
Welcome to the Prompt Guide
6 Steps to the Perfect Prompt
Creating Images with AI
Working Multimodally (Beyond Text and Image)
Next-Level Prompting (Contextual Examples and Iteration)
Summary of Part C
Assessment & Continuous Professional Development
Protected: AI Driving Licence

Next-Level Prompting (Contextual Examples and Iteration)

Once you have mastered the 6-step model we went through earlier in the course, you possess a tremendous advantage. Your texts will be sharper and your results more accurate. However, to be able to call yourself a true “AI pilot”, there are two slightly more advanced concepts within prompting that you must know. These techniques take you from being a good procurer to becoming a strategic engineer of your own workflow. The techniques are often referred to as “Few-shot prompting” and “Chain of Thought”.

Few-shot prompting (Leading by Example)

Even with a perfect instruction, it can sometimes be difficult to get the AI to hit exactly the unique tonality or structure that your specific company uses. “Few-shot prompting” solves this. Instead of merely describing what you want (which is called Zero-shot), you include a couple of examples (shots) of what a perfect final result looks like. AI models are formidable at pattern recognition, and when they see your examples, they immediately copy the “DNA” you have provided.

 

Imagine that you want to write product descriptions for your webshop. You might write: “Write a sales text about our new wool socks.” The AI will do a decent job. But if you instead write: “Here are two examples of how our product descriptions usually sound. Example 1: (Paste a previous, good text). Example 2: (Paste another previous, good text). Now, write a text about our new wool socks in exactly the same format and tone as the two examples.” The result is now astonishingly precise, adapted to your corporate culture, and immediately ready for publication. You have not only given the AI a task; you have given it a ready-made template to mimic.

 

Chain of Thought (Thinking step-by-step)

One of the greatest problems with generative AI is, as is well known, hallucinations (when it guesses incorrectly but sounds certain). This most often occurs when one asks a highly complex question where the AI must jump directly to a difficult final answer without reasoning along the way. “Chain of Thought” is an extremely powerful method for forcing the AI to slow down and break down its own work.

 

The method is ridiculously simple: you quite simply add the phrase “Think step-by-step” or “Explain how you arrived at this before you give the answer” in your prompt. Imagine that you are planning a logistically complicated conference where several speakers, breaks, and journeys must click into each other. If you ask the AI to produce a detailed schedule, it can easily miscalculate travel times or double-book a room. However, if you conclude your prompt with “Explain your logic step-by-step for each time slot”, you force the probability engine to first calculate step one, which makes step two more correct, and so forth. You are asking the machine to show its workings, exactly as a mathematics teacher does.

 

By using contextual examples (Few-shot) to control style, and chains of thought (Chain of Thought) to control logic and reduce margins of error, you elevate your prompting to a professional level. It is about iterating, getting to know the tool’s limitations, and actively building frameworks that guarantee quality.

Previous
Next