What AI can do, how it learns, and why people still need to think for themselves
Section 1 of 10
⭐ 0/10
1. What is artificial intelligence?
Artificial intelligence, or AI, refers to computer systems that derive results from inputs, such as predictions, recommendations, decisions, texts, or images. They can perform tasks that normally require human perception, language, learning, or planning.
👀
Perception
AI can recognize patterns in images, sounds, and sensor data.
🗣️
language
It can recognize, translate, and summarize language or generate new text.
🔍
Finding patterns
It compares many examples and looks for regularities.
🎯
Producing results
It provides recommendations, probabilities, content, or control commands.
Important: AI is not a human being. It does not automatically possess consciousness, feelings,
a conscience, or genuine understanding. Even very human-sounding answers are produced through computation.
Which statement fits best?
2. How long has AI existed?
The idea of thinking machines is old. However, the scientific field developed mainly in the middle of the twentieth century. Click the milestones.
AI did not develop in a straight line: High expectations were repeatedly followed by disappointment and reduced research funding. Such periods are sometimes called “AI winters.”
Why is 1956 considered an important year?
3. How does AI work and learn?
Many modern AI systems are trained with examples. Numbers within a model are adjusted so that it recognizes patterns and calculates useful results for new inputs.
📚 Training examples
→
⚙️ learning method
→
🧠 Model
→
🔮 Prediction
Train a tiny sorting AI
Show the model examples: Which things are more likely to belong to fruit and which are more likely to be vehicles?
Real training is much more complicated: Large models can contain billions of numerical values and use huge numbers of examples. Even so, they learn statistical relationships, not infallible truth.
What usually happens during training?
4. How does generative AI create new content?
Generative AI can create new text, images, sounds, videos, or computer code. In simplified terms, a language model repeatedly calculates which piece of text is likely to fit next.
Try word prediction
Complete: “The dog chases the …”
🧩
Token
Text is split into small units. A token can be a word, part of a word, or a character.
📊
Probabilities
The model evaluates many possible next pieces of text.
🔁
Repetition
The selected text element is added and the calculation starts again.
⚠️
Not a truth machine
A sentence that sounds appropriate can still be factually wrong or invented.
Why can a language model make a false claim?
5. Where is AI used?
AI can support people, but not every task should be automated. Click an area and read its benefit and caution note.
Which statement is sensible?
6. How can AI make everyday life easier?
A good request describes the task, goal, audience, format, and important limits. Such instructions are often called a “prompt.”
Build a good prompt
Your prompt appears here.
Which request is better for learning?
7. What risks and limitations are there?
❌ Incorrect answers
AI can confuse facts, invent sources, or convincingly hide gaps in its knowledge.
⚖️ Unfair patterns
Biased data can cause people or groups to be treated unfairly.
🔐 Data privacy
Entered data may be stored or processed. Private information should not be entered into a system without careful thought.
🎭 Deepfakes
Artificial images, voices, or videos can appear real and deceive people.
🧠 Dependence
Anyone who hands over every task practices less and may lose skills.
💼 Work and power
AI changes occupations. Benefits and burdens may be distributed unequally.
🌍 Energy and raw materials
Large data centers require electricity, cooling, equipment, and raw materials.
🛡️ Misuse
AI can also be misused for fraud, manipulation, malware, or surveillance.
True or invented?
Question: Who invented the internet in 1820?
AI answer: “The internet was invented in 1820 by Professor Wilhelm Netzmann.”
What should you do with a surprising AI claim?
8. How do I use AI safely and responsibly?
A simple reminder is: STOP – Think – Check – Decide.
🛑 STOP Do not accept everything immediately
→
🧠 THINK Does the answer make sense?
→
🔎 CHECK Consult sources and adults
→
✅ DECIDE People remain responsible
🔒
Private stays private
Do not enter passwords, an exact address, identification data, health data, or other people’s secrets.
📚
Multiple sources
For important information, also consult textbooks, reputable websites, experts, or adults.
📝
Disclose AI assistance
For schoolwork and publications, openly state when AI provided substantial help.
👤
Respect people
Do not create deceptive images, insulting content, or false claims about others.
Privacy traffic light
Who is ultimately responsible for the use of AI?
9. The Great AI Knowledge Test
Answer 15 questions. With at least 12 correct answers you pass the test and unlock the reward game.
10. Reward Game: Data Rescuer
🔒
The game is still locked
First pass the Great AI Knowledge Test with at least 12 out of 15 points.
Move the small robot with the arrow keys or the on-screen buttons. Collect good data and avoid hazards.
Score: 0Lives: 3Time: 45s
🤖
Game rule: 📚, ✅, and 💡 earn points. 🔐, ❌, and 🎭 cost one life. The symbols are reminders to use good information and avoid private data and deception.