Video content
| Time | Contents |
|---|---|
| 0:00 | Important theoretical foundations |
| 2:21 | Neural networks (only for deeper understanding) |
| 4:47 | What is an LLM (only for deeper understanding) |
| 6:40 | What an LLM can and CANNOT do – Important! |
| 7:35 | Practice: Which AI to use for what (quick info on this in the cheat sheet below) – ChatGPT, Claude, Gemini, Llama and Perplexity. |
| 10:45 | General information about AI – not fundamental, but useful for getting a feel for AI's possibilities and limitations. Including the "big fails." |
Cheat sheet – Which AI to use for what?
| Model | Application case | Advantage |
|---|---|---|
| ChatGPT | For short, precise and stylish answers | Fast, smart and very reliable |
| Claude | For longer conversations or extensive documents | Thoughtful tone, good for large contexts; Opus 4 for tempo, Sonnet 4 for deep thought |
| Google Gemini | Multimodal Analysis (Images, Video, Code) | Strong in large data sets and complex research |
| Calls | Full control over the model | Locally operable – used by us when data should remain in the country |
| Perplexity AI | One for all: Automatic model selection depending on the task | Selects the appropriate model independently (e.g. GPT, Claude, Gemini) |
Questions about: Basics of Artificial Intelligence:
🎲 Question 1: How does an AI really “think”?
A: She meditates in a digital lotus position.
B: It calculates probabilities for each possible next word and selects the word with the highest probability.
C: She reads the minds of everyone on the Internet.
✅ Correct answer: B
💡 Explanation: The AI determines probabilities for each next token based on its training and selects the token with the highest probability to generate coherent and appropriate texts.
🤖 Question 2: What is an LLM (Large Language Model)?
A: An artificial intelligence that behaves like a smart friend during small talk.
B: A long-distance run for computers.
C: A machine that writes texts but has no idea what it is saying.
✅ Correct answer: A and C are both correct – depending on your perspective 😉
💡 Explanation: An LLM processes huge amounts of text, recognizes patterns and generates appropriate answers – but without real understanding.
💥 Question 3: Which of these things can an LLM not do?
A: Have feelings
B: Knowing what is true
C: Help with moving
✅ Correct answer: A, B and C
💡 Explanation: An LLM has no feelings, no real understanding of truth and definitely no arms.
Exercises for this module:
Enter this question into the following models: How can I benefit from AI?
https://chatgpt.com
https://claude.ai
https://gemini.google.com
https://www.perplexity.ai/
No time for practice? Here are some sample results:
