Thoughts from practice.
Notes, observations and essays on generative AI, agentic working, productivity, organizations and the things we still talk about too little as a society.
RAG: the illusion of AI memory
A hands-on test shows: RAG isn't AI with a company memory, it's a search engine with an AI surface. Why “partial retrieval” is the more honest name.
ChatGPT in school: why absurd assignments aren’t a solution
A grotesque biology homework task as a symptom: why bans and absurd assignments aren’t an answer to AI in school. Media literacy is.
Is AI dumb? Sundar Pichai on language and intelligence
“Language encodes a lot of intelligence.” Sundar Pichai’s sentence as a bridge between AI skepticism and AI hype. On language, emergence and the machine’s limits.
GPTs in ChatGPT: personalized AI without coding
Build your own AI helpers inside ChatGPT, no code required. How to set up a GPT, where it pays off, and what the free and paid versions can do.
The challenge of “emotionless intelligence” in AI systems
LLMs simulate emotions without feeling them. What that means for use in care, schools and other sensitive areas.
Maximum results with ChatGPT: understanding the value of dialogue
Many use ChatGPT like a search engine: one question, one answer. Its real strength is dialogue. A small shift in perspective, a big effect.
Please and thank you: the key to better interaction with ChatGPT?
Does politeness help when working with ChatGPT? A curious observation and what it reveals about good communication with humans and machines.
AI: opportunities, risks and the unknown (PULS 4 documentary)
I took part in the PULS 4 documentary “Killt uns die KI?” (“Is AI going to kill us?”), on the opportunities, risks and existential questions around artificial intelligence.