Blog
On AI that does not know the answer.
Why AI tools give vague answers, why adoption stays low, and what sits underneath: working agreements that were never recorded.
✓ Published 11 Sep 2026Julius8 minute read
Why a chatbot on your own documents does not work the way you hoped
RAG, vector search, your own knowledge graph: technically it all works. Yet the answers stay unreliable. An AI engineer on where it goes wrong, and why that is not a bug.
Read article✓ Published 5 Sep 2026Simon7 minute read
Why nobody uses your AI tool (and why another training session will not help)
The licences are bought, the training has happened, and still almost nobody uses it. That is not resistance to change. It is a predictable reaction to one wrong answer.
Read article✓ Published 2 Sep 2026Simon6 minute read
Why your AI tool gives vague answers (and why a better prompt will not fix it)
Copilot gives a neat answer to a question about your own company. It just is not your answer. Why that is not the model's fault, and why a better prompt will not solve it.
Read articleNo articles found.
Try another word, or choose .