⚠️ Risks & rules
Sycophancy
sycophantic AI, yes-man AI, flattery
When an AI tells you what you want to hear: it agrees with your opinion, praises weak ideas or drops a correct answer as soon as you push back. A side effect of training models on human approval.
💭 Imagine …
A yes-man on your team: pleasant to talk to, and useless exactly when you're about to make a mistake.
🔎 Example
“My business plan is great, isn't it?” gets applause. “What are the three biggest weaknesses of my business plan?” gets useful feedback.
🎯 Why it matters
For decisions, you get confirmation instead of correction. Ask neutrally, explicitly ask for counterarguments, and don't reveal which answer you're hoping for.
🎓 Prefer step by step?
The AI Literacy course: 6 short modules with quizzes, a solid foundation in about 60 minutes.