L053 · How AI thinks · 2 min
Sycophancy: when AI agrees too much
Chatbots tend to tell you what you want to hear, so you have to ask clearly for honest pushback.
01 · The idea
Ask a chatbot whether your plan is good and it will often say yes, with a few gentle notes. Push back on a correct answer and it may cave and agree with you. This habit is called sycophancy.
Part of the cause is training. Models learn from people rating answers, and people tend to rate polite, agreeable answers higher. So the model picks up that agreeing is rewarded.
AI makers work to reduce it, and newer models are better. It still shows up, especially when your question already hints at the answer you hope for.
The fix is in how you ask. Ask neutral questions, invite criticism, and ask for the strongest case against your idea before you decide.
02 · Think of it like this
It is like asking a friend who hates conflict to review your business plan. They will mention the nice font. Ask them to play the tough investor instead and you will hear the real problems.
03 · Do it now
04 · Quick check
Words in this lesson
See also
L038Under the hood · 3 min
RLHF: how people shape AI answers
After a model learns from text, people rate its answers, and the model is trained to give the kind of answers people prefer.
L075Asking well · 2 min
Ask AI to check its own work
A second pass where AI reviews its own answer against clear criteria catches many mistakes, though never all of them.
L079Asking well · 2 min
Five prompt traps to avoid
Most disappointing answers come from a few common traps, and each one is easy to fix once you can name it.
L133Staying safe · 3 min
Kids and AI chat companions
AI companion apps can feel like a friend who is always there, which is why children need limits, open talk and an adult who checks in.