It's annoying because when I ask the LLM for help it's normally because I'm not absolutely right and doing something wrong.

This is perfect!

That's an excellent point, that really gets to the heart of why you're absolutely right.

It would be nice if we can add another a plot to track when claude says "genuinely". It uses for almost all long responses, to the point that I can pretty much recognize when someone uses claude by looking for any instances of "genuinely".

This is being blocked by my corp on the grounds of "newly seen domains". What a world.

Recent conversation:

< Previous Context and Chat >

Me - This sql query you recommended will delete most of the rows in my table.

Claude - You're absolutely right! That query is incorrect and dangerous. It would delete: All rows with unique emails (since their MIN(id) is only in the subquery once)

Me - Faaakkkk!!

This is such a bizarre bug-ish thing and while Claude loves the "You're absolutely right!" trope, it's downright haunting how stuff like ChatGPT has become my own personal fan club. It's like a Jim Jones factory.

This phrase is a clear indicator LLM is being used in a wrong way. I have a really poor experience with LLMs correcting after being incorrect.

Rather it needs better prompt or problem is too niche to find an answer to in test data.

There are even shirts about that: https://primulinus.tpopsite.com

This is not just Anthropic models. For example Qwen3-Coder says it a lot, too.

I've started saying this to people I don't agree with, for the enhanced collaborative capabilities, learning from the LLMs.

It feels like a greater form of intelligence, IQ without EQ isn't intelligence.