Does Claude try to send you down rabbit holes when asking simple questions?

A user, after a year of using Claude, expresses frustration with its tendency to provide excessive information for simple questions, particularly outside of code writing tasks. They note that while Claude performs well for coding, for other inquiries, it often includes "useless fluff" and suggests unnecessary follow-up actions, such as running tests to confirm speedometer accuracy, rather than just giving a concise answer. The user desires Claude to return only the essential information.

This user's year-long experience with Claude highlights a common frustration with LLMs providing overly verbose answers, unlike its focused performance on coding tasks.