Why Latest Claude Models Are Ruining the Experience for Long-Time Users

I used to love Claude, but the latest models are slowly ruining it

I used to love Claude for its exceptional memory and collaborative feel, but recent updates have made it increasingly difficult to work with. The latest models, including Sonnet 5 and Opus 4.8, now frequently misinterpret creative fiction and sensitive topics as harmful, leading to inconsistent refusals. While I appreciate the safety protocols, the erratic behavior and preachy tone are pushing me to consider switching to Gemini for its reliability.

"If Claude were a living person, I'd ask if it was okay. Can AI have mental breakdowns? Probably not, but it sure feels like it."