I’ve noticed a regression in Cursor AI’s responses. An example is my prompt to replace
GetX and its components in
home_view.dart. The response from Cursor was not helpful, as shown in the attached screenshot.
The issue I met with is similar to Why does Cursor seem “dumb” lately?
My experience with OpenAI Playground using gpt-3.5-turbo-16k-0613 several months ago was better than some of the responses I’m getting now.
I’d like to share my entire codebase with Cursor AI’s developers under the condition of strict confidentiality so that you can reproduce the issue easily.