So we were evaluating AI coding assistants and I pushed for a trial of Kling. We used GitHub Copilot for over a year.
After a month, the feedback from the dev team was overwhelmingly negative. I’m trying to understand if we just didn’t configure it right or if our use case is bad.
Main complaints were about it being overly verbose in comments, and the completions felt less accurate for our stack (mostly Python/TypeScript on AWS). It kept suggesting outdated AWS SDK v2 patterns even when our codebase was v3.
Also, the IDE extension felt slower. Latency on suggestions was noticeable compared to Copilot, especially in larger files.
Has anyone else made this switch? Did you run into similar issues? I wanted to like it, especially for the promised security features, but the dev velocity hit was too big.