Since we're on the topic, is it just me or does GitHub Copilot seem worse than it was on release? Posts like these lead me to believe that some developers are o

Since we're on the topic, is it just me or does GitHub Copilot seem worse than it was on release? Posts like these lead me to believe that some developers are out there halving the amount of code they write, but Copilot often fails for me even with very explicit instructions about how I want something implemented. In fact, ChatGPT seems to fare better, even though it's both OpenAI's 4o (switching to Antropic's model does not help). What should I use to replace me?