Why is this post flagged?

I've yet to try code stuff with AI (even Copilot). How well do local models like Code Llama work compared to GPT? When I use ChatGPT, GPT 3.5 feels a decade behind GPT 4. Half the answers are flawed. It's only GPT4 that boosts my productivity to an impressive level.

If you look at local text models for fiction or chat, they seem way behind even 3.5. See the examples table at https://docs.sillytavern.app/usage/faq/#what-do-you-mean-whe...

Code tasks are the sort of thing where I need as much accuracy as possible. Would something like Code Llama actually boost productivity or is this just "look, we can technically do it too, even if the result is awful" thing?