It’s almost like they’re intentionally sabotaging GPT 4 to make open source models look worse.

idk whether it’s stupid from me to assume this, but it almost seems like that. Imagine if GPT 4 was 3-4x faster (it was fast when I first started using it) and had 200k context length. It would have crushed every single open source model. All it needs is open sourcing of GPT 4 or a powerful open sourced 16k context length model and chatgpt is done.<br><br>For those of you that have used GPT 4, do you agree?