I foresee that the code-optimised version of GPT 4 with the 32K token context window will be amazing. GitHub Copilot was a derivative of GPT 3.0, which was pret

I foresee that the code-optimised version of GPT 4 with the 32K token context window will be amazing. GitHub Copilot was a derivative of GPT 3.0, which was pretty dumb compared to GPT 3.5, which in turn is the village idiot next to GPT 4... which IMHO is human-equivalent at many tasks. Not all, but many. Realistically, GPT 4 costs 100x as much as GPT 3.5 in inference mode, so it won't change the world just yet. There are still API rate limits, waiting lists, etc... Still... having the equivalent of a junior employee assisting with your code, but at a fraction of the cost and many times the speed, will be amazing.