gpt 4, a closed model that exhist only as an api, is ahead of stable diffusion 2, a model with the traning code, dataset and weights downloadable, ok. I unders

gpt 4, a closed model that exhist only as an api, is ahead of stable diffusion 2, a model with the traning code, dataset and weights downloadable, ok. I understand there are some other metrics in the mix, but the flat average makes absolutely no sense here.