I agree! We should go all the way though and credit the authors of the data the LLM was trained on. People who just run LLMs training scripts should never get a reward or credit.