I think if people took the time to understand how LLMs choose word weights based on training data, they would understand that these results are somewhat determi

I think if people took the time to understand how LLMs choose word weights based on training data, they would understand that these results are somewhat deterministic. Instead, the preferred heuristic is to look for a bogeyman.