for all of the capabilities ChatGPT has, it still kind of seems RLHF'd into the basin of Boringness, which seems very hard to avoid for any organization that is even the least bit risk-averse
in unrelated news reasonably controlling LLMs is Very Hard
openai.com/blog/chatgpt/
![](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="73" height="73"><rect fill-opacity="0"/></svg>)
