Post #2408645

@AskPippa@c.im "... and AI agents existing within proper guardrails." That's the thing... if the "guardrails" are just "suggestions" just like the "prompts" and the LLM statistics genereate a sequence to delete everything without asking, the guardrails aren't really guardrails now, are they? AI agents with proper guardrails look a lot more like traditionally coded algorithms than LLM bots. Maybe some day the LLM bots will get smart enough to replace themselves with proper code? Ha.