> Based on what we know about different systems in the brain it might be a stretch to expect hallucination free AGI, using only a single LLM. Based on what we

> Based on what we know about different systems in the brain it might be a stretch to expect hallucination free AGI, using only a single LLM. Based on what we know about brains, it might be a stretch to expect hallucination free AGI. I’ve yet to meet a general intelligence that doesn’t hallucinate. Lots of fun examples from chickens who turn off when they see a straight line to cats who get spooked by cucumbers and the endless fun cognitive biases humans fall for.