I haven't seen any reasons around why an AI must be conscious to cause harm or existential risk. My opinion is that an x-risk causing AI would almost certainly

I haven't seen any reasons around why an AI must be conscious to cause harm or existential risk. My opinion is that an x-risk causing AI would almost certainly not be conscious like humans. As for human stupidity, maybe you just have more faith in humans than I do (but I am not entirely pessimistic). Let me ask a related question, to try and gauge this more deeply: do you think anyone will ever release an engineered pathogen into the wild?