@nprofile1q... Yea, it's more the misapplication of LLM models that are umm 'scary', not the technology itself. It's more like dumb things people will do with it and then government adoption. For instance AI to scan for license plates and generate a list of 'suspects' that cops then go after without any thought despite that doing this at scale with a technology that is often right, but always wrong a certain percentage of the time means millions of (likely just in the US alone) incidents with cops pulling guns on innocent people and some percentage of those resulting in deaths as a consequence.