> Peppermill has this pretty sophisticated artificial intelligence software that analyzes faces of previous trespassers, and they get a 100% match on this guy… I can’t believe they took any of that seriously and arrested this guy. We’ve had movies for decades about AI making obvious mistakes and yet here we are, playing out the script to one of them. I’m sure there are loads of these cases stacking up now, but I remember another one where an AI analyzed camera footage of some guy holding a candy bar, identified it as a gun and called the cops. The cops show up and immediately pull their guns on this reported “armed suspect”. He was nearly killed by police because of a typical AI false positive. How long is it going to take for people to stop trusting these things?