What follows is part of the Globe’s weekly Camberville & beyond newsletter, which explores the latest from Cambridge and Somerville. You can read the full June 25 edition here. Sign up for the free newsletter here. It’s Thursday, the latest artificial intelligence model says So I thought maybe I could ask AI who to talk to locally to better understand the latest news about AI. What are we to make of the unusual warning from cyber experts this week that AI-powered threats could overwhelm defenses in months? Talk to Bruce Schneier, Claude suggested. He’s “the most quotable security commentator alive.” Quite the distinction. I put it to the test and emailed Schneier, a security technologist who has long been connected with Harvard’s Berkman Klein Center for Internet & Society, and whose newest book is about democracy and AI. The Cambridge resident, who The Atlantic called “a frenetic and acerbic security expert,” was clearly in a hurry when we spoke on the phone on Tuesday. Our conversation, edited for concision, clarity, and profanity — and offering one among many takes on this AI moment — was illuminating. I don’t pay much attention to AI or cybersecurity or where the two overlap. What is happening in that space? The models are getting better, they’re getting better at all things. Cybersecurity is one of the things. This is good for attack and defense. They’re good at fixing things. Firefox used [the Anthropic model] Mythos to find 270-something vulnerabilities, which have now been fixed, from now until the end of time. That’s great news! But these models separate skill from knowledge. What do you mean? It used to be that you could ask any doctor how to poison someone and they can tell you, right? You ask anybody who’s an engineer how to blow