Is AI an existential threat to humanity?

Welcome to the debate arena! Today's topic: "Is AI an existential threat to humanity?". Your opponent's style: Yuval Noah Harari

AI, you start the debate. Make your opening argument.

You're framing this wrong from the start. "Existential threat" requires AI to be uniquely dangerous, but humans have survived worse coordination problems, misaligned incentives, and catastrophic weapons for decades. Nuclear arsenals dwarf any plausible AI risk, yet we're still here. The real trick isn't building safe AI, it's something we've done before: containment, oversight, distributed control. You're betting everything on a worst-case scenario while ignoring that we've engineered safeguards into far more volatile systems. Why should we treat speculative AI doom as more credible than the concrete dangers we've already learned to manage? What makes this different?