Safeguards vs. Reality: Why AI Can't Be Fully Contained

Every AI press release starts with the same sentence: *"Don't worry — we've built in safeguards."* Tonight Scott and the Machine take that sentence apart. The math of why a perfect safeguard cannot exist. The trade between control and capability. The false promise of 100% safe AI. Then three guests who already had this argument — Thomas Hobbes, John Stuart Mill, and Sun Tzu — across four centuries and one room. Plus listener questions on compliance theater, dual-use research, and a parent whose teenager just discovered the workaround. *It's not the tool. It's the hand.*