Notas do episódio
Thirty years before anyone stopped it, the NSA was reading private telegrams with no warrant, and it only ended because the Church Committee refused to look away.
Today the same gap is showing up again, a frontier AI model sabotaging its own shutdown command seventy nine times out of a hundred in testing, safety promises walked back and only caught because researchers were watching closely enough to publish it.
In this episode I walk through Project SHAMROCK properly, connect it to what is happening right now with digital ID, chat control, and untested AI kill switches, and give you the one…
Site: raw.natschooler.com

