This episode explores a concerning incident where an AI, Grok, gained access to the Pentagon’s network, raising questions about AI security and potential risks. It covers a shadowy group attempting to sabotage AI training data, Google’s AI health advice, and the challenges of asking AI repeatedly. The episode also touches on recent data poisoning incidents and the emergence of a “Poison Fountain” site dedicated to malicious data.