How Worried Should We Be About AI? — with Alex Stamos

AI safety discourse is shifting from speculative existential extinction threats toward the immediate, practical cybersecurity risks posed by rogue actors leveraging open-weight models. While high-profile warnings about human extinction often lack the rigor necessary for effective policy, the democratization of powerful offensive tools enables small, financially motivated groups to execute state-level cyberattacks. These actors now utilize AI to automate social engineering and exploit development, bypassing the need for human-heavy conspiracies. Addressing these vulnerabilities requires moving beyond alarmist rhetoric to implement "track two" diplomacy between American and Chinese labs. Establishing self-regulatory frameworks, similar to FINRA, provides a more viable path for managing risks like deepfake impersonation and automated network breaches than broad, unenforceable government mandates. This approach prioritizes tangible defense against the rising tide of mid-market AI-driven threats.