Notes on engineering, AI, hiring, and building systems that hold up in the real world.
Anthropic's Frontier Red Team says small models may be months away from copying themselves across cheap connected devices, with no single place to turn them off. The demo is easy to laugh at. The engineering problem is not. If the agents we already ship can continue a job after the session that started them is gone, the off switch has to be designed in, not assumed.
As developers increasingly rely on AI-augmented workflows, proprietary SaaS codebases face a silent threat: IP drift. Copying code into consumer LLMs risks leaking core business logic into public training sets. This article outlines a pragmatic CTO framework to secure data using Enterprise AI Gateways, Zero-Data-Retention APIs, and local LLMs without hurting engineering speed.
