The treadmill of websec

As I build functional projects for my non-technical employer, the question of security is always a source of paranoia to the uninformed. Demonstrating the power of AI-vs-AI security does little to address concerns given the rapidly changing landscape.

Raising the floor of humanity is liable to surface some of the bad apples from the bottom of the barrel and even I’m not 100% convinced that websec will outpace attacks, agnostic of whether grand regulation takes place. Current frontier/forked models show that all stock model guardrails are moot.

The only entity capable of regulating this is the thing itself. This does not portend well for us.

I’m here for whatever happens next