Notes
Blog
How agents get authorized, and what a decision has to prove afterwards.
Intelligence Is Not Authority
Rethinking the architecture of LLMs and autonomous AI: a position paper on separating intent from authority.
Read →What If AI Never Needed to Be Trusted?
The problem may not be that AI is unreliable, but that we keep giving unreliable intelligence authority.
Read →Your AI Doesn't Need to Be Trusted. It Needs a Boundary.
Why the smartest way to use powerful AI is to stop asking whether you can trust it.
Read →Why agents need grants
An agent acts with your credentials at machine speed. A better filter does not close that. Authorization does.
Read →