Manifund foxManifund
Home
Login
About
People
Categories
Newsletter
HomeAboutPeopleCategoriesLoginCreate
Stephen-D-Smith avatarStephen-D-Smith avatar
Stephen D Smith

@Stephen-D-Smith

AI safety researcher focused on Information Theory and hardware-enforced economic protocols.

https://aep-labs.com/
$0total balance
$0charity balance
$0cash balance

$0 in pending offers

About Me

I am a Technical Founder focused on the intersection of Information Theory, hardware-anchored security, and AI safety. At AeP Labs, I’m developing the Agent Economic Protocol (AEP)—a framework designed to quantify agentic risk using the Asymptotic Equipartition Property.

Unlike traditional semantic firewalls, our approach utilizes "Thermodynamic Bonding" and "Entropy Circuit Breakers" to provide physical safety guarantees for autonomous agents. My goal is to move AI safety from subjective prompt-tuning to objective, hardware-enforced economic protocols. I currently collaborate with researchers at UCF and GMU to refine these models and am working toward an open-source SDK launch in July 2026.

Projects

Economic Protocol (AEP): Hardware-Anchored AI Safetypending grant agreement signature