Season 5Episode 221

Running LLMs Locally – Tether CEO Paolo Ardoino Reveals Decentralized AI Platform QVAC

May 7, 2026

31m

1 Guest

Listen Now

About This Episode

Paolo Ardoino, CEO of Tether, joins Narb on DevNTell to discuss the company's leap into decentralized artificial intelligence with the launch of QVAC. Paolo shares his journey from an 8-year-old coder to leading the world's most successful stablecoin, explaining how his background in hacking and research influenced his philosophy on technology serving humanity. The conversation dives deep into the vision for QVAC, which aims to provide "intelligence inclusion" by enabling powerful AI models to run locally and privately on consumer hardware. Paolo explains the technical underpinnings of the QVAC SDK and its "fabric" engine, and highlights Tether's recent medical AI advancement, MedPsy. He also explores the future synergy between AI agents and programmable money like USDT.

Key Takeaways

Tether is launching QVAC to promote 'intelligence inclusion,' paralleling its success with USDT for financial inclusion.

QVAC focuses on 'Edge AI,' allowing models to run locally on devices like smartphones to ensure privacy and reduce latency.

The QVAC SDK utilizes 'Fabric', a fork of Llama.cpp, to enable easy integration and local fine-tuning of AI models for developers.

Tether has released MedPsy, a medical AI model that outperforms larger models from major tech companies while remaining small enough to run on a phone.

Paolo Ardoino believes technology should be designed primarily to assist humans during their worst moments, a philosophy derived from his work on resilient communication systems.

AI agents in the future will require programmable money like USDT to execute complex tasks and financial transactions autonomously.