Tether has announced the public release of QVAC Genesis I, a massive synthetic dataset comprising 41 billion tokens, designed to train artificial intelligence models in advanced scientific reasoning and problem-solving. The move marks the company’s latest effort to open up AI innovation and reduce dependence on tech giants that currently dominate access to large-scale, high-quality training data.
According to an October 24 press release, the dataset was developed by Tether Data’s AI division, QVAC. It’s specifically built to strengthen AI performance in STEM fields—areas like mathematics, physics, and medicine—where open-source models often struggle to keep pace with proprietary systems.
Tether CEO Paolo Ardoino described the release as a step toward decentralizing AI knowledge and empowering individuals to build and run intelligent systems independently.
“Intelligence shouldn’t be centralized,” Ardoino said. “With QVAC Workbench and Genesis I, we’re opening the door to AI that learns and evolves locally on your own device. Intelligence, like information, should be free, accessible, and owned by everyone.”
A New Model for AI Access and Control
Beyond its sheer size, Genesis I is designed to help AI models think more critically by understanding relationships and causality—not just recognizing patterns. Ardoino noted that many current systems “sound smart but don’t truly think,” and said the new dataset aims to foster deeper comprehension in machine learning models.
To complement the dataset, Tether also launched QVAC Workbench, a cross-platform application that allows users to run a range of AI models directly on their own hardware. The software supports open-source models such as Llama, Medgemma, and Qwen, and is available for Android, Windows, macOS, and Linux, with an iOS version on the way.
One of the app’s standout features, called “Delegated Inference,” lets users connect their mobile device to a desktop or workstation for added computational power—effectively bringing desktop-grade AI performance to smartphones through a peer-to-peer link.
Decentralizing the Future of AI
Together, Genesis I and QVAC Workbench reflect Tether’s broader vision of a decentralized AI ecosystem. The company’s approach echoes its roots in digital assets, emphasizing open, peer-to-peer infrastructure that minimizes reliance on centralized intermediaries.
If successful, Tether’s latest initiatives could reshape how AI systems are built, accessed, and distributed—potentially giving independent developers, researchers, and smaller organizations tools that were once reserved for tech giants.
