You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
InfiniteZero is a live, open source, Ethereum-based protocol for a global AI commons. Validator nodes help train AI models collectively while raw data never leaves the device — privacy-preserving federated learning and differential privacy are hard-coded into the architecture. No single entity owns or controls what the network learns.
The structural risk we address: AI training infrastructure is concentrating in the hands of a handful of corporations before distributed alternatives exist. If that concentration becomes irreversible, the ability to course-correct is lost. InfiniteZero builds the alternative before that window closes.
DevNet is live with validator nodes running 24/7 from Japan to Canada. $165k raised in non-dilutive funding. Founded at Oxford under Sir Nigel Shadbolt and Emeritus Faculty Sir Tim Berners-Lee (Inventor of WWW). All code open source: github.com/InfiniteZeroFoundation
This grant funds the next phase: scaling the validator community, deploying the economic and privacy layers, and onboarding first real-world partner ecosystem pilots.
InfiniteZero is building a global AI commons — a live, Ethereum-based protocol where communities collectively train AI models while raw data never leaves the device. Participants run lightweight validator nodes that help train, aggregate, and validate models. No single entity owns or controls what the network learns.
The goal is to ensure a credible, community-governed alternative to centralised AI training infrastructure exists before the window for course-correction closes. We achieve this by scaling our validator network from 50 to 1,000+ nodes, deploying the economic and privacy layers, and onboarding real-world partner ecosystems across distributed solar, fitness wearables, agricultural robots, and home automation devices.
— Validator community onboarding and growth
— Additional contributor to focus full-time on community and communications
— Educational documentation lowering barriers for new participants
— Partner ecosystem pilot development
— Compute infrastructure (~$90/month)
— Community events and speaking engagements
Abraham Nash — Oxford PhD candidate, Human-Centered Computing, under Sir Nigel Shadbolt and Emeritus Faculty Sir Tim Berners-Lee. Founded InfiniteZero during doctoral research. Leads project direction, community, and communications.
Tudor Cebere — PhD researcher, Inria France. Harvard OpenDP associate. Former core OpenMined contributor. Lead author PETS 2026 — uncovering 13 privacy vulnerabilities across 11 differential privacy libraries. Leads privacy architecture.
Umer Majeed — PhD, Kyung Hee University, South Korea. Federated learning and distributed systems specialist. Responsible for live DevNet infrastructure.
Track record: DevNet live on Ethereum with validator nodes running 24/7 from Japan to Canada. $165k raised in non-dilutive funding. Recognised by UC Berkeley RDI, Foresight Institute Berlin node, and Decentralised Research Center (funded by the Ethereum Foundation). All engineering public and weekly: github.com/InfiniteZeroFoundation
Most likely cause: insufficient capacity to simultaneously maintain the protocol and grow the validator community. The core team is at capacity — without an additional contributor focused on community and communications, growth stalls while engineering continues.
Outcome: the network keeps running but the timing window — the period before AI training infrastructure consolidates irreversibly — closes without being fully used. The project does not die but loses its most important opportunity. Public goods infrastructure that arrives too late to matter is still too late.
~$120k from Artizen Fund Season 6 community round including matched funding from Funding the Commons. $40k from Edge City SHIFT Grants supported by Vitalik Buterin. $5k from Cosmos Institute (philanthropist Brendan McCord). Total raised to date across three years: ~$165k, entirely non-dilutive.