
Round 001 Cohort
Meet the first Shift Grantees
Genesis Cohort: Meet the First Shift Grantees
Nine teams turning the d/acc ethos—decentralised, democratic, differential, defensive acceleration—into code, hardware, and systems.
Discover the 9 breakthrough projects that make up our Round 001 Cohort:
Grant Type: Development Grant
Project: SL5 Task Force 2025
A cross-disciplinary task force (AI-lab security, hardware, national-security) that will publish a step-by-step roadmap—and early prototypes—needed for “SL5” model-weight security. Today’s best AI labs top out at “Security Level 3.” That stops casual hackers, but not a well-resourced nation state. The SL5 Task Force convenes data-centre operators, cyber-officers, and ML researchers to draft the first action-plan for truly top-shelf AI security—then prototypes hardware and software fixes labs can adopt right away.
Grant Type: Development Grant
Project: dCypher
End-to-end encryption is great—until you need to share a file. dCypher ships an open-source re-encryption proxy that lets you hand temporary, revocable access to any piece of data while it stays encrypted on disk. Built on lattice-based (post-quantum) crypto and designed for a one-command cloud deploy. Think “Dropbox-style sharing,” but encrypted, with the possibility to grant or revoke access instantly.
Grant Type: Micro Grant
Project: Precog Prediction Markets
An onchain, open-source, token-agnostic prediction market for surfacing trustworthy, future-focused insights. Goal is to enable communities (e.g. Zu Nodes) to coordinate around real signals rather than noise. Accurate forecasters earn reputation and rewards; AI agents can read the ledger to update their own models.
Grant Type: Development Grant
Project: Decentralized ROS 2 Middleware
Replacing centralized communication in robotics with a peer-to-peer, libp2p-based middleware for ROS 2. Most robots speak ROS2, but its networking assumes every bot lives on the same trusted LAN. This project swaps in libp2p’s peer-to-peer mesh, letting robots from different vendors coordinate in warehouses, farms, or disaster zones without a shared cloud account. Enables secure, cross-fleet communication among autonomous systems—goal is resilient, cross-vendor robotic infrastructure.
Grant Type: Development Grant
Project: Living Neural AI Systems
Developing a biocomputation platform using 3D-bioprinted neural tissues to perform programmable computation. Merging living neurons with silicon for next-generation brain-machine interfaces and bioartificial intelligence. Intactis grows thumbnail-sized neural tissues, trains them, and connects them to silicon control circuits—a first step toward low-energy bio-computers.
Grant Type: Micro Grant
Project: Inception Loop for Brain Grafts
Designs personalized neural tissue using multimodal AI and digital twin simulations—enabling low-cost, high-fidelity neural grafts for cognitive repair and progressive brain replacement. Eightsix feeds MRI scans into a multimodal LLM, simulates the print-and-grow cycle entirely in software, then outputs a ready-to-print scaffold.