Skip to main content
FAR Labs | Whitepaper home page
Search...
⌘K
Ask AI
Support
Dashboard
Dashboard
Search...
Navigation
Security: Proof of Compute
The Lazy Node
About FAR Labs
Core concepts of FAR AI
FAR AI
Abstract
Network Architecture: The Swarms & Triads
Dynamic Model Hosting
The Triad Topology
Geo-Clustering
Semantic Vector Streaming
The Engine: Semantic Vector Streaming
Privacy By Design
Privacy by Design
Hyper-Velocity Inference
Beyond the Speed Limit
The Model Registry
Dynamic Model Loading
The Client Selector
Security: Proof of Compute
The Lazy Node
Probabilistic Proof of Compute (PPC)
The Orchestrator
The Brain of the Grid
Ecosystem & Developer Hub
The SDK Layer
Enterprise Solutions
Protocol Stewardship
Conclusion
Conclusion
On this page
The “Lazy Node” Threat
Security: Proof of Compute
The Lazy Node
Copy page
Copy page
A decentralized AI network must ensure that every node is genuinely performing computation—not faking outputs to save electricity or GPU cycles. FAR AI solves this problem with a cryptographically enforced Proof of Compute layer.
The “Lazy Node” Threat
In open networks, malicious nodes may attempt to:
Pretend to run the model while returning random or low-effort outputs
Shortcut inference by using smaller models internally
Replay old outputs instead of generating fresh responses
Drop computations entirely to save power
The Client Selector
Probabilistic Proof of Compute (PPC)
⌘I
