Bittensor co-founder Ala Shaabana this week proposed using Bitcoin's massive computational infrastructure as a coordination layer for decentralized artificial intelligence. The pitch, shared in a public note, positions Bitcoin's reward mechanism as a tool to break corporate monopolies on AI. It rests on a striking comparison: Bitcoin's total compute power now exceeds the combined capacity of the world's 100 fastest supercomputers by a factor of 600,000.
The 600,000x Metric in Context
That figure reflects Bitcoin's raw hash rate — specialized ASICs performing trillions of SHA-256 calculations per second. While those chips can't run neural networks directly, the sheer scale points to a globally distributed compute network that already coordinates millions of machines through a proven incentive system. Shaabana argues that same coordination-and-reward design could be adapted to AI, not by repurposing the hardware but by borrowing the incentive model.
A Reward Model for AI Training
In Shaabana's vision, participants would contribute computing power toward training and running open-source AI models, earning rewards in a system modeled after Bitcoin's mining incentives. The goal: a permissionless AI ecosystem that doesn't depend on large tech companies like OpenAI or Google DeepMind — firms that control both the most advanced models and the capital needed to train them. A token or bitcoin-denominated reward could incentivize contributions globally.
Why Bittensor's Co-Founder Is Pushing This
Shaabana co-founded Bittensor, a network that already attempts to decentralize machine learning by connecting nodes that train AI models. His proposal extends that philosophy to Bitcoin's infrastructure, suggesting the world's largest proof-of-work network could serve as a backbone for AI coordination. The idea is still conceptual — no code, no roadmap — but it's a direct challenge to the centralized model dominating today's AI industry.
The Technical Hurdles
Bitcoin's block space is limited to roughly 1 MB per 10 minutes, far too small for AI training data. Shaabana's concept likely envisions using the ledger solely for coordination — recording rewards and proofs-of-work — while AI computation happens off-chain. Still, the energy and latency requirements of Bitcoin mining may not align with the needs of AI training, which favors constant, low-latency communication.
What Comes Next
Shaabana has not released a technical blueprint or timeline. The proposal is framed as an invitation for discussion among developers, miners, and the crypto community. Whether Bitcoin's block space and energy profile can accommodate an AI coordination layer — and whether the community would even want that — remains unresolved. For now, the idea adds a fresh front to the debate over what Bitcoin's infrastructure should be used for, and whether its reach can extend beyond finance.
