New paper on Proof of Usable Work

This innovative approach circumvents the limitations associated with the size of matrix multiplications while providing a verifiable measure of the computing effort expended, measured in Floating Point Operations per Second (FLOPs). The researchers argue that this mechanism could introduce a more efficient and secure way to maintain a decentralized Proof of Work (PoW) consensus without the need for wasteful energy expenditure traditionally seen as foundational to the security of such networks.

The paper challenges the conventional wisdom that the security of a PoW blockchain is intrinsically linked to its operational costs or wastefulness. Instead, it posits that the key to a secure and decentralized consensus lies in non-repurposeability; the notion that any attempt to attack the network would undermine the value of the currency to such an extent that it would outweigh the potential gains from such an attack. This principle ensures that miners are incentivized to dedicate their computational resources to the chain, as it represents their primary source of income. By contrast, the proposed Proof of Useful Work (PoUW) model allows for computational efforts to be directed towards productive tasks outside of the blockchain, provided that the economic incentives align to ensure that attacking the network remains unprofitable.

This shift from a dedicated (or "imprisoned") work model to a merge mining approach not only promises greater efficiency but also proposes a fundamental change in how blockchains can secure themselves. It suggests that the security of a network could be maintained if the cost of contributing sufficient computational power to solve the cryptographic puzzles exceeds the potential rewards from undermining the network. This approach flips the traditional relationship between hashrate and price observed in Bitcoin, where the anticipated future profits from mining limit the hashrate. In the PoUW model, the difficulty of the puzzles (or hashrate) acts as a regulator for the value transferred over time, thereby preventing double-spending attacks.

Moreover, the paper introduces a novel concept where the issuance of new coins is directly proportional to the amount of matrix multiplication FLOPs contributed to the network. This proposition not only aligns the creation of value with the provision of useful computational work but also marks a significant departure from Bitcoin’s fixed supply model. It encapsulates the potential for a new kind of blockchain that leverages the vast amounts of computational power available globally for purposes beyond securing the network, thereby offering a compelling and sustainable alternative to the current paradigms in cryptocurrency mining.