Classical mining, which occurs thanks to the Proof of Work consensus algorithm, such as Bitcoin mining, is necessary for ensuring the security and stability of the network, as well as for preventing double-spending. In general terms, the process is needed to calculate the hash of a block header in a specific cryptocurrency's blockchain, which includes, among other things, a random number (nonce), the hash of a set of transactions, and the hash of the previous block. If everything goes well for the miner, they find the correct value and receive coins for the work done—the block reward plus transaction fees.
The problem, often mentioned by proponents of alternative consensus algorithms, is as follows: to obtain the coveted reward, a miner has to perform a lot of "useless work" (the aforementioned hash calculation), which also consumes a lot of electricity. The problem can be attempted to be solved in at least two radical ways: by abandoning mining (for example, using a PoS consensus) and by making mining provide additional benefits (hence the name Proof-of-Useful-Work).
In a broad sense, Proof of Useful Work (PoUW) is an approach to a consensus model where the computational work of miners, necessary to secure the network, simultaneously solves tasks that have value outside the blockchain.
The idea of PoUW is not new, as developers have been searching for a "useful arrangement of mining" for many years. In 2017, a group of researchers from different countries—Marshall Ball, Alon Rosen, Manuel Sabin, and Prashant Nalini Vasudevan—published a paper titled Proofs of Useful Work.
Problems of PoUW
Despite all the nobility and attractiveness of the goal pursued by proponents of the Proof of Useful Work concept, achieving it in the most effective way in practice turns out to be not so simple. Of course, it would be great if the conditional brute-force search of SHA-256 hashes could easily be replaced by computations in fields like medicine/astronomy/AI, and also contribute to the stable growth of a cryptocurrency's exchange rate. However, some difficulties lie along the way:
-
The Problem of Proposing Useful Work. In Bitcoin mining, hash brute-forcing is uniform, meaning participants all over the world perform exactly the same action. The network difficulty level gradually regulates the amount of effort a miner needs at a given time, depending on the total hashrate. On the other hand, it is quite difficult to imagine a similar and uniform piece of useful work that other participants can perform in parallel. If the work is not uniform, the complexity of tasks will constantly vary, and miners will find it profitable to perform only easy tasks, ignoring more difficult ones.
-
The Blockchain Trilemma. For blockchains, there is a problem often called the scalability trilemma. The essence of the triple problem, omitting details: ensuring an equally high level of decentralization, scalability, and security of a distributed network simultaneously is either impossible or incredibly difficult. At best, you can improve a maximum of two out of three characteristics. In a network with classical PoW, block verification is trivial. In PoUW, a node must verify the solution to a useful task, which can be significantly more complex and time-consuming. If verification becomes a bottleneck, scalability and security suffer. On the other hand, if the useful work is not uniform, and several participants start deciding what is useful at the moment and what is not, decentralization may suffer.
-
Uniqueness of Solution. Just as the hash value of each new block in PoW is unique, tasks in PoUW must also be unique for each block. That is, you cannot allow something to be solved once and then present that solution multiple times later.
-
Predictable and Uniform Miner Effort. An enthusiast mining cryptocurrency must participate in a competitive and fair competition with others for the reward. Otherwise, the consensus loses its economic foundation. It is also unacceptable for minor improvements in mining equipment to provide a colossal advantage. Plus, reliable validation must be ensured so that a participant is not rewarded for submitting a formally suitable answer that does not carry real usefulness.
Finally, there is a serious problem of economic incentives and the economic underpinning of the cryptocurrency. Let's delve into it in more detail.
The Problem of the Economic Model
Is mining really exclusively "useless work"? Although mining cryptocurrencies using Proof of Work was first implemented by Satoshi Nakamoto, the creator of an early version of the concept should rightly be called the cypherpunk and founder of Blockstream, Adam Back. He proposed the idea in the Hashcash system to combat email spam: the user had to perform a small amount of computation, which would become a barrier to mass spam mailings. In other words, complex computations themselves are needed to ensure the operation of another useful function.
In Bitcoin, complex computations are complex precisely because they effectively perform their task—ensuring the stable operation of the system. In addition, miners need to bear real costs (buy expensive equipment, pay electricity bills, etc.) to mine new coins.
From this perspective, the miner essentially exchanges their money, knowledge, and time for maintaining the operability of the entire network. These expenses and risky investments are factored into the "cost of Bitcoin"—to one degree or another. If miners stop taking risks and buying expensive equipment, bearing other costs, it will also affect the cryptocurrency's exchange rate. In other words, the high cost of Bitcoin is partly explained by the enormous economic costs borne by miners.
For greater clarity, we can also give this analogy: a CAPTCHA can also be considered conditionally useless computation. Surely it would be more useful if people, solving a protective computer test, could somehow help science or medicine. But the main task of the "useless CAPTCHA" is to protect against spam attacks and prevent hacks. That is, the procedure does not necessarily have to be useful in principle.
Project Examples
Several interesting blockchain projects can be highlighted that have attempted, to varying degrees, to implement the PoUW concept:
-
Flux: originally associated with GPU computations, but transitioned to the Proof of Useful Work v2 model in October 2025. As the developers note, today Flux acts as a decentralized Web3 cloud infrastructure. Although the project moved away from the original ideas of useful GPU mining, in its current implementation, nodes still perform real computational work: hosting, rendering, AI tasks, and much more.
-
Primecoin: this once well-known cryptocurrency among enthusiasts, launched in the distant 2013, became the first project to combine mining with computations having practical value. The proof-of-work system was based on searching for chains of prime numbers ("Cunningham chains")—thus, it was useful for mathematicians. Later, the developers reported several times about achieving world mathematical records.
-
Gridcoin: this cryptocurrency also appeared in 2013. The developers used the open distributed computing platform BOINC from the University of Berkeley. The protocol is designed so that participants receive rewards for completing tasks in a number of scientific projects. However, conducting such work somewhat differs from the PoUW concept, as the network's base consensus is proof-of-stake (PoS).
It is worth noting that there are other decentralized networks that redistribute computing power from some participants to others for performing useful tasks, for example, Bittensor for machine learning or Render Network for GPU rendering. That is, in a broad sense, such architectures intersect with the aforementioned PoUW concept, although they do not reproduce it in full.
Conclusion
Proof of Useful Work strives to combine the security of classical mining with solving real-world tasks, but so far faces serious technical and economic limitations. Despite this, the idea itself continues to evolve, and many new blockchain projects confirm a steady interest in the useful utilization of computing power.
end-content
