Proof-of-Work = Proof-of-Sunk-Cost
The Proof-of-Work consensus model could be renamed Proof-of-Sunk-Cost because that's what it essentially is. The products of PoW mining are hash function outputs that satisfy agreed upon criteria that cannot be obtained faster or with less cost than calculating them using the mining algorithms used. The point is to provide proof of having wasted a certain amount of energy as a sunk cost to prove that one is properly incentivized to validate a transaction honesty. (If one does not validate a block correctly, no mining reward is obtained. Validation without Proof-of-Work does not afford the same level of security because validation without a cost associated incenvitizes dishonest validators to join a permissionless network en masse.)
For the reason mentioned above, it is untenable to harness all the computational power used to validate transactions to do useful work, at least of the sort that anyone could possibly want to pay money for. But how about work for which there is no market?
Does any blockchain do any useful computation as part of its consensus mechanism?
As far as I know, there is no Proof-of-Work coin that does useful computation for which there is no market as part of the consensus mechanism. However, there is a blockchain that rewards providing computational public goods: Gridcoin. The computation is done for scientific purposes such as protein folding simulations, astronomical calculations, and other scientific work in public health, for instance. But the computational work is not done to secure the network. Gridcoin uses a Proof-of-Stake consensus algorithm. The distribution of newly minted Gridcoins is done according to the computational power provided by each participant to whitelisted projects. The whitelist is maintained by network participants through blockchain voting according to the PoS protocol. The computing architecture used is the the Berkeley Open Infrastructure for Network Computing (BOINC). The purpose of the project is to improve the retention rate of volunteers who donate spare processing capacity for the public benefit.
Gridcoin has a relatively low market cap. Its market cap is about $2.5 million and it's currently at the 818th position at Coinmarketcap.
No way to algorithmically recognize computational work that is useful but has no market for it
The difficulty is that there is no way to algorithmically tell computational work that has market value from computational work that has not. (Computational work that has market value clearly cannot be used as Proof-of-Work a.k.a. Proof-of Sunk-Cost because it can be sold.)
As cool as it would be if a blockchain consensus algorithm existed that did valuable computational work as a pure public good, it does not seem like an easy problem to solve. The whitelisting process of acceptable projects takes human intelligence and a consensus model for voting.
I wonder whether it could be integrated with something like SETI. I'm not so sure PoW is the long-term solution to Crypto anyway, and I am skeptical of SETI's merits, but perhaps signal pattern analysis could be involved somehow.
SETI is a good candidate for useful but marketless computational work. Generally, anything only the public sector can do would work. Trouble is, there are no universally agreed upon criteria for such work.
I am skeptical of any claims that something is public-sector only. The fact that governments have monopolized a given service is not proof that the service in question can only be provided by the government, or that government is the best solution.
Twitter: