Topic: Proof of Useful Work? - page 2. (Read 584 times)

It is just another proof of work.  Proof of works can be useful if there is some group of people trying to find all solutions to a problem.  For example odolvlobo is referring to a public good.  One public good can be a list of all prime numbers.  If you need a prime number of a certain length as a proof of work, a byproduct of this would be recording a lot of prime numbers.  Prime numbers can be useful.

Another way to create useful proof of work is to make the problem incentivize coming up with solutions that are otherwise useful.  For example I could make the proof of work hash large amounts of data so that technology is invented that can hash large datasets faster.  Or I could make a cpu-only hashing algorithm so that more CPUs are produced, and those could be repurposed after their useful life to make laptops for struggling countries. Or if I make every hash hash the entire blockchain instead of   just the block  I incentivize more full nodes.  Using this endpoint bitcoin is already useful because it provides a financial reward to try to crack cryptography.  If cryptography is ever cracked we will know about it because we will see bitcoin moving when it shouldn't.

So anyway this paper is just another proof of work using orthogonal vector problem to solve problems that some people care about.

I'm just glossing over this topic but my first guess is that they are getting rid of the "useless" work such as the hashes which do not ultimately result in a block being mined, right?

How do you even do that without assigning the PoW problem to a predetermined set of nodes?

Thanks for your time checking and writing this!

Yeah, I felt this limited understanding too. I couldn't go on after the second page. There must be economic flaws in this mechanism as it sounds really promising. The paper was written in February this year and I have a feeling it will be recognized by the market.

And then I ask myself:  Shouldn't they break it down for the rest of the people to understand it? Isn't the purpose of such mechanism to be understood from a decent percentage of its users? How will it be evaluated?

After looking at the paper, I feel that I have a limited understanding of what is presented. It presents a formal methodology for adapting mathematical problems to proof-of-work and presents one such problem as an example. Whether or not solutions to the problem presented are actually useful is not clear to me. Other than my criticism below, the paper looks solid, but it is outside of my area of expertise, so I'll let someone else do the real analysis.

My general criticism of PoUW relates to the potential economic flaws in the concept, which is something that I don't see addressed in this paper.

First, PoW is necessary to the security of the Bitcoin protocol through the rule that the "true" chain is the one that required the most work. In Bitcoin, that work is not actually measured in hashes or joules. It is measured in time. The difficulty adjustment mechanism ensures that no matter how much effort is spent on mining, a block always takes 10 minutes of work on average. Thus, the difficulty adjustment mechanism is also a necessary component of the protocol, at least for Bitcoin.

Second, miners are incentivized to increase (or decrease) their mining capacities until the cost of mining approaches the revenue.

Now, I believe that the idea that converting wasted work into useful work would reduce the overall net cost of mining (or increase the overall net benefit) is flawed. If the result of mining provides additional value from useful work, then miners would be incentivized to increase their mining capacities until the cost of mining again approaches the revenue plus the additional value of the useful work. In the end, the overall cost of mining would be equal to the cost of the wasted work plus the value of the useful work. Thus, nothing has been gained.

Furthermore, if the value of the useful work is greater than the cost of wasted work, then the mining incentives would be dominated by the value of the useful work. Though I haven't thoroughly examined the result of this scenario, I suspect that it could also result in the failure of the protocol.

There is a potential way to resolve these flaws. If the useful work is a public good, then the miners would gain no value from it and there would be no incentive to increase mining capacity based on the value of the useful work. The problem is that it is not clear that a true public good can exist. The protocol would be a failure if miners discovered a way to exploit the value of the useful work that was originally thought to be a public good.


I like to point out this misconception whenever it comes up. That is not how Bitcoin (or any other PoW coin that I am aware of) actually works.

I recently received a message about a(nother) cryptocurrency that has implemented/will implement a new type of mechanism called “Proof of Useful Work”. The paper explains that PoUW aims to achieve usefulness during mining which can be used in artificial intelligence and in treatment of patients.

However, I can't seem understanding how can this work and because I didn't bother much, is there anyone who has heard of it and understood it so they can share their thoughts here? It sounds nice in theory, but in practice I'm sure it'll just lead to the creation of another pump & dump crypto.