Using captcha to democratize mining

[coinwhisperer]

What would the objections be to implement a captcha mechanism in the mining process in order to protect the network from centralization of mining power?
We could for instance say that a captcha should be filled out every time a worker produces a week's worth of hashes on a standard AMD GPU. The captcha should be part of the hash challenge mechanism so that pool operators would be excluded from being able to fill in one global captcha only when they find a block. This means that exclusive p2p mining should be inherent to the protocol.

The first thing that comes to mind is that people with big mining operations would just hire cheap labor to fill in the captchas. This, however, would make mining more expensive for big operators. For small operators it would just represent a minor nuisance.

Could this work?

[1714926035]

1714926035

[1714926035]

1714926035

[1714926035]

1714926035

[1714926035]

1714926035

[1714926035]

1714926035

[thejepper]

no, next to the fact that most captcha's can be solved either by a computer or for a very low fraction of the block payout, it's only a nuisance not a protection. The idea is good to prevent large pools from reaching 51% network power but i think p2p pool options are better alternatives

[michagogo]

You can't possibly have a CAPTCHA in a system like Bitcoin. A CAPTCHA requires, among other things, that there be a computer generating it and keeping the answer to compare to. The question of "is this right?" is asked of whatever generated it and has the answer stored. A decentralized system has, well, no central point that can issue the CAPTCHAs and validate them -- if nodes are generating their own CAPTCHA, you'd be relying on the honor system, and there's nothing to stop modified software from bypassing the CAPTCHA check.

[anti-scam]

If you could find out a guaranteed, decentralized way to mandate that one human takes on one identity in a P2P network then you basically will have solved Bitcoin, but CAPTCHAs are not the answer for the reason you stated and the reasons above.

[tromp]

There is a simple solution to "democratize" mining;
adopt a PoW that takes at least a GB of memory to solve
(with no time-memory tradeoff) and that depends on latency rather than bandwidth.
That would put an end to specialized mining hardware.

[grue]

There is a simple solution to "democratize" mining;
adopt a PoW that takes at least a GB of memory to solve
(with no time-memory tradeoff) and that depends on latency rather than bandwidth.
That would put an end to specialized mining hardware.

scrypt?

[tromp]

There is a simple solution to "democratize" mining;
adopt a PoW that takes at least a GB of memory to solve
(with no time-memory tradeoff) and that depends on latency rather than bandwidth.
That would put an end to specialized mining hardware.

scrypt?

No way, scrypt only needs a lousy 128KB, and even with that is very parallellizable.
Technically, it's not even a PoW since verification is a nontrivial computation,
and that only gets worse when you try to increase its memory footprint.

[Peter Lambert]

There is a simple solution to "democratize" mining;
adopt a PoW that takes at least a GB of memory to solve
(with no time-memory tradeoff) and that depends on latency rather than bandwidth.
That would put an end to specialized mining hardware.

How would that prevent specialized mining hardware? People would just make hardware with huge amounts of memory.

[Analyticse]

If you could find out a guaranteed, decentralized way to mandate that one human takes on one identity in a P2P network then you basically will have solved Bitcoin, but CAPTCHAs are not the answer for the reason you stated and the reasons above.

what you think about make this mechanizm fix some automatics? 

how much gigabit energy also etc this eat

[tromp]

There is a simple solution to "democratize" mining;
adopt a PoW that takes at least a GB of memory to solve
(with no time-memory tradeoff) and that depends on latency rather than bandwidth.
That would put an end to specialized mining hardware.

How would that prevent specialized mining hardware? People would just make hardware with huge amounts of memory.

If your PoW is constrained by memory latency, then you can't do much better than using commodity hardware, i.e. high-end PCs (possibly GPUs, although they tend to do worse at memory latency).
You cannot build an ASIC with dozens of GB of memory.

[ozzymax]

What would the objections be to implement a captcha mechanism in the mining process in order to protect the network from centralization of mining power?
We could for instance say that a captcha should be filled out every time a worker produces a week's worth of hashes on a standard AMD GPU. The captcha should be part of the hash challenge mechanism so that pool operators would be excluded from being able to fill in one global captcha only when they find a block. This means that exclusive p2p mining should be inherent to the protocol.

The first thing that comes to mind is that people with big mining operations would just hire cheap labor to fill in the captchas. This, however, would make mining more expensive for big operators. For small operators it would just represent a minor nuisance.

Could this work?

Why don't you buy more GPU's?
You just feel slighted that you can't find a block with your wiener rig