3 ms·

Sorry, doesn’t work. Useful work is not “useful”. It is too difficult to meter correctly for anything beyond a trivial case, and too unknown for security modeli

Sorry, doesn’t work. Useful work is not “useful”. It is too difficult to meter correctly for anything beyond a trivial case, and too unknown for security modeling. Primecoin. Riecoin. Gridcoin. It’s been tried. It’s not a new idea, it doesn’t work.

randomhodler84 5y ago
One such issue is such interesting “proof of work” work is typically general purpose compute, which leads to an incentive to steal general purpose compute for criminal minded individuals. Generally useful POW leads to botnets. By using specialty algorithms like SHA256 we both provided an accurate hw bound to security and - a continued investment in the state of the art ASICS.
- mrkramer 5y ago
  What's the definition of useful work? I would say work needs to be useful "enough" in order to be worth doing. It's still early in the POUW but I think in the future better solutions will be developed. Proof of work leads to stealing of general purpose computation power only in the short term because as difficulty rapidly raises "honest" miners are pressured to specialize fast in order to profit meaning moving their mining operations to specialized server farms. Once that happens it is no longer profitable for cybercriminals to use their botnets for mining since difficulty is so high they earn very little and because of that they are more likely to switch their botnets to other "traditional" cybercrimes such as SPAM and DDoS.
  - randomhodler84 5y ago
    It’s not early days for it, I feel it’s been done and done and failed every time so far. I think the fundamental problem is most useful work, which we can define as work that can be used in another context rather than just validating difficulty — eg protein folding, is very difficult to reason about the energy requirements of a “valid solution”, and linearly scaling the difficult to adjust to work magnitude. Useful work tends to be expensive to validate also, whereas verify for SHA256^2 POW is trivial. Ideally you want easy to reason, linearly scaling and highly asymmetric create/validate.

randomhodler84 5y ago

One such issue is such interesting “proof of work” work is typically general purpose compute, which leads to an incentive to steal general purpose compute for criminal minded individuals. Generally useful POW leads to botnets. By using specialty algorithms like SHA256 we both provided an accurate hw bound to security and - a continued investment in the state of the art ASICS.

mrkramer 5y ago
What's the definition of useful work? I would say work needs to be useful "enough" in order to be worth doing. It's still early in the POUW but I think in the future better solutions will be developed. Proof of work leads to stealing of general purpose computation power only in the short term because as difficulty rapidly raises "honest" miners are pressured to specialize fast in order to profit meaning moving their mining operations to specialized server farms. Once that happens it is no longer profitable for cybercriminals to use their botnets for mining since difficulty is so high they earn very little and because of that they are more likely to switch their botnets to other "traditional" cybercrimes such as SPAM and DDoS.
- randomhodler84 5y ago
  It’s not early days for it, I feel it’s been done and done and failed every time so far. I think the fundamental problem is most useful work, which we can define as work that can be used in another context rather than just validating difficulty — eg protein folding, is very difficult to reason about the energy requirements of a “valid solution”, and linearly scaling the difficult to adjust to work magnitude. Useful work tends to be expensive to validate also, whereas verify for SHA256^2 POW is trivial. Ideally you want easy to reason, linearly scaling and highly asymmetric create/validate.