5 ms·

Well this isn’t new is it? I recall chat from 2015 where people were pointing out that mining is just bad. Even when Ethereum started growing, one of the big ar

Well this isn’t new is it? I recall chat from 2015 where people were pointing out that mining is just bad. Even when Ethereum started growing, one of the big arguments quickly became that proof of stake would be a superior approach. I agree. I think the use cases have never materialized. I do believe there’s potential in peer to peer, or at least federated approaches. Email works. Matrix works. Irc works. But I don’t think that a globally replicated ledger is the way to go. Especially with the ambition to use it as the basis for a currency that’s expected to be used for every day purchases.

rjurney 6y ago
What is needed is an AI digital currency where training deep networks for representation learning has a side effect that mints coins without costing a lot of extra resources. Then mining coins is useful and we get more better AI to enslave humanity.
- XargonEnder 6y ago
  I'm confused by how much I agree with you
  - rjurney 6y ago
    I think its a no brainer and possible doom scenario. Is it possible? There are common operations like... the L1L2 norm that are used across machine learning. Is there a way to craft an operation like a regularizer that is both useful and profitable? Or could up you use a block as an initialization parameter and somehow make use of it? A prediction task itself might work - solve this provably hard problem and you own the cash. I don’t know but it seems like what is needed.
    - abrichr 6y ago
      There is some work in this area: https://sentient.org/ https://sentient.org/ https://arxiv.org/abs/1903.09800 https://arxiv.org/abs/1903.09800 https://www.researchgate.net/profile/Andrea_Merlina2/publication/337586007_BlockML_a_useful_proof_of_work_system_based_on_machine_learning_tasks/links/5e03b2de299bf10bc379616b/BlockML-a-useful-proof-of-work-system-based-on-machine-learning-tasks.pdf https://www.researchgate.net/profile/Andrea_Merlina2/publica... https://arxiv.org/abs/2001.09244 https://arxiv.org/abs/2001.09244 https://link.springer.com/chapter/10.1007%2F978-3-030-59595-1_4 https://link.springer.com/chapter/10.1007%2F978-3-030-59595-...

rjurney 6y ago

What is needed is an AI digital currency where training deep networks for representation learning has a side effect that mints coins without costing a lot of extra resources. Then mining coins is useful and we get more better AI to enslave humanity.

XargonEnder 6y ago
I'm confused by how much I agree with you
- rjurney 6y ago
  I think its a no brainer and possible doom scenario. Is it possible? There are common operations like... the L1L2 norm that are used across machine learning. Is there a way to craft an operation like a regularizer that is both useful and profitable? Or could up you use a block as an initialization parameter and somehow make use of it? A prediction task itself might work - solve this provably hard problem and you own the cash. I don’t know but it seems like what is needed.
  - abrichr 6y ago
    There is some work in this area: https://sentient.org/ https://sentient.org/ https://arxiv.org/abs/1903.09800 https://arxiv.org/abs/1903.09800 https://www.researchgate.net/profile/Andrea_Merlina2/publication/337586007_BlockML_a_useful_proof_of_work_system_based_on_machine_learning_tasks/links/5e03b2de299bf10bc379616b/BlockML-a-useful-proof-of-work-system-based-on-machine-learning-tasks.pdf https://www.researchgate.net/profile/Andrea_Merlina2/publica... https://arxiv.org/abs/2001.09244 https://arxiv.org/abs/2001.09244 https://link.springer.com/chapter/10.1007%2F978-3-030-59595-1_4 https://link.springer.com/chapter/10.1007%2F978-3-030-59595-...