News for hackers. Stuff that matters.

A comment on It takes a lot of energy for machines to learn. Parent comment

Re:It takes a lot of energy for machines to learn by _0ffh on Wednesday December 30, 2020 @06:36PM (#25587655)
The energy cost is really only a problem for on-chain transactions. Technologies like the lightning network can greatly mitigate that problem.
[ Parent | on HN ]
Re:It takes a lot of energy for machines to learn by paulryanrogers on Wednesday December 30, 2020 @07:19PM (#25588013)
Doesn't that mitigation introduce more centralization though?
[ Parent | on HN ]
Re:It takes a lot of energy for machines to learn by _0ffh on Thursday December 31, 2020 @11:22AM (#25594028)
Not that I'm aware of. The lightning network seems to be a fairly decentral affair to me. Still I suggest you make up your own mind by perusing https://lightning.network/
[ Parent | on HN ]
Re:It takes a lot of energy for machines to learn by paulryanrogers on Thursday December 31, 2020 @01:45PM (#25595668)
So the arbitration still occurs on the blockchain? Seems like maintaining all these different ledgers adds a lot of complexity when a better solution may be a non-Bitcoin, or even non-blockchain DB.
[ Parent | on HN ]
Re:It takes a lot of energy for machines to learn by _0ffh on Friday January 01, 2021 @07:00AM (#25602661)
Sure. But that's still potentially multiple orders of magnitude fewer on-chain transactions than without. I'd call that a quite efficient mitigation.
[ Parent | on HN ]