Content
@
https://warpcast.com/~/channel/bad-takes
0 reply
0 recast
0 reaction
Dan Romero
@dwr
Tell me why I'm wrong: Ethereum L1 or L2 gas fees make micropayments too expensive for everyday consumer apps. Less frequent payments? Sure. But 10x a day? No.
32 replies
19 recasts
94 reactions
EulerLagrangodamus
@eulerlagrange.eth
UTXOs can make things pretty cheap. https://vitalik.eth.limo/general/2023/11/14/neoplasma.html Plasma/Validiums are now possible because of advancements in ZKPs. This system doesn’t support arbitrary tx amount though, has to be some standard sizing.
1 reply
0 recast
4 reactions
Dan Romero
@dwr
When would this ship?
2 replies
0 recast
3 reactions
EulerLagrangodamus
@eulerlagrange.eth
@sin is the pioneer for this stuff. He just started a company around these ideas. I’ll let him fill in the details
0 reply
0 recast
1 reaction
sina.eth
@sin
I think you'll start seeing very simple flavors of this soontm (😜) starting with _very simple_ crypto interactions, and slowly increasing in complexity and capability. Once you introduce ZK, you unlock superpowers but introduce a lot of engineering/production complexity.
0 reply
0 recast
0 reaction
sina.eth
@sin
(from my armchair, not a ZK expert) The main outstanding production problem as I understand it is the unit economics of ZK proving. Is it worth saving a user their $0.10 txn fee if it costs our server $0.25 to produce their proof for them?
0 reply
0 recast
1 reaction
sina.eth
@sin
Things I'm thinking can improve the situation: - what can you do with simpler circuits (=cheaper proving) - batch more - more proving on the client (aggregate of all clients can have way more compute than even a giant server) - moore's law and friends
0 reply
0 recast
0 reaction
sina.eth
@sin
In sum, as a daily user I agree with @dwr's statement esp as it pertains to today. I'm very hopeful in some pretty-well-understood but need-devs-to-chew-enough-glass-to-bring-to-life ideas _eventually_ coming to life.
0 reply
0 recast
0 reaction
sina.eth
@sin
A potential corollary here is that ZKRs may be overkill/premature, as they target the extreme of "equivalent capability" and accordingly suffer from the max dimensions of the costs of ZK.
1 reply
0 recast
0 reaction
EulerLagrangodamus
@eulerlagrange.eth
I’m confident you can make fast zkps for this stuff using something like Spartan.
0 reply
0 recast
1 reaction
