Here is my Devcon talk arguing that Ethereum should not completely outsource L2 development, as this risks "Cosmosification." Instead, Ethereum should deploy 128 highly interoperable zk-based "native rollups."

Hi Martin, I really appreciated this talk and the courage it took to propose it. It would be useful to understand how the UX would look like for setting up a AMM pair using one asset from native rollup 1 and one from rollup 128 on an AMM that runs on rollup 67. Seamless?

Right, so you would need to send assets from 1 und 128 to 67. Rollup 67 can only accept the assets once the correctness of 1 and 128 is proven. With todays tech this takes minutes but there is already a clear path via ASICs specialized on proving to reduce that to seconds, ideally below 12 seconds. So ideally the user can just sign one transaction object containing: sending assets to 67 and setting up the AMM on 67. In the best case this can be executed within on block (12 sec) but it might take a bit longer.

you know what? the more I think about this, the more I actually agree with you...

Sharding with extra steps :) I am actually both for and against this idea. My main against is that L2 teams can ship much faster than EF so more of a culture issue than anything else.

and I don't want L2 teams to stop innovating and shipping faster - but I want Ethereum to provide a solid baseline for every developer that favors stability and reliability over innovation.

Been waiting for this video to come out. They said you roast base at the beginning 😆

I have high respect for Base and I think having Base is a net positive. But I also think one needs to understand the difference between building on Base and building on Ethereum.

Tornado Cash T-shirt👍🏿

Sure , if that means no fiefdoms, no L2 tokens , no censorship , no parasites , MEV for everyone ...

GOAT