I think DePIN might play an increasingly important in America’s attempted industrial rebuild (and indeed industrial rebuilds elsewhere). For that, we’d need hyper-performant chains to enhance coordination to service capital-intensive industry, and this notably makes me think of Solana. Need coordination mechanisms that are fast and cheap and credibly neutral. Ethereum and its broader ecosystem might do well with DeFi, but not sure it has the tech specs to compete with a Solana and, perhaps, a SUI over the long term for functions beyond basic DeFI.

Please don’t hate me for posting this. 

I love you.

Agree that Solana has the edge on dePIN, disagree that there's any reason but marketing when megaETH and RISE exist

RISE has fucking 11ms block latency and is currently hosting a full-ass MMORPG on their testnet in real-time so... Yeah, it's not a tech advantage it's distribution (which is arguably more important)

All good points. I guess I’m not convinced by the L2 scaling model — too many questions over data fragmentation and associated security concerns, along with sequencer centralization and ultimately the incentive for the L2s to break away from Ethereum altogether. And I’m thinking that the dominant models are going to be monolithic chains as well as app-chain setups that have a regional focus to capitalize on latency advantages, such as Hyperliquid and its validators hugging data centers in Japan

Monolithic scaling seems to have been a smashing success thus far — Solana does more transactions than all chains combined on any given day, which, again, is a huge success for that model — and there’s a lot more in the pipeline in terms of further scaling. What hasn’t proven itself out nearly as successfully are L2s, rollups, etc. Still a long ways to go for monolithic chains to fulfill their potential, too.

The counterpoint is simple: you cannot infinitely scale a L1. There is some practical limit. You can think that limit is outside the realm of common practice, but I don't think that's correct. In practice, there are pretty large scaling needs, and that's not even to mention the heterogeneity of application/regulatory limits

Monolithic sclaint is doomed