I'm in that slice of the population that is captivated by the potential of crypto while disgusted by everything that it currently is. I would like to see more

I'm in that slice of the population that is captivated by the potential of crypto while disgusted by everything that it currently is. I would like to see more discussions around the theoretical limits of the value of web3 with regards to computing efficiency. Even the world's most efficient public blockchain would see a huge replication factor if it saw widespread adoption. The problem is, the larger the replication factor, the larger the inherent cost of compute, as limited by the energy it costs to replicate that compute. In other words, if AWS Lambda is priced at X% premium over the electricity consumed executing code, even the world's most efficient blockchain would be priced hundreds of times that premium if the blockchain is widely adopted/replicated; there is no free lunch, so if you have a distributed system replicating data across thousands of machines, then the people bidding to put data on those machines are paying for that replication. In other words, the pricing of running your application backend on a popular blockchain (say, Ethereum) will never ever be as cheap as running your application backend on a public cloud. In a broader sense, this means that the only data that shows up on a public chain is data considered "valuable", such as recording the transactions of cryptocurrency or NFTs or other pyramidal investments; never something silly and worthless, like a system for recording nontradable in-game achievement system. On the other hand, not every blockchain is popular. My belief is that IF web3 has a future, it will involve inter-chain protocols between many cooperative blockchains, and that compute will go into a chain that fits the value of the compute. For example, suppose I want to make an app that uses a microtransaction scheme on the order of pennies, e.g. you can tip 1 penny to an author or something like that. I'm obviously not using Ethereum because a transaction on Ethereum costs vastly more than 1 penny to record. However, a smaller chain with a replication factor of say 10 machines can probably handle 1 penny transactions with far less than a penny fee. Right now, there's no good way to treat blockchains themselves as fungible compute objects that are secured & priced based on their replication factor. Just as you wouldn't move a single penny on Ethereum, you also wouldn't move a million dollars in tokenized assets on a chain with only a few machines replicating. I think making an ecosystem of fungible blockchains, with the ability to easily transfer funds inter-chain[0], is going to be the next evolution of web3 that actually enables real consumer application backends to be hosted on the crypto universe. [0] This is something that's being worked on in a limited sense, e.g. the "Wormhole" project that allows you to tokenize and move assets between Ethereum and Solana and a few others