The significance of Web5 lies in drawing a clear boundary visible to people: the boundary between non-custody and custody, high cost and low cost, censorable and censorship-resistant, self-sovereign and platform-sovereign.
Web5 = Web2 + Web3. Web2 is low-cost with good user experience but censorable, while Web3 is high-cost with poor user experience but censorship-resistant. Web2/Web3 represent two different goals, two different set of trade-offs, two worlds that need cooperation as well as clear boundaries. Without recognizing this point and always trying to pursue two conflicting goals within one architecture, the final result will be either getting nothing or having one devour another - usually the one optimize for short-term devour the one looking for long-term. As we see today, so-called decentralized applications endlessly compromise security and censorship-resistance for the sake of pursuing a web2-like experience. Only by acknowledging the existence of boundaries can both Web2/Web3 become better.
The boundary between Web2 and Web3 should be reflected in economic and technical architecture. The economic boundary is evident in the free experience of using Web2 versus the high costs of using Web3, including on-chain resource usage costs (such as transaction fees, on-chain computation fees, and rental for on-chain space), node operation costs, private key management costs, etc. The technical architecture boundary lies in whether censorship-resistance can be achieved. Only users willing to pay higher costs can truly enjoy the world of Web3 with fully owned, censorship-resistant assets. If user experience and low cost are prioritized over security and censorship-resistance, centralized or semi-centralized systems would be a better choice.
Explicit boundaries are what I believe differentiate Web5 from Web3. The pursuit of clear boundaries in Web5 can enable a positive collaboration between the two worlds, making both Web2 and Web3 better. In contrast, in the worldview of Web3 or that everything should be on-chain, boundaries are vague, always trying to fit different goals and trade-offs into one or many blockchains, attempting to convince people that user experience, low cost, security, and censorship resistance can all be achieved simultaneously. In fact, this approach only makes Web3 increasingly resemble Web2.
How can Web5 make both Web2 and Web3 better? In the world of Web5, web5/3 should be open, verifiable, and censorship-resistant systems, such as L1 chains or L2 (rollups, channels) that can provide security and censorship resistance matching L1. This part trades costs for high security and censorship resistance. web5/2 is not censorship-resistant but can still achieve openness and verifiability. Openness means open-source code, migratable private data, accessible and replicable public data; verifiability means any node or client can locally verify requests/transactions/data related to itself are correct, such as nostr, atproto, DHT, semi-centralized/permissioned blockchains.
Only on this basis there’re choices. Web5 developers can choose and combine web5/2 and web5/3 tech to meet the needs of their respective scenarios, providing the best experience through web2 and secure, censorship-resistant assets/contracts through web3. Users with different needs and cost preferences can choose different web5 applications respectively, without being forced to accept a Web3 that increasingly resembles Web2.
few people noticed the