I have been looking into ipfs and nostr and both are intriguing. At this point the total decentralization of web 3.0 appears to be repeating the same weakness of web 1.0. The weakness of web 1.0 is that it didn’t really have a good discovery method. Can’t say I’ve given much thought to the extent which Google search contributed to the rise of web 2.0 though. Web 1.0 dealt with the discovery issue by creating communities and link rings. A lot of the old web 1.0 websites were in rings of similar websites. Perhaps a partial solution to the discovery issue in web 3.0 would be to update the search process. An open web requires an open search. Of course to produce decent results the search would probably have to incorporate a sufficiently sophisticated AI to understand natural language requests and produce meaningful results.