This page is to explain a vision of the metaverse and refine its details according to that vision. You may be surprised by some of the conclusions if you haven't thought this through.

Metaverse Principles

• Persistent
• Synchronous
• Live (Dynamic)
• Presence
• Interactive (Agency)
• Economy
• Decentralized
• Peer-to-peer
• Self-extensible
• Self-moderated
• Private
• Shared
• Scalable
• Secure
• Composable
• Programmable
• Social
• Content-addressible

The first principle I'd like to look at is self-moderated. In combination with "decentralized" and "peer-to-peer", this means that users can ban people without going to a third party. To be fully uncensorable, there can be no global bans. But this also means there can be no global "whitelist" of people who are always seen. If every person can make changes and build without permission, but anyone else can block that person and ignore all changes they made, then it means there's also no central consensus on reality. Everyone lives in their own bubble universe, with changes made by their friends and friends' friends; realistically, there will be a market for curators who collect and maintain whitelists and blacklists of content and other people. There will also be an option to enable AI pattern recognition to auto detect and filter undesired content. There will always be a way to turn off all the filters, to allow everything, though if you do this, you're opening the door up for all sorts of offensive content.

Scopes will be broken down into narrow spaces, so there won't really be that many merge conflicts. Each client will compose multiple spaces together. Someone changing one thing won't affect something far away.

Synchronous. As J. Allen Brack said [ https://www.youtube.com/watch?v=0Wrw3c2NjeE ], "You don't want that, to do that, either. You think you do, but you don't." Matthew Ball thinks the world has to be synchronous [ https://www.matthewball.vc/all/themetaverse ] and waits for technology to allow hundreds or thousands of people in a room together to interact in real time. Bandwidth to transmit their actions and GPU resources to draw their avatars might not be worth the cost of this, even if truly desired. Jak the gorilla really doesn't like instancing [ https://www.youtube.com/watch?v=cWHVUPSgHiM ] and wants to know that when he goes to a place, that it will be the same version of that place as anyone else there. I hate to say to these people that this is impossible. It's impossible in the same way that the design of the Bitcoin network can't guarantee 100% that your transaction won't be rolled back. That is, it's for all intents and purposes usable: you can synchronize to any extent desired, it just takes effort, and at some point, you'll say good enough. It could be that in order to get 1 million people in a courtyard able to see each other and interact in realtime, that they actually have to go meet IRL. This is okay. But even if the bandwidth and GPU resources become cheap enough that it's possible to simulate a world with this many people in realtime, there's other reasons you might not want to. Everything digital possibly leaves a trace. Being "live" with someone makes your actions subject to traffic analysis: Someone could prove you were online at that time. Being proven to be a real person at a virtual place at a certain time makes it impossible to say you were somewhere else at that time. It might be best to leave some ambiguity as to whether a person or a bot is in control of your avatar. This gives you plausible deniability as to your whereabouts. The more certainty you leave with your actions, the less anonymity you give to everyone else.

Say you were in a field with 1,000,000 avatars. How many of them could you simultaneously monitor and detect if they were human? If you can't tell, then how do you know they're not bots? What's the reward to coming to a concert with 1,000,000 bots? Wouldn't it be more valuable to have the concert public domain, with a full 3-D recording available to everyone to re-watch and re-participate in anytime they want? Then the concert could never end!

There's something to say for having long-term conversations and interactions. Coming back to a forum post years later ("necro-ing") and responding constructively should be allowed. Same with coming to a virtual forest and writing your name into a tree (or a floating sign) "I was here" and the date. Correspondence chess is a valid way to play. E-mail discussion threads can have a deeper and more thoughtful conversation than on so-called social media, where things scroll by so fast and attention shifts to the next meme your voice becomes lost.