Pocket universes
640k worlds ought to be enough for anyone
Originally published as a twitter thread. I’ve added a few links out for rabbit holes.
I am a WORLD PLURALIST. The metaverse vision was disastrously monist, a product of the centralisation egregore that birthed it. We don’t need giant monolithic spaces where everyone is forced together into some uncanny valley monstrosity to fight for ad revenue…
we should expect an overwhelming plurality of worlds; perhaps each of us could own hundreds, like little POCKET UNIVERSES. This is already true of less interesting ‘worlds’, of which you own many. Not just in games but for projects, hobbies, groups, events…
it’s perhaps understandable that when hearing ‘world’ we reach for something that looks like a 1st person video game, but it’s also missing the wood for the trees… like being told about TikTok and imagining Hollywood…
instead these worlds will range from tiny single-use things like a napkin for an idea at dinner, to sprawling intergenerational spaces that do perhaps look pleasant enough that you’d pay to be uploaded there when you die…
to make these worlds, we need three things:
PHYSICS UTILITY FIT
a HARDNESS DIAL, and
TRUST INFRASTRUCTURE
what on earth am I talking about? …
PHYSICS UTILITY FIT: The physics of a world should be a good fit for whatever the world is for. Yes, on some level, first you must invent the universe etc etc. But you can have too much universe. The deeper the physics the more you need physicists…
HARDNESS DIAL: The exciting thing about AWs is our new found ability to instantiate immutable laws. But not every world needs immutability for every parameter. We need better thinking around what exactly should be fixed and changeable in different types of world…
TRUST INFRASTRUCTURE: We have to thread the needle between monolithic trustless spaces on one hand, and a retweet into inscrutable cozyweb hideouts on the other. Worlds need access control that restores our ability to safely develop real trust and real relationships…
For a much deeper look at trust infrastructure, I highly recommend reading Arnaud Schenk’s ‘On Trust Infrastructure’ and Baz’s ‘You should design trust infrastructure’