You call it the AR Cloud, but it seems to me arnet (AR+Internet) is a name that fits better with the capabilities required of this new medium.
In my opinion the challenge is this you just mentioned: a DNS for reality. We have it in 2D and it’s called an address, same as on the web. But the issue with ARnet is this needs to be a space (an area in 3D). And the reason I see this as a challenge is because it needs to be regulated world wide and then hardware restricted.
Here’s some challenging questions to think about:
1) Should 3D points belong to the owner of a property? So for example if I’m in a restaurant, should I be able to place AR objects on the walls? Intuitition says I shouldn’t, except if allowed to.
So if that space is sold/or owner dies, should the new owner also receive the rights to own / edit the existing AR objects in the space?
2) What happens when there’s too many objects? Should we be able to only add an object if there’s no other object there?
3) Should types of AR (like government AR indications on roads) have a higher priority and not be able to be modified by anyone? So basically nobody should be able to alter AR displayed on public spaces either?
Considering 1, 2 and 3, where will everyone create experiences together? Seems to be a space that’s becoming more restricted, the more you think about it.
4. Whoever may own the arnet (ideally the same institutions that own the internet) should make it possible for AR objects to be created by anyone. An open arnet, in other words.
I think the cloud should allow people to create objects that are displayed on demand (aka real time) when someone is in that area looking around. So the coordinates and an object reference could be stored in the arnet, but the actual data (the object, the data associated with it, everything else) stored on a startup’s database, on a standard server.
5. Now the fun part. You didn’t write about this in your article. AR that is not static. For example one that is following an object, or is looking at a static object, inside a moving car. Example — a car with an AR notification on top showing it’s your Uber.
The issue with this is that the Reality DNS doesn’t apply anymore. Or there’s layers on top of 3D layers with other rules.
All this seems like a huge problem to tackle. At the same time, AR that is not persistent and cross-device seems like a missed opportunity.