This most likely does not pull directly from Wikipedia. Wikipedia has rate limits on how much you can link to them. It looks like they are storing the images as a cache in the cloud. Then they are loading them into the individual rooms. There are some systems that allow multi user experiences. For example, there is the Uncensored Library in Minecraft which houses a variety of texts which multiple people can be in at once. You could easily put avatars in this which could walk around inside the museum. In my own project, I have an avatar which walks around in a 3D environment with books. I like the way this is laid out. This looks like a MVP right now. It probably will have many iterations.