Disclaimer: I'm bullish on 3d interaction because I own company related to this, so my opinion might be a bit colored. What surprised me the most is how fake a

Disclaimer: I'm bullish on 3d interaction because I own company related to this, so my opinion might be a bit colored. What surprised me the most is how fake and forced the conversations and interactions looked. It reminded me of those first blue screen effects you see in these old movies from the sixties. I would assume that Facebook/metaverse spent an enormous budget on it, but nevertheless the outcome still looks creepy. In my opinion we have mostly figured out how to avoid the uncanny valley for rendering, so I'd guess improving conversations and interactions are up next. Until the techniques are on par with what we need, I'd heavily avoid trying to model a real world representation in a virtual world. (Think about the early Pixar movies: instead of trying to create the most realistic world possible, they embraced their current constraints and turned them into a feature that made their movies instantly recognizable.) This being said, I see one big use case for AR/VR adaptation during the next decade: digital twins with simplified, tailored representations for development, marketing, education and operational purposes, as well as amusement. My best guess and personal bet is on this space. And even though we've had VR and more recently AR getting more accessible to the consumers, it's still at least 5 years too soon for mass adaptation; most people still consider it a novelty. In my experience interactive 3d will emerge next in the mass-market. Up until now I had the best conversion ratios by having a conventional 2d GUI/wizard and some very conceptually simple 3d drag and drop to create simplified 3d representations of products one could buy (think sheds, carports, verandas etc). First consumers will have to get familiar with 3d annotations, before we can make them interactive as well... So yes, I see a future for (some version of) the metaverse, but the implementation they showed right now was a bit meh, because they tried to model it just like the real world, and the techniques to do so properly still aren't there yet...