Its strange to me that people aren't talking about glasses or goggles rather than bulky uncomfortable headsets. There is a big difference between a bulky headset where you can't see or do anything else and a normal pair of hands-free glasses.

When it gets to the point of comfortable, unobtrusive and affordable AR/VR/MR glasses with good hand and head tracking built in, that will be a really different experience than the current standard. Maybe throw in a couple of small convient cameras that can just plug in and adhesive to the wall for external tracking.

And with the high speed Wifi available now, or the possibility of connecting to a smart phone (maybe in a pocket or some type of necklace-like carrier), I don't think you need a lot of compute in those devices.

So imagine you just put these glasses on and get 6 virtual wide screen monitors, turn to the right and its your favorite virtual coffee shop, with people inside with fairly realistic recreations from the capture and expression and eye tracking being done by the external cameras. There is 3d positional audio as well.

Imagine with a small camera on the desk, or built into the glasses, doing finger tracking on virtual keyboards. Suppose someone makes a little 'clicky' membrane that provides a bit of tactile feedback regardless of where you press on it.

Now I only need my smart phone and a pair of glasses. I don't need a monitor or a keyboard. Once people realize that every virtual keyboard can easily be 100% customized for any application, physical keyboards may become passé.

When you have good eye and expression tracking, realistic avatars enabled by high-performance AI models, 3d positional audio, realistic environments.. then there really isn't that much difference from being there in person. And there are a huge number of advantages such as convenience and time saving, being able to instantly "travel" to a popular hangout spot even if you are 400 miles away from your friends, etc.

Pretty much everything you can do now in VR, but make it more comfortable and remove some of the friction points, better integration between applications and with the real world, somewhat more realistic, etc.

Suppose automation and robotics continue to improve. Maybe artificial muscles like https://www.artimusrobotics.com/ are widely deployed, enabling high strength-to-weight ratio mobility and much more dexterous, faster, and more general-purpose robots.

In 20 years people may look back and be amazed at how much time and energy we wasted driving around to do everything in person, and how everyone restricted the majority of their "in-person" interactions to small localities.