What are the current unsolved problems in the VR/AR space?

AR -- headsets are too expensive but nowhere near good enough. See

https://kguttag.com/

VR -- the hardware is fine for sedentary activities. if you want to get up and walk around the problems get big. tracking is one thing, but you have to make the whole environment safe (no tripping on couches, pets, etc.)

I would love to rent out an old big box store and turn it into a "VR park" where people would play physically intense games like laser tag. You have to worry about people crashing into structural supports, taking hard falls wearing $3000 worth of gear, running past barriers you enforce with the VR system, etc.

What holds up VR now is that the software is not that compelling. VR games aren't really better than regular video games. The best applications so far are "watch a movie in an imaginary movie theater", "use your computer in an environment with many virtual screens".

User interface and modes of interaction. Current patterns still revolve around 2D content and laser pointers. We don't have decent user interfaces built around handling "physical" objects. The HW is quite diverse and it's hard to build for. Accessibility design is not yet explored (most applications just assume you have two fully functioning hands).

For locomotion around larger VR spaces, the teleportation is current state of art. The academic community has worked on "redirected walking" which makes best use of limited real space and distorts the 1:1 relationship to make wide VR areas reachable. It is still not widely used and allegedly causes nausea with users.

Big companies are also working on tele-presence being as natural as possible, with eye tracking and mirroring of facial expressions. Facebook is trying to tackle uncanny valley problem with cartoonish avatars and it seems to be working well.

Most of VR/AR development happens within walled gardens and this is preventing cross-compatibility and wider reach. Sony's PS5 force feedback controllers would be amazing for AR/VR tactile interactions, but patents make this impossible.

There is also the ethical danger of Big Tech spying on people through cameras, of filtering the AR camera feed for their purposes, of detecting sub-conscious tendencies through eye tracking, and subtly rewarding users when they make desirable choices.

1. User interaction paradigms that are intuitive and feel great. I don’t consider this a particularly insurmountable problem, I just find that too many game designers are failing at it, and I don’t know why. The general user experience or choice of controls and gestures is the main reason why I tend to buy VR games, play them for 20 minutes, get frustrated, and never come back to it. And don’t tell me to “git gud”. I’ve been an avid gamer for 35 years, and I’ve worked on UX at HoloLens and Oculus. Then again, maybe I’m too old and set in my ways and I’ll never leave mouse+keyboard. After all, I never did figure out how to FPS with a console controller, and apparently others did.

2. Business justification for creating VR-first AAA games and major MMOs. I am done with paying $20-$30 for extended tech demos. If you check the new releases for PC+console, you stand a chance of seeing titles people are excited to play. Check the VR new releases and it’s still a scant trickle of trite garbage nobody cares about.

I don't work in the space but my intuition tells me that the problems are mainly hardware related.

These technologies have a lot of promise but are currently limited by the state of the art. My personal opinion is that neither of these technologies will really take off until they are put into a stylish form factor - imagine a pair of ray-bans with embedded screens and chips that was nearly indistinguishable from the regular kind.

Naturally, there are a lot of problems in the way of developing such a thing, including optics, miniaturization, durability, energy storage, etc.

Authoring cools for content. Particularly for the non-technical, creator crowd.

VR controllers can get occluded from tracking cameras. Quest Pro supposed to have cameras on each controllers as well, but I think there could be some ideal tracking method out there.