SpaceEngine is Google Earth for the Observable Universe (Google Earth VR is pretty cool too). It places celestial objects that we know of where they belong and procedurally generates what is unknown based on how many stars / planets are estimated to exist. The sense of scale in VR lends itself really well to this experience. You can control where you move and how fast you move.
I was genuinely in awe as I pointed away from earth into deep space and started accelerating. I know space is vast but it's just so hard to conceptualize. You increase the throttle and you realize that even travelling at light speed doesn't feel all that fast. At that rate it still takes hours to leave our solar system. But the shocking part is when you increase it far beyond that. Cranking it up to 100x the speed of light, then to 1 light year /second, how about 100ly/s, faster still. I don't remember what my final cruising speed was when I realized that I wasn't seeing stars pass by - those were galaxies. Maybe 100M ly/s or something. And as you're going you can stop, pick a galaxy, in that galaxy pick a star, have a look around that system, pick a planet, land on it, advance time and watch day night cycles. There was an overwhelming sense of how big the universe is and just how small I am as I watched galaxies fly by at impossible speeds.
SpaceEngine is buggy. Everything in VR is buggy and half works. Menus are terrible. UIs are baffling. Settings are either hidden and esoteric or oversimplified with no real control. This is part of what makes it seem so promising to me though. There is so much low hanging fruit to be picked. Everything I've experienced in VR feels like a castle in the sand. And not just because DMCA's will eventually remove much of the current content. There is so much potential to grow and extend what is currently available. Platforms where users can make their own content currently exist but it feels like it's almost by accident. You can imagine future systems making it even easier / more open. I think WebVr is one trying to do this but I've never used it.
The low hanging fruit isn't just the software either. Everything I've mentioned above is with todays consumer hardware. The improvements in the pipeline for VR hardware are crazy. Things that are available as Dev kits, prototypes, and commercial hardware blow my mind.
In VRChat you can join via VR or a flatscreen. When people are not connected via VR you can sort of subconsciously tell at all times because there is no body language. It's as if communication has some sort of inherent fidelity where on the high end you have standing in a room with a real person talking to them and on the low end you have something like a text message. People who are connected via flat screen feel a little closer to the text message side of that slider scale. But there are also people who use special hardware to enable full body tracking, or finger tracking. Interacting with them feels a little higher on that fidelity scale.
Full body tracking is currently done by a piece of hardware clipped to your belt and or shoes. In the future I suspect it will be done with cameras or something a bit more seamless. But it's not just body tracking. There are haptic suits where you can feel when someone taps you on the shoulder, or throws a dodgeball at you. There are very early alpha haptic gloves about which I could write another comment of similar length (this video was incredible to me[0]). There is experimentation with tracking facial expressions and eye movement. Facial expressions will go a long way towards increasing communication fidelity but eye tracking turns out to offer even more potential than that.
The resolution of the current gen VR headsets just isn't that good. It's better than the old Rift CV1 but is still noticeably less clear that a computer monitor or television. However, with eye tracking you can do foveated rendering. This allows you to track where the eye is looking at any given moment and render very high resolution in that spot, and lower resolution on the rest of the image. I have not yet experienced this myself but people who describe testing this out report a remarkable difference. I remember one article in which a guy described getting into a BMW in VR and looking at the logo on the steering wheel he said it looked shockingly similar to seeing it in real life. Thats a special example and not first hand so I take it with a grain of salt. But with foveated rendering the resolution can be on the order of 10x higher. Really extreme improvements possible here.
I've always been more of a software guy personally. The potential of the hardware is exciting to me most of all because of the crazy possibilities it enables in what could be build in VR tomorrow. It's not quite Moore's law, but there is a sense that the hardware will get better for a while.
I once heard someone call VR/AR "the last interface" because any interface that currently exists or could exist can be built in VR. I don't know if thats entirely true but it gets at what makes VR so compelling. The limit is really only our imagination. This is true of other digital worlds too, like video games or to some degree even websites. But thats just through a screen. Here you can conjour into existence places for people to go, people to be, or adventures to experience.
It's not all adventure games and toaster people either. There may be even more low hanging fruit in building tools to do work in VR. Thats even less explored. A VR IDE would be pretty cool, especially once the resolution gets a little better. Or imagine if you could walk into a virtual datacenter and see all your AWS servers manifested as physical boxes racked and waiting for you to trip over power cables like it's the 90s again.
VR is a new medium. It doesn't feel like reading a book, playing a video game, or watching a movie. It feels like something else. And like all new mediums what exists there now is mostly ports of old mediums. I don't think we've even scratched the surface of whats possible in this space. There are probably whole undiscovered classes of ideas that really only work in VR the same way there are ideas that make great books but don't really cut it as movies. It just has so much potential.
It feels like the early web to me. An untouched frontier. Almost completely unexplored and teeming with possibility. What could be more promising than that.
If you get to try (or play everyday) one of the more common VR games like beatsaber, the lab, job simulator etc. those are definitely still fun too and I think still are a window into what VR can be. Rec Room is another popular one you may find that even has some user generated content (though with more limits on things like size). Really almost anything in VR as long as it's on one of the PC based headsets will show you what I'm talking about.
Thanks for asking this question. Looking at the other comments this is one of my favorite HN threads in a while.