Virtual reality. No, really it'll work this time. No, it will.

It is the year of virtual reality, again. This is something like the thirtieth straight year that it has been the year of virtual reality, again. I've been hearing about how VR is going to improve all our lives and revolutionise computing and entertainment since I was 8. VR is to my generation what jetpacks and robot butlers were to our parents; the distant promise of technology rendered impractical by economics and common sense.

There are almost no use cases for VR where the indignity and impracticality of the solution doesn't completely outweigh any utility. Where they do exist they are so extreme that there's simply no way a regular citizen will encounter them. Here are the perennial favourites:
  1. Firearms training for law enforcement
  2. Surgery training for remote medics
  3. Support in emergency situations, maybe.

These use cases have something in common. Practical experience is difficult and dangerous to come by and coaching rare and expensive. The impracticality of a VR headset is overmatched by the benefit derived from its use in both of the top two.

The problem with the third is that it's likely that to provide emergency support you would have to be able to have what is currently quite an expensive device lying idle, be able to keep it up to date, power it up, make people aware of its location and be able to broadcast the correct information to it with a low latency at very short notice appropriately pitched for someone probably in a high state of alarm. It's a fantasy, a VR headset is not a fireaxe. Whilst it's cocnvievable that paramedics might carry a remote support kit that included a VR headset in future the cases in which they might use it are obviously limited and are in fact likely to slow them down; they deal in immediacy and with people. Cutting them off from communication and from what's going on around them could easily prove disastrous.

I'm being a little disingenuous here because in talking about this we need to make some key differentiations. VR is normally mentioned in the same sentance as its much more desireable and harder to implement younger brother Augmented Reality. The two get mentioned together but are radically different. VR relies on virtualisation - creating entire environments and seeking to put your senses into them. AR takes the environment you are in and overlays things on it. VR is virtual not reality and AR is reality plus virtual. To add some confusion into the mix is immersive video which blurs the lines but is probably the ultimate destination of the technology currently being developed for virtual reality. In this context immersive video is probably best described as Limited Reality.

Understanding the utility of each as a consumer product really takes a bit of description of what each is best at and some examples:
  • VR - is best used in situations that it would be difficult or dangerous for people to go or in other situations where the actual environment in which things happened is unavailable to the human sensorium for some reason; it happened in the past, too quickly or slowly for us to witness, deep underwater, on Mars or it doesn't exist yet, defusing a bomb etc.
  • AR is best used where you need to be grounded in what's going on around you and additional information can add value to your experience. Walking around somewhere new and being given information on your route, surroundings and people you are about to meet
  • Immersive Video is best used where you would like to witness something but do not want or need to take an active part; attending a sportsball match from your lounge and feeling like you're in a seat on the halfway line or behind the netted arch thing. You can look around yourself and potentially tap into things like replays and player profiles but the experience is otherwise relatively passive.

There are very few use cases of VR that aren't better accomplished by well implemented AR. Even the firearms training above for the simple reason that how you move around your environment is one of the largest parts of that particular challenge. As you can't fully take your body into the virtual world an AR overlay on a regular course is probably better suited. Not only is it easier to make the physical challenge properly present but your brain is much more likely to accept the lo-res high speed graphics currently available because they're situated in a world you are actually experiencing. Also the nausea that VR can cause in giving your brain contradictory sensory information about your orientation, velocity etc would be gone.

Which brings us to the problem of how real we want the virtual world to become. Playing a first person shooter game in VR sounds like it would be amazing, I mean, that's the use case isn't it? Right, because what most gamers actually want is to be landed in a war. If you dial up the reality high enough that's where you end up, and it's not a game any more. There have already been studies carried out that using VR can increase your empathy for the situation you are in no matter how lo-fi the simulation. So what effect is a really good simulation going to have? Experiments with horror games and immersive video have already produced enough terror for some participants to tear off the headset and run away. The more real VR is the more things like violence become a problem for the user and the less the current publishing models function; the apparently unshakeable use case of gaming begins to look more like something more prosaic and sedate; darts or bowls, or more likely something we haven't seen (the appeal of) yet. In fact you could argue that things like the Nintendo Wii already demonstrate this to a degree.

The problem with VR is that its use case is not entertainment it is an educational and industrial tool that hasn't been built out or properly thought out yet. I'd argue that the ultimate end of the technology currently being developed with the VR buzzword on it is likely to be immersive video. Much less participatory - an 'on rails' experience - but allows the use case to remain quite strictly in a recognisable entertainment paradigm meaning that current publishing modes are much more likely to funnel money into to the development and distribution of the product. It is also a bit more sociable and less isolating as if for example you are watching a basketball/tennis/football match you could do so with a two way audio feed with a friend remotely or (heaven forefend) have them in the same room. You could synchronise the experience and activate replays and 3d versions of things like the snickometer or hawkeye from different angles exploring the event in a way that isn't even available to those attending in person while maintaining social contact. Expect to pay for this though. Events organisers, sports franchise owners are not going to let the rights to that out without substantial reimbursement.

I've given quite an optimistic viewpoint in the paragraph above; looking at it from the opposite perspective people wouldn't put on a pair of lightweight spectacles to watch 3d TV which is a failed idea chiefly for that reason (although among many others). The benefits of immersive video are going to have to be fairly substantial to outweigh the awkwardness of putting on a full headset and keeping it on for the full length of a sportsball match. Imagine how frustrating doing this could likely be for people who've got used to it to support people who haven't yet. It would be like using Skype with your parents, only more difficult. "How do I get to the replay?" "How do I change to the other side of the stadium?" "Which referee cam was that from?". Advertising also becomes a slight issue; what does an immersive video ad look like, how dirty would you feel after being in one, and just how powerful or alienating would they be? What happens when Ashley Maddison get their grubby mits on a immersive video advert?

All these faults though discomfort or awkwardness, lack of familiarity and a monetisation method that would have to be drastically reworked, it feels like there are solutions. Interfaces can be learnt, components made lighter and more comfortable, advertisements made more subtle and to play to the strengths of the medium. This a much more recognisable, accessible medium and product that has a reasonably solid platform to build out from, though it has a very long way to go and a lot of infrastructure to build out to get to usable.

Which brings us to AR which is almost all use case but it's often difficult to implement, can require substantial computing power to begin with and often freaks people out.

We tend to think of AR as quite visual but as with most transformative technologies the steps are much less obvious as they incremental in everyday life and considerably more simple than the future image we're often sold as a final destination. My favourite here is one you've likely used or seen used already, turn by turn navigation instructions given over audio. If you own a smartphone and a car it's pretty likely you've used this already.

Even AR which, well implemented, could deliver immeasurable benefit is probably not the final picture here. VR, immersive video and AR are steepping stones to a more integrated and subtle experience. As always I've allowed myself to get distracted by the toys and the depth of over-optimistic tech bullshit that accompanies something finally looking like it might work.

This is a long enough post already. AR is another story for another time and as its best executions are still emerging I think I have time to develop my thinking...