It's also a very niche market and you end up completely captive in their ecosystem if you do this. It's reasonable to assume that there'll still be a market for iPhone and Android apps 5 years from now, I don't even know if Magic Leap One will still exist in 5 months or if they'll ever achieve commercial success. Yeah it'll probably be vastly easier to be the top ML developer but it won't do you much good if you end up with 12 users and a dead platform.
IOs/Android development is reasonably mature now and they're backed by huge companies (and other smaller companies who make dedicated tools for that ecosystem). It would make sense to build on top of that instead of starting from scratch.
Also, price-wise you'll never be able to compete with smartphones when it comes to bang-for-the-buck. The economies of scale are huge. These AR/VR need as much processing power as you can throw at them, it's very hard for a small team to come up with a competitive offering, especially for an embedded wearable solution. If they were making their own custom SoC with specialized hardware to enhance the AR experience it might make some sense, but as far as I understand they're only using off-the-shelf components. It does seem like a strange design choice.
> It's also a very niche market and you end up completely captive in their ecosystem if you do this.
During my masters in Interaction Design I kind of "accidentally specialised" in gesture interfaces (the pointing kind, not the touchscreen kind). So obviously I have an old-school Kinect, and later got a Leap Motion[0]. This is exactly the problem with the latter.
Your only real options are either doing research, or doing some kind of custom job like designing an AR/VR installation for a museum.
You can use the Leap Motion in AltspaceVR and Bigscreen already, I hope the Tabletop Simulator developers will support it natively with intuitive gestures one day.
IOs/Android development is reasonably mature now and they're backed by huge companies (and other smaller companies who make dedicated tools for that ecosystem). It would make sense to build on top of that instead of starting from scratch.
Also, price-wise you'll never be able to compete with smartphones when it comes to bang-for-the-buck. The economies of scale are huge. These AR/VR need as much processing power as you can throw at them, it's very hard for a small team to come up with a competitive offering, especially for an embedded wearable solution. If they were making their own custom SoC with specialized hardware to enhance the AR experience it might make some sense, but as far as I understand they're only using off-the-shelf components. It does seem like a strange design choice.