The problem I see is you're going to want a camera built-in for vision reasons for your amazing reality-overlay, and at that point, well, you've got a camera built-in.

▲

wolvoleo a day ago | parent [-]

I'm sure you could do that without one. Gyro, accelerometer, compass, GPS, step counter, altimeter. Should be accurate enough for basic navigation. Especially with some smart dead reckoning algorithm that calibrates itself at known map points like when you turn a corner. Showing notifications shouldn't need any kind of AR awareness at all. You could just show them above the normal field of vision just like the Google glass did.

Again there the problem was not the display, it was the camera. And Google glass didn't even use it for any tracking purpose.

I don't think the issue is that it can't be done without the camera. I think the issue is that the whole product exists to get those cameras out there. Data is the new gold, those vision AIs need to be trained. So they've never even tried without one.

▲

Nursie a day ago | parent [-]

Yeah you could definitely have a go.

Where is the exact line - i.e. can you use Lidar? Infrared depth-sensing? Or do these provide too much data such that the scene could be recreated?

(I'm exploring this as a thought experiment, in general I agree that people shouldn't be carrying hidden cameras on their faces, and if those cameras are at all connected to Meta then it's much worse!)

	▲	wolvoleo a day ago \| parent [-]
		Well lidar in that form factor would end up just being an 8x8 laser DoF sensor like some smartphones have. There's no space or power budget for a real lidar. That would be ok I guess. That's not enough to capture much of anything even with a continuous feed.