21–30 of 325 posts
Gonna require "the cloud" so they can collect free machine learning model training data and spy on us, I assume?
Human nature for some reason often devolves to “what’s the point?”
Here’s the 2015 announcement. There’s a lot of potential. Maybe we’d all benefit from “imagine what’s now possible “
This is a beautiful piece of tech for any ad company looking to collect reaction information to ads (like google): * It works at low power. Turning on the front camera and processing the visual data generated needs large amounts of power. * It can't be blocked as easily as a front facing camera. * People are less aware that they are being surveilled, which makes them less motivated to block it. * Yes, it's not captur…
This is HN. This is cool technology. Can we just stop to appreciate what cool new interfaces or game concepts we might be able to build with this, rather than jumping on the knee-jerk Google hate train?
I honestly don't get why you'd need a motion sensor on a handheld device. Let alone the limited application of such an input, you'd literally need to keep the device stationary for the motion sensing to be effective. It's a usability nightmare. > https://www.youtube.com/watch?v=KnRbXWojW7c This ad not only fails to sell the feature but also makes using the touchscreen look easier.
Good lord... this is a really cool piece of technology and half the comments here are just complaints that it will enable better ad tracking or erode privacy or that touch screens are good enough. This is HN. This is cool technology . Can we just stop to appreciate what cool new interfaces or game concepts we might be able to build with this, rather than jumping on the knee-jerk Google hate train?
Gonna require "the cloud" so they can collect free machine learning model training data and spy on us, I assume?
No, it’s doing all the heavy lifting on device, only sending the compressed output to the cloud for spying, err, offering you a better service.
A lot of the leaps in high-fidelity human-computer interaction (voice, face, and likely this new gesture system) have been made with having enough data about real-world interactions to train the models. It's how a company finds out about the ten thousand things that happen in the real world that their lab-models missed, and gets their algorithm from 90% accuracy to 99.9% accuracy.