China has the ability to iterate on hardware the way we do with software which produces some really unique devices that basically wouldn't be made here due to the economics involved. The Rabbit R1 feels like this kind of device.
AI HARDWARE
-
Tensors Multilinear Operations Universal Approximation Machine Learning
By
–
Tensors are back! I am so glad to see this! We have been advocating that tensor or multilinear operations are all you need. The benefits: universal approximation and expressivity. Higher order BLAS operations for parallelism without the need for any transposition or data
-
AI Model Trained on BOTW Style for Vision Pro
By
–
Had to split the video into images since the Vision Pro won't record when the pano is open! This is using a model trained on BOTW style.
-
EverArt AI Launches Pano Feature for Immersive World Creation
By
–
You can now create your own immersive worlds in your favorite style on @everartai thanks to our new Pano feature. ✨🥽
— Pietro Schirano (@skirano) 5 février 2024
Design immersive scenes and dive into them with your Vision Pro!
Saved images will be recognized as a compatible Panorama. 🖼️ pic.twitter.com/wTqOzTyVZNYou can now create your own immersive worlds in your favorite style on @everartai thanks to our new Pano feature. Design immersive scenes and dive into them with your Vision Pro! Saved images will be recognized as a compatible Panorama.
-
Apple Screen Streaming Compatible with Apple and Intel Chips
By
–
According to the Apple website, streaming of the screen works with any Apple or Intel-based chips.
-
Advanced Gaze and Face Tracking in VR Hardware Evaluation
By
–
I got one, to evaluate its possible applications for our current projects. The basic UI experience is very familiar from the Quest (and internal projects within Google). What’s new is the uncompromising implementation of expensive nice to have features like gaze and face tracking
-
Google MobileDiffusion Enables Instant Text-to-Image Generation
By
–
Google released MobileDiffusion, a new approach that allows for near-instantaneous text-to-image on mobile. The announcement states the model can generate a 512×512 high-quality image in half a second on mobile… and is only 520M parameters. https://
x.com/xiaohuggg/stat
us/1754336686849626510/video/1
… -
Apple Accelerates AI Strategy with On-Device LLMs and AppleGPT
By
–
We know that back in December, Apple quietly published a breakout paper on a technique to run LLMs on iPhones. Plus, they've been running AppleGPT internally for some time now. With Vision Pro launch finally over, is Apple finally going all in on AI?
-
Why Advanced AR VR Hardware Costs Significantly More
By
–
A gaming laptop may cost 8x the price of a Switch, a professional camera may cost 8x the price of a point and shoot. This is fine. It’s very expensive to get eyetracking, gesture recognition, environmental scanning and real time rendering to work well.
-
MacBook Screen Replaced with VR Display Experience
By
–
Your macbook screen turns black, and instead you get one in the VR. There are slight tremors when you move, a bit of color fringing, and slight lag, but the quality is good, and it's very usable.