A focused editorial channel for robotics companies, products and technologies across borders.
Browse published records by market, product layer and editorial category.
This morning, Jiahui Yu, a core researcher at Meta Superintelligence Labs, MSL, announced his departure from X and plans to start a new company. This is just 8 days after...
The field of physical intelligence is undergoing a rapid transformation in model architecture and algorithms. However, at the same time as the number of models has skyroc...
Recently, a new study by Transluce shows that when Claude recognizes that you're an accurate researcher, it becomes less confident. Things start with a check-in.
Just look at a static image, and these states are almost identical. But in the VLA-RL method, which is responsible for evaluating the current state of the critical model,...
Today, it's easy to produce an image, a video. But what's really difficult is another thing: to make the AI understand and complete the creation according to the creator'...
When humanoids are able to walk, run and recognize everything, but in the face of bottle caps, pinching flexible thin films, precision screws and other delicate operation...
MLS-Bench's new work, covering 12 fields and 140 real-world research tasks, gave an unopposed answer. Even with the full implementation of the most powerful human method ...
The team consists of 28 researchers from 10 universities, and the core authors include Kim Jong-un at the University of California, Berkeley, Han Yang, Zhang Chen of Purd...
Machine minds are releasing multi-modal models, world models and body models are constantly reshaping the boundaries of AI's capabilities, but a more basic problem has lo...
The machine's heart releases the same model, weighs no parameters, but replaces one Harness, and the performance can fluctuate significantly. This looks like the "basics"...
2.5 minutes. Suzhou Aviation Micro Electric has been adhering to space-level standard research and development sensors, from 99-inch three-dimensional touch sensors and 1...
In the field of intelligent driving, Tesla made a choice that year: it didn't wait for an autonomous driving track to land, and it didn't build a data set based on a larg...
The next big thing is that the development of AI will no longer be confined to the boundaries of human knowledge, forming a whole new scaling law for intelligent producti...
Multimodal basic models can learn to "see" and "speak" with images and language through the Internet, but with physical intelligence there are no similar shortcuts. Robot...
In the field of computational advertising, real-time bidding (RTB) is a core paradigm: when users generate a single exposure opportunity in the media, the ad trading plat...
Based on the language comprehension capabilities of the latest generation of the Hy 3 major language model, the Hy ASR 3.0 preview combines high-precision speech recognit...
Imagine a teacher who has been teaching for twenty years, taking an apprentice who has just entered. The teacher is experienced, but if the apprentice is unstable with a ...
Wafer AI recently deployed the Kimi K3 on the AMD MI355X. As a result, it required 16 NVIDIA B200 models running across two servers to complete the deployment on an AMD s...
At the moment, a clear consensus is that the narrative of physically intelligent people is shifting from "will not make a move" to "can't do complex tasks stably in the r...
Why do people use the same body to complete walking, running, swimming, and moving things? The key lies in the combination of muscles and bones. The bones determine the s...