_________

The year has just begun, but it’s already shaping up to be pivotal for autonomous driving. In China and at the Consumer Electronics Show (CES) in Las Vegas, key industry players have unveiled new AI models that abandon language training and “go physical” by learning directly from videos.

On January 8, 2026, Xpeng has announced its new model VLA 2.0 which removes the language translation step and uses direct visual-to action generation. Founder He Xiaopeng said 2026 could become “the first true year for autonomous driving” in both China and the U.S., while announcing a rollout roadmap for robotaxis, conventional vehicles, and Volkswagen as a launch partner.

A similar technological route based on removing language "as a crutch" and relying on world models trained on massive amounts of video has been announced by Geely at CES in Las Vegas through its G-ASD (Geely Afari Smart Driving) system and WAM architecture.

Nvidia has also embraced this direction for autonomous driving. CEO Jensen Huang launched the company’s vehicle AI “Alpamayo” at CES, saying it was trained “end-to-end, literally from camera-in to actuation-out.”

The Nvidia founder said global tech competition is rapidly shifting from digital to physical AI, with much of this battle to be fought on the road. "The ChatGPT moment for physical AI is here", Huang announced, who found Mercedes-Benz as a first partner.

What Xpeng, Geely and Nvidia are simultaneously establishing in the industry amounts to a new paradigm for physical AI models for autonomous driving. It is a major shift away from traditional VLA models that translate human instructions into movement towards a new technology of large-scale simulation and learning via images. It marks the difference between memorising rules and living by them. ➤➤➤Read more.

_________

Artificial intelligence has taken over mooring operations for container ships longer than 300 metres at Qingdao Port. A large model calculates factors like wind, waves and currents in real time, while AI guide powerful vacuum suction pads that attach to the vessel’s hull.

The system blends multi-sensor fusion perception, intelligent decision-making algorithms, and real-time motion tracking with physical technologies like hydraulic actuation and high-negative-pressure vacuum suction.

These pioneering AI-driven docking operations are part of China’s ‘smart port’ blueprint. The mooring time per vessel is reduced from 20–30 minutes to under 30 seconds, allowing operators in control rooms to handle more than ten additional vessels per berth per year, generating significant economic benefits.➤➤➤Read more.

_________

Lenovo and Nvidia have teamed up to launch a joint “Lenovo AI Cloud Gigafactory” initiative, presenting it at CES in Las Vegas on January 6, 2026. They aim to help cloud service providers and enterprises speed up the deployment of AI.

With Nvidia delivering the GPU backbone for these new AI gigafactories, for example its new Blackwell/Vera Rubin racks, and Lenovo bringing its considerable hardware resources to the table, for example its Lenovo Neptune liquid cooling technology, the partnership certainly has the attention of the cloud industry and large enterprises in need of hyperscale AI solutions.

Beijing-based Lenovo became the world’s largest PC maker by shipment volume after first acquiring IBM’s PC business and later its x86 server unit.

Nvidia, meanwhile, is often called the world’s most important maker of AI chips. “When the ‘king of devices’ meets the ‘king of computing power,’ a transformation in how AI moves from the cloud into real-world use officially begins,” one commentator in China remarked.

While the most recent wave of attention on artificial intelligence has focused on algorithmic breakthroughs by the likes of OpenAI and Google, the next wave could be dominated by companies that best solve the practical challenges of deployment, and this might well be infrastructure vendors like Lenovo and Nvidia.➤➤➤Read more.

_________

Chinese neuroscientists see promise in brain-computer interface (BCI) technology based on ultrasound. Chinese investor Peng Lei, a successful serial entrepreneur who was once acqui-hired by Alibaba, has founded the new company "Gestala" in Chengdu, according to Chinese media reports.

"We believe that ultrasound-based, whole-brain readable and writable BCIs represent the next core technological direction," Peng Lei said at the recent GeekPark Innovation Conference. He also said the name of his new company was inspired by the German word "Gestalt".

By relying on ultrasound waves rather than electrical signals recorded by electrodes, as used in clinical trials currently underway in the U.S., scientists hope to gain access to wider regions of the brain. “When it comes to the brain, the whole is greater than the sum of its parts,” Peng said.

By venturing into this cutting-edge field of neuroscience, the Chinese scientists and startup founders are competing head-on with Elon Musk and Sam Altman in the United States, who have founded competing startups in this field.

Except for bringing hope to millions of patients who are paralysed or suffer from neural diseases like ALS worldwide, BCI is also considered as a promising new application of physical AI, eventually promising all kinds of scientific progress..➤➤➤Read more.

_________

China has begun modernising its electrical grid with artificial intelligence, aiming to implement a comprehensive “Source-Grid-Load-Storage”System. This means that the country's power technologies are experiencing the most profound transformation in a century.

The initial success of China’s green transformation, with a rapidly rising share of renewables in the energy mix, has created an urgent need to better control volatility in the grid. Through holistic system planning, combined with AI monitoring, the government has now begun to better integrate wind and solar power.

Some municipalities got a head start. Shanghai has already implemented an “AI + Guangming” initiative, combining virtual power plants, a fully digitalised energy management platform, and large models to manage its grid.

The new “AI+ Energy” initiative from the central government in Beijing, underway since September, aims to integrate “more than five large models” into the national energy system by 2027.

As China leads the world in installing wind and solar power plants, it urgently needs to create a more flexible grid to ensure that this daily and seasonally unstable supply is matched with equally fluctuating demand. Central planners in Beijing are now employing AI to completely modernise the energy system within a few years.

➤➤➤Read On.