Advertisement
Vincent Chow
Senior AI reporter, Tech
Vincent Chow is a technology reporter covering frontier AI models, AI governance and US-China competition. He previously covered Chinese society and was awarded a SOPA award for his culture reporting.
Vincent Chow is a technology reporter covering frontier AI models, AI governance and US-China competition. He previously covered Chinese society and was awarded a SOPA award for his culture reporting.
Areas of Expertise:
AI models, AI governance, Chinese societyLanguages Spoken:
Cantonese, English, Mandarin
UC Berkeley’s RDI centre earlier this month introduced Agents’ Last Exam, a new benchmark that tests how well AI agents perform.
Financial Secretary Paul Chan discusses probe on three brokerages, HKIC’s Clara Chan reveals investment vision, at closed-door event.
China’s advantages in energy and industrial applications could erode America’s AI ‘moat’, according to University of Hong Kong academic.
AI models that understand the workings of the physical world and can simulate it will give the technology more human-like powers and drive embodied intelligence.
Advertisement
In just a few months, Chinese AI models have risen from near-zero ‘evaluation awareness’ to within striking distance of their US counterparts.
The commerce department’s mandate forces a shutdown of top-tier systems after officials reportedly found ways to bypass safeguards.
Anthropic says its Mythos model is edging closer to ‘recursive self-improvement’ as Chinese developers seek to close gap.
Eight Unitree G1 humanoid robots performed an elaborate dance routine on America’s Got Talent, alongside a Chinese dancer from Sichuan.
The tech giant’s paywall gamble may have backfired after teasing its paid-subscription options for flagship chatbot Doubao saw users bolt.
Gu Quanquan helped oversee the tech giant’s release of its latest flagship AI model, Seed 2.0.
The tech trio team up to create a state-of-the-art robot reference design allowing researchers to build, fine-tune and deploy skills faster.
MuleRun lowers adoption hurdles, positioning Alibaba’s agentic AI as a safer alternative, with reach across 43 countries.
Its global user base reaches 300 million while ARR more than doubles in two months, though company reports large total losses.
US AI companies have restricted access for users in Hong Kong and mainland China, forcing banks to navigate a compliance minefield.
Model ranks among world’s best for intelligence per dollar thanks to now permanent price cut, according to Artificial Analysis.
In the AI era, the problem with being a technological follower is the constant risk of having the rug pulled from under you, Liu says.
Move underscores start-up’s efforts to build harnesses – software that turns an AI model into a functional and autonomous AI agent.
A wave of demand for agentic AI will prove to be a pivotal ‘inflection point’ and vindicate high capital expenditure, Eric Sheridan says.
American firm’s warning about China’s tech capability seen by some as fearmongering as hopes of bilateral cooperation on AI safety rise.
The funding round was led by venture capital firms GV and Greycroft, with ‘major participation’ from Nvidia and AMD.
IDC projects that China’s AI cybersecurity industry will be valued at US$8.7 billion by 2030, a more than 37-fold increase from 2025.
China and US have different goals for artificial intelligence and ways of developing it. Washington’s curbs on tech exports have widened the gap between them.
On DeepSeek’s chat interface, a new ‘image recognition mode’ has been added alongside the ‘expert’ and ‘flash’ chat modes.
Case shows US-China competition has expanded beyond chips and models to areas of engineering and product design, analyst says.
Researcher Chen Deli is emerging as DeepSeek’s new public face as speculation over the whereabouts of the company’s founder and CEO lingers.
Economic planner asks both parties to withdraw from deal for AI firm, in move with wider implications for other start-ups operating overseas.
Chinese state broadcaster CCTV said on its influential social media account that Mythos showed ‘unprecedented cyberattack capabilities’.
DeepSeek says its V4 model will have throughput issues until the second half of the year, until Ascend 950PR supernodes ‘ship at scale’.
The company says its cost-efficient new V4 model is competitive with top closed-source models from OpenAI and Google DeepMind.
