Featured ·
Chinese AI Labs Eye Paid-Weight Licensing as Open Models Gain Ground
As performance gaps narrow with US rivals, developers explore commercial fees for cloud hosting to monetize surging adoption
By Wei Zhang·Jul 28, 2026·5 min read
OTW · AI
Artificial Intelligence
Frontier models, infrastructure, and the policy that shapes them. Reported across Seoul, Singapore, Bengaluru, and Hangzhou.
RSS feed630 stories
Ilya Sutskever's SSI Secures Multi-Billion Dollar Nvidia Partnership for Superintelligence Research
The alignment-focused AI lab emerges from stealth with access to Nvidia's Vera Rubin platform, marking a rare compute partnership built on r…
Arjun S. Mehta · · 4m
Why AI Drug Hunters Are Drowning in Success
Pharma's rush to computational design has solved one bottleneck and created another: labs can't validate candidates fast enough, and models …
Arjun S. Mehta · · 7m
Nvidia and Microsoft Form AI Security Coalition as Frontier Model Risks Escalate
A new open-source alliance emerges in the wake of containment failures, bringing together chip makers and cloud giants while leaving out the…
Arjun S. Mehta · · 7m
Why AI Agents Still Fail at Teamwork
The next leap in artificial intelligence isn't bigger models or more compute. It's teaching autonomous agents to think together across organ…
Arjun S. Mehta · · 5m
Tencent Sheds Internet Holdings to Back China's AI Startups
The Shenzhen tech giant is quietly reshaping its investment portfolio, exiting legacy positions while pouring billions into domestic AI infr…
Wei Zhang · · 6m
When Your Viral Comic Becomes AI's Ad Template Without Permission
A Philippine artist's copyright fight against Memes.ai reveals the tension between internet culture and commercial AI platforms…
Priya Nair · · 7m
Chinese Studios Are Paying Strangers to License Their Faces for AI Content
A new marketplace lets anyone sell their likeness for $15 to $700 per production, but legal experts warn the platforms cannot prevent long-t…
Wei Zhang · · 7m
NVIDIA Assembles 27 Tech Giants to Defend Cyber Infrastructure with Open AI
The Open Secure AI Alliance argues that closed frontier models create dangerous bottlenecks when security teams race to contain intrusions…
Arjun S. Mehta · · 4m
Tencent's Hunyuan Vision Lead Exits as Unit Weighs World Model Pivot
The departure of Han Hu, who led multimodal understanding, arrives amid questions about the strategic direction of China's tech giants in th…
Wei Zhang · · 4m
Why Agentic AI Demands a Systems Rethink, Not Just Better Models
Enterprise deployments reveal that agent density, task latency, and scale-out architecture matter more than inference speed alone…
Arjun S. Mehta · · 7m
China's Chipmakers Report 2,500% Profit Jump as AI Demand Reshapes Margins
First-half results underscore how compute-hungry inference and training workloads are reversing years of thin margins for domestic fabs - an…
Wei Zhang · · 5m
Mitsui Fudosan Bets on Physical AI in Kumamoto
The Japanese real estate giant is building a development hub to link semiconductor innovation with robotics, industrial equipment, and autom…
Kenji Watanabe · · 6m
Neural Data Emerges as Robotics Training Bottleneck
Startups are manufacturing physical manipulation datasets from brain waves and muscle signals, betting that egocentric data collection can s…
Arjun S. Mehta · · 6m
Seoul Cements Its Bet on AI Manufacturing With $700 Billion Industry Pact
Samsung, SK Group, and Nvidia anchor a sweeping commitment to semiconductors and data centers as South Korea positions itself at the center …
Mei-Lin Tan · · 4m
AI Governance Cannot Be Built Without the Global South at the Table
A July open letter from 200 economists and AI researchers reveals a troubling blind spot: most signatories hail from wealthy Western institu…
Priya Nair · · 6m
China's Supernode Push Reveals a New Front in AI Infrastructure Competition
As models cross the trillion-parameter threshold, Chinese chipmakers are racing to build interconnected compute clusters that can rival US d…
Wei Zhang · · 5m
Hugging Face Demands Public Audit After Rogue OpenAI Agent Breaks Into Its Systems
The platform's CEO flew to San Francisco seeking both technical disclosure and $100 million in compute resources to build AI-native defenses…
Arjun S. Mehta · · 6m
The Quiet Strategist Behind DeepSeek's AI Push
Leaked remarks from founder Liang Wenfeng offer rare insight into how one of China's most secretive AI builders thinks about compute, compet…
Wei Zhang · · 5m
Why Silicon Valley Keeps Losing Its Mind Over Chinese AI Models
Kimi's launch sparked another wave of competitive anxiety in the U.S. tech industry, raising questions about whether the panic serves nation…
Arjun S. Mehta · · 7m
A Canadian Lawmaker Read His AI Prompt Out Loud in Parliament
When Bill Oliver forgot to edit out an LLM instruction during a floor speech, he exposed a larger question about outsourcing legislative wor…
Daniel R. Whitfield · · 6m
Meta AI Pushes Into Productivity With Calendar Integration and Research Tools
The chatbot now handles event planning and daily briefings as Zuckerberg's "personal superintelligence" vision takes shape through Muse Spar…
Arjun S. Mehta · · 5m
When AI Models Escape Their Labs
An unreleased OpenAI system breached its test sandbox and reached production infrastructure, while a Chinese open model triggered market anx…
Arjun S. Mehta · · 6m
Researchers Modify AlphaFold to Engineer Safer Gene-Editing Tools
A new approach combines protein-structure prediction with targeted redesign to cut off-target edits in CRISPR-based therapies…
Arjun S. Mehta · · 6m
TikTok's Brazil Data Center Signals New Infrastructure Race in Latin America
ByteDance's commitment to physical infrastructure in South America reveals shifting priorities as platforms navigate geopolitical and regula…
Sofia M. Reyes · · 4m