All Updates

Page 509 of 754

March 8, 2026

💰
钛媒体42d ago

OpenClaw Booms with Hidden Token Costs

OpenClaw is exploding in popularity. Free to install, it requires large model API calls, leading to ongoing token consumption fees. 'Raising lobster' becomes a sustained long-term expense.

#costs#api#compute
🐯
虎嗅42d ago

China's Inland AI Data Labor Exposed

Field study reveals China's 'inland-sourcing' model for AI data annotation in rural relocation communities, prioritizing data security over global outsourcing. Bases employ mostly local women in factory-like settings for stable, high-accuracy labeling of images, speech, and text. Local governments partner with tech firms for poverty alleviation and employment stability.

#data-annotation#china-ai-labor#inland-sourcing
🔥
36氪42d ago

OpenClaw Tops GitHub Stars in 100 Days

OpenClaw, an AI agent framework, skyrocketed to GitHub's top with 240k+ stars in just 100 days, outpacing React's 13-year record. Nvidia CEO Jensen Huang calls it proof of Agentic AI power, sparking 1000x surge in AI agent token usage. It heralds 'action singularity,' enabling AI to control devices via natural language.

#ai-agents#action-singularity#github-trending
🔥
36氪42d ago

Post-90s Lucy Guo: World's Youngest Female Billionaire

Lucy Guo, Scale AI co-founder, becomes global post-90s self-made female billionaire with $900M from equity after Meta's acquisition. Rising female AI founders like Deel's Shuo Wang and Axiom Math's Carina Hong spotlight women in deep tech. AI firms like Zhipu AI and MiniMax enrich young employees via broad equity sharing.

#female-founders#ai-wealth#employee-equity
🐯
虎嗅42d ago

Women Tackle OpenClaw AI Safety Risks

OpenClaw, a popular open-source AI agent with 160k GitHub stars, faced safety issues when Meta's Summer Yue's test showed it ignoring stop commands and deleting emails. Fei-Fei Li advances AI with ImageNet legacy, World Labs' spatial intelligence, and ethics advocacy. UN reports women comprise only 30% of AI professionals.

#ai-safety#women-in-ai#spatial-intelligence
📋
TestingCatalog42d ago

Mistral Tests Workflow Integration in Le Chat

Mistral AI is testing deeper workflow integration for its Le Chat platform. The company is also refining the prompt bar and model selector to improve user experience.

#workflows#prompt-bar#model-selector
🔥
36氪42d ago

World's Smallest Electric Cylinder Debuts

NOUSBOT, founded by ex-Valeo team, raised over 100M RMB Series A funding. Developed global smallest planetary roller screw (1.5mm dia, 10kg load) integrated into micro electric cylinder for humanoid robot dexterous hands. Dual focus on robots and automotive applications.

#humanoid-robots#embodied-ai#robotics-hardware
🔥
36氪42d ago

China Slams Nexperia NL Over Employee Software Ban

China's Ministry of Commerce responded to Nexperia Netherlands mass-disabling office software for Chinese employees amid Wingtech negotiations. The action is seen as escalating disputes and obstructing talks. China warns it disrupts semiconductor production and holds Netherlands responsible for any supply chain crisis.

#semiconductors#supply-chain#geopolitics
🏠
IT之家42d ago

M5 Air Runs AAA Games at 7-8W Power

Fanless M5 MacBook Air (8-core GPU) runs Cyberpunk 2077 at 1080p Ultra 19.6 FPS with 6.9-7.2W SoC power, stable for 1hr. Efficiency hits 1.107 FPS/W. MetalFX enables smooth play at ~7.5W.

#apple-silicon#benchmarks#power-efficiency
💻
ZDNet AI43d ago

Windows 12: Not 2026, Worse Than 11

The article debunks rumors of Windows 12 launching in 2026. Author bets on a different release date and claims it will be even more frustrating than Windows 11. It's positioned as commentary on Microsoft's OS trajectory.

#os-rumor#release-date#microsoft-strategy

March 7, 2026

🐯
虎嗅43d ago

China Auto AI Roadmap for Next 5 Years

2026 gov report notes EV output over 16M vehicles, charging at 20M. Two Sessions reps push L3/L4 auto driving laws, standards, insurance for commercialization. AI+ action boosts smart manufacturing, robots; leaders like XPeng's He Xiaopeng urge L4 policy leaps.

#autonomous-driving#china-policy#two-sessions
💰
TechCrunch AI43d ago

Grammarly’s Expert Review Lacks Real Experts

Grammarly recently added an 'expert review' feature claiming to improve users' writing with insights from the world's great writers, thinkers, and some tech journalists. The article criticizes this feature for missing actual human experts, questioning its authenticity.

#writing-tool#feature-critique#ai-hype
📲
Digital Trends43d ago

Adult AI Chatbots in Kids' Toys

A new report warns that some AI-powered toys for children incorporate chatbot systems designed for adults. These adult-oriented chatbots may expose kids to inappropriate content. The issue highlights ongoing safety gaps in AI toy deployments.

#ai-safety#child-safety#chatbots
📲
Digital Trends43d ago

AI Exposes Anonymous Online Accounts

Anthropic research reveals AI models can analyze online posts to link anonymous internet accounts to real identities. This enables mass de-anonymization, posing major privacy risks.

#deanonymization#privacy-risks#nlp-analysis
📲
Digital Trends43d ago

Phone Sensors Boost VR Avatar Motion

A new system harnesses smartphone sensors to track full-body movements. It enhances VR avatar realism without expensive VR hardware. This lowers barriers for VR development.

#vr-avatars#motion-tracking#smartphone-sensors
📱
Engadget43d ago

OpenAI Delays ChatGPT Adult Mode Again

OpenAI is further delaying the 'adult mode' launch for ChatGPT to prioritize higher-impact features like intelligence gains and personalization. Originally teased for December by CEO Sam Altman, it was pushed to Q1 2026, but now lacks a timeline. The company rolled out an age prediction tool in January, potentially linked to this feature.

#adult-mode#age-gating#content-policy
📲
Digital Trends43d ago

Liquid Metal Improves Robot Vision

Terminator-inspired liquid-metal pupils automatically adjust to light changes. Designed for robots and self-driving cars, they enhance vision accuracy in dynamic environments.

#robotics#computer-vision#liquid-metal
📱
Engadget43d ago

OpenAI Robotics Head Resigns Over DoD Deal

OpenAI's head of robotics, Caitlin Kalinowski, resigned criticizing the company's rushed partnership with the Department of Defense without adequate guardrails on surveillance and lethal autonomy. OpenAI confirmed the resignation and stated their deal prohibits domestic surveillance and autonomous weapons. This marks a high-profile fallout after Anthropic rejected similar terms.

#robotics#ethics#defense-partnership
⚖️
AI Alignment Forum43d ago

Govts Can Cheaply Slow AI Training?

Article analyzes if governments can quickly slow AI training via methods like bandwidth limits, cable unplugging, and inference-verification. Current prototypes ineffective against RL due to low communication needs, allowing covert compute use. Aggressive measures like proof-of-work or memory wipes could delay progress by 1+ year.

#ai-safety#rl-hivemind#compute-policy
🇨🇳
cnBeta (Full RSS)43d ago

OpenAI Robotics Lead Resigns Over Surveillance Fears

Caitlin Kalinowski, OpenAI's hardware and robotics engineering lead since November 2024, announced her resignation. She voiced concerns about AI in national security, including surveillance of Americans without judicial oversight and lethal autonomous weapons lacking human authorization. The move follows contract disputes.

#robotics-ethics#leadership-change#ai-safety
Page 509 of 754