2026-08-19
AI Briefings Daily Report.

Xiaomi's new-generation humanoid robot debuts at World Robot Expo: large-model-driven real-time interaction ↗

Mi official just released a video of the new-generation Xiaomi humanoid robot interacting with the audience in real time. According to the introduction, from handing flowers, shaking hands, to fist bumps and finger hearts, behind these seemingly simple actions, a large model drives the robot to autonomously understand, judge, and complete the actions.

VentureBeat appoints Rob Strechay as first chief analyst, expanding enterprise AI research layout ↗

Rob Strechay, until recently managing director and principal analyst at theCUBE Research, has joined VentureBeat as our first Lead Analyst and a founding analyst of VentureBeat Research. His arrival is the next step in a deliberate move at VentureBeat toward deeper specialization: analysis built for the technical decision-makers — the directors, VPs, CIOs, and CTOs — who are evaluating, buying, and deploying enterprise AI.

Calendly Launches Meeting Recording Tool, Joining Competition in AI Note-Taking Apps ↗

Calendly is also releasing a meeting scheduling assistant called Callie.

China allows small batch of Nvidia H200 chips to flow in, helping local AI enterprises catch up with the US ↗

China is letting small batches of Nvidia's H200 chips onto the mainland to help domestic AI firms in the race with the US.

Liquid AI releases LFM2.5 series QAD Q4_0 quantized checkpoints, recovering 97% precision loss ↗

Li Feifei: Tech practitioners should more clearly explain to the public what value AI can bring ↗

Fei-Fei Li, the 'Godmother of AI', believes that tech practitioners including herself still need to better explain to the public what value artificial intelligence can bring. She also warned that rising anti-AI sentiment in American society may pose risks globally.

Anthropic Revenue Surpasses OpenAI for the First Time ↗

Anthropic has passed OpenAI on revenue for the first time in the AI race.

After only half a year in office, Microsoft AI division CMO Andrea Mallard reportedly to depart ↗

Two informed sources said Mallard's departure is also related to the recent strategic adjustment of Microsoft's AI department.

Vivodyne uses robotic lab HIVE to generate causal biological data, calling it key to AI curing cancer ↗

A biotech startup called Vivodyne says the AI drug-discovery industry has a data problem, and that it has built a machine to fix it. HIVE, modular robotic labs built by the company, can grow 20 kinds of human tissue, then autonomously dose and monitor them, generating the kind of causal biological data that today’s AI models are missing — data that today mostly comes from animal testing, or studies of single cells or proteins, not living tissue.

First assessment by non-profit organization Guidelight: AI labs such as Anthropic and OpenAI have not fully implemented internal system controls ↗

No AI company fully applies basic control measures to its own internal AI systems.

Anthropic Has Claude Agents Run Full Protein Design Process, Drug Discovery Hit Rate Exceeds Industry Average ↗

Anthropic had its Claude models design small proteins on their own that dock onto target structures in the body, a key step in early drug development. The hit rate reached up to 35 percent, far above the industry average of 10 to 15 percent. Claude only steered existing specialized tools, and an independent review is still pending.

AI Chip Startup Etched's Valuation Soars to $21 Billion; First Rack Shipments Delivered ↗

The lead investor of its funding round and chip customer Jane Street is satisfied with the preliminary test results of Etched's product.

Surge in AI Chip Demand; Samsung May Raise Advanced Process Foundry Prices by Up to 15% ↗

According to industry estimates, the price increase also means Samsung's wafer foundry business, which has been losing money since 2022, is beginning to turn around. The surge in prices of memory chips used in AI systems has helped Samsung set profit records, but the wafer foundry business has still not significantly narrowed its gap with TSMC.

China's first AI customer service national standard takes effect Sept 1: regulates human-intelligent customer service collaboration process ↗

To standardize the collaboration process between human and AI customer service, China's first national standard focusing on the collaboration mechanism between human and AI customer service, 'Requirements for Customer Contact Service Human and Intelligent Customer Service Collaboration', will be implemented from September 1.

Y Transformers: When Paper meets Partner, how many more Possibilities does AI have? ↗

When global young researchers meet AI entrepreneurs: talking tech, capital, products, and also choices in the AI era

Turing Award winner Sutton warns: Excessive reliance on synthetic data in the AI industry is a big mistake ↗

Turing Award winner and father of reinforcement learning Rich Sutton warned on Sequoia Capital's podcast that the AI industry's excessive reliance on synthetic training data is leading the industry in the wrong direction. He believes the real world far exceeds any simulation, and AI should learn through interaction with the environment. His startup Oak Lab is putting this philosophy into practice, aiming to develop AI agents that can learn from their own experiences.

Durov: Telegram has applied for".gram"Top-level domains can provide dedicated subdomains for one billion users and support"Let AI Create a Complete Website from One Sentence" ↗

After approval, users will get a dedicated domain name based on their username, and can also use AI to create a personal website with one sentence; the domain name still needs ICANN review.

Anthropic Says Claude Desktop Background Startup Speed is Approximately Twice as Fast as One Month Ago ↗

@ClaudeDevs posted on the X platform today (August 19), stating that in 95% of use cases, the launch speed of Claude Desktop is approximately twice as fast as it was one month ago.

French government to introduce AI tools to detect cybersecurity vulnerabilities, considering hiring Mistral ↗

To respond to consecutive cyberattacks and data breaches, the French government decided to introduce AI tools to detect cybersecurity vulnerabilities and has commissioned a domestic AI company with the task. Recent multiple attacks led to leakage of millions of taxpayers' information, and the security situation is severe.

Anthropic extends 50% bonus on weekly usage limits for Claude Code until August 31 ↗

Anthropic announced an extension of the 50% increase in weekly usage limits for Claude Code, extending it from the originally scheduled August 19 to August 31, and exploring adjusting it to a fixed quota scheme.