I. Industry Development
[2025 World Artificial Intelligence Conference (WAIC) Opens, Humanoid Robots Showcase Their Capabilities]
The 2025 World Artificial Intelligence Conference (WAIC), held in Shanghai from July 26 to 29, is reportedly poised for a major explosion in embodied intelligence. According to statistics, last year’s WAIC saw only a dozen or so robotics companies participating, but this year, the number has grown to over 80.
The “most impressive” robots at the exhibition—they “can do it all,” taking on roles such as salesperson, assembly line worker, minimally invasive surgeon, popcorn dispenser, and cheerleader—showcased their skills in a way that resonated with everyday life. Humanoid robots were undoubtedly the shining stars of the conference, which was unprecedented in scale and showcased China’s formidable capabilities in the field of humanoid robotics.
[Musk’s xAI plans to raise $12 billion, aiming to have 50 million H100-equivalent computing units within five years]
XAI, Elon Musk’s artificial intelligence startup, is reportedly preparing to raise $12 billion to maintain its lead in the AI arms race. Musk revealed that xAI’s goal is to have AI computing units equivalent to 50 million H100s within five years, with higher energy efficiency.
[Alibaba (BABA) Open Sources its Most Powerful Programming Model! 480 Billion Parameters, Agent Scores Crushed Kimi K2]
On July 23rd, Alibaba’s Qwen team open-sourced its latest generation flagship programming model, Qwen3-Coder-480B-A35B-Instruct. The Qwen team claims this is their most powerful open-source agent programming model to date, boasting 480B parameters, 35B activation parameters, native support for 256K contexts, and scalability to 1 million contexts (inputs) through extrapolation. Its maximum output is 65,000 tokens. In benchmark tests, Qwen3-Coder demonstrated impressive performance in programming and agent-based tasks, achieving open-source state-of-the-art results in three categories: agent coding, agent browser-use, and agent tool-use. This performance surpassed open-source models like Kimi K2 and DeepSeek V3, as well as closed-source models like GPT-4.1, and rivaled Claude Sonnet 4, a model renowned for its programming capabilities.
Qwen3-Coder will be available in multiple sizes, with the most powerful variant now open-sourced. Its parameter count exceeds the 235B (235 billion) of Alibaba’s flagship Qwen3 model and is smaller than the 1T (1 trillion) of Kimi K2. According to Alibaba, with Qwen3-Coder, a novice programmer can complete a week’s work for a seasoned programmer in just one day, and can generate a brand’s official website in as little as five minutes.
Amazon (AMZN) Acquires AI Wearables Company Bee, Boosting its Investment in Personal “AI Companion” Market
According to a LinkedIn post by Maria de Lourdes Zollo, co-founder of AI wearables startup Bee, Amazon has acquired the company. Amazon confirmed the acquisition to TechCrunch but noted that the deal has not yet been finalized.
An Amazon spokesperson revealed that Bee employees have received job offers to join Amazon. The acquisition signals Amazon’s interest in developing wearable AI devices, a distinct area beyond its voice-controlled home assistant products such as the Echo line of speakers.
OpenAI, the developer of ChatGPT, is developing its own AI hardware, while Meta is integrating its AI into smart glasses. Apple is also rumored to be developing AI-powered smart glasses, suggesting Amazon is also poised to take new steps.
Meta’s (META) Super AI Team Revealed: 50% Coming from China, Annual Salaries Could Reach $100 Million! According to foreign media reports, tech giant Meta is investing heavily in recruiting AI talent to stay ahead in the artificial intelligence (AI) race. The company is focusing its expertise at its Meta Superintelligence Labs to research various foundational models. A recently released list shows that 50% of the lab’s 44 employees are from China, and 40% previously worked at OpenAI.
[Google (GOOG) Releases Q2 Earnings, Highlighting Optimistic Trends in AI Glasses]
Alphabet recently released its Q2 2025 financial results, reporting $96.43 billion in revenue, a 14% year-over-year increase and exceeding analysts’ expectations of $93.97 billion.
During the earnings call, Google CEO Sundar Pichai specifically highlighted his views on AI glasses, stating that every advancement in AI brings new experiences, including hardware experiences. Therefore, AI will particularly empower glasses and other devices, which have long been expected to usher in a new wave of innovation. We are very excited about our investment in glasses and have seen a significant improvement in the experience compared to the previous generation.
[Alibaba (BABA) is about to release AI glasses, and the AI+AR glasses may be officially launched during this year’s Double 11 shopping festival.]
According to reports, the Alibaba team will officially launch its first self-developed AI glasses at the Shanghai World Artificial Intelligence Conference (WAIC 2025). These AI glasses will include basic functions such as voice assistant, music playback, phone calls, real-time translation, and meeting minutes, and will integrate with Alibaba’s ecosystem features such as maps, payment, and shopping.
Related media reports indicate that the products will be divided into AI smart glasses without a display and AI+AR smart glasses with a display. The latter is a higher priority and is targeted for official launch during this year’s Double 11 shopping festival.
[Elon Musk’s brain-computer interface company announced that it has completed two surgeries in one day for the first time! Neuralink, the brain-computer interface company owned by billionaire Elon Musk, recently announced, “We successfully completed the P8 and P9 surgeries. This marks the first time we have performed two surgeries in one day. Both patients are currently recovering well and in good spirits.” Musk stated that Neuralink will ultimately change the lives of millions, perhaps even billions, of people.
Earlier this month, Neuralink founder Elon Musk held a press conference to unveil the company’s new three-year plan for brain-computer interface technology. The plan outlined the goal of achieving a full-brain interface by 2028, while gradually increasing the number of electrode channels from the current thousands to tens of thousands.
According to Neuralink’s plan, its expansion of brain-computer interface technology follows a roadmap spanning motor, language, vision, and the entire brain. The number of electrodes implanted in the brain is planned to increase from 3,000 in 2026 to over 25,000 by 2028, with the company exploring deep integration with AI technology.
II. Market Dynamics
[Apple (AAPL) Launches AppleCare One Subscription Service, $19.99 Per Month]
Recently, Apple launched a new subscription service, AppleCare One, which can be considered an upgraded version of Apple Care+. For just $19.99 per month, users can provide coverage for up to three Apple devices, including the entire Apple product line, including iPhones, iPads, Apple Watches, and Vision Pros.
According to feedback from some Vision Pro users, AppleCare One subscriptions can only be linked to one Vision Pro device per ID, which may be related to the high cost of Vision Pro repairs.
[Wimi Hologram Cloud Inc. (WIMI) Integrates Blockchain and Digital Twins to Explore Blockchain-Based Distributed Collaborative Computing]
The rapid development of connected vehicle technology is profoundly changing the landscape of intelligent transportation systems. It not only promotes interoperability between vehicles but also accelerates the diversification and personalization of in-vehicle applications.
Distributed collaborative computing, as an emerging computing paradigm, aims to improve overall system performance and responsiveness by integrating distributed computing resources. This concept is particularly critical in the connected vehicle (IoV) sector, as it can dynamically adjust computing loads based on real-time demand, thereby optimizing the user experience. However, distributed computing also raises issues of trust and collaboration.
To address these challenges, WiMi is exploring an innovative solution: a blockchain-based distributed collaborative computing architecture specifically designed for in-vehicle digital twin networks. This architecture cleverly combines the decentralized nature of blockchain with the simulation and predictive capabilities of digital twins to address the challenges of resource collaboration and decision-making in the IoV.
Blockchain technology, with its immutable and decentralized nature, plays a central role in the IoV. It not only ensures the security of data exchange between vehicles but also promotes transparent resource management. In WiMi’s blockchain-based distributed collaborative computing architecture, each vehicle is treated as a node, and its resource information, such as computing power and storage space, is recorded on the blockchain, enabling real-time resource allocation and optimized utilization.
WiMi’s blockchain-based distributed collaborative computing architecture promotes open resource sharing and precise management decisions enabled by digital twins through blockchain. This architecture not only overcomes existing challenges but also further drives the development of the IoV towards greater efficiency and intelligence. In future research, WiMi will continue to explore blockchain-based distributed collaborative computing, further reducing system latency, improving resource allocation efficiency, and strengthening privacy protection measures to ensure user data security, thereby enabling seamless connectivity and collaborative operation of connected vehicles.
[Qualcomm (QCOM) Hosts XR Day Event in India and Collaborates with Lenskart on AI Glasses]
Recently, Qualcomm held an event in India called “Snapdragon for India XR Day.” The event aims to promote the development of spatial computing in the Indian market, and Qualcomm used the opportunity to showcase a range of its innovations in content capture, entertainment, fitness, and education.
Qualcomm also revealed partnerships to provide AI-powered smart glasses for Indian consumers, including a collaboration with Lenskart, which uses Snapdragon chipsets. Lenskart is a well-known Indian eyewear manufacturer founded in 2010 and has become one of Asia’s largest omnichannel eyewear retailers.
Meta (META) Unveils EMG Wristband Control Technology, Revealing Details of AR Glasses Gesture Control and HUD Glasses
Meta published a paper titled “A Generic Non-invasive Neuromotor Interface for Human-Computer Interaction” in the journal Nature, systematically and scientifically detailing its multi-year development of electromyographic (sEMG) wristband interaction technology for the first time.
The research was developed by CTRL Labs, a neural interface startup acquired by Meta in 2019. CTRL Labs was founded and led by computational neuroscientist Thomas Reardon, who continues to lead the project.
