Tobias

Microsoft’s 2025 Work Trend Index reveals that excessive work hours are hindering productivity. The study highlights a blurring of work-life boundaries, with many experiencing extended workdays. Peak productivity is observed between 9 AM and 3 PM but is often disrupted by meetings. Microsoft suggests AI could streamline tasks, potentially leading to a more focused, six-hour workday.

Aleksandar Janković, the U20 coach, has been appointed interim coach of the Chinese national football team to lead them in the East Asian Cup. This follows the termination of the previous coach’s contract due to the team’s failure to advance in World Cup qualifiers. With a history of success leading youth teams, including recent tournament wins, Janković faces the challenge of stabilizing a team marked by frequent managerial changes, the current appointment is the 14th in 23 years.

JD.com founder Liu Qiangdong revealed his intent to enter China’s food delivery market, signaling a major competitive challenge to Meituan. He hosted a dinner with key industry figures, including Meituan’s founder Wang Xing, to announce his plans. JD.com launched its food delivery service focusing on quality and lower commissions. It also provides full social security and insurance for its delivery workforce, aiming to attract talent and disrupt the market.

Dorvi, a feminine hygiene brand co-founded by Huang Zitao, addressed consumer concerns about black specks in its products. The brand clarified these are harmless byproducts of the manufacturing process, exceeding industry visual inspection standards. Dorvi is offering replacements, emphasizing transparency and inviting feedback. Huang Zitao also issued a statement, inviting media and customers to observe the production, including online oversight for all existing customers.

JD.com’s CEO, Richard Liu, discussed parallels between JD.com’s operational philosophy and a successful retail chain’s employee-focused approach. He highlighted prioritizing employee well-being, fair treatment, and customer service. This retailer’s model emphasizes no sales targets, significant profit allocation to employees, and generous benefits like short workdays, ample paid leave, and store closures, fostering a positive work environment.

JD.com CEO Xu Ran revealed impressive earnings for the company’s food delivery riders. Full-time riders in Beijing, Shanghai, Guangzhou, and Shenzhen now average nearly RMB 13,000 monthly, a significant increase. This increase comes despite comprehensive benefits. JD.com’s delivery arm is rapidly expanding, with over 120,000 riders and processing over 25 million daily orders, highlighting the company’s commitment to sustainable growth in the food delivery sector.

In June 2025, reports emerged of fraud within the tech sector, exploiting a new government subsidy program for digital devices. “Scalpers” are reportedly using students to purchase subsidized items and resell them, netting profits of up to 300 yuan per device. This illegal activity involves dedicated chat groups and circumvents in-person verification. Legal experts warn that this scheme, even with small amounts, could constitute fraud, leading to penalties.

JD.com, led by Liu Qiangdong, plans to aggressively enter the hospitality and travel sectors. Leveraging its logistics network and financial strength, JD.com aims to reduce supply chain costs significantly in food, travel and hospitality, potentially lowering consumer costs by two-thirds while providing new channels for partners. They will expand from convenience stores and food delivery, with CEO Xu Ran confirming a focus on high-growth local services.

DeepSeek R1-0528 surpassed Claude Opus 4 in web programming, showcasing strong coding capabilities, especially with detailed prompts. Though performing well in web app development, it faltered with a Tetris game. R1 is also the top-performing open-source text model but has been recently surpassed by Kimi-Dev in open-source code.

Moonshot AI introduced Kimi-Dev-72B, an open-source code LLM designed for software engineering. The 72-billion parameter model tops the SWE-bench Verified benchmark, surpassing larger models like DeepSeek-R1. Kimi-Dev-72B utilizes RLHF to autonomously repair code within Docker. Key features include BugFixer, TestWriter, mid-stage training, and test-time self-play. The model is available on Hugging Face and GitHub.