Author: lethuphuong

📌 The AI war is shifting from performance competition to cost competition. DeepSeek and Xiaomi are not just cutting prices by a few percent but are driving AI costs down by up to 98–99% compared to many leading US models. With performance nearing GPT and Claude but prices being dozens of times lower, enterprises deploying AI Agents, document processing, and large-scale automation have strong incentives to switch to open-source or Chinese models to significantly reduce operating costs.

📌 China is testing an unprecedented robot development model: centralized training for over 100 robots from multiple brands and turning collected data into a shared “super brain” for the entire industry. With 10 million data points annually and hundreds of thousands of practice sessions daily, the project aims to drastically reduce training costs, accelerate the commercialization of humanoid robots, and create a competitive edge for the Chinese robotics industry in the coming decade.

📌 AI is shaking the traditional consulting model that relies on an army of junior consultants and billable hours. AI-native startups, backed by private equity, are leveraging Agentic AI to scale rapidly at much lower costs. While the Big Four still hold advantages in capital and global networks, they are under immense pressure from AI automation, changing fee models, and the risk of talent draining toward more flexible AI-native firms.

📌 The confrontation between AI and knowledge workers is escalating sharply in the U.S. media industry. The New York Times is accused of using AI tools like DX and Glean to monitor employee performance, creating quantitative pressure and supporting labor discipline. This controversy reflects a broader trend where AI is increasingly integrated into newsrooms but lacks mechanisms for transparency, control, and worker protection.

📌 This event shows that the power of Big Tech in Washington remains immense, even after Elon Musk and David Sacks left their official roles in the White House. At the same time, it reflects the deepening conflict in the U.S. AI race: between the need to accelerate innovation to compete with China and the fear that frontier AI could create unprecedented risks in cybersecurity, unemployment, and social instability.

📌 The UAE is becoming one of the first countries to deploy AI agents on a government-wide scale rather than just local pilots. The goal of automating 50% of public services in two years shows that AI is no longer just a support tool but is becoming the new operational layer of the state. Alongside the technology, the UAE is also investing heavily in human resource training to build an “AI-native government” model that could serve as a blueprint for many other nations.

📌 Russia is increasingly dependent on China’s AI ecosystem and chips as it is cut off from advanced GPU access by the West. The competition for Huawei chips between Sberbank and Chinese Big Tech shows that AI demand in China is so massive that even a strategic ally like Russia must fight for supply. Simultaneously, AI is becoming a new pillar in Russia-China relations, not only in the economy but also in defense and strategic technology. This reflects the increasingly clear formation of a “non-Western” AI ecosystem centered around China.

📌 The U.S. and ASEAN are accelerating a strategic AI alliance with a focus on computing infrastructure, data centers, semiconductors, and trusted technology supply chains. The participation of corporations like Amazon and Google shows that Southeast Asia is becoming a new hotspot for the global AI economy, while simultaneously expanding cooperation into healthcare, agriculture, climate, food security, and space technology.

📌 The finance industry is entering a “post-AI-hype” phase, where speed is no longer as important as accuracy and risk control. The article shows that Hybrid AI, combining language models with mathematical and logical verification systems, is becoming a mandatory architecture for banks. In an environment where a small error can cause millions of dollars in damages, “approximate AI” is no longer safe enough to manage customers’ money.

📌 Wira-LLM demonstrates that Malaysia aims to build a complete sovereign AI ecosystem rather than relying on foreign models. With its air-gapped capability, support for over 100 languages, and a score of 89.20% on MalayMMLU, Wira is positioned for government and corporate environments requiring high security. Gamuda is simultaneously expanding into sovereign cloud and AI agents to provide domestic end-to-end AI solutions.