Author: lethuphuong

📌 The UK government has appointed two “AI Champions” to lead the deployment of AI in financial services, aiming to boost growth, productivity, and investment while protecting consumers and maintaining system stability. The roles take effect from January 20, 2026, are direct ministerial appointments, and are unpaid. With 75% of financial businesses already using AI and a potential to create an additional $25–38 billion USD in value by 2030, the UK is shifting from encouragement to strategic coordination. The appointment of AI Champions demonstrates an effort to balance accelerated innovation with risk governance. If successful, this will be a lever…

📌 AI is not a single bubble but multiple stacked bubbles. The AI ecosystem is divided into 3 layers: wrapper companies, foundation models, and infrastructure. Wrapper companies are at risk of collapse within 18 months, foundation model companies will consolidate within 2–4 years, while infrastructure companies, despite short-term excess, will retain long-term value. Advice for builders: the biggest risk is not building a wrapper, but remaining just a wrapper; one needs to own the workflow, data, and distribution channels.

📌 ChatGPT Free currently meets basic needs and light usage well, while Plus is the most balanced choice for frequent users wanting fewer restrictions and early access to GPT-5.2, Codex, Agent, and Sora for $20. Pro is only truly reasonable for research experts, programmers, or businesses needing maximum power, as the $200 price tag mainly trades for high limits, fast speed, and the GPT-5.2 Pro model.

📌 Mistral’s CEO and co-founder argues that the company’s biggest competitive advantage in Europe comes not from smarter AI models, but from being a non-US option that aligns with the need for control and technological sovereignty. With a $14 billion valuation, military contracts for the French Ministry of Defense, and an open-source strategy, Mistral exploits the demand for sovereignty, control, and trust from governments and regulated enterprises. Their long-term advantage is not just the model, but their geographical location and their approach to building AI.

📌 Canada has never officially defined “sovereign AI,” allowing the concept to form loosely through data center funding, domestic cloud partnerships, and data residency requirements. Consequently, “sovereignty” is applied inconsistently, relying on infrastructure location rather than true control. Businesses optimize for rapid deployment on foreign platforms to meet geographic requirements while lacking technical and legal control. For example, despite government investment in Cohere, the compute infrastructure is operated by US-based CoreWeave, subject to US law. If Canada wants to build a sustainable AI ecosystem, sovereignty standards need to shift from “where” to “who has the power,” so that value, knowledge,…

📌 Experts urge Malaysia to soon issue clear regulations to manage foreign AI platforms like DeepSeek, aiming to protect data privacy and prevent external influence. Malaysia faces a strategic choice regarding foreign AI like DeepSeek: prohibition or smart governance. Experts agree that risks regarding data, bias, and political influence are real, but the sustainable solution is a clear legal framework, risk assessment, and the development of domestic AI. With the AI law expected to be presented in June 2026, how Malaysia balances innovation and data sovereignty will determine the safety and reliability of the national AI ecosystem.

📌 The strategy relying on AI and high-tech helps China build a long-term foundation but lacks sufficient power to pull the economy in the short term. The real estate slump, job losses due to automation, and increasing reliance on exports make the 5% growth target difficult to achieve, as new sectors would need to expand 7-fold in the next 5 years to meet this level. Automation and robotization could eliminate up to 100 million jobs over a decade, while urban unemployment exceeds 5% and youth unemployment is roughly three times higher.

📌 Sergey Brin, co-founder of Google, admits the company is hiring many employees without bachelor’s degrees because they “tinker and solve problems from very strange angles.” This trend is not unique to Google but has spread to Microsoft, Apple, and Cisco. This reflects a strong shift towards skills-based hiring in the AI era. As technology changes the nature of entry-level work, a degree is no longer the sole reliable metric. This opens up huge opportunities for self-taught Gen Z while placing pressure on universities to redefine their value in the new labor market.

📌 Microsoft’s report indicates that generative AI is spreading rapidly but unevenly. The global generative AI adoption rate in the second half of 2025 reached 16.3%, an increase of 1.2 percentage points compared to the first half; equivalent to about 1/6th of the world’s population having used AI for learning, work, and problem-solving. Vietnam recorded positive progress: the AI usage rate increased from 21.2% to 23.5%, higher than the global average and surpassing many developing economies, but a significant gap remains compared to leading ASEAN nations.

📌 After 2 years, Malaysia’s Ministry of Digital has transcended a symbolic role to become the true architect of the national AI future. With fortified foundations in governance, law, and trust, along with its leadership role in ASEAN and the clear AI Nation 2030 goal, Malaysia is shifting from “digital transformation” to “smart nation building.” The focus is not just on technology, but on ethics, governance, and sustainable competitiveness in the AI era.