24-hour AI news overview: Domestic accelerated implementation of agents and content applications, overseas AI computing power and financial security are heating up simultaneously
In the past 24 hours (April 15 to April 16, 2026), domestic AI hotspots are concentrated on the production of intelligent products, industrial data and content, showing that the application side continues to accelerate; overseas, focusing on the expansion of the AI computing power chain, strong chip demand and Financial risk control is unfolding, and industrial investment and regulatory attention are increasing at the same time.
1. MiniMax launched MaxHermes, a cloud self-evolution assistant,
MiniMax released MaxHermes, which focuses on sandboxes, self-learning and skills accumulation capabilities in the cloud. This means that domestic manufacturers are further shifting from "model capabilities" to "agent products that can perform tasks sustainably."
2. The average daily number of word calls in China exceeded 140 trillion
The National Bureau of Statistics disclosed at the press conference that as of March, the average daily number of word calls in China had exceeded 140 trillion, an increase of more than 40% from the end of the previous year. This shows that the commercial invocation of large models is still rapidly increasing, and the real demand on the enterprise side continues to expand.
3. Jiangsu Implementation AI Agent Data Support Plan
Enterprise Chacha announced that it will open up agent data platform services to entrepreneurs and form a linkage with relevant Jiangsu policies. After the data base is completed, the training, retrieval and execution costs of enterprise-level agents are expected to continue to decline.
4. AI generates more than 2 billion pieces of video and audio content a year
The latest online audio-visual development report shows that AI will generate more than 2 billion pieces of video and audio in 2025. AI is moving from assisted creation to large-scale content supply, and platform governance and copyright issues will also receive more attention.
5. TSMC's first-quarter profit increased by 58%
TSMC announced a record profit growth in the first quarter. One of the core drivers is still the demand for AI chips. The boom in upstream wafer manufacturing shows that the global computing power construction heat has not cooled down yet.
6. ASML raised its 2026 performance forecast
ASML said that AI-driven new orders and equipment demand were stronger than expected, and raised its full-year outlook. The strengthening of the equipment side further verifies that advanced processes and high-end production lines are still the focus of global capital expenditure.
7. The European Central Bank pays attention to the financial risks of the new generation of AI models
News said that the European Central Bank will ask financial institutions questions about the banking system risks that the new AI model may bring. The regulatory perspective has shifted from "whether AI can be used" to "how to prevent AI from amplifying systemic risks."
8. Air Liquide adds to Japan's AI chip supply chain
Air Liquide announced an investment of US$236 million in Japan to serve the production of next generation AI chips. In addition to chip design and manufacturing, supporting links such as materials and industrial gases are also sharing the AI infrastructure dividends.
Frequently Asked Questions (Q A)
Q: What are the most noteworthy themes in these 24 hours?&
A: One is the accelerated implementation of domestic agent applications, and the other is the continued expansion of overseas AI computing power industry chain.
Q: Is domestic AI more applied or underlying model?
A: Judging from the latest developments, the current trend is more towards application implementation, data infrastructure and content production improvement.
Q: Why overseas talk about growth and risk at the same time?
A: Because AI drives up demand for chips and devices, it also brings new risks in the fields of finance, compliance and security.
Q: What is the most direct inspiration for entrepreneurs?
A: It is easier to form near-term opportunities around agents, industry data, content production and enterprise services than simply building models.
