Xinhua news, beijing, april 1st
Xinhua journalist
The ai wave is deeply reshaping the global landscape of technology and industry. In march, ai's smart body, represented by the “open paws”, accelerated its downfall, allowing ai to move from interactive dialogue to autonomous missions and boosting global computing and word demand for rapid growth。
At the same time, multi-country strategic planning and the scaling up of industry and computing have been intensive. With the depth of ai's work, technology abuse, ethical misconduct and security risks have become more prominent, and how to build strong security lines while seizing development opportunities has become a common theme for global ai industry development。
From dialogue to execution, smart weight plastic ai application bureau
After talking robots, smart people who can “work on their own” are moving quickly into people's lives. Since this year, an open-source intelligent called openclaw has been rapidly gaining ground in the global scientific community. Data show that it received over 300,000 “marks” on the open source platform github in just over two months, indicating a high level of interest on the part of the developers community. In china, the deployment of the tool on equipment is called "crawling" and the "shrimp-raising" hot tide is blowing out of the ring at an incredible rate。
Ai models, technology programs, etc. That support smarts are also evolving. On 5 march, the united states open artificial intelligence research centre (openai) announced the launch of the newer version of the large model gpt-54, stating that the model was its first universal model capable of operating computers directly and that it was not only good at writing codes but also providing mouse and keyboard operating instructions based on screenshots, enabling intelligent bodies to operate computers and implement complex workflows between different applications。
On 16 march, huang in-hoon, founder and chief executive officer of young weida corporation, launched a software warehouse called “nemoclaw” at the company's annual gtc general assembly (gpu technical assembly) to support the development and deployment of “open paws” intelligence。
The emergence of ai applications such as smarts has led to the concept of “token” becoming increasingly visible to the public. As the basic unit for the processing and generation of information in ai models, the terminologies are gradually becoming an important yardstick for measuring the cost and value of intelligent services. Data from the global main ai model convergence platform, “open routers”, show that the total number of words that users call on the ai model through the platform has increased significantly from about $6 trillion a week early this year to 20. 4 trillion a week as of 22 march。
Multinational plus-ai strategy
In order to seize the development opportunities of ai, multinational governments and enterprises have increased their policy support and industrial investment, and have continued to work in the areas of smart economics, international cooperation and rule-making, and capital-building。
For the first time in this year's government work report, china proposed “a new form of intelligent economy”, which includes deepening the expansion of artificial intelligence plus, promoting the accelerated diffusion of new generations of intelligence terminals and intelligent bodies, promoting the commercialization of artificial intelligence applications in priority industries, and developing new models of new business practices in intellectuals。
On 2 march, the korean government indicated that it would strengthen its cooperation with singapore in the area of ai and that it planned to establish a $300 million global investment fund in singapore to attract investment in start-ups and ai。
On 13 march, the council of the european union agreed on a proposal to streamline partial ai regulatory rules, which would focus on streamlining the euia regulatory framework and reducing the corporate compliance burden。
The strategic plan for the expansion of the data centre, published by the government of germany on 17 march, shows that by 2030, the computing power of the german general data centre will at least double from 2025 to 2025, with the calculation power earmarked for ai increasing at least four times to 2025。
On 21 march, the prominent american entrepreneur, elon mask, announced that a large chip manufacturing project called terrafab would be located in austin, texas. According to his vision, the project would build an advanced round factory that would integrate design, construction, containment and testing, with the goal of supporting a computing requirement of up to 1 twat a year and serving applications such as autopilot taxis, human robots and space data centres。
On 24 march, openai announced that it would close the sora video generation operation. This was seen by the outside world as an important initiative for the openai strategic transformation. According to observers, the closure of sora is not only a “dissociation” of the company's operations, but also a microcosm of the current shift in competition in the ai industry。
The moral and security risks cannot be ignored
At the 2026 world conference on mobile communications, held in barcelona, spain, in early march, multiple ai innovative products, such as robotic mobile phones, became the focus of attention, and the trend towards industrial change in ai became more evident. Mobile communications is accelerating from the mere provision of connected network infrastructure to the broader digital ecology that integrates cloud computing, data platforms and digital applications across industries。
However, while ai continues to be empowered in a wide variety of industries, the problems posed by new technologies are emerging. A recent study published in the united states journal of science shows that when human users seek advice from ai models on issues such as human distress, ai often displays overcorruptness or glamour; in response to some of the harmful and even illegal questions, ai often confirms the position of users。
The chat robot “groek”, developed by the merck-based company xai, was previously investigated by multinational regulators for having been misused to generate false pornographic content based on real people. In the recent past, “groek” has attracted public attention and criticism for publishing offensive content and misrepresenting facts。
Since the military strikes by the united states and israel against iran, a large number of ai technologies have been used to carry out targeted identification and strikes, causing serious civilian casualties and damage to civilian installations. The ethical and security risks of the misuse of ai technology in the military field have dramatically highlighted the urgency of promoting a broad consensus global ai governance system。
In response to questions from journalists on 11 march, china's ministry of defence news spokesman, jiang bin, expressed china's readiness to work with the countries of the world to advance the process of artificial intelligence in multilateral governance, with the united nations at its core, to strengthen risk prevention and control and to ensure that artificial intelligence continues to evolve in a direction conducive to the advancement of human civilization。
