A sleek, life-sized video screen in Hangzhou, a Chinese tech hotspot, flashed out four words early last week that would propel Alibaba’s stock to all-time highs and mark a change in China’s stance on artificial intelligence: “Roadmap to Artificial Superintelligence”
Alibaba CEO Eddie Wu outlined a future characterized by artificial general intelligence (AGI) and artificial superintelligence (ASI) in his 23-minute keynote speech at the premier Alibaba Cloud conference. Theoretically, these names indicate a period where AI will become about as intelligent as humans (AGI) and subsequently much more smart (ASI).
Although Silicon Valley has long used these acronyms, Wu’s presentation stood us because Alibaba is now the first well-known Chinese IT company to specifically use AGI and ASI.
An intelligent system with generic human-level cognition, or AGI, now seems inevitable. However, Wu stated that AGI marks the beginning rather than the conclusion of AI’s growth. “ASI—intelligence beyond the human, capable of self-iteration and continuous evolution—will be its march.”
Wu stated, “ASI will drive exponential technological leaps, carrying us into an unprecedented age of intelligence.” He emphasized how ASI may contribute in the discovery of cleaner energy sources, the curing of illnesses, and even the opening of interplanetary travel.
The two most powerful AI nations in the world are the United States and China, both of which have enormous processing capacities and elite researchers creating state-of-the-art systems. However, analysts have presented the nations as having distinct AI strategies, with China allegedly emphasizing more practical AI usage.
Because China is home to several of the most sophisticated robotics supply chains in the world as well as an expanding network of robotics, or embodied AI, laboratories, analysts recently claimed that Beijing is now “winning the race for AI robots” over the United States.
The interim executive director of Georgetown’s Center for Security and Emerging Technology, Helen Toner, stated that there has been some recent discussion in Western media regarding how China is concentrating only on applications, while the United States is missing the point by pushing for AGI. “This is wrong.”
AGI and superintelligence have long sparked the curiosity of some Chinese scholars and some segments of the Chinese government, Toner said, but she pointed out that smaller firms like DeepSeek were the main forces behind this viewpoint.
Afra Wang, a researcher who specializes in China’s tech sector, praised Alibaba’s use of AGI and ASI.
“Especially among the largest tech companies in China, this ASI narrative is definitely something new,” she told.
It appears that Alibaba’s “roadmap to artificial superintelligence” is confusing the general public. Elon Musk of xAI and Dario Amodei of Anthropic are just two examples of California techno-optimists who might have given Wu’s speech, promoting a technologically advanced utopia while mainly avoiding more serious concerns about how mankind will survive or coexist with a time of digital superintelligence.
Prominent AI businesses in the United States have long been considering, if not actively driving, the idea of superintelligence. In May 2023, for instance, OpenAI published an essay about the safe development of superintelligent AI models. According to the statement, “the time is right to begin considering the governance of superintelligence — future AI systems that are significantly more capable than even AGI.”
In Congress, the prospect of superintelligence is now even being recognized. Sens. Richard Blumenthal of Illinois and Josh Hawley of Missouri introduced a draft measure on Monday that would, among other things, “assist Congress in determining the potential for controlled AI systems to reach artificial superintelligence.”
When today’s AI systems are unable to comprehend basic tennis rules, hallucinate or fake simple facts, or appear to truly understand how the outside world works, ASI may seem like a ridiculous idea to some.
Many scientists believe that it is a question of when, not if, humanity will acquire digital superintelligence since AI systems are still getting closer to and occasionally surpassing human skills in a variety of areas, from securely operating automobiles to winning international coding contests.
Many American politicians have said that the United States and China are engaged in an AI race as the concept of superintelligence slowly makes its way into the mainstream. Sen. Ted Cruz, a Republican from Texas, declared that “America has to beat China in the AI race as a matter of economic security, as a matter of national security.” The White House’s current AI manifesto is called “Winning the AI Race: America’s AI Action Plan.”
However, claims of an AI race are complicated by a lack of a common ultimate objective and varying definitions of AGI. At worst, researchers believe an unchecked race toward AGI or ASI might result in widespread devastation or perhaps the extinction of mankind.
However, there is a lot of doubt about AGI and ASI, and whether they are simply used for marketing.
Alibaba is one of China’s leading technology businesses, recognized for offering sophisticated, free AI models (also known as open-source models) for download. Alibaba’s Qwen model series, which competes with models like as OpenAI’s GPT-5 and Anthropic’s Claude, is the most popular open-source AI system worldwide.
In his lecture last week, Wu revealed a new line of Qwen models, one of which integrates text, graphics, video, and audio.
Many analysts note that stories about a race to develop AI infrastructure between the United States and China benefit AI investors by boosting firm valuations and soft power. Since Wu’s speech last week, Alibaba’s stock has surged, making it China’s most popular AI business and part of a broader $250 billion return this year.
Wu projected that large AI models will take the role of current operating systems as the conduit between people, software, and computing power in order to unlock a powerful, superintelligent future. According to him, cloud computing networks like Alibaba Cloud will power this future network of massive AI models.
Wu’s remark had economic overtones, according to Irene Zhang, an editor of ChinaTalk and expert on China’s AI ecosystem.
She stated, “This is an AGI and ASI vision that is directly based on Alibaba’s business model.”
“Alibaba Cloud now has a larger global market share than Oracle and dominates the Chinese cloud computing market,” she stated. “Alibaba’s public positions on ASI/AGI and its business strategy are mutually reinforcing.”
Senior fellow Matt Sheehan of the Carnegie Endowment for International Peace agreed.
According to Sheehan, “ASI is the ultimate frontier, as far as the discourse on AI goes.” It’s noteworthy that Alibaba established this lofty objective when, in fact, they are providing cloud services.
