At the Yunqi Summit held in China Hangzhou, Wu Yongming, CEO of Alibaba Group and Chairman of Alibaba Cloud Intelligence Group, delivered a significant speech. He clearly stated that achieving Artificial General Intelligence (AGI) is no longer a hypothesis but an inevitable trend. However, this is merely the starting point; the ultimate goal is to develop Artificial Superintelligence (ASI) capable of self-iteration and comprehensively surpassing human intelligence. For the first time, Wu Yongming elaborated in detail on the three major evolutionary phases towards ASI.
The first phase is “Intelligent Emergence,” where AI, through learning vast amounts of human knowledge, acquires generalized intelligent capabilities. The second phase is “Autonomous Action,” where AI masters the use of tools and programming capabilities, beginning to assist humans in completing complex tasks. This is the stage where the industry currently resides. The third phase is “Self-Iteration,” where AI, by connecting with the physical world and achieving self-learning, ultimately surpasses human intelligence.
To achieve this goal, Alibaba Cloud has defined a clear strategic path. As a “full-stack artificial intelligence service provider,” Alibaba Cloud will advance its AI strategy through two core pathways. First, Tongyi Qianwen will adhere to the open-source and open route, committed to building the “Android of the AI era.” Second, Alibaba Cloud will construct a super AI cloud as the “next-generation computer,” providing a global intelligent computing power network.
Wu Yongming revealed that Alibaba is actively promoting a three-year plan for AI infrastructure construction with a total investment of 380 billion RMB and plans to continuously increase investments. The latest news is that this large-scale investment marks one of the most ambitious AI infrastructure initiatives in China, signaling Alibaba Cloud’s determination to prepare for the ASI era. According to long-term planning, by 2032, the energy consumption scale of Alibaba Cloud’s global data centers will be ten times that of 2022, preparing for the arrival of the ASI era. In his speech, Wu Yongming specifically thanked global developers for their contributions to the development of cloud computing, AI, and the technology industry. He pointed out that the Yunqi Summit originated from Alibaba Cloud’s developer conference, and over the past decade, developers have driven technological progress in China and globally.
The world is currently undergoing an intelligence revolution driven by artificial intelligence. Over the past few hundred years, the industrial revolution amplified human physical capabilities through mechanization, and the information revolution amplified human information processing capabilities through digitization. This time, the intelligence revolution will far exceed imagination. AGI will not only amplify human intelligence but also liberate human potential, paving the way for the advent of ASI. In recent years, AI’s intellectual level has rapidly increased, jumping from a high school level to a doctoral level, even winning gold medals in the International Mathematical Olympiad. AI chatbots have become the fastest function to penetrate the user base in human history, with industry penetration rates surpassing all previous technologies. The total global investment in the AI industry has exceeded $400 billion, and cumulative investment over the next five years will exceed $4 trillion, which will accelerate the birth of more powerful models and promote the widespread penetration of AI applications.
Wu Yongming believes that the goal of AGI is to liberate humans from 80% of daily work, allowing people to focus on creation and exploration. ASI, as a system that comprehensively surpasses human intelligence, may create a group of “super scientists” and “full-stack super engineers” capable of solving scientific and engineering challenges at an unimaginable speed, such as conquering medical problems, inventing new materials, solving sustainable energy and climate issues, and even achieving interstellar travel.
The path to ASI will go through three stages. The first stage is “Intelligent Emergence,” where AI acquires generalized intelligent capabilities by understanding the world’s collection of knowledge. The second stage is “Autonomous Action,” where AI gains the ability to act in the real world, capable of decomposing complex tasks, using and creating tools, and autonomously completing interactions with both the digital and physical worlds. The third stage is “Self-Iteration,” where AI, by connecting with the physical world and achieving self-learning, ultimately surpasses human intelligence.
In the “Autonomous Action” stage, key AI capabilities include Tool Use and Coding. Through Tool Use, AI can call external software, interfaces, and physical devices like a human to execute complex real-world tasks. The improvement of Coding capabilities helps AI solve more complex problems and digitize more scenarios. In the future, natural language will become the source code of the AI era; anyone will be able to create their own Agent using natural language to accomplish almost all tasks in the digital world. In the “Self-Iteration” stage, AI needs to obtain more comprehensive and raw data directly from the physical world. Currently, the areas where AI is advancing fastest are content creation, mathematics, and Coding – areas where knowledge is 100% defined and created by humans. However, for the broader physical world, what AI encounters is mostly knowledge summarized by humans, lacking extensive, raw data from interacting with the physical world. Therefore, for AI to achieve breakthroughs surpassing humans, it needs to acquire data directly from the physical world.
As AI penetrates more physical world scenarios and understands more physical world data, AI models and Agent capabilities will become increasingly powerful, having the opportunity to build training infrastructure, optimize data processes, and upgrade model architectures for their own model upgrades and iterations, thereby achieving Self-learning. Future models will, through continuous interaction with the real world, acquire new data and receive real-time feedback, leveraging reinforcement learning and continuous learning mechanisms to autonomously optimize, correct biases, and achieve self-iteration and intelligent upgrades.
Wu Yongming also made two important judgments. First, large models are the next generation’s operating system, replacing the current status of OSs and becoming the tool interface linking the real world. In the future, almost all user needs and industry applications will execute tasks through tools related to large models; LLMs will become the OS of the AI era. Second, the ASI cloud is the next generation’s computer, meeting anyone’s needs and carrying massive computational resources.
Based on these judgments, Alibaba Cloud has made strategic choices: Tongyi Qianwen chooses the open route to build the Android of the AI era. Alibaba Cloud believes that in the LLM era, the value created and the scenarios penetrated by open-source models will far exceed those of closed-source models. Simultaneously, Alibaba Cloud operates China’s leading and globally top-tier AI infrastructure and cloud computing network, being one of the few super AI cloud computing platforms globally capable of vertical integration of software and hardware. Alibaba Cloud is fully committed to building a new AI supercomputer that possesses both the most advanced AI infrastructure and the most advanced models, enabling collaborative innovation in basic architecture design and model architecture, thus ensuring the highest efficiency when calling and training large models on Alibaba Cloud. To meet the industry’s long-term demand for AI infrastructure, Alibaba Cloud is actively promoting the three-year 380 billion RMB AI infrastructure construction plan and plans to continuously increase investment.
At the end of his speech, Wu Yongming discussed the collaborative relationship between humans and AI after the arrival of artificial superintelligence. He believes that in the future, AI will exponentially amplify the leverage of human intelligence, making humans more powerful than at any time in history. Each technological revolution, after unlocking more productivity, creates more new demands. AI will reconstruct the entire infrastructure, software, and application system, becoming the core driving force of the real world and setting off a new wave of intelligent revolution.
