On September 19th, Wu Yongming, CEO of Alibaba Group and Chairman and CEO of Alibaba Cloud Intelligence Group, delivered a keynote speech at the 2024 Yunqi Conference. He believes that the development speed of AI has exceeded any historical period, and the investment threshold for advanced models in the next stage is in the billions of dollars. Robots will be the next industry to undergo tremendous changes, and in the future, all movable objects will become intelligent robots.
In his speech, Wu Yongming shared his three judgments on the future development trend of AI:
Although the speed of AI development has exceeded any historical period in the past 22 months, it is still in the early stages of AGI transformation.
In his opinion, with the rapid iteration of big model technology, the usability of the technology has greatly improved. The large model already possesses multimodal capabilities of text, speech, and vision, and can begin to complete complex instructions. Last year, big models could only help programmers write simple code, but today they can directly understand requirements and complete complex programming tasks. Last year, the mathematical ability of big models was only at the level of middle school students, but today it has reached the gold medal level of international Olympic competitions and is close to the doctoral level in physics, chemistry, biology and other disciplines.
In addition, the inference cost of the model has decreased exponentially, far exceeding Moore's Law. He took Tongyi Qianwen as an example. In the past year, the call price of Tongyi Qianwen API on Alibaba Cloud Bailian has decreased by 97%, and the lowest cost of calling one million tokens has dropped to 50 cents. There is still a trend of further decline in the future.
The open source ecosystem is also thriving. On the open-source model library Huggingface, Alibaba's Qwen large model has nearly 50000 native and derivative models, ranking second in the world. There are over 10000 models on the Alibaba Cloud Magic Community, serving over 6.9 million developers.
In Wu Yongming's view, all of this is just beginning. To achieve true AGI, the next generation of models needs to have a larger, more general, and more generalized knowledge system, as well as more complex and multi-level logical reasoning abilities. The investment threshold for advanced model competition worldwide will reach billions or even tens of billions of dollars.
Secondly, AI's greatest imagination is not in mobile phone screens, but in taking over the digital world and changing the physical world.
In the past three decades, the essence of the Internet wave is connectivity, connecting people, information, commerce and factories. Through connectivity, the collaboration efficiency of the world has been improved, great value has been created, and people's lifestyle has been changed. But generative AI creates new value through the supply of productivity, thereby creating greater intrinsic value for the world and overall improving the productivity level of the entire world. This kind of value creation may be ten times or dozens of times of the value of mobile Internet connection.
Generative AI will gradually penetrate and take over the digital world, and most things in the physical world will have AI capabilities, forming the next generation of new products with AI capabilities, and connecting with the cloud AI driven digital world to generate synergies.
In the past, the focus of AI was mainly on simulating human perceptual abilities, such as natural language understanding, speech recognition, and visual recognition. However, the rise of generative AI has brought a qualitative leap, as AI is no longer limited to perception, but for the first time demonstrates the power of thinking, reasoning, and creativity.
AI models can understand various aspects of the real world through tokenization of physical world data, such as human walking, running, driving vehicles, using tools, painting, composing, writing, expressing, teaching, programming skills, and even starting a company. After understanding, AI can mimic humans to perform tasks in the physical world. This will bring about a new industrial revolution.
The automotive industry is undergoing such a transformation, where AI models directly learn massive amounts of human driving visual data, giving cars the ability to drive beyond most drivers.
Robots will be the next industry to experience tremendous changes. In the future, all movable objects will become intelligent robots. It can be a robotic arm in a factory, a crane on a construction site, a porter in a warehouse, a firefighter on a firefighting site, including pet dogs, nannies, and assistants at home. There will be many robots in the factory, producing robots under the command of AI models. Nowadays, every urban household has one or two cars, and in the future, each household may have two or three robots to help people improve their efficiency in life.
In Wu Yongming's view, the AI driven digital world connects the physical world with AI capabilities, which will greatly enhance the productivity of the entire world and have a revolutionary impact on the operational efficiency of the physical world.
Furthermore, AI computing is accelerating its evolution and becoming the dominant force in computing systems.
The reconstruction of the digital and physical worlds by generative AI will bring about fundamental changes in computing architecture, and this trend is already very evident. In the past few decades, CPU dominated computing systems have been accelerating their transition to GPU dominated AI computing systems. In the future, almost all software and hardware will have reasoning ability, and their computing cores will become a computing mode dominated by GPU AI computing power and supplemented by CPU traditional computing.
In the new computing power market, over 50% of new demand is driven by AI, and AI computing power demand has already taken the mainstream position, and this trend will continue to expand. In industries such as automotive, biomedicine, industrial simulation, weather forecasting, education, enterprise software, mobile apps, and gaming, the new industrial revolution is quietly evolving.
Wu Yongming stated that based on historical experience, people often overestimate and underestimate the new technological revolution in the short term and in the long term. Because in the early stages of new technology application, the penetration rate is still relatively low, and people's inherent experience has not experienced such events. Most people's instincts will naturally have doubts, which is normal. But the new technological revolution will grow in people's doubts, causing many to miss out in hesitation.