Artificial general intelligence (AGI) remains a lofty aspiration, but artificial intelligence itself continues to break new ground. On September 19, Alibaba Group CEO Eddie Wu shared his vision of AI’s future at the Apsara Conference, making it clear that the technology’s reach is set to far exceed current expectations.
Wu opened by asserting that AI’s greatest impact won’t just be about refining digital tools or rolling out the next generation of apps. Instead, it will rewire the physical world. Robotics, he emphasized, is on the cusp of a major breakthrough. Wu predicted that nearly every movable object—whether it’s a factory arm or a household appliance—will soon be an intelligent robot. This transformation, he said, will spark a wave of productivity that spans industries, positioning AI-driven robots to be as common as today’s family car.
The real frontier, according to Wu, isn’t just about AI mimicking human perception but also mastering how to operate and think within the physical world. This shift will unlock new forms of value, with the potential to boost global productivity by tenfold or more.
AI’s influence goes beyond reshaping industries—it’s changing the core of computing itself. For decades, CPUs have been the backbone of computing, but Wu suggested this era is waning. The future, he argued, will be defined by AI-first GPU systems, with nearly all software and hardware integrating AI inference capabilities. It’s a shift that reflects how AI’s needs are overtaking traditional computational demands.
Over half of new computational requirements today stem from AI, and this figure is climbing. Wu acknowledged Alibaba Cloud’s substantial investments in AI computing, but even those efforts are lagging behind the soaring demand. The road ahead will require scalable infrastructure tailored to AI’s needs, touching industries from automotive to biomedicine, and weather forecasting to gaming.
In his address, Wu highlighted the breathtaking speed of AI’s development. In the past 22 months alone, AI has progressed at an unprecedented rate, accelerating faster than any other technological advancement in history. Despite this momentum, he cautioned that the journey has just begun. Achieving true AGI, he warned, will demand significant financial commitment and sustained effort from the global AI community.
A key driver of AI’s recent surge in adoption is the plummeting cost of inference. Wu pointed out that Alibaba’s Bailian platform has slashed inference costs by 97%, making it more affordable for businesses and developers to access advanced AI tools. This price drop has triggered widespread adoption, expanding AI’s reach across industries.
Alibaba’s open-source AI ecosystem is flourishing as a result. The company’s Qwen-2 model, which was made open-source in mid-2023, rapidly became a favorite on Hugging Face. The open-source community has since exploded, with over 50,000 models available and serving more than 6.9 million developers globally. However, Wu emphasized that more powerful and complex models will be essential to fully realize AI’s potential.
Already, AI is learning from vast amounts of data—surpassing human drivers in tasks like autonomous navigation. In the near future, it will extend into intelligent robots that will transform fields such as manufacturing, logistics, and home management.
Wu concluded by reminding companies that technological evolutions are often underestimated at their inception. His message was clear: the future is both exciting and fraught with challenge. Fail to embrace AI’s potential, and companies may find themselves at a serious disadvantage in the years ahead.