First, no victor in the AI era can afford to sideline the Chinese market. This reality alone casts a long shadow over the prospects of Nvidia and OpenAI.
Second, the technologies currently in the spotlight—whether it’s the so-called foundation models or the GPU accelerators—face an inevitable erosion of their technological moats. In the future, these businesses are likely to depend more on economies of scale and rigorous cost control. History tells us that while such ventures can generate steady profits, they rarely yield exceptional margins.
Finally—and most critically—the outdated business models underpinning these companies preclude them from becoming the true leaders of the AI revolution.
Consider the early days of the personal computer era. The media’s spotlight shone brightly on hardware giants like IBM, WANG, and Compaq, the vanguards of that period. Yet, in hindsight, they were not the ultimate winners. It was the then-overlooked Microsoft and Intel that ultimately defined the era.
Similarly, during the internet boom, companies such as Cisco in network hardware, AOL and Yahoo in the portal space, and even Netscape once enjoyed the limelight. But these too faded into relative obscurity, while the unheralded Google and Alibaba emerged as the definitive powerhouses of the digital age.
What, then, distinguishes transient players from enduring titans? The answer lies in business model innovation. In the PC era, the breakthrough came with software licensing—a model that fundamentally separated ownership from usage rights. In the internet age, the disruptive innovation was the platform model, which effectively decoupled the roles of users and customers.
Likewise, we can boldly predict that the AI era will once again witness the old guard being swept aside by the new. Even luminaries such as Nvidia and OpenAI—firms still anchored in traditional business models—may ultimately prove to be mere passing phases. The true champion of this era will be the one that successfully redefines the AI business model.