Nvidia CEO Jensen Huang Warns Investors That the AI Market Is Bigger Than They Realize With Over ‘One and a Half Million AI Models in the World’

Bottom Line Up Front: Nvidia (NVDA) CEO Jensen Huang, in his role as the leader of the primary company behind the ongoing technology movement, says that many people don’t realize just how big the AI revolution is. He says there are currently over 1.5 million AI models throughout the world, from healthcare and drug discovery to more commonly known names like Elon Musk’s Grok and Sam Altman’s ChatGPT. 

The Details: Nvidia founder and chief executive Jensen Huang has consistently framed artificial intelligence as an infrastructure-scale transformation rather than a single technological breakthrough. In a fireside chat hosted by the Center for Strategic and International Studies (CSIS) with President and CEO Dr. John J. Hamre, Huang continued outlining his layered view of AI, emphasizing the systems, capital, and breadth of applications that underpin the technology’s long-term impact.

Building on earlier discussion of energy, chips, and systems, Huang described the current infrastructure buildout of AI as being multi-layered. The first layer is simple: energy. Huang frames electricity buildout as the most fundamental part of the entire AI revolution. 

The next is Nvidia itself: Chips. Without chips, there’s nothing to use electricity and process the AI compute needed to run the AI models. 

Huang then highlights the third layer, which “includes financial services, because it takes an enormous amount of capital to do what we do.” His remarks highlight that AI development at scale is not only a technical challenge, but also a financial one, requiring sustained investment in data centers, networking, and long-lived computing assets.

Huang then turned to the fourth layer that receives the most public attention: the models themselves. He acknowledged the prominence of well-known systems, describing how “this is where people largely focus on when they talk about AI,” citing examples such as ChatGPT, Anthropic’s Claude, Google’s Gemini, and xAI’s Grok. However, he placed those systems in a much wider context, pointing out that “those are four of the one and a half million AI models in the world.” The statement reframes popular generative models as an incredibly small subset of a far larger and more diverse ecosystem. This statement diversifies the risk away from the larger, more broadly known AI models, and instead frames Nvidia as an infrastructure play amongst a technological revolution being implemented across every industry in the world. 

The broader point of Huang’s remarks was that artificial intelligence is not confined to language or consumer interaction. He emphasized that “AI is not just intelligence that understands English or language,” but includes systems that “understand genes, proteins, chemicals, the laws of physics,” as well as AI that “understands quantum,” physical movement and robotics, long-term patterns, financial services, and healthcare across multiple data types. By listing these domains, Huang underscored AI’s role as a general-purpose technology applicable across science, industry, and services.

This perspective is consistent with Nvidia’s evolution over the past decade. The company’s hardware and software platforms are widely used not only in consumer-facing AI, but also in drug discovery, climate modeling, industrial automation, and financial analysis. Huang’s authority on the subject stems from Nvidia’s position at the center of these applications, supplying the computing infrastructure that enables a wide range of AI workloads beyond high-profile chat interfaces.

In a broader market and policy context, Huang’s comments speak to a recurring challenge in technology cycles: public focus often narrows on the most visible applications while underestimating the underlying infrastructure and the diversity of use cases. As investment flows into AI-related companies, distinctions between model developers, infrastructure providers, and domain-specific applications become increasingly important. His layered framework offers a way to evaluate AI development that accounts for capital intensity, specialization, and long-term deployment rather than short-term novelty.

By describing AI as a platform with multiple layers and millions of specialized models, Huang positions the technology as a foundational shift comparable to previous industrial transformations. His remarks suggest that the long-term impact of AI will be shaped less by any single model and more by how broadly intelligence is applied across disciplines, industries, and economic systems.