Jensen Huang on move from generative AI to agentic AI

Key Concepts

• Agentic AI: AI systems capable of reasoning, planning, and executing complex tasks using external tools (browsers, software, compilers) rather than just generating content.
• Token Generation: The fundamental unit of output for AI models; in this context, it represents the "product" being sold.
• Compute-as-Revenue: The economic model where investment in computational power directly correlates to the ability to generate profitable AI tokens.
• Inference: The process of running a trained AI model to make predictions or perform tasks.

---

The Shift from Generative to Agentic AI

Jensen Huang highlights a fundamental evolution in the artificial intelligence landscape. While the industry previously focused on "Generative AI"—which primarily creates content—it has now transitioned to "Agentic AI."

• Capabilities: Agentic AI systems possess the ability to understand complex prompts, reason through problems, and formulate plans.
• Tool Integration: These systems can actively utilize software tools, including web browsers, spreadsheet applications, PowerPoint, simulators, and C compilers.
• Productivity: The core distinction is that AI has moved from being a novelty to a tool that performs "productive work," effectively writing and executing code to deliver tangible value.

The Economic Engine: Compute and Tokens

The transcript outlines a direct causal link between computational infrastructure and financial performance.

• The Token Economy: AI model builders are now generating tokens that have inherent market value. Because these tokens are profitable, there is an insatiable demand for the infrastructure required to produce them.
• Compute-as-Revenue: Huang posits that in the current AI era, "compute is revenue." This explains the parabolic demand for NVIDIA’s hardware; companies are aggressively scaling their compute capacity to maximize their token generation capabilities, which in turn drives their revenue growth.

Financial Performance and Market Position

The discussion underscores the rapid growth of the company, supported by specific financial indicators:

• Revenue Growth: The company reported an 85% increase in revenues.
• Future Outlook: The forecast stands at $91 billion, reflecting the massive scale of the current demand.
• Market Share: The company is rapidly expanding its share in the "inference" market, which is the operational phase where AI models perform their tasks and generate the aforementioned tokens.

Synthesis and Conclusion

The primary takeaway is that the AI industry has reached a critical inflection point where AI is no longer just a creative assistant but an autonomous agent capable of executing work. This shift has transformed AI from a cost center into a revenue-generating engine. Because the generation of AI tokens is now a profitable business model, the demand for high-performance compute has become a strategic necessity for AI developers. Consequently, the company’s growth is being fueled by the industry's need to scale infrastructure to meet the demands of this new, agentic-driven economy.