Conference Presentation, Fireside Chat, Interview, Keynote
A Conversation with Nvidia CEO Jensen Huang | Global Conference 2025
- AI as an Industrial Revolution: Jensen Huang defines AI as the "next industrial revolution," distinct from past IT tools because it automates tasks and augments the digital workforce rather than just replacing labor.
- New Economic Layer: The AI economy is positioned within a $100 trillion scope, with "AI factories" emerging as a new industry layer.
- These factories produce "tokens" (numbers) that are reformulated into words, images, video, chemical combinations for drug discovery, or motor skills for robotics.
- Current AI factories are scaling to gigawatt sizes, costing approximately $50–$60 billion each.
- Forecasts suggest the construction of tens of gigawatts of AI factory capacity globally over the next decade.
- Intelligence Infrastructure: AI is characterized as a foundational infrastructure layer similar to energy and information systems, affecting every industry from finance and healthcare to manufacturing and logistics.
- Labor Market Dynamics:
- While some jobs will be lost, Huang asserts that no one loses their job to AI, but rather to someone else using AI.
- AI is positioned as the primary tool to close the global technology divide, allowing the 7.5 billion non-programmers to interact with technology via natural language rather than code.
- Huang projects AI could reintegrate 30–40 million workers into the global workforce, addressing current labor shortages and boosting global GDP.
- NVIDIA's Strategic Origins: The company spent 33 years executing a mission to solve problems "normal computers cannot," a path that initially lacked customers and competitors because the market preferred solvable problems.
- This strategy mirrored Peter Thiel's "Zero to One" concept, creating a unique, isolated position until the technology matured.
- Intel previously rejected NVIDIA's approach due to their success in their existing domain.
- Corporate Culture: NVIDIA maintains a culture of "perpetual going out of business," fostering resilience, efficiency, and the willingness to endure long-term struggles without taking success for granted.
- Chip Manufacturing Scale: NVIDIA operates as an end-to-end AI infrastructure company, building the entire stack from architecture and software to systems and networking.
- Each AI system weighs roughly 1.5 tons, costs $3 million, and requires liquid cooling and supercomputer-level testing.
- The supply chain involves 200 global manufacturers and suppliers, with an R&D budget of $20–$30 billion per generation.
- Huang notes a critical shortage of U.S. talent to build factories, estimating a need to import 7,000 workers from Taiwan.
- Export Control Stance: Huang argues against restricting chip exports to adversaries like China on national security grounds.
- He contends that restrictive policies do not limit military advancement as adversaries already possess millions of chips or can secure their own resources.
- Restricting sales forfeits a $50 billion market (comparable to Boeing's total value) and allows competitors like Huawei to establish rival global standards.
- The strategic goal is to establish American technology standards globally by serving all markets.
- Customer-Driven Innovation: NVIDIA's architecture evolves based on direct feedback from diverse customers, including virtual protein modeling, virtual cells, and robotics.
- Future demand is shifting from consumer internet applications to "Physical AI" in healthcare, life sciences, and manufacturing.
- The future factory is envisioned as a "giant robot" orchestrating nested layers of robotic workers.
- Hiring Philosophy: NVIDIA prioritizes domain expertise (e.g., digital biology, quantum chemistry, robotics) and general intelligence alongside technical skills.
- The company values "resilience" and the willingness to endure hardship, citing a professor's observation that character is revealed by how long it takes for unmerited advantages (relationships) to be displaced by hard work (ability).