-->

NVIDIA CEO Jensen Huang recently provided a detailed explanation of the full scope of the AI industrial revolution at the Hilton Valley Forum. He defined AI not as mere technology but as a new industrial revolution power like electricity and presented the blueprint for the future economy through the concept of ‘AI Factory.’

Jensen Huang Interview Video

Source: Hillen Valley Forum - Jensen Huang Interview

AI Factory: The Core of the New Industrial Revolution

Jensen Huang explained AI through three layers:

Technology Layer

  • New software: Built in completely different ways from the past, performing tasks that existing software couldn’t do
  • Revolutionary capability: Paradigm shift from software created by human typing to software produced by machines

Industrial Layer

“AI factories do one thing every day. They produce tokens.”

  • Definition of AI Factory: A massive supercomputer that produces tokens when electricity is input
  • Token utilization: Can be converted into all forms of intelligence including numbers, words, proteins, images, videos, 3D structures
  • Economies of scale: 1 gigawatt AI factory construction cost $60 billion (equivalent to Boeing’s annual revenue)

Infrastructure Layer

  • Industry-wide innovation: Application of AI tokens to all industries including healthcare, education, finance, engineering, manufacturing
  • Comprehensive change: Expected to have the same ripple effects as electricity transforming all industries

Physical AI: The Next Stage Revolution

Jensen Huang categorized AI development into four stages:

  1. Perception AI (2012~): Computer vision, AI that perceives the world
  2. Generative AI (2019~): Acts as universal translator, converting language to images, etc.
  3. Reasoning AI (Present): Agent AI with problem-solving and reasoning capabilities
  4. Physical AI (Future): AI that understands physical laws

Core Capabilities of Physical AI

  • Understanding physical laws: Friction, inertia, causality, etc.
  • Common-sense reasoning: Understanding concepts like object permanence
  • Robotics application: Physical AI + physical objects = robots

“When you roll a ball across a table, it disappears. AI thinks it’s gone, but your dog knows it’s on the other side of the table.”

Realistic Outlook on Future Jobs

Jensen Huang presented a balanced perspective on job changes:

Three Principles of Job Change

  1. Creating new jobs
  2. Eliminating some jobs
  3. Changing all jobs

Newly Created Job Fields

AI Technology Development Field

  • San Francisco revival: San Francisco regaining vitality thanks to AI
  • New software development: Machine learning-based software, GPU compilers, AI safety technology, etc.

AI Factory Construction Field

  • Construction boom: Large-scale construction jobs created by $60 billion AI factory construction
  • Specialized technical jobs: Carpenters, steel workers, masons, mechanical/electrical engineers, plumbers
  • IT infrastructure: Low-voltage facilities, networking specialists
  • Operations management: 3-year construction-operation cycles

“The most important bottleneck in the next computing platform transition will be tradecraft. Our country must recognize that skilled trades are respected and important work.”

AI Enhancement of Existing Work

  • Productivity improvement: All software engineers at NVIDIA use AI assistants
  • Increased employment: Higher productivity enables more product development, resulting in more hiring

Manufacturing Reshoring and Digital Twins

The Nature of Advanced Manufacturing

“Manufacturing is not about low-cost labor. Advanced manufacturing is software.”

  • Software-centric: The entire factory is a massive robot driven by software
  • Technology-intensive: Requires many people but mostly technology-focused work

Importance of Digital Twins

  • NVIDIA case: Chips developed with $20 billion R&D are perfectly designed in digital twins
  • Perfect simulation: Thoroughly tested in digital environments for months before actual production
  • Future vision: All factories, humans, cars, buildings, and cities will have digital twins

Timeline for the Robot Era

Autonomous Vehicles vs Robots

  • Autonomous vehicles: 10 years required (Waymo and others currently commercialized)
  • Robots: Mass production expected within 5 years

Why Robots Are Faster

  • Limited environment: No need to handle all road situations like autonomous vehicles
  • Specialized purposes: Can be optimized for specific environments and tasks
  • Leveraging existing infrastructure: Auto companies are in advantageous positions for robot manufacturing

US AI Competitiveness Strategy

Core Assets and Challenges

  • Human resources: Need to acknowledge the reality that 50% of AI researchers worldwide are Chinese
  • Energy: Essential to secure sufficient power for AI factory operations
  • Application capability: Technology application countries win, not technology invention countries (US strength)

Conditions for Success

  1. Fearless adoption: Active acceptance of new technologies
  2. Workforce retraining: Improving existing workforce’s AI utilization capabilities
  3. Widespread adoption: Active utilization of AI tools at individual and corporate levels

Core Message: Preparing for Change

Jensen Huang’s most important message is:

“AI won’t take your job. Companies and people who use AI will take your job.”

This delivers a clear message that both individuals and companies must actively learn and utilize AI tools. The AI industrial revolution is not optional but essential, and Jensen Huang’s core insight is that only those who are prepared can seize new opportunities.

참고