Robots will become part of everyday life around the world within 5 years! Nvidia's Huang Renxun: Physical AI will become the key to the next generation of industries, and AI factories will lead to an explosion of blue-collar job vacancies

NVIDIA CEO Jensen Huang delivered a speech on 5/5, comprehensively analyzing how AI is reshaping modern economy, industrial architecture, and the job market. Using the "AI Factory" as a core metaphor, he discussed everything from technological fundamentals to policies, and even highlighted the global "labor force transformation challenge".

Three-Layer Structure Decoding AI: From Technology, Factory to Application

Huang stated that AI is bringing an industrial revolution similar to the "birth of electricity", which can be broken down into three levels:

• Technology: AI is no longer traditional software written by humans, but generated through supercomputers and massive electricity, producing intelligent Tokens that can be converted into text, images, proteins, structures, etc.

• AI Factory: These supercomputers operating AI are actually "AI Factories", continuously producing intelligent content daily.

• Application: AI will penetrate all industries including education, healthcare, finance, supply chain, and manufacturing, just like electricity permeated various sectors in the past.

AI Will Coexist with Traditional Factories: Every Company Must Operate on 'Dual Tracks'

Future companies will not only manufacture physical products but also operate an "AI Factory":

"For example, companies making lawnmowers or excavators must simultaneously produce the AI control system for that machine," Huang stated.

AI Three-Stage Evolution: From Perception to Generation, Then Entering Reasoning and Robotics

Huang divided AI development into three stages:

1. Perception AI: Starting from 2012, beginning with image recognition, such as computers understanding images, sounds, and temperature information

2. Generative AI: Enabling AI to translate languages, convert to images, videos, etc., "understanding and then producing"

3. Reasoning AI: Entering a phase "capable of solving unknown problems", with the emergence of AI agents with autonomous action capabilities

These AI Agents will integrate tools, execute tasks, and become "digital employees" within enterprises, with the IT department functioning like "HR for AI employees" in the future.

The Next Wave is "Physical AI": Understanding Common Sense, Physics, and Mobility

The AI that can truly transform large industries is one that "understands the physical laws of the real world".

• Understanding that objects cannot pass through tables, balls will fall, and items have inertia

• Having the concept of "object permanence" (for example: a dog knows the ball hasn't disappeared, it just rolled to the other side of the table)

"When AI has these physical common sense principles, and is placed in a robot, it becomes the next-generation 'Physical AI Robot'."

These AI robots will become the main force in new factories and logistics systems, solving global labor shortages.

Who Can Lead This AI Game? The Key is "Energy, Talent, and Application Speed"

Huang Renxun pointed out that for the US to win in the global AI race, it must control three key resources:

1. Talent (Intellectual Capital): Currently, about half of global AI researchers are from China

2. Energy: AI factories are essentially "converting electricity into intelligent products", with energy being the raw material

3. Application: The winner will be the "country that applies new technologies the fastest"

AI Revives Old Industries! New Factories Create Massive Blue-Collar Job Opportunities

A 1GW AI factory costs up to $60 billion, equivalent to Boeing's entire annual revenue.

• Factory construction requires traditional workers like electricians, plumbers, steel workers, and pipeline workers

• Future growth bottlenecks will not be software engineers, but a shortage of technical workers

Huang Renxun directly stated: "We need to re-respect and cultivate these traditional technical professions."

It's Not AI Stealing Your Job, But Those Who Can Use AI Will Take Your Job

He emphasized: "Some jobs will disappear, some will be newly created, but every job will be changed." For example:

• Software engineers now work with AI agents to increase productivity

• Company overall output and revenue increase, naturally creating more employment opportunities

• AI is not a threat, the real risk is falling behind in AI usage speed.

Digital Twin Technology: Building Virtual Versions of Factories Before Physical Construction

Huang Renxun emphasized that NVIDIA itself designs chips using "digital twin" technology:

• All chips undergo digital simulation and verification before physical manufacturing

• He calls for establishing "digital factories" before construction to significantly reduce costs and errors

In the future, every individual, city, and factory will have a "digital twin" for management and simulation.

How Long Until Robots Become Commonplace? Huang Renxun: Widespread Within 5 Years

Although self-driving cars took 10 years to truly hit the road, Huang Renxun believes robots will be faster:

• Because they can be limited to fixed spaces (warehouses, factories), the requirements are not as complex as self-driving cars

• Estimated to be mass-produced and enter daily life and workplaces "within 5 years"

Finally, he solemnly reminded everyone:

"This is not a 60-minute game, this is an infinite game with no endpoint."