Technology & Power
NewsOnScale Staff
October 25, 2026
Every AI system running today runs on Nvidia chips. Every large language model. Every surveillance camera with facial recognition. Every autonomous weapons system. Every military targeting platform. Every data center training the AI that will define the next century of human civilization.
Nvidia controls more than 80% of the market for the chips used to train and deploy AI models. The company provides chips for over 75% of the world's top 500 supercomputers.
One company. One man. The infrastructure of the AI age.
Who Jensen Huang is
Jensen Huang was born in Taiwan in 1963. His family moved to the United States when he was nine years old. He has described his early years in America as difficult — he and his brother were sent to a boarding school in Kentucky that turned out to be a school for troubled youth, where he learned to survive by mopping floors and making the beds of older students.
He worked as a dishwasher at Denny's as a teenager. He met his wife at a Denny's where he was bussing tables. He studied electrical engineering at Oregon State University and then earned a master's degree from Stanford.
After Stanford he worked at LSI Logic and then AMD before co-founding Nvidia in 1993 with Chris Malachowsky and Curtis Priem. The company's name comes from the Latin word invidia — envy.
Nvidia's origin story is one of the more genuine ones in this series. Huang came from modest circumstances. His family was not wealthy. He worked service jobs. He earned his engineering credentials and built a company from technical expertise and vision.
The company he built is a different matter.
What Nvidia actually is
Nvidia began as a graphics chip company — making the processors that rendered video games. That market made it profitable. What made it dominant was a pivot that nobody outside the company fully anticipated: in the early 2000s researchers discovered that the parallel processing architecture of Nvidia's graphics chips was extraordinarily well-suited for the mathematical operations that machine learning requires.
The chip built for gaming became the chip that trains artificial intelligence. The same hardware that renders a video game at 60 frames per second can train a neural network on billions of data points. Nvidia did not plan this. But when the opportunity became clear Huang moved faster than anyone else.
Today Nvidia's H100 and B200 chips are the primary infrastructure of the global AI buildout. Every major AI company — OpenAI, Google, Meta, Microsoft, Amazon — depends on Nvidia chips to train and run its models. The waiting list for Nvidia chips has at times stretched to a year or more. Countries are fighting over access to them.
The government relationship
Nvidia has received DARPA contracts for embedded processor research going back years. It delivers AI supercomputers to US military institutions — Huang personally traveled to the Naval Postgraduate School in Monterey to commission a DGX GB300 system for the military academy. It is building seven AI supercomputers with the Department of Energy.
Huang has publicly supported the US military's use of AI for national security purposes. At the Milken Global Conference he argued that private companies should not block the government from deploying AI for lawful national security purposes. He said elected governments have the authority to determine how such technologies are used in defense and security.
The man who controls 80% of the AI chip market is publicly advocating for government authority over AI deployment — while his company holds the government contracts to supply the chips for that deployment.
The China problem
This is where Jensen Huang's story becomes significantly more complicated.
Nvidia has been one of the largest suppliers of advanced AI chips to China. Chinese companies including Alibaba and Tencent — the same Tencent that owns a stake in Discord — have been among Nvidia's largest customers.
A US Congressman accused Nvidia and Huang of actively supporting the Chinese military by selling advanced AI chips to Chinese military-linked firms. Huang responded that the Chinese military does not need Nvidia's chips because they can build their own.
Meanwhile China is currently constructing massive data centers to house over 115,000 Nvidia AI GPUs. The chips that Huang says China doesn't need for its military are being installed in Chinese data centers at scale.
The Trump administration has been negotiating with Nvidia over the details of chip sales to China — reportedly taking a 25% cut of revenues on approved sales. Nvidia was also reportedly attempting to sell chips to ByteDance — TikTok's Beijing-based parent company — before hitting a regulatory sticking point over rules preventing American products from reaching the Chinese military.
The company that supplies 80% of the world's AI training chips has been simultaneously supplying American military institutions and Chinese technology companies with potential military connections.
What the chip dominance means
Nvidia's position is unlike anything in the history of technology infrastructure. Standard Oil controlled oil. AT&T controlled telephone lines. Google controls search. Each of these monopolies controlled access to one critical resource.
Nvidia controls access to the computational substrate of artificial intelligence itself. Every AI system — military, commercial, surveillance, medical, financial, governmental — runs on chips that Nvidia makes in quantities no competitor can match at quality no competitor can equal.
That is not a technology company. That is infrastructure. And the man who controls that infrastructure sits at the center of every decision about AI that matters — in America, in China, and everywhere else.
The question nobody is asking
When one man's company controls 80% of the chips that power every AI system on earth — including the AI systems used by the US military, the AI systems used by Chinese technology companies with military connections, the AI systems used to conduct mass surveillance, and the AI systems that will increasingly make decisions affecting every American's life — that is a concentration of infrastructure power that the founders of this republic had no framework to anticipate.
They warned about monopolies on land. On money. On political power. They did not warn about monopolies on the computational substrate of intelligence itself because such a thing did not exist.
It exists now. It is called Nvidia. Its chairman is Jensen Huang. He went to Denny's as a dishwasher and built the most strategically important company in the world.
The question is not whether he deserves credit for what he built. He does.
The question is whether any single person should control the infrastructure of artificial intelligence for the entire planet — selling chips to the American military and to Chinese companies with military connections simultaneously, with no meaningful oversight and no democratic accountability.
That question is not being asked. Not in Congress. Not in the media. Not in the boardrooms of the companies that depend on his chips to operate.
It should be.
NewsOnScale is owned by AMILLI AI, Corp., founded by JJ Johnson. JJ Johnson is a declared 2028 presidential candidate. FEC filing: JJJOHNSON2028.