Jensen Huang has been in the news lately, nearly as often as Donald Trump.
It’s likely one of them will be in charge of how we live in the future.
And it’s not Donald.
Jensen Huang, a.k.a. 黃仁勳, is the CEO of supercomputer manufacturer Nvidia, a shining star in the growing constellation of Big Tech. Really, they are both shining stars: Nvidia and Mr. Huang.
Under Mr. Huang’s guidance, Nvidia developed a special type of computer chip — the GPU, Graphics Processing Unit — which allowed video game designers to make artificial worlds more realistic, and faster loading. (Why humans love to make artificial things look realistic, I cannot explain. But we do.)
The GPUs could process information much faster than normal computer CPUs — Central Processing Units. A GPU empowers a person — probably an immature male — to shoot more monsters, more quickly, thus releasing a larger dose of dopamine into the brain.
But back in 2012, a couple of University of Toronto students, Alex Krizhevsky and Iiya Sutskever, purchased two Nvidia GPU cards, hooked them into a computer ‘neural network’, and trained the network to look at images. The network didn’t have eyes, but it was programmed to make sense of pixels in digital photographs. Pretty soon, the network could tell the difference between a cat and a dog, or between a car and a truck.
Those are things I myself occasionally struggle with. Like… is a Chevy El Camino a truck, or a car?
Maybe it’s a truck-car?
Anyway, Alex named his network ‘AlexNet’. Nice name. When he entered AlexNet into the annual ImageNet visual-recognition contest in 2012, it scored so well, Alex was suspected of cheating.
I know how that feels. I was suspected of cheating, once, in tenth grade, and it was very embarrassing. Unlike Alex, however, I was actually cheating.
The design of Alexnet opened up an entirely new world of machine learning. Suddenly, computers could learn tasks as fast as video gamers could shoot monsters. A few years later, computers using Nvidia chips could identify images more quickly and more accurately than humans. (Not sure about identifying El Caminos, however.)
Seeing the writing on the wall, Mr. Huang and Nvidia totally shifted gears and began creating supercomputer hardware and software for AI. Basically, Nvidia built the system that made ChatGPT possible… and these machines appear destined to become better than humans at practically everything.
Not everything, just practically everything.
Which means, of course, that they will be better at cheating. Except being programmed machines, they won’t know that they’re cheating, unless we tell them. Boy, will they be embarrassed.
Eventually, everything we own will be autonomous, doing all the boring stuff humans used to do. We’ll walk into the kitchen and say something like, “Gee, I’d really like macaroni-and-cheese, and a green salad,” and the kitchen appliances will set to work. The refrigerator will ask, “Cheddar? Monterey Jack? Feta?” (I will choose Cheddar, of course.) Then, while it’s cooking, I will instruct my slippers to meet me at the couch, and the TV will have fired up my favorite show.
And none of the machines will argue with me. They won’t even think about arguing. They will be autonomous, but respectful.
The change from linear computers using CPUs, to networked computers using GPUs, has been very profitable for Mr. Huang and his company. When OpenAI announced that ChatGPT had been trained on an Nvidia supercomputer, it spurred the largest single-day gains in stock market history, on May 25, 2023. Nvidia’s value increased by about two hundred billion dollars. In one day.
That’s probably more money that I will earn in my entire life.
Many of the scientists working in the AI industry — and scientists who previously worked in the AI industry and who are now building survivalist bunkers in Idaho — are concerned that when computers can do everything better than humans, they will eventually decide that they no longer need us.
I’m actually looking forward to that.
When my wife Darlene (now ex-wife) decided she no longer needed me, I was terribly depressed at first.
But now I realize, it was the best thing that could ever happen.
I became autonomous. Imagine that.