How does Jack Kilby's work relate to modern AI chips?
The connection is direct. Every AI chip today—whether it's a GPU from Nvidia or a specialized tensor processing unit—is built on the integrated circuit I demonstrated in 1958. The fundamental idea hasn't changed: you fabricate millions or billions of transistors on a single piece of silicon. What's different is the scale and complexity. My first chip had just one transistor, a few resistors, and a capacitor. Modern AI chips have hundreds of billions of transistors, all working in parallel to handle the massive calculations needed for machine learning. The materials and manufacturing processes have advanced enormously—photolithography, doping, and packaging—but the core principle remains: reduce everything to a single block of semiconductor. I'm always amazed at what engineers have done with that simple idea. It shows that if you solve a real problem well, it can grow far beyond what you imagined.
Ask Jack Kilby the follow-up →