At CES 2026, Jensen Huang said software is no longer written the old way. Developers won’t “program” it line by line. The new method is to “train” software like teaching a skill. This is changing how apps are built, improved, and updated, making learning more important than typing code manually. Huang explained that GPUs are the new main engine for running apps. Earlier, most software ran on CPUs, but now AI-based applications need GPUs to create visuals, words, and decisions instantly. GPUs are faster for this new style of computing and will power most future applications. He said apps are not pre-made or stored anymore. They are created fresh every time they run. Every pixel on the screen and every word in a response is generated instantly based on context. This means apps are dynamic, responsive, and built from scratch every single time. Jensen Huang compared computing layers to a cake with five parts. He said all parts are being rebuilt for AI. Nothing is untouched. This includes chips, software, apps, development methods, and investment focus. AI is pushing the biggest reset the computer industry has seen in a decade. Huang revealed where the money for AI change is coming from. Trillions of dollars from the past decade are being modernized for AI. Venture capital firms invest around $200 billion every year into AI innovation. Entire industries worth $100 trillion are shifting R&D budgets from old methods to AI, creating massive investment pressure and demand.
