Nvidia CEO Jensen Huang is presenting a radical and striking vision for the future of software engineering: a world with no coding. In a recent podcast appearance, Huang revealed that every engineer at the $3 trillion chip giant now uses an AI coding assistant called Cursor. His ultimate goal, however, is to completely free engineers from the burden of writing code, a key part of the Jensen Huang AI coding philosophy.
The New Era at Nvidia: Jensen Huang’s AI Coding Mandate
Huang states that he would be overjoyed by a scenario where engineers write zero code, focusing solely on solving undiscovered problems. This approach is based on his frequently mentioned distinction between “Purpose and Task.” According to him, coding is merely a tool to complete a task, whereas an engineer’s true purpose is to identify and solve problems.
Furthermore, Huang is insistent on AI adoption within the company. When he learned that some managers were pressuring their teams to reduce AI usage, he reacted strongly. Asking them, “Are you crazy?” the CEO made it clear that he wants every task that can be automated to be handled by artificial intelligence. He promises employees they won’t lose their jobs; instead, they will take on more ambitious work.
To clarify his point, Huang used the example of radiology. He noted that contrary to predictions from AI pioneer Geoffrey Hinton, radiologists did not become unemployed. As machines began reading scans, the number of radiologists didn’t decrease; in fact, their salaries have increased by 48% since 2015. For Huang, reading a scan was just a task; the real purpose is to diagnose illness and guide treatment.
A Cautious Reality: Is the No-Code Future Here Yet?
However, not everyone in the tech world shares Huang’s enthusiasm. Even Michael Truell, the CEO of Cursor—the very tool Nvidia has adopted, now valued at $29.3 billion—speaks cautiously. Truell warns against scenarios where developers let AI build software without reviewing the code. He points out that blindly trusting artificial intelligence could undermine foundational principles and lead to project failures.
Similar skepticism comes from Andrej Karpathy, the former Director of AI at Tesla. Karpathy admitted that AI agents do not yet work well enough and that he had to write most of his own project by hand. Additionally, a study from July 2025 revealed that experienced developers using AI tools actually worked 19% slower, despite feeling 20% faster. These data points suggest that the code-free future Huang envisions might be a bit further away than anticipated.
So, what are your thoughts on Jensen Huang’s vision? Share your opinions with us in the comments!
