At Computex 2025, Nvidia CEO Jensen Huang shared a powerful insight into the evolving nature of artificial intelligence: programming AI is no longer about writing code—it’s about teaching and training, much like educating a human being. Huang’s comment reflects the dramatic transformation in how machines are built to learn, reason, and interact.
A Shift from Coding to Coaching
Traditionally, programming meant meticulously crafting lines of code that machines would follow. But modern AI, especially large language models (LLMs) and generative AI, learns through exposure to data rather than hard-coded instructions. Huang explained that instead of programming specific behaviors, developers now “train” AI systems by feeding them massive datasets and allowing them to learn patterns, nuances, and decision-making processes.
This approach mimics how humans learn: through examples, trial and error, feedback, and context. Just as a child learns language by listening and mimicking, AI models are trained using vast amounts of text, images, and interactions.
Nvidia at the Center of the AI Boom
Nvidia is playing a central role in this AI revolution. Its GPUs (Graphics Processing Units) power nearly all leading AI systems today—from OpenAI’s ChatGPT to Google’s Gemini and Meta’s Llama. Huang likened the GPU to a modern-day AI factory, where intelligence is being “manufactured” using data and computing power instead of physical materials.
At Computex, Huang also announced new GPU architectures and partnerships, including the highly anticipated Rubin platform and the launch of Blackwell Ultra chips in 2025. These developments are expected to further accelerate the training and deployment of AI models across industries.
Implications for Developers and Businesses
This shift from programming to training AI alters the role of developers and business leaders. Success in AI now depends more on data quality, training techniques, and ethical considerations than on traditional software engineering skills. Developers are becoming AI teachers, curating datasets, designing training objectives, and fine-tuning behavior based on performance.
For businesses, this means adopting a mindset focused on continuous learning and adaptation. AI isn’t static—it evolves with data, user interaction, and fine-tuning. Companies that can manage, refine, and govern AI training will gain a significant edge in the digital economy.
Final Thoughts
Jensen Huang’s statement captures a key turning point in the tech world: AI is no longer just software—it’s a system that can learn, adapt, and reason like a human. As Nvidia continues to drive AI infrastructure forward, the future will increasingly rely not on writing code, but on teaching machines to think. In this new era, the best developers may look more like educators—and the most powerful technology may feel more like a student than a tool.
