Nvidia CEO Jensen Huang has a simple answer to the fear that AI will make engineers obsolete: engineering is not the same thing as typing code.
That distinction matters. AI can already generate code, test fragments of software, summarize documentation, and automate many narrow technical tasks. But Huang’s broader point is that the purpose of engineering has never been only code production. It is problem discovery, system design, judgment, trade-offs, testing, safety, and turning messy real-world constraints into working technology.
In January 2026, Huang was named the 2026 IEEE Medal of Honor recipient, recognized for his leadership in GPU development and its application to scientific computing and artificial intelligence. IEEE described the award as its highest honor, and the recognition places Huang’s comments about engineering inside a larger story: the AI boom is not only about software. It is also about infrastructure, energy, chips, data centers, robotics, safety, and the physical world. 0
Huang has described engineering as one of society’s foundational professions. That is not just motivational language. His argument is practical. If AI removes some of the repetitive parts of engineering, good engineers may become more valuable, not less, because they can spend more time on the hard parts: identifying the right problem, asking better questions, designing the system, and checking whether the answer actually works.
This is especially clear in software engineering. Coding is a task. Solving problems is the job. A model can write a function. It cannot automatically know which product constraint matters most, which failure mode is dangerous, which customer need is misunderstood, or which system assumption will break at scale. That kind of work still requires technical judgment, domain knowledge, and responsibility. 1
The same logic applies beyond software. AI infrastructure is becoming a huge engineering project. Huang has described AI as a multi-layer system involving energy, chips, cloud infrastructure, models, and applications. The World Economic Forum summarized his view as “the largest infrastructure build-out in human history,” because each layer must be built, powered, maintained, secured, and integrated. 2
That means the AI era does not only need prompt users. It needs electrical engineers, mechanical engineers, software engineers, chip designers, data center specialists, robotics engineers, network engineers, cybersecurity specialists, and people who understand how technical systems behave when they meet the real world.
One of Huang’s more provocative ideas is that future engineers will not work alone. They will work with many AI agents, using AI compute almost like a new industrial tool. In March 2026, reports from Nvidia’s GTC discussions noted Huang’s view that high-value engineers should use large amounts of AI “tokens” to expand their productivity, much as chip designers would use modern CAD tools instead of pencil and paper. 3
There is a self-interested side to this, of course. Nvidia sells the hardware and infrastructure that make this AI-heavy future possible. But the underlying point still holds: the valuable engineer of the next decade may not be the person who manually writes the most code. It may be the person who can direct machines well, understand a domain deeply, verify outputs rigorously, and connect mathematics, physics, software, hardware, and human needs into one working system.
This also explains Huang’s focus on “physical AI.” The next phase of AI is not limited to chatbots and text generation. Nvidia has been pushing into robotics, autonomous systems, factories, logistics, transport, and machines that interact with the physical world. In March 2026, Nvidia said Huang framed physical AI as a shift where industrial companies increasingly become robotics companies. 4
For young engineers, the lesson is not “ignore AI.” That would be naive. The lesson is also not “AI will do everything.” That is too simple.
The more useful lesson is this: learn AI fluency, but do not confuse tool fluency with engineering depth. Learn how models work, how to use them, how to test their outputs, and how to automate repetitive work. But also keep building the harder foundations: mathematics, physics, computer science, systems thinking, debugging, communication, and domain expertise.
At InsightArea, Costin Liculescu often writes about exactly this kind of intersection: science, mathematics, computer science, artificial intelligence, and the way complex ideas become understandable when we connect them carefully. Huang’s argument fits that broader theme. AI changes the tools of engineering, but it does not remove the need for disciplined thinking.
In fact, it may make disciplined thinking more important.
When code becomes cheaper, judgment becomes more valuable. When answers become easier to generate, verification becomes more important. When tools become more powerful, the responsibility of the engineer grows.
That is why engineering may not disappear in the AI era. It may become more demanding, more interdisciplinary, and more central to how society builds the next layer of technology.
Comments are closed.