Jensen Huang is correct that engineering will thrive in the AI era, but the version of engineering that survives will look almost nothing like what we teach today.
That’s my read after sitting with Huang’s repeated, emphatic claim that engineering is the most noble career path a person can choose — and that AI, far from displacing engineers, will trigger a new Industrial Revolution that creates more engineering roles than it destroys. As someone who spends most of her working hours thinking about how agent systems are actually architected, I find Huang’s thesis directionally right but dangerously underspecified.
What Huang Is Actually Saying
Huang’s position is not subtle. He believes AI will transform every job on the planet, and that engineering sits at the center of that transformation — not as a casualty, but as the engine. He has called engineering the most noble career, and Nvidia’s own three-decade arc is his evidence. The company he built has fundamentally redefined what an engineer does, moving from chip design into the infrastructure layer that now underpins nearly every serious AI system in production.
His broader argument is that AI creates new categories of work rather than simply eliminating old ones. This is consistent with how major technological shifts have played out historically. The Industrial Revolution didn’t end labor — it reorganized it, scaled it, and in many cases made it more specialized. Huang is betting the same dynamic applies here.
Where the Analysis Gets Interesting
From an agent architecture perspective, Huang’s framing raises a question that I don’t think gets enough attention: which kind of engineer are we actually talking about?
There is a meaningful difference between the engineer who builds AI systems and the engineer who works alongside them. Both will exist. Both will be valuable. But the skill profiles are diverging fast, and the educational pipeline has not caught up.
Consider what it means to build a production-grade agentic system today. You need people who understand how to structure multi-step reasoning chains, how to manage tool-use reliability, how to design fallback logic when an agent hits an ambiguous state. That is engineering work — deeply technical, genuinely hard, and not something a language model can fully automate away. The demand for people who can do this well is real and growing.
But there is also a second category: engineers in traditional domains — civil, mechanical, electrical — who will increasingly work with AI tools as collaborators. Huang has gestured at this too, noting that AI will transform all roles. For these engineers, the question is not whether they will be replaced, but whether they will adapt quickly enough to stay effective. The ones who treat AI as a capable but fallible junior colleague — one that needs clear instructions, structured outputs, and human verification — will do well. The ones who either ignore it or over-trust it will struggle.
The New Industrial Revolution Framing
Huang’s “new Industrial Revolution” language is doing a lot of work, and I think it deserves scrutiny rather than applause.
The original Industrial Revolution created enormous wealth and enormous disruption simultaneously. It expanded opportunity at the macro level while causing serious dislocation at the individual level, often for decades before the benefits distributed broadly. If the AI era follows a similar arc — and there are structural reasons to think it might — then Huang’s optimism about job creation is probably correct in aggregate and potentially cold comfort for specific workers in specific roles during the transition period.
This is not a reason to dismiss his thesis. It is a reason to be precise about it. Engineering as a category will grow. Individual engineers who do not update their skills will face real pressure. Both things are true.
What This Means for Anyone Thinking About Career Bets
- Systems thinking is the durable skill. Engineers who understand how components interact — whether in a physical system or an AI pipeline — will find their mental models transfer across the transition.
- Prompt engineering alone is not enough. The engineers who will matter most are those who can reason about failure modes, not just happy paths.
- Domain expertise plus AI fluency is a stronger position than either alone. A structural engineer who understands how to use AI-assisted simulation tools is more valuable than a pure AI generalist with no domain grounding.
Huang has spent thirty years building the physical infrastructure that makes the AI era possible. When he says engineering will drive what comes next, he is speaking from a specific vantage point — and that vantage point is worth taking seriously. The engineers who thrive will be the ones who treat AI as a tool they understand deeply, not a force they simply adapt to.
That distinction is small on paper and enormous in practice.
🕒 Published: