Jensen Huang says his software engineers at Nvidia have fundamentally shifted their work. In a company interview published Wednesday, the Nvidia CEO stated that his engineers now prefer building AI agents to writing traditional code .
“If you ask me, every one of my software engineers prefers to be building agents than to be writing Python code.” — Jensen Huang, Nvidia CEO
Jensen Huang says this shift represents an evolution from mundane tasks to creative, high-value work. Engineers are spending less time on what he calls “typing” — the mechanical act of coding — and more time on building agents, benchmarks, and guardrails .
The Shift: From Coding to Agent Building
Jensen Huang says AI has transformed the software engineering role at Nvidia. The distinction he draws is between coding as a task and engineering as a craft :
| Before | After |
|---|---|
| Writing Python code | Building AI agents |
| Manual syntax work | Creating guardrails |
| Repetitive coding | Designing benchmarks |
| Individual tasks | Agent orchestration |
| “Typing” work | Creative problem-solving |
“You’re taking all the mundane work, and you’re trying to get this agent to do it. That requires imagination, that requires creativity, a lot of technology.” — Jensen Huang
Jensen Huang says AI agents break down a task into multiple smaller steps, each tackling a specific part to achieve a broader objective . This allows engineers to focus on higher-level architecture and system design rather than syntax.
AI Agents: What They Actually Do
Jensen Huang says AI agents are fundamentally different from traditional software. An AI agent :
- Breaks a large goal into a sequence of smaller steps
- Plans and acts rather than simply answering a prompt
- Automates mundane work
- Requires ongoing engineering to maintain guardrails
Jensen Huang says Nvidia is mass-deploying agents across every division to improve productivity . This includes:
- Autonomous verification agents for chip design (reducing verification cycles by 40x)
- Engineering simulation agents
- Security operations agents
- Software development agents
Jobs: The Huang vs. Industry Debate
Jensen Huang says AI creates jobs — and he’s not backing down from that position .
Huang’s Position
“The amount of work that we have to do to bring AI into the world is really quite incredible. So it’s creating a whole bunch of jobs. And, my software engineers love this.” — Jensen Huang
Jensen Huang says AI is the “United States’s best opportunity to re-industrialize ourselves” .
The Counterargument
Not everyone agrees. Other tech leaders have warned about job displacement:
Reality Check
Jensen Huang says his engineers are still employed — they’ve simply shifted what they do. The distinction matters: Nvidia is hiring 2026 graduates for System Software Engineer roles specifically focused on building AI agents and GPU-accelerated systems .
The Nvidia Hiring Picture
Jensen Huang says the shift to agent building is reflected in Nvidia’s hiring priorities. The company is actively recruiting 2026 graduates for roles requiring :
- Strong programming skills (Python, C++, CUDA)
- Deep learning frameworks (PyTorch, TensorFlow)
- Agentic AI frameworks (LangChain, LangGraph, CrewAI)
- Understanding of LLMs and autonomous agents
Jensen Huang says these new roles represent a shift in what software engineering looks like — not its elimination.
The Infrastructure Behind the Vision
Jensen Huang says Nvidia is building more than just chips. The company has unveiled a complete stack for agentic AI :
| Component | Purpose |
|---|---|
| Nvidia Agent Toolkit | Collection of software components for building agents |
| OpenShell | Secure runtime with governance and security controls |
| Nemotron 3 Ultra | 550B-parameter model optimized for agent frameworks |
| Vera CPU | Standalone processor for agentic AI workloads |
Jensen Huang says this infrastructure is being adopted by major partners including Cadence, Siemens, Dassault Systèmes, CrowdStrike, Palantir, Microsoft, and Red Hat .
The Bigger Picture
Jensen Huang says the era of agentic AI is here, but the transition is not without tension:
The Optimism
- Engineers are doing more creative, high-value work
- AI is creating new roles and new industries
- Nvidia is investing heavily in agentic infrastructure
The Reality
- AI coding tools still struggle with complex, production-ready code
- A METR study found AI assistants actually decreased experienced developers’ productivity by 19%
- “Vibe coding” works for prototypes but fails for maintainable code
The Bottom Line
Jensen Huang says his engineers prefer building agents to writing code. Whether this shift scales across the industry — and whether it creates more jobs than it eliminates — remains to be seen. What’s clear is that Jensen Huang says Nvidia is betting big on agentic AI as the next phase of computing.
What do you think about Jensen Huang’s vision? Would you rather build AI agents or write traditional code? Share your thoughts in the comments.
This response is AI-generated and for reference purposes only.
xplore More on Coggnix.io
- Best AI Tool for Proposal Writing: 7 Tools Tested & Compared (2026 Guide)
Best Free AI Image Generator With No Restrictions: 7 Tools That Actually Work (2026) - Best Free AI Workflow Automation Tools: 8 Tools That Save Hours Every Day (2026)
- Best AI Video Generator Free No Sign Up No Limits
This article contains affiliate links. Coggnix.io may earn a commission if you purchase through these links, at no additional cost to you. We only recommend tools we have tested and believe deliver value.
Follow us one Facebook for more Educational Content
