Coding or the Trades? Decoding Jensen Huang’s Career Advice in the AI Boom

Jensen Huang’s “skip coding” vs “become a plumber” advice: what’s actually going on?

In February 2024, Nvidia’s CEO Jensen Huang was quoted as saying that AI means “nobody needs to program anymore” and that the “programming language is human now”. In late 2025, he’s pushing a different line: Gen Z should skip coding and go into skilled trades like plumbing, electrical work and carpentry because the AI boom is driving demand for data centre build-outs.

This Reddit post pulls the two together and argues it’s not career advice – it’s sales. More AI users means more GPUs sold. More data centres mean more tradespeople to wire, cool and maintain them. It’s a neat theory, and worth unpacking.

Here are the source links referenced by the poster: Fortune on the trades comment and TechRadar on the Dubai remarks. You can read the original Reddit discussion here.

What Huang actually said – and what it probably means

“Everyone’s a programmer now” is about natural-language programming

“Nobody needs to program anymore. AI handles it. Programming language is human now.”

Strip out the hype and this points to a real shift: large language models (LLMs) can turn natural-language instructions into working code. That lowers the barrier to automating tasks, especially for non-developers. It doesn’t erase software engineering as a discipline; it changes the interface. Think less syntax, more system design, data quality, testing and integration.

“Become a plumber” is about physical bottlenecks in the AI boom

“If you’re an electrician, a plumber, a carpenter we’re going to need hundreds of thousands of them.”

AI at scale runs on data centres. Those need power, cooling, fibre, switchgear, and a lot of hands-on installation and maintenance. Even if prompts write code, they don’t pull cable trays, install chillers or commission substations. Huang is highlighting the limiting factor: infrastructure and skilled trades.

Are these statements contradictory? Not really – but incentives matter

Both claims can be true at once:

• AI will automate chunks of coding and widen who can “program”.
• AI growth will intensify demand for electricians, plumbers and other trades.

The Redditor’s point about incentives is fair. Nvidia benefits if more people use AI tools and if data centres get built faster. That doesn’t make the statements false, but it does make them incomplete as life advice. CEOs speak to their markets; you should plan for your skills, interests and local opportunities.

UK context: where does this leave British students and career-changers?

The UK has long-running shortages in skilled trades – a point widely discussed by industry bodies – driven by decades of university-first messaging and underinvestment in vocational routes. AI-driven data centre demand may amplify that, but it didn’t create it.

On the software side, UK employers are not ditching developers; they’re expecting them to deliver more with AI tooling, and to care about security, data protection and regulatory compliance. Whether you’re in London fintech, the NHS, or a regional SME, the work increasingly sits at the intersection of software, data and operations.

Should you “skip coding”? No – but learn the right kind

AI reduces the value of boilerplate coding and increases the value of architecture, integration and verification. If you like software, lean into:

• System design and data modelling – how components fit and scale.
• AI-assisted development – prompt fluency, code review with LLMs, test generation.
• Integration skills – APIs, event-driven systems, data pipelines.
• Security and compliance – privacy by design, logging, auditability.
• Product thinking – tying features to measurable outcomes.

The Redditor notes a human beat an OpenAI model in a coding competition; source not disclosed. The broader point stands: human judgement, problem framing and debugging still matter.

If you want a practical start, see my guide on connecting ChatGPT to Google Sheets with a custom GPT. It shows how non-trivial automation emerges from clear prompts, guardrails and small bits of glue code.

Should you “become a plumber/electrician”? Maybe – if you like the work

Skilled trades offer solid, resilient careers with real autonomy. They’re physical, licensed, safety-critical and increasingly tech-infused (think building management systems, heat pumps, EV chargers and smart metering). Data centres are one client segment among many.

In the UK, apprenticeships and vocational qualifications let you earn while you learn and avoid heavy student debt. If you’re curious, speak to local employers and training providers, visit sites, and understand the day-to-day before committing.

Practical decision checklist for UK readers

• Aptitude and appetite – Do you enjoy hands-on physical work, or abstract problem-solving and digital systems?
• Training path – Apprenticeship vs degree vs bootcamp; time-to-earning; cost.
• Local demand – What are employers hiring for within 25 miles? Look at job boards and speak to recruiters.
• Regulatory and safety requirements – Especially for electrical and gas work.
• AI as a multiplier – In either path, plan to use AI to augment your work, not replace it.

Why this matters beyond the headlines

For developers

Take the hint without the hype: coding is not “dead”, but rote coding is shrinking. Use AI to move faster and shift your value to design, integration and assurance. Keep a tight handle on data privacy and provenance to avoid accidental leakage of sensitive information.

For trades

Demand looks firm with or without AI. The AI boom may add a cyclical boost through data centre projects, but domestic retrofit, commercial fit-outs and energy transition work are longer-term pillars.

For everyone

Beware of soundbites. Huang’s comments highlight real trends, filtered through Nvidia’s interests. The smart move is to pick a path you can sustain, then compound it with AI-enabled productivity.

Bottom line: don’t build your future on a keynote

Huang’s two lines can both be right and still be self-serving. Treat them as signals, not instructions. If you’re UK-based and early in your career, the best hedge is simple: choose work you’ll stick with, focus on durable skills, and make AI your co-pilot rather than your competitor.