The man selling the pickaxes just told the electricians they're the real gold rush.
The Summary
- Nvidia CEO Jensen Huang told Carnegie Mellon grads there's "no better time" to start a career, even as AI-driven layoffs hit companies like Cloudflare and Snap
- Huang's message extends beyond software engineers to electricians and plumbers, whose work is essential to the physical infrastructure of AI datacenters
- AI is "closing the technology divide," allowing anyone to build useful tools, according to Huang's commencement address
- The optimism clashes with Pew Research findings of widespread public anxiety over AI's impact on work
The Signal
Jensen Huang stood in front of Carnegie Mellon's class of 2026 and delivered a message that cuts against the zeitgeist. While headlines scream about AI displacement and companies justify thousands of layoffs by citing automation, the CEO with a $186 billion net worth told new grads this is the best moment in history to start working. The dissonance is instructive.
His reasoning centers on AI democratizing creation. The "technology divide" is closing, he argued, meaning you don't need a Stanford CS degree to build something valuable anymore. But here's the twist that matters for people actually trying to plan their lives: Huang isn't just talking to the programmers in the room.
"The physical infrastructure buildout for AI creates as much opportunity for skilled trades as it does for ML engineers."
Fortune reports Huang specifically called out electricians and plumbers, saying "this is your time." That's not throwaway politeness. Someone has to wire the datacenters. Someone has to manage the cooling systems that keep H100 clusters from melting. Someone has to build the physical layer of the agent economy. The demand is real, and it's growing faster than the supply of people who know how to do the work.
The timing of this message matters. Huang launched Nvidia in 1993, right as the internet was taking off. He's drawing a parallel: the people who built fiber optic networks and server farms in the late '90s did just fine, even if they weren't coding Java applets. The infrastructure builders won too. Now we're doing it again, but at datacenter scale that makes the dot-com buildout look quaint.
Key threads in Huang's argument:
- AI lowers barriers to building, expanding who can create value
- Physical infrastructure demands skilled labor at unprecedented scale
- Career timing matters, and generational technology shifts create asymmetric opportunities
But let's acknowledge the contradiction head-on. Huang is making this optimistic case while companies like Cloudflare and Snap are laying off thousands and explicitly citing AI as the reason. Customer support, content moderation, basic coding tasks—all getting compressed or automated. A Pew study found widespread anxiety about AI's impact on work, and that anxiety isn't irrational.
The Implication
Watch where the money flows for infrastructure, not just software. If Huang's right, the next five years will see massive demand for people who can build and maintain physical systems. That means electricians, yes, but also HVAC specialists, construction managers, fiber optic technicians, and industrial equipment operators. These jobs can't be done remotely and can't be automated by an LLM. They require hands, tools, and expertise built over years.
For new grads trying to figure out their path: the choice isn't just "learn to code" or "get displaced." There's a third option Huang is pointing at—become essential to the physical layer of the digital economy. The agents need infrastructure. Someone's going to build it and get paid well to do it.