Your Company or Government Isn't Coming to Save You From AI

The anxiety is everywhere right now. Slack threads, LinkedIn posts, conference hallways. Every industry event I attend has the same undercurrent: What happens to us when AI gets good enough?

Qasar Younis landed on something important on Lenny's Podcast this week. The biggest AI revolution will play out in mining, farming, construction, and trucking over the next 5-10 years...not in software. "Any time there's a new technology, there's a lack of understanding," he said. The fear isn't about the technology. It's about the unknown.

He's right. And the industries with the most to gain from AI aren't the ones with too many workers. They're the ones that can't find enough.

81,000 Openings and Nobody to Fill Them

The BLS projects 81,000 openings for electricians per year through 2034...mostly replacing retirements, not growth. Employment is projected to grow 9%, much faster than average. Nearly 30% of union electricians are between 50 and 70. The apprenticeship pipeline? Around 40,000 active apprentices in IBEW programs with a completion rate below 35%. The math doesn't work.

Training takes 4-5 years. And we're layering on massive new demand from the Inflation Reduction Act, the Infrastructure Investment and Jobs Act, and a data center boom that shows no signs of slowing. Goldman Sachs projects data centers will consume 8% of US power by 2030...up from 3% today. A single hyperscale data center project requires 2,000-3,000 electricians...and there are over 400 under development right now. Microsoft's president estimates the US may need 500,000 new electricians in the next decade. NECA estimates the electrical construction industry needs 80,000 new workers annually just to keep pace...while only 7,000 new electricians enter the industry each year and 10,000 retire from it.

This isn't a hypothetical labor crisis. It's here. The same pattern holds in farming...average US farmer is 58.1...trucking, mining, and dozens of other physical industries. AI isn't the threat to these workers. It's the lifeline.

Leaning In Instead of Pulling Back

In the smart building space, I see this convergence daily. Class 4 Fault Managed Power now lets low-voltage technicians...trained in 6-18 months, not 4-5 years...install power distribution that used to require licensed electricians. PoE-based lighting systems cut installation labor 25-50%. Converged IP infrastructure reduces the number of specialty contractors on a commercial build from 7-9 down to 3.

That's not displacing electricians. There aren't enough electricians to displace. It's making buildings possible that otherwise wouldn't get built on time or on budget.

The knowledge worker version of this is the same story with different tools. The displacement risk is real, but it's not evenly distributed. It concentrates on people who stop learning. The ones doing repetitive, template-driven work and calling it expertise. The ones who see AI as something to defend against rather than build with.

Learn, Build, Ship

I spend a lot of personal time and capital learning...for me, not for my employer. It's a strategy and genuine curiosity.

In early 2025 I completed Cornell's AI Strategy Certificate through the Johnson Graduate School of Management...not because my employer required it, but because I wanted to understand the landscape well enough to build in it. That investment directly shaped what I built next.

I built Elevation...a strategy platform for smart building technology that uses AI to turn assessment data into prescriptive roadmaps. I built MomentumEQ...a career development platform because the existing tools didn't work the way coaching actually works. I built OasisAgent...an open-source autonomous infrastructure agent that uses tiered AI reasoning to detect, classify, and remediate failures.

None of that was assigned to me. All of it made me better at my day job. And all of it required leaning into the tools everyone else is debating whether to touch.

The pattern is the same whether you're a CTO, a consultant, or a cabling technician: learn the tools, apply your domain knowledge, build things that solve real problems. The people who do that aren't worried about displacement. They're too busy shipping.

Don't Wait for Permission

Yes, government needs to regulate AI. Guardrails matter. Ethical frameworks matter. I genuinely believe that.

But your company isn't going to hand you a roadmap for staying relevant. And if we're waiting for government to save us, we might be waiting a very long time.

The best hedge against AI disruption isn't a policy paper or a corporate training module. It's a side project. Something you built with your own hands...or your own prompts...that proves you can use these tools to create value. Something that compounds your domain expertise instead of competing with it.

Nobody is coming to tell you it's safe to start learning. The industries with the biggest labor gaps already figured that out...they're adopting AI fastest because they have no choice. The rest of us have the luxury of choosing. Don't waste it.