TL;DR
- Big companies are shrinking entry-level headcount in finance, marketing, and software engineering because AI can now do junior-level work. Senior and oversight roles are growing instead.
- Small businesses (under 50 employees) are stepping in... expected to hire close to a million new grads in 2026, up about 5% year over year.
- The talent market has quietly shifted in favor of small businesses: the candidates that used to go straight to Fortune 500 are now available.
- Two job paths are actually growing: AI and tech-native roles (engineers, startup operators) and hands-on AI-proof trades (service technicians, field workers).
- If you run a small business, now is a real window to hire someone hungry and AI-forward. If you're a new grad, your best move is to find a small business owner and show them what you can build.
Two articles from CNBC and Fortune landed in the same week with the same headline in different words: small businesses are picking up the entry-level hiring that big companies are quietly dropping. Fortune's title put it best... "and some of the hottest roles are gloriously AI proof."
This isn't a fluke. It's a structural shift that's been building for a while, and it matters differently depending on which side of the hiring table you're on.
Why big companies are pulling back
The short version: AI can do what a lot of junior hires used to do.
Entry-level roles in finance, marketing, and software engineering at large companies are shrinking. Not because those companies are struggling... because they don't need someone to do four hours of research that a model now does in four minutes. What they do need is experienced people who can review the output, catch the mistakes, and make judgment calls.
That's a senior-level need. So the senior roles are growing and the junior ones are contracting.
Layoff data for 2026 so far puts corporate cuts at roughly 300,000 to 430,000 just in the first half of the year. If that pace holds, it's close to a million by December. Right about where the small business hiring number lands. The net might be roughly flat... but the distribution is changing fast.
What small businesses actually want right now
A carpet cleaning business owner I talked to last week said something that stuck: "I wouldn't hire somebody from another carpet cleaning company. I'd rather hire someone green who doesn't have bad habits yet."
That framing matters more than it sounds. Small businesses aren't just filling seats. They're looking for people who:
- Can contribute on day one without a 90-day orientation
- Know AI tools well enough to actually use them, not just talk about them
- Don't bring the clock-in-and-hide-behind-paperwork energy from corporate
The CNBC piece called out something worth noting: these businesses are actively aggressive right now while the big players pull back. They're not waiting to "see how AI plays out." They're hiring people who can help them figure it out today.
That's a different ask than what big companies are making. And it's an opportunity.
If you want a quick read on where AI actually fits in your business before you start building a team around it, this AI tools checklist is a good starting point.
The two paths that are actually growing
The article flagged two job categories that are expanding, not contracting:
AI and tech-native roles. AI engineers, startup operators, people who can wire up workflows and build things. The interesting wrinkle here isn't just "know how to code." It's the person who can take a vague business problem and use AI to solve it without needing a full engineering team around them. Jackson raised a fair question on the episode: does a law firm hire a computer science major over a law major because they can actually move the business forward faster? That's a real question that's going to play out across every industry.
Hands-on AI-proof trades. Service technicians, field workers, anyone whose job requires showing up in person and doing something physical. These roles are growing partly because there's no model that can fix your HVAC unit from a server room. Figure AI is running live robot vs. human experiments in warehouses right now... the robot is winning on volume. But the trades aren't a warehouse. The complexity is different.
Pretty much: learn to use AI, or learn to use your hands. Both paths are real.
The John Henry moment we're living through
There's a piece of American folklore that fits here better than any market analysis.
John Henry was a steel driver... the guy who hammered steel rods into rock to blast tunnels for the railroad. When a steam-powered drilling machine showed up claiming it could out-drill him, he said he could beat it. He did. And then he died of exhaustion at the finish line.
The metaphor isn't "don't compete with machines." It's: if you're going to win, win smart. John Henry tried to out-brute-force a tool built to do exactly that. The better play was probably to learn how to run the machine.
That's the move right now. Not panic, not pretending the machines aren't getting better... just learning to be the person who runs them and reviews what they produce.
What this means if you're hiring
The talent pool just got more interesting for small businesses.
The candidates who might have gone straight to a big company are now available, and some of them are genuinely motivated to do real work instead of getting lost in corporate onboarding. The ones worth betting on:
- Show up already using AI tools in their own projects
- Can point to something they built or shipped, even if it's small
- Want to help grow something, not just fill a role description
The Gen Z cohort that used ChatGPT to get through college isn't lazy. They're efficient. The ones who are motivated can do work in a day that used to take a week. That's the upside. The thing you manage is making sure the output is real... checking sources, verifying facts, reviewing before anything goes out under your name. Same as you'd do with any new hire, honestly.
Only 18% of businesses actually measure the ROI on their AI use. If you hire someone to run AI for your business, track what changes. Hours saved, leads responded to, tasks that used to pile up. That's the only way to know if it's working.
What this means if you're job hunting
The default track... degree, recruiter queue, big company offer letter, three years to a decent salary... that path peaked a while ago. It's not dead, but the math has changed.
What works now: get good at AI tools, find a small business owner somewhere, and demonstrate what you can do. Work for free for a little while if you have to. Build something real for them. Show the output.
The interview of the future isn't "tell me about yourself." It's "here's a business problem, you have an hour, show me what you build." Some tech companies are already moving this direction. Small businesses will get there too, probably faster, because they can't afford to wait 90 days to find out if someone's actually useful.
If you're not sure where your skill gaps are or what move to make first, this strategy quiz is a fast way to find out what's actually blocking your next step.
The window is open right now
The college-to-big-corporate pipeline isn't going away. But it's thinning. And the talent that used to flow entirely into Fortune 500 onboarding programs is now available to every carpet cleaner, electrician, HVAC company, and marketing agency with the appetite to hire someone hungry.
That's new. It didn't used to work this way.
If you run a small business and you've been thinking about bringing in someone to help you with AI... this is probably the best hiring environment for that you're going to see for a while. The candidates exist. They're motivated. And they're not waiting on a corporate offer anymore.
Go find one.