Why it is harder to land a first job despite AI

HostIt feels like everyone is talking about how AI is going to change work without actually taking our jobs. We keep hearing that the total number of roles will stay the same, or even grow, but if you talk to anyone finishing school right now, the story on the ground sounds a lot more stressful.

HostWhy does it feel like the front door to the workforce is being locked even while companies say they're still hiring?

GuestIt's a really strange time because the numbers we see in big reports don't match what people are feeling. You're right that most big bosses say they're not planning to cut their total staff count because of AI. But what they're not saying as loudly is that they're changing who those people are. The entry level is where the floor is falling out.

GuestThink about what a new hire used to do in their first year at an office. They did the grunt work. They took the notes, they wrote the first drafts of basic emails, they cleaned up data, or they wrote simple bits of code. Those tasks were the training wheels of a career. Now, a senior manager can just ask a chatbot to do those things in ten seconds. So, the tasks that used to justify a junior salary are just gone.

HostSo the job itself is still being done, but we don't need a person to do it. That makes sense for the company, but it feels like we're cutting off the bottom of the ladder. If you don't have those junior roles to learn the ropes, how does anyone ever become a senior?

GuestThat's the big worry. We're seeing a hollowing out of the middle. Companies are becoming top heavy. They want people who can already do the hard stuff, the high level thinking, and the strategy. They want people who can manage the AI tools, not people who are still learning what the business even does.

GuestIt has created this bar that's way too high. I was looking at some job posts lately for entry level roles that ask for three years of experience. It sounds like a joke, but it's actually a sign of this new gap. Because the AI handles the easy stuff, the work that's left for humans starts at a much harder level. There's no more shallow end in the swimming pool. You have to jump straight into the deep end on day one.

HostThat sounds like a recipe for a massive talent shortage ten years from now. If nobody is hiring the beginners today, there will be no experts tomorrow. Surely companies see that they're eating their own future?

GuestYou would think so, but most businesses are focused on the next three months, not the next ten years. Right now, they're obsessed with a specific kind of speed. If an experienced person using AI can do the work of three juniors, the company saves a huge amount of money on health insurance, desk space, and training time.

GuestThere's also this sense that the old way of learning by doing is too slow now. We used to let a new hire mess up a little bit on a small task because that's how they learned. But when an AI can do that small task perfectly and instantly, our patience for human learning just vanishes. We have started to expect everyone to produce at the speed of software from their very first day.

HostI wonder if this changes what we even mean when we say a company isn't cutting headcount. If they keep a hundred people but they stop hiring twenty new grads every spring, the total number looks stable, but the culture of the place must shift. Does it just become a room full of older pros talking to machines?

GuestIn some ways, yes. And that creates a very lonely or sterile workplace. But there's another side to this. Some companies are realizing they can't just stop hiring young people. Instead, they're looking for a totally different set of skills. They don't care if you can write a basic report anymore. They want to know if you can spot when the AI is lying to you.

GuestThey want people with a high level of what some call soft skills, though I think that name makes them sound less important than they are. I mean things like empathy, or the ability to navigate a tricky conversation with a client, or knowing how to ask the right question to get a useful answer out of a machine. The problem is that our schools are still mostly teaching people how to do the grunt work that the machines already took over.

HostIt feels like a massive mismatch. We're still training people to be the engines of the office when they actually need to be the drivers. But even a driver needs to understand how the engine works, right?

GuestExactly. You still need that base of knowledge. If you have never written a line of code or a news story yourself, you won't know when the AI gives you something that's actually junk. We're seeing this new kind of worker who's a bit of an editor. But you can't be a good editor if you have never been a writer.

GuestThis is why some smaller firms are actually starting to pull back. They realized that their senior staff were spending all day fixing AI mistakes because they didn't have any juniors coming up who understood the craft from the ground up. There's a small movement back toward a kind of apprenticeship, where you pay a person to watch and learn even if they're not as fast as a bot yet. It's an investment in the person, not just the output.

HostI hope that catches on, because the idea of a world where you need experience to get your first job is a loop that nobody can win.

GuestThe real test will be whether we value the human spark of a beginner more than the cheap speed of a prompt.

HostThose training wheels might have been slow, but they were the only thing keeping the whole bike from falling over.

GuestCompanies are finding out the hard way that you can't build a house by only buying the roof.

HostThe ladder is still there, but we have to make sure the bottom rungs aren't made of thin air.