Graduating Into the AI Job Market: What I'd Tell My Younger Self

I'm going to write this as someone in his 40s who's been through a redundancy and now consults on AI transformation. I'm not going to pretend I know what it's like to be 22 right now. But I've got some perspective that might be useful, or at least more honest than what your university careers service is telling you.

Your careers adviser probably said something like "AI creates new opportunities." Which is true in the aggregate and completely useless to you personally as you stare at job listings wondering why entry-level positions require three years of experience and fluency in tools that didn't exist when you started your degree.

The entry-level problem is real

Let's not pretend it isn't. AI is hitting entry-level roles hard. The whole point of junior positions used to be that someone needed a human to do the grunt work. Research. First drafts. Data entry. Basic analysis. Administrative support. That grunt work is increasingly done by AI.

Companies that used to hire five graduates now hire two graduates and give them AI tools. The work still gets done. The headcount doesn't.

i see this directly in my consulting work. When a company implements AI in a department, the first roles that go are the most junior ones. Not because anyone's targeting young people. Because the junior tasks are the most standardised and the easiest to automate.

This isn't me being gloomy for the sake of it. This is the reality you're entering. Understanding it clearly is better than being blindsided by it.

What your degree is actually worth now

Your degree proves you can learn, commit to something for three years, and produce work under pressure. That still matters. What it doesn't prove is that you can do anything AI can't do.

If your degree taught you to write reports, AI writes reports. If it taught you to analyse data, AI analyses data. If it taught you to research, AI researches.

So what's left?

Judgement. Relationships. The ability to be in a room with another human and read what they're actually thinking. The ability to take ambiguous instructions and turn them into something useful. The ability to disagree productively with your boss. The ability to handle situations that don't have a template.

You've probably got less of this than you think (you're 22, no offence) but you've got more capacity to develop it than you think. And it's the stuff that will still be valuable in ten years when half of today's "AI proof skills" turn out to be automatable after all.

What I'd actually do if I were graduating now

Get into a role where you're close to the decision-makers. Not for the status. Because the further you are from where decisions get made, the more replaceable your work tends to be. Assistants to senior leaders, for all the unglamorous nature of the role, learn more about how organisations actually work in a year than most people learn in five. And that knowledge compounds.

Learn AI tools properly, not performatively. Don't just put "proficient in AI tools" on your CV. Actually use them. Build things with them. Have opinions about them. Know which ones are good at what and which ones are rubbish at what. Employers can tell the difference between someone who's used ChatGPT a few times and someone who's integrated it into how they work.

Stop looking for the "AI proof career." It doesn't exist. Anything anyone tells you is safe could be disrupted in three years. Instead, build transferable capabilities. Communication. Problem-solving under ambiguity. The ability to work with people who are annoying. That last one is genuinely underrated.

Consider smaller companies. Big graduate programmes are shrinking because big companies are cutting junior roles first. Smaller companies still need people who can do a bit of everything. You'll learn faster, have more responsibility earlier, and be closer to the people making decisions. The pay might be lower initially. The experience will be worth more.

The generational tension you're going to encounter

Here's something nobody tells you. You're entering a workplace where your older colleagues are terrified of AI too. They might express that by being dismissive of AI ("it's just a fad"), by being excessively enthusiastic about it ("we must transform!"), or by being resentful of you because you're younger and presumably more comfortable with technology.

None of that is your problem to solve. But understanding it helps. The 45-year-old marketing manager who seems hostile to you using AI in your work isn't hostile to you. They're scared about what it means for them. Maybe read about what that anxiety looks like from their side.

Be diplomatic. Use AI tools to do better work, not to make your colleagues feel obsolete. This is political advice disguised as career advice.

The comparison trap

Instagram and LinkedIn are going to show you people your age who seem to be thriving. Got an amazing graduate role. Already promoted. Building AI startups.

Some of that is real. A lot of it is performance. And the people who are genuinely thriving at 23 are usually in specific circumstances (family money, right connections, got lucky with timing) that aren't replicable by copying their behaviour.

Your career is going to be 40+ years long. Where you are at 23 is almost completely uncorrelated with where you are at 40. I know because i was mediocre at 23 and... well, I'm doing alright now. After a redundancy. Which sounds like a contradiction but isn't.

The financial reality

You probably have less financial obligation than anyone else reading this site. That's your actual advantage. Not your tech skills. Your freedom to take risks.

You can afford to take a lower-paying role that gives you better experience. You can afford to try something and fail. You can afford to move cities for an opportunity. Your over-40 counterparts can't do any of this without significant personal cost.

Use that flexibility while you have it. It's a genuinely finite resource.

And start building financial resilience now, even on a small salary. The financial planning stuff might seem like it's for older people but the habits start early.

One thing nobody tells graduates

The best career move you can make right now isn't learning a tool or getting a certification. It's finding one person ten years ahead of you in a career you find interesting and asking them to have coffee with you once a month. Not a "mentor" in the formal, cringe sense. Just someone who'll tell you how things actually work.

Every good career decision i made came from someone telling me something I couldn't have figured out alone. Every bad one came from following generic advice.

The one thing to do today: message someone whose career you find interesting and ask for 20 minutes of their time. Be specific about why you picked them. Most people say yes. And one good conversation is worth more than a hundred articles, including this one.