Adam, founder · Pivotum · Fall 2026
I started building The Career Index because my daughter is getting close to choosing hers. That's the honest origin story. I was staring down a decision worth four years and a serious amount of money, and I couldn't find anything that told me what that decision actually looked like in an AI economy.
But the reaction I didn't anticipate came from a different group of parents entirely. The ones whose kid is already two years into a degree. Or graduated last spring and is six months into a job search that isn't going the way anybody promised. Their question is sharper than mine, and it usually arrives with an edge of panic: is it too late?
No. But your child is much closer to the coalface than mine, and that changes what's useful.
A parent in my position is planning against a market four years out. A parent in your position is watching your child walk into the one that exists today — where the shifts aren't a projection at all, because they've already started. Same analysis, far shorter fuse. What is still a forecast for me is this year's reality for you.
So this piece is a map of the three things I think matter most for you. Each one gets its own essay over the next few weeks, because each deserves more room than I can give it here.
AI is eating the bottom rungs of the ladder. The document review, the first-pass analysis, the basic build, the research summary — that's the most automatable work in almost every profession, which is why it goes first. It is going now.
But nobody reaches the higher rungs without having climbed the lower ones. That entry-level work was never really about output. It was how judgment got built. You did a hundred mediocre versions of a thing and slowly learned what a good one looked like.
That's the trap. Employers want experience. The jobs that used to manufacture experience are being automated. Both are true at once, and your child is standing exactly where they collide.
And this isn't a forecast any more. The clearest signal I can see is this: the youngest workers — the ones a year or two out — are starting to lose ground in exactly the jobs AI is best at, while their older, more experienced colleagues in the very same roles hold steady or grow. Junior openings are thinning while senior ones rise. And the quiet part is being said out loud now — companies admitting they've simply stopped hiring for the entry-level work a machine can already do. The bottom rung isn't wobbling. In places it's already gone.
And one detail matters more than all the rest, because it's the thread that runs through this whole piece: the damage concentrates where AI automates the work — does it instead of the person. Where it augments the work — makes the person better rather than replacing what they did — those entry-level jobs held on, and some of them grew. That distinction is the entire game. It's what I built the Index to measure, and it's the difference between a path that's disappearing and one that's about to get a tailwind.
You'll hear people say this self-corrects — that companies will notice the pipeline drying up and rebuild the junior roles. I don't think that's right, and why is the whole subject of the next piece.
Your child's advantage is not experience. They don't have any, and chasing it is a losing race.
Their advantage is that they can be native to these tools in a way most people twenty years into a career never quite will be. Not "has used ChatGPT." Native — comfortable with AI as a digital worker. Something you brief, direct, check, correct and hold to a standard, the way you would a person. Not a gadget you occasionally reach for. A worker you manage.
For someone entering work now, that can simply be normal. The way it has always been. My own generation will struggle to get there, myself included, and I do this for a living. It isn't how we were brought up and it isn't how work flows in our heads. Most of us end up either loving AI or hating it, and very few of us ever get to being genuinely at one with it.
As AI accelerates into companies — faster inside organizations than most people outside them realize — they are going to be desperate for people who are truly comfortable in this way of working. Not people who tolerate it. People who think in it.
There's a line that gets repeated so often now it's nearly a cliché, and it's a cliché because it's true: you won't lose your job to an AI. You'll lose it to someone who uses AI better than you do.
But the defensive version of that misses the point entirely. The people who work well with these tools aren't the nervous ones, quietly hoping to be overlooked — they're the busy ones. More in demand, not less, because they get more done and spend the time they save on the parts that genuinely need a person. Give a young worker real fluency and they start operating years above their experience. That's the posture worth passing on. Not AI is coming for you, protect yourself, but get so comfortable, so early, that fluency is simply how your child works.
Universities are splitting into two camps right now, and the gap between them is widening fast.
Some are leaning all the way in — rebuilding whole degrees to put AI inside the actual discipline, and, more importantly, changing what they grade. Instead of marking the finished essay, they mark the process: the drafts, the annotations, the places a student caught the tool being wrong or overconfident. That is almost exactly what I'd argue for, because it assesses the one thing that will still matter — judgment applied to a machine's output.
Others are running the opposite way — back to handwritten exams in silent, supervised rooms, the whole apparatus of catching cheats. Some of that is fair enough; there are things a person should be able to do unaided. But where it's the entire strategy, your child is being assessed against a way of working that won't exist by the time they graduate.
It's worth knowing which camp your child's university sits in, because it tells you how much of this you'll need to supply yourself.
But the more useful question still isn't what their university permits. It's how your child uses AI when nobody is watching — because that's what decides what they're actually building.
There's a version of AI use that builds nothing. Ask, receive, submit. The work looks fine and the student is hollow. The people worried about cognitive decline are seeing something real, and this is it: some struggle is load-bearing, and a student who has only ever received finished answers hasn't built anything underneath.
And there's a version that builds precisely the skills that will matter. Same tool, opposite outcome. Here's what I'd say to them directly:
That's the whole distinction, and it's worth saying to them in one sentence: use it to go further, not to skip the distance.
It's also perfectly assessable, which is why bans are the wrong remedy and something better is available. That's the third essay.
Here's the part that doesn't need to wait for any of those pieces. I've written it to them rather than to you, so read it out loud, or just forward it.
Your degree does not define you and it does not define your career. What matters is what it installed in you. Strip the subject away and look at the raw material underneath: Where are you actually curious? Where are you good at making a judgment call without complete information? How do you handle a decision under real pressure and a real clock?
Those are the things that hold up. Not the credential, not the subject, and not the job title you assumed came attached to them.
Then look sideways. Almost every field of study sits adjacent to a dozen careers nobody mentions in a course catalog, and the adjacent one is often better positioned than the obvious one. Don't discount anything until you've properly vetted it as a dead end — actually looked, actually spoken to someone doing it. Ruling things out on a vague impression is how you get funneled into the single most crowded path your degree points at.
I've written a lot of alarming sentences here, so let me be straight about the other side.
AI offers as much upside as downside. New fields, new work, real leverage for someone who arrives fluent in it. The reason I'm cautious is that the upside and the downside don't land evenly: the pain concentrates hardest on people at the low-experience end, which is exactly where your child is standing. That's a distribution problem, not a doomsday.
Companies will need them. That doesn't stop being true. But what they're hired for is shifting, and if your child's primary assets are adaptability and real AI fluency, they won't just survive it — they'll be visibly better than the alternative, early, when it counts most.
That's as far as I can tell, and I'll say so plainly when I turn out to be wrong.
Your job in this is not to have the answers. It's to make sure they aren't working from a map that expired.
— Adam, founder, Pivotum
I ended on that word on purpose. The map I'd actually hand you is free: The Career Index scores all 28 careers for AI exposure, and the pack it comes in includes a short guide written for exactly the situation you're in — the kid who's already chosen. You can build yours in about a minute at pivotum.ai/map. The next essay in this series will land there first.