How Do I Prepare for a Future I Can't Yet See?

Listen to this article

How Do I Prepare for a Future I Can't Yet See?

0:00

Four fears, ten slides

Save it, share it, or send it to whoever’s asking themselves this same question right now.

Swipe to preview · 10 slidesDownload the PDF

The question underneath the question

“How do I prepare for a future I can’t see?” sounds like one question. It is really four, depending on who is asking it, and the honest answer changes each time. A 23-year-old asking it is really asking whether they will ever get the first real chance to prove themselves. A manager fifteen years into a career is asking whether the specific expertise they built is still worth anything. A business owner is asking whether the model that has fed their family for a decade still works next year. A CEO is asking it in front of a board that expects a confident answer regardless of whether one exists yet.

None of these are the same fear, even though they get compressed into the same three words: AI anxiety. What ties them together is the structure of the worry, not its content. Each of these people is being asked to commit real resources, time, money, or a career, against a picture that keeps changing before the ink dries.

Previous waves of disruption at least had the decency to move at a pace a person could plan around. The shift from typewriters to word processors took years to touch most offices. E-commerce reshaped retail over roughly two decades. This one is different mainly in tempo. It is touching hiring, product design, customer service, finance, and the boardroom simultaneously, and the tools themselves are still meaningfully different from what they were a year ago. Pew Research found that 51% of US adults are now more concerned than excited about AI’s growing role in daily life, up from 37% in 2021, and the American Psychiatric Association’s 2026 Annual Mental Health Poll found that nearly half of Americans feel more anxious than they did a year earlier, with AI now named among the contributing causes. That is not a fringe reaction. It is close to a majority position.

Every stage of a career shows up somewhere in this table. Nobody is exempt from the uncertainty, and nobody has a clean read on it either.

The fresh graduate must prove value with no track record to point to

A graduate’s problem has a particular cruelty to it. Every generation before this one could point to a resume, an internship, a grade point average, evidence that they had already done some version of the work. That evidence used to be enough to get a foot in the door, where judgment and real skill got built on the job over the following few years. Now the door itself is narrower. Nearly half of hiring managers say they would rather invest in AI tools than hire and train a recent graduate, and college seniors’ pessimism about their own prospects has jumped from 46% to 62% in a single year.

What makes this worse is that the usual advice, work harder, get better grades, add another certificate, does not address the actual problem. A graduate is not competing against AI on raw execution speed; nobody wins that race. They are competing to be the person a company trusts to notice when the AI’s confident answer is wrong, and that kind of judgment is normally built through years of exactly the entry-level reps that are disappearing first. I wrote about the shape of that gap, and where it is opening a different door, in The Entry-Level Job Market Broke. Here’s the Opening It Left Behind.

There is no shortcut around needing real reps. What seems to help is not waiting for the perfect first role to start getting them. A smaller company, a messier project, a task nobody senior wants to own, all of these still build the pattern-recognition that a resume alone cannot fake, and they build it faster than another course does.

The mid-career professional must unlearn what used to work

The mid-career fear is quieter than a graduate’s, but it runs deeper, because it attacks something that felt earned. You spent a decade or more getting good at something specific, and for most of working history, that specificity was the whole plan. Now 40% of employees say they fear losing their job to AI, up twelve points in two years, and Microsoft’s 2026 Work Trend Index found something sharper underneath that number: 65% of AI users say they are afraid of falling behind if they don’t adapt quickly, while 45% say it still feels safer to keep doing their current job the way they already know how.

That gap, between knowing you should change and choosing not to yet, is the actual trap. That gap is a rational response to genuine risk, not laziness: redesigning how you work takes time and looks uncertain, while doing the job you already know how to do feels safe, right up until it isn’t. I went into this specific dynamic, and what tends to replace raw tenure as the thing that earns pay now, in Why Experience Alone No Longer Guarantees Higher Pay.

The honest reframe here is that expertise was never really the asset. Judgment was, and expertise was just how you used to build it. That means the years already spent are not wasted. They just need to be pointed at a slightly different target: not defending the specific way you’ve always done the work, but using what you already know to catch what a model gets wrong.

The enterprise owner must rewrite the plan while running the business it describes

Business owners carry a version of this anxiety that graduates and employees don’t have to deal with: there is no separate planning department, and no pause button. Sixty-eight percent of small businesses already use AI regularly, but 77% have no formal AI policy, and only 23% of the businesses using it have given anyone formal training on how. That is not recklessness. It is what happens when you have to decide and act at the same time, with no slack in the schedule to do the redesign properly first.

What’s notable is that this gap coexists with genuine confidence. Ninety percent of small business owners say they feel capable of adopting digital tools and AI, and only 9% describe themselves as pessimistic about it. The anxiety, where it exists, is quieter and more specific than dread: it is the worry that you are building the plane while flying it, and that a competitor with a cleaner process gets to the same customers first. I wrote about how to pressure-test the assumptions underneath a multi-year plan before a board, or a competitor, does it for you, in Building Resilience: Why Your Strategic Assumptions Are Dangerous

The C-suite leader is accountable for a strategy no one can fully verify yet

At the top of an organisation, the anxiety compounds because it stops being private. Seventy-three percent of CEOs report stress or anxiety about their company’s AI strategy, and 38% describe it as high or “crippling.” Nearly four in ten say, in PwC’s own words, that their company will not be viable in ten years if it stays on its current path. What makes this harder than any other seat in this piece is that a CEO is expected to hold a confident public position on a question the entire industry is still arguing about, and that position gets judged by a board long before the results are in.

The instinct under that kind of pressure is to reach for false certainty, to pick a bold narrative and defend it, because admitting “we don’t fully know yet” feels like it undermines the confidence the role is supposed to project. That instinct is usually the mistake. I wrote about the specific failure pattern this produces, chasing an ambitious AI story while skipping the unglamorous sequencing that determines whether it lands, in The Avenger Tower Fallacy: Why Your AI Strategy Is Failing Before It Starts

What seems to help

I said at the start that this is not a five-step plan, and I want to hold to that, because I think the pretence of a clean plan is part of what makes the anxiety worse. Certainty about the future was probably never really available, at any career stage, in any era. It used to feel available only because change moved slower than a career did. The real difference now is the gap between how fast the picture shifts and how fast any of us can personally verify what’s true.

A few things seem to help close that gap, across every person in this piece. Shortening the planning horizon to something you can check, a quarter rather than five years, and revising honestly when it’s wrong. Getting real reps handling AI output directly, because judgment only builds through contact with the actual work. And staying close to people at a different stage than you, because a graduate, a manager, an owner, and a CEO are each seeing a different slice of the same shift, and none of them has the whole picture alone.

Where this leaves us

None of this promises the anxiety goes away, and I would not trust anyone who told you it should. A future you cannot see is uncomfortable to sit with regardless of your title or years of experience. What seems to change, for the graduates and the CEOs both, is not the discomfort itself. It’s what you’re willing to do while it’s still there.

I run a two-day programme for owners and mid-market leaders that starts from exactly this problem: how to make real decisions about AI without waiting for a certainty that isn’t coming. Details are on the masterclass page.