The Quiet AI Miscalculation T&T Professionals Are Making (Part 1)
The most disruptive technology in a generation is the one we're least prepared for
In the spring of 1943, a young Black woman named Miriam Mann started work at the Langley Memorial Aeronautical Laboratory. She was a mathematician. The job title on her personnel card said computer, because back then that is what they called people whose job was to compute.
She sat at a desk with a slide rule and a mechanical calculator and produced, by hand, the aerodynamic calculations engineers needed to test aircraft for the war effort. The pay was good. The work carried status. And Miriam Mann was excellent at it.
So were the dozens of other women in the West Area Computers, the segregated, Black-only unit inside Langley, most with mathematics degrees, all recruited on the same promise: do the calculations accurately, do them on time, and the government will reward you with a career you can build on. For nearly 20 years, that was enough.
Then an IBM 7090 arrived. It filled a room the size of a small ballroom and could run trajectory calculations in minutes when it took the women days. No memo was sent. The machine simply arrived and began doing the work.
Within 18 months the computing pool was dissolved. Some women learned to program the new machines and kept their jobs. Others were reassigned or retired. Their job title, computer, lingered on personnel cards for years, a ghost word inside an agency that had already moved on.
The ones left behind had done everything the institution asked of them. The degrees, the skills, 15 years of reliable service. Miriam Mann was among them. It was not enough. The world had changed on them while they were busy doing everything right.
That story is yours
If you are a working professional in Trinidad and Tobago in 2026, with a growing sense that something about the last twelve months has been off, you have something in common with the women in that story.
AI just learned to do 80 percent of the work your career was built on. Your employer is already recalculating what your role is worth. And the rules you followed to get here cannot help you with what comes next.
Truth be told, you followed the playbook your parents handed you. The degree from UWI. The post-grad credentials. Show up early, stay late, build the network, wait for the hierarchy to move.
It worked for 40 years, under specific conditions, inside a specific labour market.
Those conditions do not hold now.
The sharp, sudden death of knowledge work
Think about what you did this week.
The slides. The report. The policy document. The spreadsheet you reconciled line by line. How much of that required your judgement, your relationships, your ability to read a room?
And how much was execution you could automate?
Your employer just asked the same question. And the answer they arrived at is reshaping every job description, every headcount budget, and every hiring decision in the building.
The contract the legal secretary spent three years learning to draft? AI writes it in 40 seconds. The financial model the analyst spent a decade perfecting? A prompt. The campaign plan, the vendor evaluation, the compliance audit, the system scope — every skill in the book took the best employee years to develop. AI now replicates them for less than a monthly parking spot.
Everyone is using AI now. But the data on where it hits hardest inverts everything the old career advice promised. A recent study mapped AI exposure across 342 occupations in the U.S. Here is what it found.
Bachelor’s degree: 6.7 out of 10 for AI disruption.
No degree: 4.1.
Read that again. You’re worse off with a degree.
The most exposed roles are not factory jobs. They are not trades. They are financial analysts, compliance officers, HR managers, bookkeepers, and customer service representatives. The jobs that require a computer, a login, and countless hours of knowledge work a week. The jobs the degree was supposed to protect.
But don’t get me wrong…
The degree is not the problem in and of itself. But the degree put you behind a desk. And the work done at that desk is exactly where AI hits hardest.
I suspect some part of you knows this technology is not some shiny new object you play with then dismiss. You can tell this is transformative. What to do about it is another story.
The miscalculation
I run a careers company, so I talk to professionals across Trinidad every week. I have worked on their resumes, reviewed their career strategies, and listened to how they think about the market. And what I keep seeing is that too many of them are looking at AI through the wrong lens.
They know what AI can do with a prompt.
What most of them have not grasped is the leap that happened in October 2025 with the launch of Claude Code for the web. The AI they formed their opinions about, the one that hallucinated, needed heavy editing, and produced impressive but unreliable work, that version is already obsolete.
Six months ago, AI crossed from generating text to running near-autonomous workflows: writing and deploying production code, managing multi-step projects end to end, making judgement calls that previously required a senior employee and a week of meetings. The shift from generative AI to agentic AI erased nearly every limitation that gave professionals comfort, and it happened so fast that for most people outside the technology industry it just did not register.
Local professionals with opinions about AI, even experience using it, are judging a 2026 threat using a 2024 mental model. The distance between what they think AI can do and what it can actually do right now is where careers will be lost.
That is the miscalculation this essay is about.
What this means inside companies right now
In the United States and other developed markets, unleashing AI within the firm is already company policy. Case in point: this recent memo from Shopify’s CEO:
Every team must demonstrate why they cannot get what they want done using AI before the company will approve a new hire.— Tobi Lütke, CEO of Shopify, April 2025
Shopify is not an outlier. Jack Dorsey, the co-founder of Twitter, just announced he is shrinking his company Block because AI can now do the work that previously required more people. Not the manual jobs. The corporate ones.
Amazon recently announced its corporate workforce would shrink. Not the warehouse workforce. The corporate one. The degree jobs.
That is not a future scenario. That is a Tuesday in corporate America in 2026. And we should know what starts there does not stay there.
The hiring data tells the same story at scale. A Harvard Business School study published in February 2026 analysed nearly every job vacancy posted in the United States between 2019 and March 2025, covering more than 19,000 tasks across 900 occupations. Job postings for structured, repetitive desk work fell 13 percent after ChatGPT launched in November 2022. Postings for roles requiring human judgement, interpersonal skills, and creative problem-solving grew 20 percent. Finance and technology took the largest cuts.
The repricing
What is happening is a great reset. First the work changes. Then the employer reprices each role, which is to say, figures out how to pay less for the same work. Can you blame them?
The repricing always favours the firm.
They are in business to make money or save money. If what a company paid a full salary for three years ago now costs a fraction to produce, the business would be foolish not to act on that.
A chief financial officer in Port of Spain told me this last month. She is already using AI to do work that previously required two analysts. She has not fired anyone. She simply stopped replacing people when they left.
The headcount shrinks one resignation at a time. In fact, I don’t doubt that in unionised environments, roles could be hollowed out from the inside while the org chart stays the same. The title is still there. The salary is still hitting the account. But the ground underneath has moved. This is what I have seen with just about every major technology shift.
And the economists whose job it was to provide reassurance have stopped providing it. For decades, the standard response to any automation scare was that technology has never destroyed more jobs than it created. That consensus held through every wave, from ATMs to spreadsheets to the internet.
In April 2026, the New York Times reported that it is breaking. Labour economists who built their careers on the “technology creates jobs” position are publicly reconsidering. One University of Chicago economist called the leap to agentic AI “potentially an industrial revolution-scale event, if not more.” A Brookings Institution fellow who studies AI and the workforce put it in terms any professional in Trinidad would understand: if you can do your job locked in a room with a computer and nobody else, that job is in trouble.
That is always the order:
The market adjusts first.
Then the employer adjusts the role.
The employee is the last to find out.
What happens next
There are two roads from here. Not ten. Not five. Two.
Option 1: Do nothing
Your role gets hollowed out one task at a time. The title stays, the salary stays (for now), but the distance between what you earn and what you contribute grows wider every quarter until someone in finance notices.
It is a slow, quiet erosion. And it is already underway for thousands of professionals across Trinidad who have not yet felt it in their pay.
Option 2: Lead the orchestra
Reposition your career with this mindset: AI can play every instrument in the building. What it cannot do is lead the orchestra.
It cannot decide which instruments to play, in what order, for what purpose, or why the performance matters at all.
That is the job your employer will pay a premium for.
Not the person who can produce the work. AI does that now. The person who can look at a room full of data and see the one thing that matters. Who can scope a problem the machine cannot define. Who can walk into a meeting and set the direction everyone else follows.
The career that survives this reset is built on what you can see, scope, envision, and build. Not on what you already know.
The person who sees this shift and moves first does not just keep her job. She rewrites it.
AI is the first technology in history that gives a single professional the productive capacity of an entire team. The person who understands that does not wait for a promotion. She creates her own leverage, inside the company or outside it.
At the same time, this is not about abandoning your career. It is about repositioning it.
The professionals who will thrive understand there is a part of almost every job that still requires human judgement and instinct, the “court sense” to see the play and help the company unlock better outcomes.
There is no grace period
I hear you agreeing with me and promising to get on this soon. But if you hold one of these AI-vulnerable jobs, time is not on your side. Even the boardroom is being warned: move faster on AI or risk obsolescence.
The failure to understand the speed with which AI can reshape a career is the most dangerous part of this miscalculation.
Remember COVID? In March 2020, the virus was a news story from somewhere else. By the end of that same month, the borders were closed and the economy looked nothing like it did four weeks earlier.
One week you were at your desk. The next week you were home, wondering if your job still existed.
There was no grace period then. There will not be one now.
AI is the next shock on that scale. The difference is that COVID closed your office and eventually you went back. AI is not going to close your office.
AI is going to walk in, put its feet up on your desk, and do most of your job before you arrive on Monday morning.
Most professionals in Trinidad will not see this shift until it arrives on their desk.
You are seeing it now.
Where to start
Most of the women at Langley waited for someone to tell them what came next and lost their jobs. Many never recovered.
One did something different.
Before the IBM 7090 was operational, before anyone at NASA had asked her to, Dorothy Vaughan found a manual for FORTRAN, the programming language the new machine spoke and taught herself to use it. Then she went back to her unit and taught the women on her team.
When the transition came, she was the only person on the floor who could operate the machine and bring her people with her. Her unit transferred intact into the programming division.
These women kept their careers because Dorothy Vaughan chose augmentation over automation before anyone told her she had to.
The next two essays are about this transformation, and what it means for a Caribbean professional making the same decision right now.
To your success,
Kerry 👋
P.S. This is Part 1 of 3. Here’s what’s coming next.
Part 2 — diagnose where your career sits on the exposure map, what’s safe, what’s already being automated.
Part 3 — the playbook: specific moves to reposition yourself as the person who leads the orchestra, not the one who gets replaced by it.
Both go out to ShortlistHR subscribers only.
ShortlistHR is career intelligence for Caribbean professionals. Job roundups, salary data, and exclusive playbooks to level up your career. 👇
