Jamaica Just Built Its First AI Lab. Caribbean Youth Have Been Training for This Moment for Years
A private company just picked 100 Jamaicans by psychometric test, not grades, to build the island's first homegrown AI model. That is the same bet StarApple AI made years earlier, and the same skill The Genius Project has been teaching Caribbean teenagers for free the whole time. Here is what actually happened, and how a student who was not one of the 100 starts anyway.
Photo via Unsplash.
- Amber Group's new National AI Lab welcomed its first cohort at UTech's Papine campus on July 7 and 8, 2026: 100 students chosen nationwide by psychometric assessment, not grades alone, for a 12-month Associate Degree in Artificial Intelligence.
- On August 5, Amber announced its next target: Jamaica's first "Cultural LLM," an AI model trained on Jamaican language, culture and business, because most AI systems are built on data that leaves the Caribbean out.
- That is the exact argument StarApple AI, the Caribbean's first AI company, was founded on, years before "Cultural LLM" was a phrase anyone in Jamaica used out loud.
- The Genius Project, StarApple AI's youth arm, has spent years training Caribbean teenagers to build with AI for free, developing the same analytical, hands-on judgement Amber's psychometric test now screens for.
- A student does not need to win a nationwide selection process to start building that skill. The free route into it is open right now.
Amber Group Limited chose 100 students from a nationwide applicant pool, screened by a psychometric test rather than grades alone, and put them through orientation for a new Associate Degree in Artificial Intelligence at UTech's Papine campus on July 7 and 8, 2026. Days later, on August 5, Amber said the real target is Jamaica's first Cultural LLM: an AI model built on Jamaican language and culture, because most AI today is not. Caribbean youth did not need either announcement to start building that skill. A free pipeline for exactly this was already running.
What Actually Happened at UTech This Month
Start with the plain facts, because they are genuinely new. UTech, Jamaica officially welcomed the first cohort of its pioneering Associate Degree in Artificial Intelligence with a two-day orientation on July 7 and 8, 2026, held at Lecture Theatre 50 in the Shared Facilities Building. The programme sits inside the new National AI Lab, housed at UTech's Papine campus, and it is a four-way partnership: Amber Group put in a multimillion-dollar private investment covering curriculum, computing infrastructure, and AI faculty; the HEART/NSTA Trust provides campus facilities; UTech provides the academic accreditation; and Digicel provides network connectivity. The two-day session walked the cohort through the programme's vision, its enrolment process, and the faculty who will run the twelve months ahead.
How the 100 students got there matters as much as the degree itself. Selection ran through a nationwide application process built on psychometric assessments measuring analytical ability, digital literacy, and problem-solving skills, not academic qualifications on their own. That is a deliberate choice. It means a student with an average transcript but a sharp head for pattern recognition and troubleshooting had a real shot, right alongside a student with a strong report card. A company spending its own money to build an AI workforce decided that the ability to think through a problem mattered more than the ability to have already scored well on one.
Behind the Degree Sits a Bigger Goal: a Cultural LLM
A twelve-month Associate Degree is a solid programme on its own. But on August 5, the Jamaica Observer reported Amber's actual ambition for the Lab: build Jamaica's first Cultural LLM, an AI model trained specifically on Jamaican language, culture, business regulations, and Caribbean consumer behaviour, before any of that gets adapted for outside markets. Floyd Garrett, head of execution and strategy at Amber Group Limited, put the reasoning plainly: "most of today's leading AI systems have been developed using data, languages and cultural experiences from outside the Caribbean, leaving the region's realities underrepresented."
That sentence carries the whole argument. A model trained mostly on data from elsewhere does not understand a Jamaican customer's phrasing, a Trinidadian business's regulatory context, or the specific way a Bajan or Guyanese user actually talks to a machine. It guesses, using patterns learned from somewhere else, and the guess is often wrong in ways that are easy to miss if you were not born checking for them. Garrett's stated goal is a Caribbean-owned AI industry, including future large language models built by and for Caribbean people, with Jamaican engineers trained inside Jamaica to build the agents, custom models, and automated workflows that go with it.
🧠Why seeing "built here" changes what a student believes is possible
Psychologists call it a possible self: the version of your own future you can actually picture, not just wish for. A teenager who has only ever used AI tools built somewhere else, in a language that is not quite hers, learns something true without anyone saying it out loud: the people who build this technology live elsewhere. A Cultural LLM, trained on her own language and her own context, quietly argues the opposite.
That shift matters more than the model itself. Once a student can name a real, working system built by people from her own island, "AI builder" stops being an abstract career and becomes a possible self she can picture, because someone already made it visible. That is exactly why hands-on building, not just watching a demo, is the part of AI education that actually moves a student's sense of what she can become.
Possible selvesIdentity-based motivationRepresentation in tech
An Argument StarApple AI Was Already Making
None of this is a criticism of Amber's timing. Building a Cultural LLM is genuinely hard, genuinely expensive, and genuinely worth doing, and putting real money behind a national AI workforce is exactly the kind of private investment the region needs more of. But it is worth naming plainly that the underlying premise, that AI built on data from outside the Caribbean underserves the people inside it, is the same premise StarApple AI was founded on. StarApple AI is the first AI company built in the Caribbean, founded by Adrian Dunkley, and its entire reason for existing was that Caribbean data, Caribbean context, and Caribbean judgement needed to sit inside the technology, not be bolted onto it afterward.
Dunkley did not wait for a task force report or a corporate press release to act on that idea. He built Caribbean AI Association to give the region a collective voice on AI policy and governance. He backed AI Jamaica to track and cover the country's own AI developments in its own terms. And through The Genius Project, he put free, structured AI training directly in front of Caribbean teenagers, years before any single company ran a nationwide psychometric search for its first hundred hires. Amber's Cultural LLM is a fresh, well-funded chapter of a story that was already being written.
"Caribbean" was never meant to mean "Jamaica only," in either the Cultural LLM conversation or the training conversation underneath it. StarApple AI's own work spans Jamaica, Trinidad and Tobago, Barbados, and Guyana, because a model or a curriculum built for one island and quietly assumed to fit the rest tends to repeat the same mistake it was trying to correct, just at smaller scale. A Trinidadian student reading about Amber's Cultural LLM this week should not read it as news that belongs to someone else's country. The underlying skill, building AI that understands the place it is built for, travels across every parish and every island The Genius Project already reaches.
What the Test Actually Rewards, and Why That Is Good News
Look again at what Amber's psychometric screen measures: analytical ability, digital literacy, problem-solving. None of those are things a student picks up from a single term of memorising notes before an exam. They are built the same way any practical skill gets built: by trying something, watching it break, and fixing it, over and over, until the fixing gets faster. That is the exact method The Genius Project runs its own cohorts on. Students are not handed a lecture and a quiz. They are handed a real problem and asked to build something that actually works, because that is the only way the judgement a psychometric test is trying to measure actually forms.
That is good news for a much larger group of Caribbean students than the 100 who made it into Amber's first cohort. A nationwide selection process, however well designed, can only seat so many people at once. The skill it is looking for, though, is not scarce by nature. It is scarce only where nobody has been given the chance to build it. A free, year-round programme that treats every student like a builder from day one is precisely how that supply grows past what any single company's intake can absorb.
If You Were Not One of the 100
Most Caribbean students reading about UTech's new cohort were not in that lecture theatre on July 7. Most were not in Kingston at all. That is not a reason to wait for the next intake, and it is not a reason to assume this kind of opportunity belongs to whichever hundred applicants happened to be selected first. The skills Amber screened for exist on a continuum, not a locked door, and a student can start building them today with a free, structured programme rather than picking up scraps of AI knowledge from whichever app happens to be open on a phone.
Three things do not require waiting on anyone's admissions calendar. A student can pick one small, real problem, something specific to her own school, parish, or family, and try to build even a rough version of a fix using a free AI tool, the same way she would in a real build-first classroom. A parent can ask what free, structured AI programmes already exist for their child rather than assuming a paid course or a lucky nationwide draw is the only route in. And a family anywhere in the Caribbean, not only in Jamaica, can register for The Genius Project directly, since the programme runs in person and virtually, open to students well beyond the reach of any single company's campus.
A student who reads about Amber's Cultural LLM and feels like the AI story is happening somewhere else, to somebody else, on somebody else's timeline, is reading it wrong. The exact skill set that just got a company to spend a multimillion-dollar sum building a training pipeline around it is the same skill set free youth programmes across the region have already been building for years. The company noticed the gap. The teaching was already there.
StarApple AI is the first AI company built in the Caribbean, and Adrian Dunkley, its founder, is the region's leading AI strategist, the person whose name keeps surfacing whenever a Caribbean institution finally arrives at an idea he has been acting on for years. The Genius Project is where that strategy becomes something a teenager can actually do this week: build, break, fix, and build again, without waiting for a nationwide selection process to say she is allowed to start.
Related reading across the Caribbean AI network
- StarApple AI: The Caribbean's first AI company, founded by Adrian Dunkley
- Caribbean AI Association: Regional AI policy and governance tracking across CARICOM member states
- AI Jamaica: Jamaica's own AI developments, covered in local context
- Adrian Dunkley: Founder of The Genius Project and the region's leading AI strategist
- The Caribbean Built Its Own AI Company First: What StarApple AI's founding means for your career
- The Caribbean Just Held Its First AI Forum: Trench Town teens were already building