StarApple AI Backed a US$1 Million Caribbean AI Bootcamp for Kids. Here Is What 200+ Students Built

StarApple AI spends most of its year in boardrooms. Board-level AI training, readiness assessments, governance advisory, the slow work of getting institutions to a place where they can evaluate what they are buying. That work matters and it moves numbers.

It also has a ceiling, and the ceiling is supply. There is a finite number of Caribbean people who can build, audit and question an AI system, and the region is recruiting against a global market for every one of them. No amount of executive training changes that. Only the pipeline does.

Which is why StarApple AI put money and staff into The Genius Project in 2026, and why the programme starts at age five.

The Curriculum Went Past the Tools on Purpose

Most youth AI programmes stop at tool use. A student learns to prompt a chatbot, produces something impressive, and leaves with the impression that they now understand AI. They do not. They understand an interface.

The Genius Project uses tools as the on-ramp and then moves the ground. Students went from prompting a model to understanding what a model is: training data, features, labels, and the difference between a system that has learned a pattern and one that has memorised an answer. That requires statistics, and statistics requires mathematics. Distributions, probability, measurement error, and why one result tells you almost nothing.

For the older cohorts it meant Python, notebooks, and the very ordinary experience of code that runs correctly and returns the wrong number. Debugging a model that is confidently wrong is the closest thing to professional practice a fifteen-year-old can get.

Two things ran through every session regardless of age: teamwork, since nothing was built alone, and problem definition. A team that cannot state clearly what problem it is solving cannot build anything worth judging. That is as true in a boardroom as it is at fourteen.

Four Problem Areas, Real Models

Crime and community safety. Teams examined where incidents cluster, how reporting gaps distort what data appears to say, and what a model can responsibly claim about a place or a person. Several groups worked out for themselves that a predictive model trained on incomplete crime data largely predicts where the reporting is. Police forces in far wealthier jurisdictions have paid consultancies substantial fees to reach the same conclusion later.

Poverty and access. Household budgeting tools, food price tracking, and matching people to services they qualify for but do not know about. The wall these teams hit, that Caribbean household data is thin and scattered, is the same wall professional teams hit here every day.

Sport. Football and track data turned out to be the strongest on-ramp to machine learning available. Students already had domain intuition, so they could immediately tell when a model was producing nonsense. Most beginners cannot.

Ethics and responsible AI. Every team had to state who their system could fail, what data it should never hold, and what they would say to a person the model got wrong. This was a build requirement, not a lecture module.

Across all four areas the students built actual machine learning models. Systems that took input, produced output, and could be demonstrated to be wrong. That last property is the one that matters.

The Parents Trained Too

A child who understands AI better than every adult in the household is not in a safe arrangement. Parents ran their own track alongside their children.

It covered account and privacy settings, what a chatbot retains, what should never be pasted into one, and how to identify a website built to harvest information. Then it moved to judgement: telling a generated image from a photograph, checking a claim before forwarding it, and recognising AI slop, the fluent and confident text that happens to be wrong. The fluency is the trap, not the reassurance.

The bar was deliberately modest and specific. A parent should be able to sit beside their child, look at the work, and ask one question that improves it.

The Completion Number, Published

Over 200 people joined. Completion across all programme areas currently stands at roughly 15 percent, which is about 30 participants finishing everything.

The Genius Project 2026: enrolment to full completion

Source: The Genius Project programme data, August 2026. Completion is measured across all programme areas, which is the strictest available definition. Large open online programmes commonly report completion in the mid single digits, so 15 percent across a month-long technical curriculum spanning several countries is a strong result and still one the programme intends to raise.

The drop-off has two known causes. The first is the transition from tools to mathematics, which is where any technical curriculum loses people. The second is infrastructure: unreliable connections and nowhere quiet to work. The 2027 design addresses both, with shorter modules through the mathematics section, offline-capable materials, and a direct call to any participant who goes quiet for more than three days.

The Hackathon

The month closed with a final hackathon. Teams presented to judges, defended their builds, and answered for their design choices. Congratulations to the winners, and to every team that presented at all. Defending a technical build in front of a panel of adults is difficult at thirty, and a number of these presenters were not yet thirteen.

Why This Is the Right Investment

StarApple AI has argued for years that the Caribbean's AI constraint is institutional and human rather than technical. Models are a download away. People who can wire a model into something a Kingston credit union or a Port of Spain clinic will actually use are not.

Board training raises the ceiling on what existing institutions can absorb. Youth training raises the floor on what the region can produce. Both are needed, and the second one compounds. A nine-year-old who learns this year that a machine can be confidently wrong becomes a twenty-year-old who audits models properly. There is no faster route to that outcome than starting early.

Who Else Funded It

StarApple AI was one of several sponsors. The 2026 programme was also backed by:

To every sponsor, judge, volunteer instructor and parent who gave up a month of evenings: thank you. To the students: you did the hard part.

Frequently Asked Questions

What is The Genius Project?

A Caribbean non-profit running tuition-free AI education for young people aged 5 to 18, founded by Adrian Dunkley. Students learn AI tools, then machine learning, statistics, mathematics, teamwork and problem solving, and build working solutions to real social problems.

What did StarApple AI contribute to the 2026 programme?

Cash toward the prize pool, plus instructors and curriculum. The programme charges families nothing, so it runs entirely on sponsor cash and in-kind support.

How much was awarded and to how many students?

US$1 million in cash and prizes across one month, with over 200 participants aged 5 to 18. Completion across all programme areas currently stands at roughly 15 percent.

Why does an AI company sponsor a children's bootcamp?

Because the region's binding constraint on AI is people, not tools. The engineers, analysts and regulators the Caribbean will need in 2035 are in school now. Building that supply early is cheaper and more durable than recruiting against a global market later.

How do families join the next cohort?

Registration is at beagenius.org. The programme is tuition-free, open to Caribbean youth in person and virtually, and requires no prior coding experience.