The question students ask me most, more than any business owner or engineer does, is "How do I get into AI?" This guide is my answer for Jamaican students from third form to final year at university, and for the parents and teachers behind them.
I have been building AI in Jamaica since 2011. I founded StarApple AI and have run free weekly AI training sessions for more than seven years, and thousands of Jamaicans have come through them. Everything below can be done from Jamaica, on a basic laptop or a phone, and most of it costs nothing.
Where the line sits between learning with AI and cheating
If AI does the thinking and you hand in the result as your own, that is academic dishonesty. If AI helps you understand the material and the work you hand in reflects your own understanding, you are using a tool well.
| Use | Verdict |
|---|---|
| Pasting a ChatGPT essay and putting your name on it | Cheating |
| Asking for a second explanation of a CAPE Physics topic your textbook covers badly | Fine: it is what a tutor does |
| Asking for twenty practice questions on integration before Additional Mathematics | Fine |
| Asking it to find the error in your chemistry calculation | Fine |
| Using any AI tool in an exam room | Cheating, and grounds for disqualification |
A simple test: if your teacher asked how you prepared the work, could you tell them without discomfort? If you would hide the AI use, you have probably crossed the line. Many Jamaican schools still have no written AI policy. If yours has one, follow it. If it does not, ask your subject teacher what is allowed before the assignment, not after.
Six study techniques that work with any AI tool
These work for CSEC, CAPE and university courses. The common thread is that you tell the tool what you already know and where you are stuck.
- Explain it differently. Do not type "explain photosynthesis". Type "Explain photosynthesis to a CSEC Biology student who understands cell structure but is confused by the light-dependent reactions. Use short steps."
- Generate practice questions. "Give me 10 CSEC Mathematics Paper 2 style questions on simultaneous equations, mixed difficulty, with worked solutions." The style will not match CXC exactly, so use these alongside past papers, not instead of them.
- Teach it back. Explain a topic to the AI in your own words and ask it to list what you got right, what you got wrong and what you left out. Teaching forces you to organise what you know.
- Build a revision plan. Give it your subjects, your weak topics, your exam dates and the hours you can study each day. Ask for a timetable with rest days built in.
- Connect subjects. "How does the probability in my Additional Mathematics course explain the ratios in Mendelian genetics?" A link between two subjects makes both easier to remember.
- Diagnose a wrong answer. Paste a past paper question you got wrong, with your working, and ask where your reasoning broke down. That tells you which gap to fill, which the mark scheme alone will not.
AI tools make confident mistakes, especially in maths and in anything specific to the CXC syllabus. Check every answer against your textbook, the syllabus document or a worked mark scheme.
A roadmap from high school to university
Using AI to study is one skill. Building AI is another, and it starts with mathematics.
High school (roughly ages 14 to 17)
- Subjects: Mathematics, Additional Mathematics if your school offers it, Information Technology and Physics. Statistics, probability and algebra are the working language of machine learning.
- Python: start with freeCodeCamp or Codecademy's free introductory course. Google Colab runs Python in a web browser, so you do not need a powerful computer.
- Sceptical use: when ChatGPT or Claude gets something wrong, work out why. That habit builds more intuition about how these systems behave than any video.
Sixth form (roughly ages 16 to 19)
- Subjects: Pure Mathematics, Applied Mathematics (the statistics units), Computer Science where offered, and Physics. If your school does not offer CAPE Computer Science, cover the ground online.
- One structured course: Harvard's CS50 Introduction to Artificial Intelligence with Python (free on edX), Google's Machine Learning Crash Course (free), or Andrew Ng's Machine Learning Specialization on Coursera (auditing is usually free; the certificate is paid, with financial aid available). Finish one before starting another.
- Projects: train an image classifier in Google's Teachable Machine, which needs no code. Then analyse a dataset you care about, such as Jamaican rainfall, cricket scores or crime statistics from JCF releases. Put each project on GitHub.
University (roughly ages 18 to 22)
- UWI Mona: the Department of Computing's BSc in Computer Science covers algorithms, data structures, databases and software engineering. Choose machine learning and data science electives and add statistics courses from the mathematics department.
- Research experience: approach a research group or lecturer working on applied AI in your first or second year and offer to help with a project. The IMPACT AI Lab at UWI is one place to start.
- UTech Jamaica: its computing programmes are more industry-oriented, with fewer AI-specific courses than UWI at present. Pair them with online specialisations.
- Depth: fast.ai's Practical Deep Learning for Coders is free and teaches by building working models first.
Free resources, grouped by what you need
| Need | Resources | Cost |
|---|---|---|
| Mathematics | Khan Academy; 3Blue1Brown (YouTube) for linear algebra and calculus; Professor Leonard (YouTube) for university maths | Free |
| Python | freeCodeCamp; Codecademy intro course; Automate the Boring Stuff with Python (free online book); Google Colab | Free |
| Machine learning | CS50 AI (edX); Google Machine Learning Crash Course; fast.ai; Microsoft Learn's AI Fundamentals path | Free to learn; Microsoft's certification exam is paid |
| Data science practice | Kaggle datasets, competitions and short courses; Google Data Analytics and IBM Data Science certificates on Coursera | Kaggle free; Coursera certificates paid, with financial aid |
| In Jamaica | StarApple AI's free weekly training sessions; UWI Department of Computing seminars; Jamaica Computer Society events; tech meetups in Kingston and Montego Bay | Free |
Most of these resources cost nothing. The students who build real skill are the ones who work through one course a week, every week, for a year.
AI careers open to Jamaican graduates, with local pay ranges
| Role | What you do | Entry pay in Jamaica |
|---|---|---|
| Machine learning or AI engineer | Build, train, test and deploy models. Needs strong Python and ML fundamentals. | About J$2–4 million a year; remote roles abroad can pay US$60,000–150,000 |
| Data scientist | Find patterns in data that change business decisions. Banks, telecoms and BPO firms hire for this. | About J$1.5–3 million; J$5–10 million with experience |
| Data analyst | Clean data, build reports and dashboards. The most accessible entry point. | About J$1.2–2 million |
| AI product manager | Decide what AI products to build and why. Suits technically literate people strong in communication. | Varies widely |
| AI researcher | Advance the field. Usually needs a master's or PhD. | Mostly abroad or in university labs |
| AI consultant | Help businesses decide where AI fits. Needs technical knowledge plus business sense. | Varies with clients |
| AI ethics and policy | Advise organisations and government on legal and social effects. Jamaica's National AI Task Force is one example of where this work happens. | Varies |
Remote work for overseas companies is open to Jamaican graduates because of English fluency and a time zone close to North America's. LinkedIn and Upwork carry remote AI and data roles. Your GitHub portfolio will count for more than your grades when you apply.
What parents and teachers can do this term
Your child or student is almost certainly using AI already. Banning it at home drives it out of sight. Ask them to show you what they use it for, and talk through the difference between using it to learn and using it to avoid learning.
For teachers, the value you add shifts from delivering information to building understanding. CXC syllabuses already weight application, analysis and evaluation through their profile dimensions, and those are the skills an AI answer cannot demonstrate for a student. Assignments that ask for local context, personal reflection, oral defence or work built on a class discussion are harder to outsource, and they assess understanding better anyway.
A school AI policy needs three columns: prohibited uses (submitting AI output as your own), permitted uses (explanations, practice questions) and encouraged uses (building AI skills). Review it every year. Individual schools should not have to work this out alone, and the Ministry of Education and Youth is the right body to issue national guidance.
Prompt templates you can copy
I am a CAPE [subject] student sitting exams in [month]. I am strong in [topics] and weak in [topics]. Build a 4-week revision plan that gives more time to my weak topics, lists past paper questions by topic for each session, and fits around study for [other subjects].
I am a Jamaican student in [form or year], strongest in [subjects], interested in AI. My family cannot pay for tutoring or courses. Build a free learning path that gets me ready for a university computer science programme, with specific free resources, monthly milestones and simple projects that show my skills.
Act as my study partner for CSEC [subject]. I will explain [topic from the syllabus]. Check my explanation for errors, name any misconceptions and gaps, then ask me three follow-up questions. After I answer, tell me what to review next.
I know basic Python and use Google Colab. Suggest 5 beginner AI projects relevant to Jamaica or the Caribbean. For each, name the dataset, the technique I would learn, how long it would take a beginner, and the build steps.
What to do next
- Students in CSEC or CAPE years: this week, pick your weakest topic in one subject and run the teach-back prompt above on it. Check the AI's corrections against your syllabus document.
- Students who want to build AI: enrol in one free course (CS50 AI or Google's Machine Learning Crash Course) and set a fixed weekly slot, for example three hours every Saturday, until you finish it. Most students can complete either in 8 to 12 weeks at that pace.
- Every student: open a free GitHub account and publish your first small project, even a notebook that charts one dataset, within a month. Add one project each term.
- Parents: sit with your child for 20 minutes and have them show you how they use AI for one assignment. Agree the rules together and write them down.
- Principals and heads of department: draft a one-page AI use policy with the three columns above before the next assessment period, and share it with students and parents.
Frequently Asked Questions
Can I use AI on my CSEC or CAPE School-Based Assessment?
Only in the ways your teacher and CXC's current rules allow, and you must be able to show the work is yours. Your teacher has to confirm that the SBA is your own work before it is submitted, so keep your notes, drafts and data as evidence of how you produced it. Using AI to understand a method is different from having it write your analysis. Check CXC's latest guidance with your teacher before you start the SBA, not at the deadline.
Do I need a paid AI subscription to study well?
No. The free tiers of ChatGPT, Gemini and Claude handle explanations, practice questions and revision plans. Paid plans cost about US$20 a month and mainly give higher usage limits. Check the age rules first: each provider sets a minimum age in its terms, some require parental permission for under-18s, and Claude's consumer app is limited to users aged 18 and over.
How do I check whether an AI answer is correct?
Compare it with the CXC syllabus, your textbook and past paper mark schemes, which are the standard your exam uses. For maths and physics, rework the answer by hand or check it with Wolfram Alpha. If the tool cites a source, open the source and confirm it says what the AI claims. If two tools give different answers, treat both as unverified until you check.
What should a beginner's AI portfolio contain?
Two to four small, finished projects on GitHub beat ten half-finished ones. Each should have a README that states the question, the data source, the method, the result and what did not work. A project that uses local data, such as Jamaican weather or school results, stands out to Jamaican employers and to admissions tutors abroad.
How can I pay for AI study outside Jamaica?
Look at postgraduate scholarships open to Jamaicans, such as Chevening for UK master's degrees, Fulbright for graduate study in the United States and the Commonwealth Scholarships. Georgia Tech's online Master of Science in Computer Science, which has a machine learning specialisation, costs a fraction of an on-campus degree and can be studied from Jamaica. For undergraduate study at home, the Students' Loan Bureau finances tertiary tuition. Deadlines for the international awards usually fall 9 to 12 months before the course starts.
Do I need to know programming to work in AI?
For engineering and data science roles, yes, mainly Python. Product management, policy, ethics, sales and training roles need less code, but basic Python and an understanding of how models are trained and tested will set you apart in interviews. A useful benchmark is being able to load a spreadsheet in Python, chart it and explain what the chart shows.